STRENGTHENING HYDROMET AND EARLY WARNING SYSTEMS How to Improve Weathering Hydromet Services the Change in Developing Countries? Authors David P. Rogers Vladimir V. Tsirkunov Haleh Kootval Alice Soares Daniel Kull Anna-Maria Bogdanova Makoto Suwa Global Facility for Disaster Reduction and Recovery How to Improve Weathering Hydromet Services the Change in Developing Countries? Authors David P. Rogers Vladimir V. Tsirkunov Haleh Kootval Alice Soares Daniel Kull Anna-Maria Bogdanova Makoto Suwa Global Facility for Disaster Reduction and Recovery © 2019 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org The text in this publication may be reproduced in whole or in part and in any form for educational or nonprofit uses, without special permission, provided acknowledgement of the source is made. GFDRR’s Secretariat would appreciate receiving a copy of any publication that uses this report as a source. Copies may be sent to the Secretariat at the above address. No use of this publication may be made for resale or other commercial purpose without prior written consent of the Secretariat. All images remain the sole property of the source and may not be used for any purpose without written permission from the source. Notes: Fiscal year (FY) runs from July 1 to June 30; the financial contributions and expenditures reported are reflected up to June 30, 2017; all dollar amounts are in US dollars ($) unless otherwise indicated. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 Unported license (CC BY 3.0) http://creativecommons. org/licenses/by/3.0. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Citation Please cite the report as follows: World Bank, 2019. Weathering the Change: How to Improve Hydromet Services in Developing Countries?. Washington, DC: World Bank. Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Cover photo: Elen11. Design: Miki Fernández/ULTRA Designs, Inc. Table of Contents Acknowledgements ................................................................................................................................ vii List of Acronyms...................................................................................................................................... viii A Guide to Key Terminology .................................................................................................................. x Executive Summary................................................................................................................................. xii Chapter 1 Backdrop for the Guide........................................................................................................ 1 Introduction............................................................................................................................................. 2 A Snapshot of the Global Weather Enterprise...................................................................................... 2 World Bank Experience in Modernizing NMHSs.................................................................................. 5 Modernization Cannot Be Piecemeal.................................................................................................... 8 Improving Service Delivery.................................................................................................................... 10 Chapter 2 Meteorological and Hydrological Services......................................................... 15 Introduction............................................................................................................................................. 16 Category 1: Delivery Systems................................................................................................................. 17 Moving toward a Service Culture................................................................................................... 18 A Common Approach to Service Delivery..................................................................................... 20 Category 2: Production Systems............................................................................................................ 26 Monitoring and Observing Systems.............................................................................................. 26 Meteorological and Hydrological Modelling Systems................................................................. 27 Forecasting and Warning Systems................................................................................................ 29 Category 3: Support Systems................................................................................................................. 31 Quality Management Systems....................................................................................................... 31 Information and Communication Technology Systems............................................................... 32 Technology Infusion Systems........................................................................................................ 33 Capacity Building............................................................................................................................ 33 Summary of Services and Systems............................................................................................... 35 iv Weathering the Change: How to Improve Hydromet Services in Developing Countries? Chapter 3 Key Considerations for Creating Effective NMHSs....................................... 37 Introduction............................................................................................................................................. 38 Business Plan........................................................................................................................................... 38 Business Models...................................................................................................................................... 39 Strategic Planning................................................................................................................................... 44 Concept of Operations (CONOPS).......................................................................................................... 45 Section 1: Scope...................................................................................................................................... 47 Section 2: Referenced Documents......................................................................................................... 47 Section 3: Current System...................................................................................................................... 49 Section 4: Justification for and Nature of Changes.............................................................................. 50 Section 5: Concepts of the Proposed System....................................................................................... 51 Section 6: Operational Scenarios.......................................................................................................... 54 Section 7: Summary of Impacts............................................................................................................. 54 Section 8: Analysis of the Proposed System........................................................................................ 55 Summary of Changing Roles.................................................................................................................. 55 Chapter 4 Recommendations on Modernizing NMHSs...................................................... 57 Introduction............................................................................................................................................. 58 How to Carry Out NMHS Modernization Programs.............................................................................. 58 Component 1: Institutional Strengthening................................................................................... 58 Component 2: Enhancement of the Service Delivery System..................................................... 62 Component 3: Modernization of Observation and Telecommunication Infrastructures, Modelling, and Forecasting Systems............................................................................................. 62 Component 4: Project Management.............................................................................................. 64 How to Make Modernization Programs Sustainable............................................................................ 64 Conclusion................................................................................................................................................ 73 ANNEX 1 Progress Model for Service Delivery................................................................................... 74 ANNEX 2 Progress Model for Observations, Telecommunications, and Forecasting...................... 75 ANNEX 3 Progress Model for Climate Services................................................................................... 78 ANNEX 4 Progress Model for Hydrological Services.......................................................................... 79 ANNEX 5 Annotated Concept of Operations....................................................................................... 80 Weathering the Change: How to Improve Hydromet Services in Developing Countries? v List of Figures Figure 1.1 Top Global Risks Include Major Environmental Issues................................................... 3 Figure 1.2 2010-2018 World Bank Hydromet Program Growth...................................................... 5 Figure 2.1 Schematic of an NMHS as a “System of Systems”.......................................................... 16 Figure 2.2 A Matrix of the “System of Systems”................................................................................ 17 Figure 2.3 Service Delivery Systems................................................................................................... 21 Figure 2.4 Monitoring and Feedback Systems on Services.............................................................. 22 Figure 2.5 Monitoring and Observing Systems.................................................................................. 26 Figure 2.6 Modelling Systems............................................................................................................. 28 Figure 2.7 Forecasting and Warning Systems.................................................................................... 29 Figure 2.8 Quality Management Systems........................................................................................... 31 Figure 2.9 Information and Communication Technology Systems.................................................. 32 Figure 2.10 Technology Infusion Systems............................................................................................ 33 Figure 2.11 Capacity Building............................................................................................................... 34 Figure 3.1 China Meteorological Administration (CMA) Switches Operations to a Hybrid Model 42 List of Boxes Box 1.2 The Russian Federation’s Roshydromet Modernization................................................. 8 Box 2.1 A Major Role for ICT Developments.................................................................................. 19 Box 3.1 WMO and the Private Sector............................................................................................. 44 Box 3.2 Avoiding Common Mistakes.............................................................................................. 48 Box 4.1 Initial Steps to Match NMHS Modernization Needs with Opportunities...................... 60 Technical Insight 2.1 Measuring User Satisfaction............................................................................ 23 Technical Insight 2.2 Putting the Focus on Weather Impacts........................................................... 30 Technical Insight 2.3 Training Program to Strengthen Forecasting................................................. 35 Technical Insight 3.1 A Sample CONOPS............................................................................................ 46 Technical Insight 4.1 Types of Twinning Arrangements.................................................................... 67 List of Tables Table 1.1 Examples of the World Bank Hydrometeorological Projects......................................... 6 vi Weathering the Change: How to Improve Weather Services in Developing Countries? Field training on snow and avalanche risk assessment for Central Asian specialists near Almaty, Kazakhstan. Source: M. Schaer, Swiss Institute for Snow and Avalanche Research (SLF), 2019. Weathering the Change: How to Improve Hydromet Services in Developing Countries? vii Acknowledgements The authors are grateful to the many institutions and individuals that have contributed to the improvement of national meteorological and hydrological services throughout middle- and low-income countries. We are especially grateful for the hard work and persistence of the many current and former World Bank staff who have led these efforts, including: Greg Browder, Saurabh Dani, Francis Ghesquiere, Tafadzwa Randriamarolaza (Direction Générale de la Météorologie, Irvine Dube, Suranga Kahandawa, Ditte Fallesen, Haris Khan, Madagascar), Hanneke Schuurmans (HaskoningDHV), Daryl Fields, Shelley McMillan, Poonam Pillai, Anil Pokhrel, Govindarajalu Srinivasan (Regional Integrated Multi-Hazard Frederica Ranghieri, Claudia Sadoff, Prashant Singh, Lia Early Warning System for Africa and Asia), Michael Tanu Sieghart, Ulan Torobekov, Tolkun Jukusheva, Xiaolan Wang, (Permanent Representative of Ghana with the WMO), Moleni Lucy Hancock, Naohisa Koide, Tetiana Shalkivska, Suiko Tuuholoaki (Tonga Met Service). Yoshijima, among many others. The authors would like to thank the World Meteorological We are also indebted to the many discussants at the GFDRR- Organization, ECMWF, USAID, UK Met Office, National sponsored event on the modernization of NMHSs, which Weather Service, ZAMG (Austria), MeteoSwiss, Roshydromet, was co-hosted by InterMET Asia 2018 in Singapore. In JMA, and other NMHSs for the collaboration in strengthening particular, we would like to recognize the insights of Patrick of the national meteorological and hydrological services. Benichou (Meteo France International), Mike Bergin, (Bureau of Meteorology), John McHenry (Baron), Marcel Marchand The team would like to thank Yoko Kobayashi, Erika Vargas, (Deltares), Khanmany Khoun Phonh (Director General of David Tucker, and Laura Wallace for their help in the prepa- the Department of Meteorology and Hydrology, Lao PDR), ration of the book. The team thanks Julie Dana, Manager of Harri Pietarila (Finnish Meteorological Institute), Andi Sakya the Global Facility for Disaster Reduction and Recovery who (Permanent Representative of Indonesia with WMO), Luc supported this work. viii Weathering the Change: How to Improve Hydromet Services in Developing Countries? List of Acronyms ACAS Agriculture and Climate Advisory Service EPS Ensemble Prediction System AFD French Development Agency EUMETNET European Meteorological Services Network AFDB African Development Bank EUMETSAT European Organization for the Exploitation of AFR Africa Region Meteorological Satellites AI Artificial Intelligence FIF Financial Intermediary Fund AMDAR Aircraft Meteorological Data Relay GCF Green Climate Fund API Application Programming Interface G2B Government-to-Business ARSO Slovenian Environment Agency G2G Government-to-Government ASECNA Agency for Aerial Navigation Safety in Africa G2P Government-to-Public and Madagascar GDP Gross Domestic Product AWS Automatic Weather Station GDPFS Global Data-processing and Forecasting System BMKG Indonesia Meteorology, Climatology and GFDRR Global Facility for Disaster Reduction and Geophysical Agency Recovery BOBER Better Observation for Better Environmental GFCS Global Framework for Climate Services Response GIS Geographic Information System CAHMP Central Asia Hydrometeorology Modernization GPC Global Prediction Center Project GPC Global Production Center CAP Common Alerting Protocol GPS-RO Global Positioning System – Radio Occultation CAREC Central Asia Regional Economic Cooperation GTS Global Telecommunication System CMA China Meteorological Administration GWE Global Weather Enterprise CONOPS Concept of Operations HMEI Association of Hydro-Meteorological Equipment CREWS Climate Risk Early Warning Systems Manufacturers DCRMP Disaster and Climate Risk Management Project HPC High Performance Computing DMH Myanmar Department of Meteorology and IA International Advisor Hydrology IaaS Infrastructure as a Service DRM Disaster Risk Management IABM International Association of Broadcast EAP East Asia and Pacific Region Meteorologists ECA Europe and Central Asia Region IBRD International Bank for Reconstruction and ECMWF European Centre for Medium-Range Weather Development Forecasts ICAO International Civil Aviation Organization ECOMET Economic interest grouping of the National ICT Information and Communication Technology Meteorological Services of the European IDA International Development Association Economic Area Weathering the Change: How to Improve Hydromet Services in Developing Countries? ix IMO International Maritime Organization QPF Quantitative Precipitation Forecast IoT Internet of Things QTF Quantitative Temperature Forecast ISO International Standards Organization RCC Regional Climate Center LAC Latin America and Caribbean Region R&D Research and Development LAM Limited Area Model RSMC Regional Specialized Meteorological Centre LMIC Low- and Middle-Income Countries RTC Regional Training Center MENA Middle East and North Africa Region SAR South Asia Region MoldATSA Moldovan Air Traffic Services Authority SAWS South African Weather Services MOS Model Output Statistics SDG Sustainable Development Goals MOU Memorandum of Understanding SHS Moldova State Hydrometeorological Service MTR Mid-Term Review SI Systems Integrator NCEP National Centers for Environmental Prediction SLA Service-Level Agreement NFCS National Framework of Climate Services SOLAS Safety of Life at Sea NHS National Hydrological Service SWFDP Severe Weather Forecasting Demonstration NMA Romania National Meteorological Project Administration UCAR University Corporation for Atmospheric NMHS National Meteorological and Hydrological Research Service UN United Nations NMS National Meteorological Service UNDP United Nations Development Program NWS National Weather Service UNISDR United Nations Office for Disaster Risk NWP Numerical Weather Prediction Reduction O&M Operation and Maintenance WB World Bank PMD Pakistan Meteorological Department WBG World Bank Group PMU Project Management Unit WFP World Food Program PPCR Pilot Program on Climate Resilience WIS WMO Information System PRIMET Association of Private Meteorological Services WMO World Meteorological Organization PWS Public Weather Service WRMA Weather Risk Management Association QA/QC Quality Assurance/Quality Control WWW World Weather Watch QCBS Quality- and Cost-Based Selection ZAMG Austrian Central Institution for Meteorology and Geodynamics QMS Quality Management System QPE Quantitative Precipitation Estimate x Weathering the Change: How to Improve Hydromet Services in Developing Countries? A Guide to Key Terminology Academic Sector: Public and private high education Weather, climate, and water: A tag used frequently instead establishments and non-profit research institutions. of meteorology and hydrology. Meteorology is inclusive of weather and climate, and these terms are interchangeable. Business Plan: A roadmap for an organization outlining goals Water refers to hydrology and occasionally to oceanography. and how those goals will be achieved – it can be used to The term meteorological embraces both meteorological and justify increased financing. climatological phenomena. Business model: Describes how an organization creates, NHMSs: An abbreviation that encompasses both National delivers and captures value. Meteorological Services (NMSs) and National Hydrological Concept of Operations: A document describing the Services (NHSs). The abbreviation NMHS also refers to a characteristics of a proposed system from the view point of an national hydrometeorological service (if hydrology and individual who will use that system. meteorology are combined in a single institution). Global Weather Enterprise: A term coined to describe the Hydrometeorology: is a branch of meteorology and hydrology totality of activities by individuals and organizations to enable that studies the transfer of water and energy between the land weather information to be created and provided to society. It surface and the lower atmosphere. encompasses the public, private and academic sectors. Objective Forecast: is one made without the personal Meteorology: The scientific study of the Earth’s atmosphere judgement of the forecaster. as it relates to short-term weather and long-term climate Operating model: Describes the underlying arrangements of variations. people, processes, systems and information needed to execute Hydrology: The scientific study of the Earth’s water system. the business model. Impact-based Forecasts and warnings: Forecasts and warnings Private sector: Means that part of the economy run by designed to express the expected impacts as a result of the individuals or groups. expected weather. They require information on the hazard and Public Sector: means that part of the economy run by the state. the vulnerability of those affected. Strategic plan: A document that articulates the decisions made Impact Forecast and Warnings: Forecasts and warnings about an organization and the organization’s goals and the designed to provide detailed information down to the ways it will achieve those goals. individual, activity or community level. They require information on the hazard, and the vulnerability and exposure World Meteorological Organization: An intergovernmental of those affected. organization with 191 Member states and territories with the purpose of facilitating worldwide cooperation in the Meteorological and hydrological hazards: Flash floods, river establishment of stations for making meteorological, floods, thunderstorms, tropical cyclones, and other extreme hydrological and other geophysical observations; promoting weather-related events, as well as slow-onset hazards, such as the establishment and maintenance of systems for the droughts. rapid exchange or meteorological and related information; Weathering the Change: How to Improve Hydromet Services in Developing Countries? xi promoting standards; furthering the application of Synoptic meteorological network—a network of stations at meteorology to aviation, shipping, water problems, agriculture which surface and upper-air observations (at locations and other human activities; promoting activities in operational that give meteorological data representation of the area in hydrology and close cooperation between meteorological and which they are situated, that could range many hundreds of hydrological services; and encouraging research and training. kilometers) are made at standard times (i.e., main synoptic times: 0000, 0600, 1200, and 1800 UTC (Universal Time Forecasting: The application of science and technology to Coordinated); and intermediate synoptic hours: 0300, predict the state of the atmosphere for a given location on 0900, 1500, and 2100 UTC) for the purpose of presenting a timescales of hours to years. Forecasts are often referred to comprehensive and nearly instantaneous picture of the state as nowcasts (from 0 to 6 hours), very-short-range weather of the atmosphere forecasts (up to 12 hours), short-range weather forecasts (from 12 to 72 hours), medium-range weather forecasts (from 3 to 10 Mesoscale—in meteorology, it is the study of weather systems days), extended-range weather forecasts (from 10 to 30 days), of horizontal resolutions generally ranging from around 5 and long-range forecasts (from 30 days to 2 years). There are km to several hundred kilometers, and temporal resolutions also monthly, trimonthly, and seasonal outlooks (covering, typically ranging from 1 hour to 1 day (WMO 1992). Examples of for example, December to February, March to May, June to mesoscale weather systems are: sea breezes, squall lines, and August, or September to November) and longer term climate large convective cells. Vertical velocity often equals or exceeds predictions (from years to centuries). horizontal velocities in mesoscale weather systems due to non- hydrostatic processes such as buoyant acceleration of a rising thermal or acceleration through a narrow mountain pass. xii Weathering the Change: How to Improve Hydromet Services in Developing Countries? Executive Summary Climate change is impacting the lives and livelihoods of people everywhere. More people are vulnerable and exposed to the effects of extreme weather, especially through the impact of floods, landslides, droughts, heat waves and strong winds. One way to cope is by reducing exposure to harm by provid- and Tsirkunov (2013a) introduced some of the basic concepts ing more accurate, reliable, and actionable weather, climate, of how and why NMHSs function and the role of the WMO as and hydrological information. This is amply demonstrated in the facilitator of worldwide cooperation. With that work as a high-income countries, where the impact of extreme weather starting point, this guide focuses on modernization of NMHSs events is mitigated by their ability to take early action based with the aim of helping ensure the services are fit-for-purpose on meteorological and hydrological warnings. This is possible in the 21st century. because these countries have invested in their publicly financed Providing meteorological and hydrological services is com- National Meteorological and Hydrological Services (NMHSs), plex. The main purpose of meteorological and hydrological encouraged the development of complementary private ser- services is to enable the public and economic sectors to make vices, and invested heavily in research and development. appropriate decisions when faced with weather, climate, and The gap between the most and least advanced meteorolog- hydrological hazards. Regardless of the level of development, ical and hydrological services is too large. Despite efforts the NMHS and its partners need to be able to: to address the disparity between the more capable and less ■■ Make meteorological and hydrological observations. capable in the most vulnerable countries, a significant gap in the availability of reliable meteorological and hydrological ■■ Combine this information with products generated by the information remains. Given the risks to lives and livelihoods, it WMO community—generally in the form of gridded numerical products, which assimilate observations from everywhere is imperative that all countries have access to, and can use, the into numerical weather prediction systems. best available information wherever it originates. The World Meteorological Organization (WMO) encourages cooperation ■■ Make accurate and timely forecasts and warnings relevant to among its Members, which goes a long way to addressing the their national users. gaps. However, significant investment is still needed in many ■■ Disseminate this information, using diverse means to match low- and middle-income countries (LMICs) to enable them to different sectors of society. This information has to be build on this cooperation by sustaining their own national mon- useful to the users if it is to support appropriate behaviors, itoring, forecasting, and service delivery infrastructure. especially during extreme weather events. Modernizing NMHSs. Equipping countries with the tools Often, too great an emphasis is placed on buying observation and techniques they need has been a pillar of World Bank equipment and not enough on the delivery of services using the strategy for many years. Considerable knowledge has been “best available” information. Information and Communication gained through this experience. In an earlier work, Rogers Technology (ICT) is central to a modern NMHS. It is the means Weathering the Change: How to Improve Hydromet Services in Developing Countries? xiii by which all data are managed, and products and services It is useful to adhere to some basic practices—road mapping, generated. We cannot overestimate the importance of a central business planning and modelling, and strategic planning. Tools, system for data management and the need for data policies, such as the development of a Concept of Operations provide which require all suppliers of equipment, especially observing the means of developing consensus among stakeholders so that systems, to comply. there is a common understanding and support for the proposed Investments must be sustainable. Many development projects operational system. The rigor imposed allows the NMHS and focus on meeting the NMHSs’ demands for more observations; stakeholders to understand the implications of any change in however, sustaining this capability is difficult and the benefit the current system—which may provide new capabilities, but of the investment often short-lived. Convincing governments also increase operating costs, in turn affecting the overall sus- to follow borrowing with investment to sustain their pubic tainability of the organization. services is a challenge—one for which there is no easy remedy. Reference material. Throughout the guide, we refer to many Convincing development partners that donating equipment WMO publications that serve as foundation for the provision without providing the resources to maintain that equipment is of meteorological and hydrological services worldwide. In equally challenging. addition to the referenced documents, teams are encouraged to The approach. A systematic approach is recommended. access the many relevant WMO publications available through Preparation of complex projects requires considerable effort. the WMO e-library.1 https://library.wmo.int 1 Purpose of this guide This guide aims to help World Bank task teams and development practitioners—as well as NMHSs, which are, or may be, involved in working with national governments—to improve the delivery of national meteorological and hydrological services to their citizens and economies. It touches on all actors involved in the production and delivery of these services, with an emphasis on the role of the public sector. The guide provides insights into how to improve the skill, efficiency, and cost-effectiveness of publicly funded NMHSs so that they can carry out their mandate to protect lives, livelihoods, and property, and are able to support economic development. xiv Weathering the Change: How to Improve Hydromet Services in Developing Countries? Overhead trolley at a hydrological post on the Djety-Oguz River, Kyrgyz Republic. Source: U. Torobekov, 2018. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 1 Chapter 1 Backdrop for the Guide Climate change is impacting the lives and livelihoods of people everywhere. More people are vulnerable and exposed to the effects of extreme weather, especially through the impact of floods, landslides, droughts, and winds. > Extreme weather, natural disasters, failure of climate mitigation and adaptation, and water crises are among the top five risks that will have the biggest impact on societies and economies in the next ten years. > The Global Weather Enterprise (GWE)—comprising public, private, and academic sectors—is crucial for the provision of accurate and reliable weather information and services that save lives, protect infrastructure, and enhance economic output. > It is recognized that meeting the challenges of a 21st century society—especially, the 2030 Agenda for Sustainable Development—now requires significant growth in all areas of the GWE. > While much of the growth will be realized through the private sector, the public sector institutions—notably, National Meteorological and Hydrological Services (NMHSs)—are central to meeting each government’s responsibility to warn and protect people from harm caused by extreme weather. > Over the past decade, investments in NMHSs have been growing, but sustaining the improved services remains a big challenge, with many instances of equipment falling into disrepair. 2 Weathering the Change: How to Improve Hydromet Services in Developing Countries? Introduction The societal need for more accurate and reliable weather, This chapter examines the current state of the GWE, the need climate, and hydrological information is growing fast as pop- for modernizing National Meteorological and Hydrological ulation density and migration increases and climate change Services (NMHSs), and the difficulties of sustaining improve- takes place. Nowhere is this need more acutely felt than in ments once they occur. In effect, it sets up the rest of this low- and middle-income countries (LMICs). The Global Risk report. Chapter 2 takes an in-depth look at NMHSs, high- Perception Survey 2017–2018 (WEF 2018) lists extreme lighting the unacceptably big gap between the capabilities weather events, failure of climate change mitigation and of the most and least advanced ones; Chapter 3 examines adaptation, and water crises among the top five risks that why NMHSs need to focus more on providing the services will have the biggest impact in the next 10 years (Figure 1.1). that their stakeholders need and want; and Chapter 4 pro- The four environmental risks all have a higher-than-average vides detailed recommendations for modernizing NMHSs. likelihood of occurrence and are tangibly affecting human well-being, including health and economic prosperity. Similarly, the World Economic Forum highlights the essential A Snapshot of the need to build a shared future that better predicts extreme Global Weather Enterprise risks and fosters the resilience required to mitigate these What exactly is the GWE? It is an enabling environment risks. More than 1 billion people have lifted themselves out fostering global engagement between public, private, and of poverty in the past 15 years, but climate and disaster risks academic sectors that shares the common goal of providing threaten these achievements (World Bank Group 2017). Global accurate and reliable weather information and services that asset losses from disasters are now reaching an average of save lives, protect infrastructure, and enhance economic more than US$300 billion a year, which is more than the com- output. It includes the scientific research, technology, bined gross domestic product (GDP) of Sri Lanka, Ethiopia, observations, modelling, forecasting, and forecast products Myanmar, and Costa Rica. A recent World Bank report finds that need to come together to reach this goal—a goal that is that the impacts of disasters on well-being are equivalent to fully aligned with the requirements of the universally agreed a US$520 billion drop in consumption—60 percent more than Sustainable Development Goals (UN 2015a), the Sendai the asset losses usually reported—and force some 26 million Framework for Disaster Risk Reduction (UNISDR 2015), and people into poverty every year (Hallegatte et al. 2017). the Paris Agreement (UN 2015b). How can these environmental risks be reduced, and losses So far, the GWE has been successful. But it now needs to shift avoided? Because weather, climate, and the water cycle into a higher gear with a better and more frequent dialogue know no national boundaries, international cooperation is and codesigned initiatives between the various actors in the essential for people and society to get timely access to high public, private, and academic sectors (Thorpe and Rogers, quality and actionable information to mitigate the threat of 2018). Here, a key element will be a continuing improve- meteorological and hydrological hazards. This international ment in the NMHSs—which form the public backbone of the cooperation takes place in what is called the Global Weather GWE—as well as more private sector investment. A failure Enterprise (GWE)—a term coined to describe the totality to manage these changes to the GWE may have detrimen- of activities by individuals and organizations to enable tal consequences, as competition between the public and weather information to be created and provided to society. private sector could become the norm, instead of mutual The GWE is a supreme exemplar of the value of international cooperation—with significant implications for NMHSs’ mod- cooperation, public-private management, and scientific ernization projects. technological know-how. But it now must be strengthened to match the needs of a 21st century society (Thorpe and For the World Bank, modernizing NMHSs is a relatively new Rogers 2018). area of technical assistance and investment support. Until Weathering the Change: How to Improve Hydromet Services in Developing Countries? 3 FIGURE 1.1 Top Global Risks Include Major Environmental Issues (The Global Risks Landscape, 2018) 4.0 Weapons of mass destruction Extreme weather events Natural disasters Failure of climate-change mitigation and adaptation Water crises Cyberattacks Biodivesity loss and Food crises ecosystem collapse Large-scale involuntary migration 3.5 Spread of infectious Interstate conflict diseases Man-made environmental 3.40 disasters ▲ average Critical information Profound social Failure of national infrastructure breakdown instability governance Fiscal crises Terrorist attacks Failure of regional or Unemplyment or Underemplyment Data fraud or theft global governance Asset bubbles in a major economy Failure of critical State collapse or crisis infrastructure Failure of financial Energy price shock mechanism or institution Adverse consequences of Failure of urban planning technological advances 3.0 Impact Illicit trade Unmanageable inflation Deflation 2.5 3.0 ▲ 4.0 4.5 3.48 plotted area average Likelihood 5.0 Top 10 risks in terms of Top 10 risks in terms of 1.0 5.0 Likelihood Impact Categories 1 Extreme weather events 1 Weapons of mass destrcution 2 Natural disasters 2 Extreme weather events Economic 3 Cyberattacks 3 Natural disasters 4 Data fraud or theft 4 Faliure of climate-change mitigation and adaptation Environmental 5 Failure of climate-change mitigation and adaptation 5 Water crises 6 Large-scale involuntary migration 6 Cyberattacks Geopolitical 7 Man-made environmental disasters 7 Food crises Societal 8 Terrorist attacks 8 Biodivesity loss and ecosystem collapse 9 Illicit trade 9 Large-scale involuntary migration Technological 10 Asset bubbles in a major economy 10 Spread of infectious diseases Source: World Economic Forum Global Risks Perception Survey 2017–2018. 4 Weathering the Change: How to Improve Hydromet Services in Developing Countries? IMPACTS OF INCONSISTENT WEATHER WARNINGS ON PUBLIC BEHAVIOR T he coexistence of public and private weather providers The results showed that most people regularly consult weather can lead to a confusing multiplication of warnings, information and do so from several weather providers. Half of sometimes undermining the official public authority. the people who had received warnings from different providers In many countries the public can easily access inconsistent for an event indicated that these were inconsistent. A further information—increasingly so on their mobile devices—that evaluation of warning quality and intended response actions can negatively impact effective reactions and responses to in a decision scenario characterized by two severe rainfall warnings. warnings showed the negative impacts of these inconsistencies. Not surprisingly, consistent warnings were least confusing and On a rainy day in November 2017, different weather information fully inconsistent warnings were considered most confusing. providers issued a variety of extreme weather warnings for the southernmost part of Switzerland around Lake Lugano. As Figure 1 shows, perceived warning quality and the likelihood Warnings were issued by the national meteorological service to engage in risk minimizing behaviors were significantly higher (MeteoSwiss), the national broadcasting service (SRF Meteo), for consistent warnings than for inconsistent warnings. There and two commercial providers (MeteoNews and Meteocentrale). were no significant differences in the impacts of inconsistent They all had different visualizations (colors of warning levels), textual information compared to inconsistent visual information based also on different levels of forecasted rainfall. As on warning quality and intended actions. Enhanced cooperation between multiple weather providers is clearly warranted, in compared to the ultimately observed rainfall, only the warning particular to agree on consistent delivery of information to delivered by MeteoSwiss could be considered “correct.” incentivize risk mitigating behavior by the public. Following the event, a survey was conducted to examine whether the inconsistency of the warnings made people doubt Authors the whole information package (Weyrich et al. 2019). A simple Philippe Weyrich,1 Anna Scolobig,2 and Anthony Patt1 matrix was developed to categorize the level of inconsistency 1 Climate Policy Group, Department of Environmental Systems Science, Swiss Federal Institute of Technology (ETH Zurich), 8092 Zurich, Switzerland. across multiple warnings, considering both visual (warning 2 Environmental Governance and Territorial Development Institute, University colors) and text information. Four combinations were assessed: of Geneva, 1205 Geneva, Switzerland. consistent warnings (“consistent”); inconsistent visual Reference warnings (“inconsistent visual”); inconsistent textual warnings Weyrich, P., Scolobig, A., and Patt, A. 2019. Dealing with inconsistent weather (“inconsistent text”); and inconsistent visual and textual warnings: Effects on warning quality and intended actions. Meteorological warnings (“fully inconsistent”). Applications, Doi: 10.1002/met.1785. FIGURE 1 Intended actions and evaluation of warning quality 4.5 Take protective 4 action Mean likelihood The warnings are in agreement 3.5 Take information- seeking action 3 The warnings are easy to understand 2.5 Consistent inconsistent visual inconsistent text fully inconsistent Note: Mean likelihood represents how much participants agreed or disagreed with each statement on a five-point Likert scale from ‘not at all’ to ‘very’. Intended actions are represented by dark and blue lines. Warning quality variables are represented by yellow and orange lines. Error bars = 95% confidence intervals (CI). Weathering the Change: How to Improve Hydromet Services in Developing Countries? 5 World Bank Experience in Modernizing NMHSs the mid-1990s, investments in NMHSs were structured as hydrometeorological investments and a gradual shift toward small-scale activities within water resource management, developing more complete systems—upgrades that serve all agriculture, or emergency operations—typically motivated or many users and explicitly aim at being sustainable. In 2010, by flooding and flood control (Hancock and Tsirkunov 2013). there were about 25 investment projects, or projects with The aim was to improve forecasting for more than one user, components directed to hydrometeorological modernization, but it did not address the needs of all users. The approach with total funding of US$197 million (50 percent of which was fragmented, contributing to weather and climate ser- was the Russia Hydromet Modernization Project). In 2018, vices designed to serve a single sectoral user and partial these figures increased to almost 70 projects or components, systems. A systemic problem appears to have been the with total funding of US$888 million—and for 2019, another failure to integrate observations and forecasting into warn- US$147 million is in the pipeline. At the regional level, while ing systems and services that benefit all weather-sensitive Europe and Central Asia led at US$80 million in 2010, by socioeconomic sectors. 2018, South Asia was leading at US$453 million, followed by Africa at US$239 million (Figure 1.2). In addition, a review by the World Bank Independent Evaluation Group indicates that the sustainability of hydrome- Some of the initial major systems projects, which are now teorological investments was an issue, especially for Africa, complete, were implemented in Poland, Turkey, and Russia where the fragmented approach was more common until (Table 1.1). At the time, Russia’s project was ranked as the recently (Chomitz, Akhmetova, and Hutton 2012). It notes World Bank’s largest hydrometeorological modernization that in Africa, the network of hydrometeorological stations is project at US$173 million (2005–13). But since this project sparse and deteriorating, hydrometeorological data are often could not address all problems accumulated during more spotty and inaccurate, and existing stations are often not func- than 20 years of underfunding, a second modernization tioning or fail to communicate with the global meteorological program is now under way (Box 1.2). In Poland and Turkey, network. Only 4 of 12 completed African projects reported hydrometeorological modernization was triggered by a attention to maintenance, and only in the Senegal River Basin major disaster that convinced governments of the impor- did the self-evaluation report consider sustainability to be tance of a well-functioning NMHS for early-warning systems likely (Chomitz, Akhmetova, and Hutton 2012). and mitigating natural hazard risks. Since the mid-1990s, there has been an overall rising trend in FIGURE 1.2 2010-2018 World Bank Hydromet Program Growth 500 450 400 350 300 US$ million 250 200 150 100 50 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 AFR EAP ECA LAC MENA SAR Source: GFDRR Assessment. 6 Weathering the Change: How to Improve Hydromet Services in Developing Countries? TABLE 1.1 Examples of the World Bank Hydrometeorological Projects Total Project Hydromet/EWS Status (as of Project title Funding (US$) Funding (US$) Description January 2019) Mexico: Water Resources Management 186.5 million 41.0 million Overall modernization of meteorology Complete Project (1996–2002) and hydrology Poland: Emergency Flood Recovery Project 200.0 million 62.0 million Observation networks, forecasting, Complete (1997–2006) service improvement Peru: El Niño Emergency Assistance 150.0 million 7.0 million Forecasting Complete (1997–2002) Dominican Republic: Hurricane Georges 111.1 million 10.0 million Observation networks and forecasting Complete Emergency Recovery Project (1998–2002) Turkey: Emergency Flood and Earthquake 369.0 million 26.0 million Observation networks and forecasting Complete Recovery Project (1998–2005) Russian Federation: Hydromet 173.0 million2 173.0 million Institutional strengthening, observation Complete Modernization Project I (2005–2013) networks forecasting, and service delivery Albania: Disaster Risk Mitigation and 9.16 million 1.8 million Strengthening of hydrometeorological Complete Adaption Project (2008–2012) services Moldova: Disaster and Climate Risk 10.0 million 5.2 million Observation networks, forecasting, and Complete Management Project (2010–2016) agrometeorology Sri Lanka: Dam Safety and Water 143.0 million 8.5 million Observation networks (mostly Complete Resources Planning (2008–2018) hydrological) and forecasting India: Bihar Koshi Basin Development 250.0 million 30.0 million Observation networks and flood Under Project (2011–2023) forecasting implementation Afghanistan: Irrigation Restoration and 219.7 million 28.9 million Establishing hydromet service Under Development Project (2011–2020) implementation Central Asia: Hydrometeorology 39.2 million 39.2 million Institutional strengthening, observation Under Modernization Project (2011–2021) networks and forecasting, and service implementation delivery Mozambique: Climate Resilience: 21.0 million3 21.0 million Institutional strengthening, observation Under Transforming Hydro-Meteorological networks and forecasting, and service implementation Services (2013–2019) delivery Nepal: Building Resilience to Climate- 31.0 million 31.0 million Institutional strengthening, observation Under Related Hazards (2013–2019) networks and forecasting, and service implementation delivery Second Russia Hydromet Modernization 179.4 million4 179.4 million Institutional strengthening, observation Under Project (2013–2021) networks and forecasting, and service implementation delivery DRC: Strengthening of Hydro- 8.0 million 8.0 million Institutional strengthening, observation Under Meteorological and Climate Services networks and forecasting, and service implementation (2017–2022) delivery Ayeyarwady Integrated River Basin 100.0 million 30.0 million Institutional strengthening, observation Under Management (AIRBM) Project (2013-2020) networks and forecasting, and service implementation delivery Bangladesh Weather and Climate Services 113.0 million 113.0 million Institutional strengthening, observation Under Regional Project (2014-2022) networks and forecasting, and service implementation delivery 2 Including government counterpart funding 3 Including co-financing 4 Including government counterpart funding Weathering the Change: How to Improve Hydromet Services in Developing Countries? 7 TABLE 1.1 Examples of the World Bank Hydrometeorological Projects (cont.) Total Project Hydromet/EWS Status (as of Project title Funding (US$) Funding (US$) Description January 2019) Lao PDR Southeast Asia Disaster Risk 31.0 million 10.0 million Institutional strengthening, observation Under Management Project (2016-2022) networks and forecasting, and service implementation delivery Pakistan Hydromet Modernization and 210.0 million 210.0 million Institutional strengthening, observation Under Climate DRM Services Project (2017-2023) networks and forecasting, and service implementation delivery Burkina Faso: Strengthening Climate 31.0 million 31.0 million Institutional strengthening, observation Under Resilience (2018-2024) networks and forecasting, and service implementation delivery country’s hydro-meteorological, early warning and response systems and services in targeted areas. Mali Hydrological and Meteorological 47.5 million 47.5 million Institutional strengthening observation In preparation Services Modernization Project networks and forecasting, and service delivery Sri Lanka Climate Resilience Improvement 100.0 million 30.0 million Institutional strengthening, observation In preparation Project networks and forecasting, service delivery Since 2012, a new wave of hydrometeorological modern- Thus, innovative ways need to be found for staffing NMHSs ization projects have been initiated in Mozambique, Nepal, in these countries. Such methods may include: Vietnam, the Republic of Yemen, Myanmar, Bangladesh, ■■ Establishing special, higher remuneration categories of em- Bhutan, Sri Lanka, Pakistan, Laos, Tonga, Samoa, Mali, ployment for technical posts in the public sector. Burkina Faso, Niger, Democratic Republic of Congo, and Federated Republic of the Marshall Islands. Now many, ■■ Using contract personnel. albeit not all, of these projects are examples of an end-to-end ■■ Providing incentives through specialized training and estab- approach, focusing on three key areas of activities—institu- lishment of career paths. tional strengthening, service delivery, and modernization of ■■ Using public-private partnerships. the observation and forecasting systems required to improve services (Rogers and Tsirkunov 2013a). ■■ Outsourcing functions. A big problem for many of these countries is retaining highly Another systemic problem is ageing of qualified staff in the skilled employees, such as forecasters and information NMHSs. One of a few countries that found a solution to this technology personnel. Competition with the private sector issue is Indonesia, where the Meteorological, Climatological is substantial as economies develop, and the capacity and and Geophysical Agency (BMKG) provides targeted support flexibility to reward is much higher in the private sector. In and opportunities for growth for young qualified staff. The addition, there is government resistance to higher remu- agency proudly reports that 70 percent of its workforce is neration for those with high qualifications—often technical younger than 40 years. workers are paid at the same level as unskilled workers. 8 Weathering the Change: How to Improve Hydromet Services in Developing Countries? BOX 1.2 The Russian Federation’s Roshydromet Modernization In the mid-1980s, the capacity of Roshydromet to provide ■■ Increased reliability of seasonal flow forecasts in the pilot services to Russia and globally was steadily declining, mostly river basins. owing to inadequate funding, which was below 50 percent of ■■ Introduction of client satisfaction surveys in Roshydromet the basic operational needs. The decline affected all elements performance evaluation. of the system: (a) observation networks; (b) data transmis- ■■ Development and government approval of a long-term sion, archiving, and processing facilities; (c) forecasting; Roshydromet strategy, which led to a massive increase in (d) research and development facilities; and (e) workforce government support to the agency—from US$76.6 million quality. After more than 20 years of decline, the government in 2003 to US$570 million in 2011. recognized the challenge of modernizing such a complex hydrometeorological system and requested World Bank The Second Hydromet Modernization project started right assistance in preparing and implementing a modernization after the end of the first one in 2013 to ensure continuity of project. This request was processed as the first self-standing the modernization process. Its objective is to further enhance and integrated hydrometeorological modernization project in the capacity of the national hydromet to deliver reliable and the World Bank’s portfolio. timely weather, hydrological, and climate information to the Russian public and economic sector. The project amount The project’s objective was to increase the accuracy of fore- is US$179.4 million, of which the International Bank for casts provided to the Russian people and economy by mod- Reconstruction and Development (IBRD) loan accounts for ernizing key elements of the technical base and strengthening US$60 million. Roshydromet’s institutional arrangements. The project began in 2005 and was finished in 2013. At US$173 million, it was Several key activities are focused on hydrology and rehabilita- the World Bank’s largest hydrometeorological modernization tion of the hydrological observation network along the Volga project. The project was a success, reaching or exceeding the River. Russia has also upgraded the computing capacity by main agreed performance indicators: procuring a 1.2 petaflop supercomputer for the Main Computer Center of Roshydromet, along with smaller high-performing ■■ Increased lead time and accuracy of global and regional computers (HPCs) for regional centers and research facilities. forecasts. And in addition to this significant infrastructure upgrade, the ■■ Improved data collection and transmission. project supports the institutional strengthening and capacity ■■ Drastic reduction of response time for requests of archived building of Roshydromet institutions by helping to develop data. decision-making systems and methodological guidelines and bringing the experiences of advanced NMHSs. Modernization Cannot Be ■■ Growing number of development partners’ and donors’ programs. Piecemeal ■■ Growing interest in the private sector to provide services. While there is no definitive methodology to modernizing ■■ Typically limited financial and human resources and NMHSs in LMICs, there is now a sufficient body of knowl- management skills at NMHSs. edge from the World Bank Group (WBG) hydrometeorolog- ical investment portfolio, along with diverse development ■■ Challenges of applying World Bank and other development partners, to understand the strengths and weaknesses of partners’ processes for designing and implementing different approaches. It remains an axiom that modernizing large-scale integrated modernization projects. NMHSs cannot be piecemeal (Rogers and Tsirkunov 2013a, A big issue is what level of improvement is appropriate. Pilon and Asefa 2011). The process should be integrated and Sometimes modernization efforts have focused more on the transformative, ensuring that NMHSs can deliver the services technical details and infrastructure without paying sufficient that their users and stakeholders expect within a landscape attention to the expected level of services or the absorption of multiple suppliers of information. Modernization efforts capacity of the recipient NMHSs. If NMHSs are to contribute must account for a number of factors: to saving lives and increasing prosperity, modernization Weathering the Change: How to Improve Hydromet Services in Developing Countries? 9 LONG-TERM SUPPORT INCREASES IMPACT AND SUSTAINABILITY I n 2008, the World Bank published Weather and Climate Services in Europe and Central Asia: A Regional Review, which identified serious capacity deficiencies in many NMHSs, including those in Central Asia. It also identified the significant benefits that relatively modest but sustained invest- ments in these NMHSs could deliver. This study initiated a strategic engagement of the World Bank with the Central Asian NMHSs that continues to deliver even more results to this day. Following the above assessment and recognizing the need for a deeper understanding of the needs of Central Asian NMHSs, a more detailed analysis was undertaken in the Kyrgyz Republic, Tajikistan, and Turkmenistan in 2009, financed by GFDRR in support of the Central Asia and Caucasus Regional Economic Cooperation Initiative on Disaster Risk Management, coordinated by the World Bank, UNISDR, and WMO under the umbrella of the Central Asia Regional Economic Cooperation (CAREC). The assessment laid out clear economic arguments, strategies and investment scenarios for strengthening Central Field training on flash flood, mudflow and landslide assessment for Central Asian specialists near Almaty, Kazakhstan. Source: Y. Kubakov, Regional Asian NMHSs. Center of Hydrology (RCH), 2018 Informed by these assessments, the Central Asia Hydro- meteorology Modernization Project (CAHMP) was developed users, the project helped integrate Turkmenistan better into and approved by the World Bank in 2011, constituting a US$27.7 regional efforts, which was not participating in CAHMP. million IDA and PPCR investment across a regional component and two national ones (Kyrgyz Republic and Tajikistan). By 2018 the benefits of investing in the NMHSs in Central Asia This represented one of the first World Bank investments in through this combined national and regional approach had hydrometeorological services that pursued a three-pronged become obvious to the involved governments and the World approach at the national level: building institutional technical Bank. As such, additional financing of US$11.5 million was and management capacity; rehabilitating and modernizing approved for CAHMP, with Turkmenistan now participating. observation networks and forecast production systems Both CAHMP and the technical assistance project are currently (ICT); and strengthening service delivery through improved scheduled to continue until 2021. engagement with users and product delivery. This is being While the road to modernizing the NMHSs of Central Asia has supported by regional investments in strategic coordination, not been without challenges, the ability for the World Bank information exchange, regional forecasting systems, and a to stay engaged for over a decade has ensured that issues shared training system to support professional development could be solved in a thorough and sustainable manner. While through peer-to-peer exchanges. there are always project timelines that need to be met, larger After a somewhat challenging start, CAHMP began to produce operational challenges could still be tackled in a more robust results three to four years into the project. At the same time, and strategic manner. In addition to this long-term engage- it became clear that more “soft” capacity building support ment, partnership with many highly qualified and experienced was required. As such, in 2016 the World Bank initiated the institutions such as WMO, Roshydromet, ECMWF, and several Strengthening of Early Warning of Mountain Hazards in Central other advanced NMHSs has meant that the NMHSs of Central Asia technical assistance project, financed by three GFDRR grants Asia have always had expert support available to them. At the for a total of US$2.1 million. In addition to providing regional same time, coordination with parallel support from other World training and exchange, on-the-job coaching, improved access Bank projects, the Asian Development Bank, UN agencies, and to global forecast products, study tours, and support to develop bilateral donors such as Switzerland and Finland have ensured outreach strategies and materials for better engagement with more holistic support. Weather Services 10 Weathering the Change: How to Improve Hydromet Services Developing inin Countries? Developing Countries? Monitoring station on the Syr Darya River at Kairakum, Tajikistan, Source: GFDRR, 2018 projects should result in NMHSs being able to deliver the economic outcomes (WMO 2015a). When available informa- services their users require. How this is achieved is likely to tion is underutilized, value can be increased by improving vary from country to country. It could be accomplished by the forecast, enhancing communication, and refining the improving all of the functional systems within an NMHS, or decision-making process. by teaming up with other NMHSs, outsourcing, or engaging Thus, effective service delivery is about providing products with the private sector. and services that bring utility to users. It is essential to understand the users’ value chain in order to gain knowl- Improving Service Delivery edge about users, the decisions they must make, and how weather-, climate-, water-, and environment-related infor- Given that NMHSs must compete for scarce public resources, mation is applied to minimize risk and provide benefits, not they need to demonstrate their value by realizing cost effi- only for specific user groups but also for society as a whole. ciencies while delivering high-quality and useful products With this knowledge, service providers are able to develop, and services. 5 Policy makers and the public continually produce, and deliver services that are useful, relevant, and assess the effectiveness of NMHSs based on their ability responsive. NMHSs should be able to measure the value of to meet the service delivery standards of the nations they their information to society and continually evaluate and serve. Even the best forecast, issued on time, will have little improve these services. Adopting a more collaborative effect if it does not generate the desired response from those approach provides everyone in the service delivery pro- at risk. This means that the utility of weather-, climate-, cess—providers, users, and partners—with a clear under- water-, and environment-related information depends on standing of service needs. the degree to which it has a beneficial effect on societal and The WMO Strategy for Service Delivery and its Implementation Plan, WMO 2014a, WMO-No.1129. 5 Weathering the Change: How to Improve Hydromet Services in Developing Countries? 11 CAN WE DO A BETTER JOB PREDICTING FLOODS AND MANAGING WATER RESOURCES? B etween 21 and 22 December 2018, parts of northern Sri Did the forecasters miss something? The European Centre for Lanka received over 300 mm of rainfall, resulting in Medium-Range Weather Forecasts (ECMWF) model output issued extensive flooding and flash floods. In Kilinochchi dis- on 17 December predicted heavy rains for the north of Sri Lanka trict, the flooding was exacerbated by spilled water from major between 21 and 22 December (Figure 2). However, these events and minor tanks. Nine of the sluice gates of the Iranamadu are often associated with high dynamical instability in the atmo- water tanks were opened, causing major water logging in the sphere. This means that there was a large spread by the members paddy fields downstream in the Kandawalai division (Figure of an ensemble forecast—i.e., a wide range of possible outcomes. 1). Overall, the flooding affected more than 45,000 people. Other members showed much lower rainfall amounts. Coordinated by the Disaster Management Center, more than Without looking further, it might be reasonable to assume 2,000 people were rescued by troops and sent to safety centers. this event was improbable. However, inspecting the Extreme Why was the warning lead time so short? The meteorological Forecast Index (EFI) and the Shift of Tails (SOT)—measures of forecast issued by the Department of Meteorology gave little the potential for extreme weather compared to climatology for indication of the severity of the situation. The summary of their a given location and time of year—indicated differently. A high forecast for December 22 issued at noon on 21 December stated EFI value indicated that an extreme event was more likely than “Showery condition is expected to enhance over the island usual, and the SOT provided information on how extreme the event could be. Specifically, it compares the tails of the distri- (particularly during 21st and 22nd), especially in the Northern, bution of the ensemble members (the extremes) and the model North Central and Eastern provinces due to the low level atmo- climatology (so-called M-climate distribution). spheric disturbance in the vicinity of Sri Lanka.” It would be difficult to conclude from the summary that a major flood would The EFI/SOT charts from 19 December showed that the EFI occur. The synopsis mentioned up to 75 mm of rain. ranged from 0.5 to 0.7 in the northern part of Sri Lanka, which FIGURE 1 Flood-affected paddy field in Kandawalai, Kilinochchi Source: World Food Programme, Photo credit: Indu Abeyratne, WFP. 12 Weathering the Change: How to Improve Hydromet Services in Developing Countries? CAN WE DO A BETTER JOB PREDICTING FLOODS AND MANAGING WATER RESOURCES? (cont.) FIGURE 2 Twenty-four hour accumulated precipitation 21 December 00UTC to 22 December 00UTC forecast on 17 December 12UTC Precip ECMWF 20181217 12 Mem:1 0.5 10 20 30 40 50 60 90 253 80°E 10°N 10°N 80°E indicated that an extreme event was more likely than usual. Use of the warning matrix would have enabled, permitting Also, high SOT values started to appear from 18 December early action, or at the very least, raising awareness among in the north of Sri Lanka, indicating that some members had decision makers and those at risk. extreme rainfall (Figure 3). This suggests that the possibility The situation highlights the importance of an integrated of this event should not be dismissed, and coupled with the approach to forecasting and warnings of the impact of floods, potential for high impact, an alert of warning could have been landslides, and extreme weather. In Sri Lanka, this requires issued as early as 19 December. close cooperation among the Disaster Management Centre, Combining the likelihood of a meteorological event happening the Department of Meteorology, the Irrigation Department, with the impact of that event enables a risk matrix to be gen- which is responsible for flood forecasting, and the National erated – an approach developed by the UK Met Office for warn- Building Research Organization, which is responsible for ings. This can be done for individual assets, for communities, managing landslide risks. at district levels, and so forth. In this case, we are focusing on A new initiative from the World Bank is helping the Sri the Kandawalai water tank. The tank was vulnerable because it Lankan government address these issues. was near capacity, and it would take several days to spill water safely without causing a downstream flood. Consequently, Acknowledgments the impact would be rated as high in the matrix (Figure 4). A With thanks to Linus Magnusson at ECMWF for providing the low likelihood coupled with a high impact would generate an ECMWF model images. amber/orange warning. This should be interpreted as recom- mending precautionary measures to reduce risk. In general, it is important to pay attention to medium and high impacts even when the likelihood is low or very low. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 13 FIGURE 3 Extreme Forecast Index and Shift of Tails for total precipitation forecast on 19 December, valid 21–22 December (ECMWF) Wed 19 Dec 2018 00UTC @ECMWF VT: Fri 21 Dec 20018 00UTC – Sat 22 Dec 2018 00UTC 4872h Extreme forecast index and Shift of Tails (black contours 0,1,5,10,15) for: total precipitation 0.5 0.6 0.7 0.8 0.9 1 70°E 80°E 90°E 10°N 10°N 0°N 0°N 70°E 80°E 90°E FIGURE 4 Using a warning risk matrix, the meteorological forecast indicates a low likelihood of an extreme precipitation event; the impact, however, on a dam that is near capacity would be high. Therefore, the warning would have been “amber/orange” HIGH LIKELIHOOD MEDIUM LOW ✔ VERY LOW VERY LOW LOW MEDIUM HIGH IMPACT Doppler weather radar at the Chisinau International Airport, Moldova. Source: POPs Sustainable Management Office, 2015 Weathering the Change: How to Improve Hydromet Services in Developing Countries? 15 Chapter 2 Meteorological and Hydrological Services > The main purpose of meteorological and hydrological services is to enable the public and economic sectors to make appropriate decisions when faced with weather, climate, and hydrological hazards. > Effective service delivery is underpinned by a set of functions common to all meteorological and hydrological organizations, which we refer to as a “system of systems.” > These systems contribute to the overarching goal of improving services—each must reach an acceptable standard that meets all or, at least most, of the needs of users and stakeholders. > Major challenges facing the public sector meteorological and hydrological services include operating and maintaining modern observing systems and embracing the advances in information and communication technologies (ICT). 16 Weathering the Change: How to Improve Hydromet Services in Developing Countries? Introduction Another way to visualize the system of systems is to view it as a matrix (Figure 2.2). Here services flow from left to Meteorological and hydrological services can be considered right, cutting across the various product activities. This can as a “system of systems,” which are grouped in three cate- be used to illustrate how more than one organization could gories: delivery systems, production systems, and support cooperate. For example, three separate entities may provide systems. This is often referred to as the value chain of meteorological, hydrological, and agricultural services—but meteorological and hydrological services, which links the they have the opportunity to share common observations, production and delivery of services to user decisions and models, and means of delivering services. They could the outcomes and values resulting from those decisions develop entirely separate systems or integrate their sys- (WMO 2014a). These are underpinned by support systems. tems. This might include developing a common architecture One way to visualize this system of systems is illustrated in for observations, common standards, and operating proce- Figure 2.1, which shows a schematic using different colors. dures—or even transferring responsibility for components of The categories and subcomponents are: the observing networks to one of the entities. In the case of flood forecasting and warnings, it might be preferable to col- Delivery systems locate the hydrologists and meteorologists that are involved Comprising service delivery systems; in numerical prediction, and those involved in forecasting and actions, services monitoring, and feedback systems. and warning preparation could collocate with disaster man- Production systems Comprising monitoring and observing agers during an operational event. systems; modelling systems; and objective and impact This chapter takes an up-close look at the “system of sys- forecasting and warning systems. tems”—and its main objective is effective service delivery. It also explores the major challenges facing the public sec- Support systems Comprising ICT systems, quality man- tor, such as operating and maintaining modern observing agement systems and technology infusion systems—along systems and embracing the advances in information and with capacity building. communications technology (ICT). FIGURE 2.1 Schematic of an NMHS as a “System of Systems” Objective Monitoring and Impact and Actions, Services, Forecasting Observing Monitoring and Warning Systems Modelling Service and Feedback Systems Systems Delivery Systems Systems Quality Management Systems ICT Systems Technology Infusion Systems Capacity Building Note: Delivery services. Production systems. Support systems: and which underpin the delivery services and production systems. Note: Green objects are delivery services; blue ones are production systems; items in bottom portion (brown and orange background) are support systems, which underpin the delivery services and production systems. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 17 FIGURE 2.2 A Matrix of the “System of Systems” ■■ Available and timely: at time and space scales that the user needs. Production Functions ■■ Dependable and reliable: delivered on time to the required user specification. Monitoring and Observing Systems Objective and Impact Forecasting ■■ Usable: presented in user specific formats so that the client Weather Services can fully understand. and Warning Systems Modelling Systems ■■ Useful: to respond appropriately to user needs and require- Hydro Services ments. Services ■■ Credible: for the user to confidently apply to decision making. Climate Services ■■ Responsive and flexible: to the evolving user needs and requirements. Other Services ■■ Sustainable: affordable and consistent over time. ■■ Expandable: to be applicable to different kinds of services. Quality Management Systems Service delivery comprises seamless delivery of information in a tailored, interactive manner in an operational mode to ITC Systems the general public and specialized users. For NMHSs, service delivery takes on an emergency assistance role during extreme weather conditions to reduce disaster risk in weather-, climate-, Technology Infusion Systems (R&D) water-, and environment-related events. Here, seamless implies integrated information relevant to the users across all Capacity Building appropriate timescales for their decisions. Support Functions It is useful to define some generic principles applicable to all types of services provided by NMHSs and private sector providers. In this way, service providers have the incentive Note: This matrix can apply to a single service organization or several separate to prove that their service is compliant with internation- entities, each having a common set of production systems and support functions. ally agreed generic principles, namely: (i) science based; (ii) quality assured; (iii) user oriented; (iv) adding value to Category 1: Delivery Systems users; and (v) the best possible services that could be pro- vided under the particular conditions of the NMHS or private To understand service delivery, it is first necessary to pro- sector provider. vide a common definition of service and service delivery. The WMO Strategy for Service Delivery and Its Implementation To be “useful,” services should be user driven. This means Plan (herein referred to as “the Strategy”) (WMO 2014a), that services should be based on the requirements of users defines service as “the delivered product and the activi- and produced and delivered in response to those require- ties associated with the people, process and information ments. However, sometimes services evolve around the technology required to deliver it, or an activity carried out research interests of service providers or have simply grown (advice, interpretation, etc.) in order to meet the needs of out of practices and traditions without consultation with the the user or which can be applied by a user.” Service deliv- stakeholders or clients. Therefore, the first consideration ery is defined as “a continuous process for developing and needs to be, who are the users of the services prescribed delivering user-focused services, defined in terms of user by their programs and projects? The next most important engagement, service design and development, evaluation considerations will be whether there is coherence and and improvement.” To be effective, services should have the structure in the approach to service delivery, and whether following attributes: those services possess the attributes and general principles 18 Weathering the Change: How to Improve Hydromet Services in Developing Countries? cited above. If the responses to these questions are not ■■ Risks inherent in applying information to decision making. overwhelmingly positive, this could be interpreted as a lack If appropriate, NMHSs may want to explicitly outline the of consistent service culture among those programs. agreement reached with the user in a service-level agreement Those NMHSs that provide commercial services involving (SLA). The same principles also apply to the private-sector contractual obligations are no doubt acutely aware of the service providers. need for high standards in service and service delivery, but In the public sector, countries benefit from a uniform and struc- high-quality delivery should also apply to weather-, cli- tured approach for service development and delivery, which mate-, water-, and environmental-related services that are builds upon and institutionalizes existing practices to strengthen provided to the public and to the government agencies and service delivery aimed at protecting the public from harm. “The departments. Strategy”— which draws heavily from the experiences of the Good service delivery relies on the organization’s internal WMO Public Weather Services (PWS) Programme—provides processes, and thus those processes should be adjusted to an adaptable tool and initial step for NMHSs. An emerging need is to introduce a service culture across the entire NMHSs’ match the service delivery systems’ needs. Effective and structures and projects through common applicable steps and efficient internal processes directly impact the quality of activities to apply a joined-up approach, with the aim of creat- service delivery, the value of products and services, and ing a service discipline that is applicable to all programs. It is the cost-effectiveness of an organization’s day-to-day oper- fully recognized that the content will remain the responsibility ations. The measurements of an organization’s internal pro- of the individual NMHSs’ programs. A service-oriented culture cesses should be driven by the quality management system requires the use of accuracy measures from the user’s perspec- (QMS) used and the key processes defined therein. tive, which differs from some of the accuracy measures widely The focus on understanding the users’ value chain is vital applied, for example, within the Numerical Weather Prediction because it provides knowledge about the users, the deci- (NWP) community. A service-oriented organization should use sions they must make, and how weather-, climate-, water-, forecast parameters that have a direct impact on the users’ and environment-related information is applied to minimize activities and operations. risk and provide benefits—not only for specific user groups but also for society as a whole. With this knowledge, service Moving toward a Service Culture providers are able to develop, produce, and deliver services What is behind the growing push for NMHSs to focus on that are useful, relevant, and responsive. Another important effectively delivering the services that users and stakehold- reason is that the public sector needs to ensure that the ers want? The key drivers are many. user requirements fall within the NMHSs’ mission and that Shift toward open data. This shift in many countries stems NMHSs have the capability to meet those requirements—in from national or regional policies that open opportunities other words, that the services are “fit-for-purpose.” Having a for service provision by different entities (such as public service “fit-for-purpose” implies that an agreement has been entities, private companies of different sizes, and research reached, either implicitly or explicitly, with users, providers, institutions). The resulting competition among public and and partners. It should consider: private service providers is seen as a major driver to move ■■ Current and evolving user needs. toward providing high-quality services in a very efficient manner. A successful service provider should be able to ■■ Provider capabilities, including strengths and limitations. demonstrate the benefits for the user through informed ■■ What services will be provided and how they will be decision making, reliability, and cost efficiency. For NMHSs, provided. this comes on top of their general mandate, which is related to reducing the risk to lives and adverse impacts on society. ■■ How services will be used. Greater need for efficiency. A lack of public funds and com- ■■ Expectations of acceptable outcomes and provider petition with the private sector requires NMHSs to maximize performance. their efficiency. Hence, a more streamlined approach is criti- ■■ Acceptable costs or levels of effort. cal in the future work of NMHSs, especially service delivery. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 19 This is an area where cooperation among the public and pri- value-added information. In agriculture, PPEs would be criti- vate sector has potential, with each exploiting the strengths cal for the range and specificity of information products (like and responsibilities of the other to provide cost-effective advisory services to farmers) and their market penetration. and efficient services tailored to the specific requirements In ICT, the use of telecommunication masts for installing of different user groups. automatic weather stations could minimize the need for the NMHS to acquire land, and reduce O&M costs, vandalism, Social and technological changes. These pose broad chal- and other overheads. lenges for NMHSs in terms of the ways in which people and activities are vulnerable to weather, climate, and water Disaster risk reduction. Evolving national policies, legisla- influences, along with how they use NMHS information to tion, and legal frameworks in areas like disaster risk reduc- reduce risks and vulnerabilities and seize opportunities. tion provide opportunities for a greater recognition of the Thus, the challenge for NMHSs is much more than mustering NMHSs by governments and stakeholders. This could result resources and achieving stable funding for infrastructure. in strengthened partnerships and increased resources and They must also foresee and plan for a wide range of social opportunities for providing products and services. However, and technological changes, including how these changes it could also increase demand and liabilities for the provision will affect their ability to provide services and enable of high-quality, reliable, and timely products and services to users to realize the benefits. In many countries, the NMHSs support disaster risk management (DRM) operations. face major hurdles in ensuring their capacity to meet the ever-growing demand for their services, while maintaining Cost-recovery services. In many instances, the NMHS is the integrity of the science that is the basis for these ser- required to recover costs for its services. But the quality of vices—as well as providing authoritative information and those services may not have been established and, given the advice for decision making. growth of the commercial sector, may not be the most com- petitive in terms of quality and cost. Faced with competi- The bottom line is that there needs to be major investments tion, the NMHS has a few options: one is to restrict access to in many aspects of the operation of NMHSs in delivering ser- data, thereby tilting the field in its favor; a second is to use vices—such as effective tailoring of services to user needs; legislation to delegitimize any competition. In an extreme efficient public and private service delivery arrangements; case in one southeast Asian country, the meteorological law improved methodologies and algorithms for use of meteorolog- explicitly calls for destroying non-approved observation ical, hydrological, and related information in decision making; stations. Numerous studies have shown that restrictive and state -of-the-art ICT and computing facilities (Box 2.1). data practices do not serve the national interest—and Public-private engagement (PPE) arrangements could disempowering private sector providers and the ability of complement and enhance NMHS functions by developing public and private businesses to choose their service pro- cost-sharing and revenue-generating activities, or sales of viders, may damage economic performance. For example, if BOX 2.1 A Major Role for ICT Developments ICT developments in the last decade have brought enormous As a result, the ICT capabilities of the service provider have new opportunities in accessing data and products, and in com- become a key factor in the successful delivery of services. municating between providers and users. Traditional ways This is an area of direct competition with both national and of disseminating information (like hard copy newspapers) international private sector service providers and other are being rapidly replaced by web and mobile applications. NMHSs that have developed sophisticated and user-friendly Moreover, mobile devices are quickly becoming the main plat- mobile applications. NMHSs are increasingly becoming aware form for delivering information to end-users (both corporate of this competition, as well as recognizing that in today’s fast and individual). pace of technological advances, the user/customer has many options to choose from and does not need to make the NMHS the first choice for service provision. 20 Weathering the Change: How to Improve Hydromet Services in Developing Countries? weather-sensitive business decisions cannot be made based by default NMHSs as the major service providers. Internal on the best available analysis and forecasts, this is likely to standardization of services, based on a standardization impact profitability and consequently tax revenues. framework, will facilitate sharing and applying best prac- tices among sectors, flexibility of staff use, and resource A third option for NMHSs is to focus exclusively on their leveraging. It would also contribute to the branding of the core public task—public weather services—and adopt the service as an important part of market competitiveness. most efficient and effective business models. Met Norway, the Israel Meteorological Service, US National Weather Furthermore, this needs to be complemented with a prod- Service and Japan Meteorological Agency have embraced ucts and service development and delivery process that this approach, making their data widely available and reus- would incorporate: able1. The latter requires the government to adequately fund ■■ Identifying and prioritizing the target users. this task and to recognize that the activities of the public sector can enhance economic development by making their ■■ Establishing working arrangements to understand the target data openly available for reuse by others. At the same time, users’ needs and requirements. discipline is needed in the private sector to respect the ■■ Developing products and technical support to meet the government’s decision on how to provide public services in needs and requirements. the interests of safety of life. ECOMET, created by European ■■ Developing the proper service delivery model appropriate NMHSs, focuses on maintaining a level playing field for for this target user group. commercial activities in Europe—such as operating and maintaining an administrative framework to increase and ■■ Delivering the service. improve access to data and products, along with acting as ■■ Establishing feedback mechanisms for improving products advisors on the exchange of meteorological data and prod- and identifying new product opportunities with the target ucts. PRIMET is a pan-European trade association for mete- user. orological services operating in the private sector. Together, The expected benefits of such an approach for the NMHSs these organizations try to constructively resolve problems are: increased effectiveness and efficiency, increased com- that arise between public and private sectors within Europe. petitiveness with other service providers, and sustaining New markets for services. In many areas, the users of and expanding their position as a public service provider. the services of NMHSs may be driven by national policies, The benefits for the end-users are: receiving more quality legislation, and mandates requiring delivery of services for and value, stronger focus on their primary needs, efficiency, certain applications. On the other hand, NMHSs may also be and affordability. realizing other opportunities by identifying and developing other markets. Developing these services should be based In addition, there is a dynamic push now toward establishing on identification, segmentation, and prioritization of target a “user interface mechanism,” instead of programs sticking user groups within and across sectors. It is important to to the traditional building blocks, which focus on a linear understand the needs and requirements of the target users movement from data to information to services to users. In as they vary across the different user groups within a sec- doing so, it is recognized that all programs and activities torial value chain. Thus, the types of products and services of NMHSs’ systems are interdependent. Successful imple- may range from basic over-the-counter services (products mentation of the user interface concept in service delivery that meet the needs and requirements of a variety of users) requires much work, both on the side of technology and on to tailored services. the side of standardization. It is not going to be efficient and effective to develop individual user interfaces for each type A Common Approach to Service Delivery of service (such as climate, weather, and water). What is vital is that all services should be consistent and Thus, special efforts are needed to create a common concept compliant with the attributes and general principles of and initiate work on setting up service standards that are services and related delivery methodology—recognizing flexible to cover different types of services. Key issues to MeteoWorld, 2016. N0.3 ISBM 1818-7137, September 2016, World Meteorological Organization, 4pp. 1 Weathering the Change: How to Improve Hydromet Services in Developing Countries? 21 FIGURE 2.3 Service Delivery Systems scope to cover users—based on the principles of service de- livery and QMS. ■■ Setting competency requirements for the NMHSs to inform Public services their staff development activities for optimum service deliv- system G2G Disaster ery—like those developed by the WMO for aeronautical me- Management teorological services, PWS, climate services, hydrology, and services system Public-Private marine meteorology services. cooperative services to key business system As for the service delivery systems, these include: (a) pub- Service G2B & G2G lic weather services; (b) government to government (G2G) Transport services Delivery system disaster management services; (c) G2G agricultural services; Systems (d) G2G water and power management services; (e) G2G and G2G & G2B Climate services government to business (G2B) aviation services; (f) G2G and system G2G Agricultural G2B climate services; and possibly (g) public-private coop- services system erative services (Figure 2.3). G2G Water and power management Who are the key users and stakeholders2 of NMHSs? Certainly, system water resources, agriculture and food security, and DRM services would top the list. Surveys in many countries have Note: This matrix can apply to a single service organization or several separate confirmed that: entities, each having a common set of production systems and support functions. ■■ These sectors need real-time updates of forecasts, longer lead times, clear information (not vague), more accurate consider include the following: forecasts, information on impacts, and more site-specific information. ■■ Weighing the necessity of all weather-, climate-, water-, ■■ Regular dialogue with the users and stakeholders is an es- and environment-related services to further develop mech- sential step to develop awareness of user requirements. anisms for interacting with users; identifying user require- ments would be a first step. ■■ Color-coded messages are a better means of communicating warnings. ■■ Establishing regular consultation and information sharing sessions among NMHSs’ programs to better plan activities. ■■ The government is responsible for the safety and security of its citizens, and thus the provision of meteorological and ■■ Expanding the WMO Technical Regulations to cover generic hydrological warnings—the so-called authoritative voice or aspects of service delivery, based on relevant standard oper- source. ational procedures (SOP), in addition to the specific aspects covered by the existing regulations (like aviation). ■■ There is a general consensus that the conduit for these warn- ings is the NMHS—but views differ on whether the NMHS ■■ Improving services and service delivery through fast uptake should do this alone, or it should be the conduit for collect- of science and technology developments. ing the best available data (public and private) to inform the ■■ Setting priorities for delivering weather-, climate-, water-, warning services. and environment-related services to meet the rapidly chang- As for aviation and marine services, these are a somewhat ing needs of society—including impact-based forecasts and special class of service because they are regulated, and the warning services. government must designate a meteorological authority, ■■ Investing in training and capacity development—primari- which may or may not be the NMHS. In some countries, avi- ly for NMHS forecasters and managers, then extending the ation services are separate from the NMHS. In francophone A distinction is made between users, representing everyone who uses meteorological and hydrological services, and stakeholders, which may have a formal 2 relationship with the service provider—in the case of an NMHS, this would include other government agencies. The stakeholder relationship may extend to a formal oversight role as representatives of a class of users. 22 Weathering the Change: How to Improve Hydromet Services in Developing Countries? FIGURE 2.4 Monitoring and Feedback Systems on Services are available to download via the WMO website. The WMO’s survey of how NMHSs are performing (WMO 2015b) high- lights the need for improvement—with smore than one-third Monitoring and reporting that they either did not have a service delivery feedback system for public services strategy, or that it had just been initiated.3 One area where such a strategy is required, as part of quality management, is civil aviation. Actions, service Issues related to service delivery should be considered monitoring, and with even more attention than is given to observing and feedback systems forecasting systems. The profile of many NMHSs in LMICs as the main source of weather, climate, and water infor- mation is low among the public and stakeholders. This is Monitoring and Monitoring and feedback system attributed to a lack of active outreach and engagement with feedback system for for national businesses and provincial the user community, leading to insufficient awareness of government user requirements. The limited capability of many NMHSs’ to provide products and services required by stakeholders Note: Monitoring and feedback systems represent all of the actions, service mon- has contributed to lower levels of satisfaction by users. itoring, and feedback mechanisms from users and stakeholders to the service providers. Surveys are still the best means of measuring public satis- faction. Examples of indicators of user satisfaction are: Africa, the Agency for Aerial Navigation Safety in Africa ■■ Has the product fully met users’ requirements? and Madagascar (ASECNA), based in Senegal, is providing ■■ Is the format of the products appropriate (layout, colors, aviation services on behalf of its member governments. graphics, animation)? Regulations require that the service provider be supported by airport and overflight charges on the industry. But while ■■ Is the means of dissemination and delivery adequate this should be a significant source of revenue, many coun- (easy to access, fast)? tries do not reimburse the NMHS directly for this service. ■■ Is it easy to understand the content of products and The International Maritime Organization (IMO) is respon- services? sible for Safety of Life at Sea (SOLAS) regulations, which include marine meteorology. However, it does not have the ■■ Is the frequency of updates of the products sufficient? same level of enforcement as the International Civil Aviation ■■ Are the phenomena and features shown on a forecast Organization (ICAO) Annex III (Meteorological Service for relevant (for example, use of isobars on a map, which may Air Navigation) and does not require payment from the confuse people if no explanation is given and users do maritime industry to support NMHSs marine forecasting not see the relevance of such features to the weather’s operations. impact)? Measuring customer or user satisfaction is both necessary ■■ Is the wording used appropriate (too technical for some, and useful in assessing performance and areas for further or too simple for some others)? development. In many NMHSs, there is no system in place ■■ Are cultural norms observed in delivering the service (too to monitor the quality and delivery of services, which may formal for some cultures, or too casual for the liking of some be somewhat surprising given that this is the output of the organization (Figure 2.4). Thus, there needs to be a process others)? to gather quantitative and qualitative information on perfor- Sectorial users’ satisfaction with services is easier to deter- mance from the users’ perspective (Technical Insight 2.1). mine and is usually done by personal contacts or relatively Survey tools have been developed by several NMHSs and small workshops. https://www.wmo.int/pages/prog/amp/pwsp/documents/REPORT_SSD_SURVEY.pdf 3 Weathering the Change: How to Improve Hydromet Services in Developing Countries? 23 What can be done to address the issues raised above and and decide on how to create the required capabilities. determine the current situation of NMHSs regarding service Allocating resources to implement the agreed actions will delivery? As a first step, the Service Delivery Progress Model be essential. This would be followed by a review of the prog- of the Strategy (Annex 1) should be applied to map the exist- ress of activities against the action plan, and finally sharing ing service delivery capability of client countries’ NMHSs. best practices and knowledge with other NMHSs (perhaps A next step would be to create an action plan, with short-, through twinning or other collaborative activities). medium-, and long-term goals to address the weaknesses TECHNICAL INSIGHT 2.1 Measuring User Satisfaction Met Office User Satisfaction Surveys following selected severe weather warnings. The Met Office carries out annual and ad-hoc surveys to gain ■■ Telephone interviews of 500 people in the affected area. insight into the public’s requirements and levels of satisfac- ■■ Monitors awareness and usefulness of the warning against tion with our forecasts and severe weather warning services. targets. These surveys are carried out for the Public Weather Service ■■ At least six surveys are carried out each year. Customer Group (PWSCG) by independent market research companies to ensure they are unbiased and representative ■■ Conducted independently by DJS Research Ltd. of the views of the UK public. They also help us to identify Usefulness of warnings new requirements and ensure that the Met Office is providing ■■ Respondents are asked “Overall how useful would you services that meet public need. have said this severe weather warning was?” Results ■■ Answers included: Very useful, fairly useful, not very These are updated annually after a series of surveys are car- useful, not at all useful, and don’t know. ried out. ■■ 92% of respondents found their warning very or fairly useful in 2015/16. Only five surveys were conducted during National Severe Weather Warning Service (NSWWS) surveys this period due to a smaller number of amber warnings It is important to ensure that the warnings the Met Office being issued at times when it was appropriate to survey. issues reach the people who need them and that they find the ■■ The target for 2016/17 was 84% (average for at least six warnings useful. The Met Office, therefore, carries out surveys surveys). Indicator 1.1—Public Usefulness Warnings 100 90 80 70 Storm Aileen 60 Percentage Ex-hurricane Ophelia Storm Doris 50 % Very useful % Fairly useful 40 Benchmark 30 Snow and Wind 20 Snow and Ice 10 Snow Snow Snow Snow Wind Wind Wind Wind Wind Wind Wind Rain Rain Rain Rain Rain 0 14 14 15 15 15 15 15 16 16 17 17 17 17 18 18 18 18 18 2/ 9/ 0/ 2/ 1/ 1/ 8/ 1/ 2/ 7/ 2/ 1/ 1/ 1/ 1/ 2/ 3/ 5/ /0 /0 /1 /0 /1 /0 /1 /0 /0 /0 /0 /0 /1 /0 /0 /0 /0 /1 13 10 04 09 07 16 28 17 23 18 27 28 16 31 13 13 01 31 24 Weathering the Change: How to Improve Hydromet Services in Developing Countries? TECHNICAL INSIGHT 2.1 Measuring User Satisfaction (cont.) Reach of warnings period due to a smaller number of amber warnings being ■■ Respondents were asked “Did you see or hear anything issued at times when it was appropriate to survey. about this severe weather warning?” ■■ The target for 2016/17 was 70% (average for at least six ■■ 80% of respondents had seen or heard their warning in surveys). 2015/16. Only five surveys were conducted during this Indicator 1.2—Public Reach of Warnings 100 90 80 70 60 Percentage % Snow event 50 % Non-snow event 40 Wind Ex-hurricane Ophelia Benchmark snow (80%) 30 Benchmark Wind Storm Aileen Wind Storm Doris non-snow (70%) Snow and Wind 20 Snow and Ice 10 Snow Snow Snow Snow Wind Wind Wind Wind Rain Rain Rain Rain Rain 0 14 14 15 15 15 15 15 16 16 17 17 17 17 18 18 18 18 18 2/ 9/ 0/ 2/ 1/ 1/ 8/ 1/ 2/ 7/ 2/ 1/ 1/ 1/ 1/ 2/ 3/ 5/ /0 /0 /1 /0 /1 /0 /1 /0 /0 /0 /0 /0 /1 /0 /0 /0 /0 /1 13 10 04 19 16 08 28 17 23 18 27 28 16 31 13 13 01 31 Public perception survey Results It is important to ensure that the forecasts the Met Office These are updated each year when the next survey is carried issues reach the people who need them and that they find the out. forecasts useful. It is also important that we know how people Accuracy of forecasts are accessing forecasts (e.g., mobile, television, internet, and ■■ Respondents are asked “Generally speaking, how accurate where from). To do this we carry out a survey every year in the or inaccurate do you think most weather forecasts are?” first week of October. ■■ Answers include very accurate, fairly accurate, neither ■■ In street, face to face, computer-assisted personal accurate or inaccurate, fairly inaccurate, very inaccurate, interviewing (CAPI) with 2,089 people across the UK. or don’t know. ■■ Monitors perceptions of accuracy and usefulness, plus ■■ 78% of respondents felt that forecasts were very or fairly frequency of use. accurate in 2017. ■■ Conducted independently by DJS Research Ltd. ■■ The target for 2017 was set at 73%. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 25 TECHNICAL INSIGHT 2.1 Measuring User Satisfaction (cont.) Theme 2a: Public Services—Direct Research Indicator 4—Public Perceptions of Forecast Accuracy 100 90 80 70 60 Percentage % Very accurate 50 % Fairly accurate 40 Benchmark (73%) 30 20 10 0 07 8 08 8 9 0 1 2 3 4 5 16 17 1 1 1 1 –0 0 1 0 1 v– v– v– v– v– v– v– v– v– l– v– v– b No No No No No Ju No No No No No No Fe Usefulness of forecasts ■■ 87% of respondents felt that forecasts were very or fairly ■■ Respondents are asked “Overall how useful would you say useful in 2017. weather forecasts are these days?” ■■ The target for 2017 was set at 80%. ■■ Answers include very useful, fairly useful, not very useful, not at all useful, and don’t know. Theme 2c: Public Services—Products and Content Indicator 16—Trends in Usefulness of Public Forecasts 100 New survey added Nov–14 90 80 70 60 Percentage % Very useful 50 % Fairly useful 40 Benchmark 30 20 10 0 07 08 08 08 09 10 1 2 3 14 4 15 16 17 1 1 1 1 v– v– v– v– v– v– c– v– v– v– b– l– v– v– No No No No No Ju De No No No No No No Fe Source: Met Office https://www.metoffice.gov.uk/about-us/who/accuracy/your-say 26 Weathering the Change: How to Improve Hydromet Services in Developing Countries? Category 2: Production Systems FIGURE 2.5 Monitoring and Observing Systems The second category of the system of systems centers on the production systems, which include: Global Data Systems (i) monitoring and observing systems; Monotoring and Observing (ii) modelling systems; and Systems (iii) impact forecasting and warning systems. National Data Systems In situ observing systems Monitoring and Observing Systems Radar and other remote sensing systems The monitoring and observing systems basically comprise External data systems—social me- two subsystems—the global data system and the national dia, private networks and subnational data system (Figure 2.5). ■■ The global data system. This includes all of the data avail- Note: Monitoring and observing systems have two distinct subsystems. The first able through the WMO Information System (WIS) and the is the global data system, which includes all of the information received via the global telecommunication system (GTS) that is accessible by WMO Information System or the Global Telecommunication System and data the NMHS—including data from satellite systems available from unique sources, such as satellite products from specific providers. The sec- ond is national data, which is a combination of observations networks supported through systems such as CMACast or EUMETCast or any oth- by different entities—public and private, and crowdsourced data associated with er GEONETCast for other satellite data providers. social networks—commonly associated with big data. ■■ The national data systems. These comprise the data that are produced by, or on behalf of, the NMHS and any other na- tional or local data supplied by other agencies or the private environment. However, while automation can save costs sector or academia or through crowdsourcing (Krennert et in developed countries where the labor cost is high, it may al. 2018a, 2018b). come with a higher price tag in developing countries where labor is cheaper. Increasingly big data4 are being exploited—such as from social networks to improve warnings, or from drones for Making sure observation networks are fit for purpose. precision agriculture to maximize yields and minimize It is important to ensure, given the high cost of supporting inputs of water and fertilizers. A typical national network observing networks, that they are fit for purpose, can be may include synoptic stations (manual and automatic), maintained, operated effectively, and are not acquired for radar, rainfall networks, lightning detection, river gauges, the wrong reasons. Radar data, for example, are very useful climate stations, and upper air and oceanographic stations. for nowcasting and very short-range forecasting, which typ- All of these stations need to be networked so that the data ically uses rapid refresh systems based on remote-sensed are available centrally within the NMHS for subsequent observations from radar or satellite, blended with NWP. processing and use in analyses and forecasts. Automation They provide very high-resolution quantitative precipitation of observing networks has the advantage of providing high estimates (QPE), which are very relevant to rainstorm mon- fidelity and high temporal resolution (minutes rather than itoring and flash flood forecasting. They can also contribute hours), which is critical for nowcasting and very short- to 24 to 48 hour forecasts—if they are assimilated into range forecasting, along with rapid updates of the forecast convection5-permitting (~1 km or less) numerical models 4 Data of a very large size such that its manipulation and management present significant logistical challenges and is too complex for processing by traditional data management tools. Big data often comprises structured and unstructured data from traditional and nontraditional sources. Size is less of an issue than the rate at which these data must be processed. For example, extracting meaningful information on the impact of a meteorological or hydrological hazard on a city by interpreting social media data. 5 Convection—organized motions within a layer of air leading to the vertical transport of heat, momentum, etc. (WMO 1992), leading to thunderstorms, heavy precipitation, etc., typically within ~1 km or less horizontal resolution and timescale of ~1.5 days or less. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 27 (Soares and Mylne 2013), which provide very high-reso- small fraction of the data they collect on the misunderstand- lution quantitative precipitation forecasts (QPF) that are ing that because these data are not designated as essential critical for flood forecasting. However, attention to quality within the meaning of WMO resolution 40 and WMO general control is absolutely vital for any type of observational regulations, they are not important to the global network. data, especially for radar data assimilation due to the The high resolution of global models and the need for verifi- effect of clutter and various non-meteorological sources of cation of model output makes all data valuable and essential back-scatter. Another area where local observations are to provide the best possible NWP and forecasts. Close atten- critical, but often overlooked, is forecast verification, which tion must be paid to data redistribution rights if the NMHS is an essential part of improving the forecast (including for pursues a data services option. model calibration, post-processing, and downscaling). Monitoring and observing progress model. Similar to the Forecast verification. This is done to monitor the quality of a service delivery model, a progress model for monitoring forecast by measuring its accuracy and whether the forecast and observing can be adopted as a means of mapping the is improving over time; to discover what is wrong with the capabilities of a client country’s NMHS to maintain or access forecast; and to determine to what extent one forecasting meteorological, hydrological, and related observations system gives better forecasts than another.6 Different types (Annex 2). An approach, similar to service delivery, in terms of forecasts require different verification methods—ranging of an action plan to address weaknesses in the monitoring from dichotomous for deterministic forecasts to ensemble and observing system could be implemented. methods for probabilistic forecasts. While the methodolo- gies are mature, many NMHSs do not apply these methods Meteorological and Hydrological Modelling consistently. Systems NWP is rapidly evolving—with the modelling systems now O&M of technical equipment. The addition of highly comprising global, regional, limited area, nowcasting, and technical equipment to an institution that may be used to hydrological models (Figure 2.6). Global models provide a operating manual stations is likely to raise the operation wide range of products for the preparation of forecasts on and maintenance (O&M) costs significantly, and thus should timescales of hours to months. They are less and less seen as be done with care and due diligence to avoid the situation just low-frequency, low-resolution models providing a gen- where new equipment rapidly becomes nonfunctional, eral large-scale outlook and boundary conditions for high owing to lack of maintenance or access to spare parts (Lynch resolution regional and national models. Frequently, NMHSs and Allsopp 2008). A contributor to the increased costs is request support for their local limited area deterministic the need to hire, train, and retain higher level technical NWP efforts that use in-house high-performance computing. expertise than traditional observers, which can be chal- These NMHSs tend to run outdated versions of determin- lenging in limited labor markets and with a not-so-flexible istic NWP models on workstations at spatial resolutions government pay scale. Quality control of all of the observa- much lower than those currently available from the Global tional data is essential. In some circumstances, it may be Production Centers (GPCs), and with no ability to assimilate more cost effective to supply data to the NMHS as a service. local observational data into the models. Rarely is forecast While many NMHSs may be reluctant to pursue this option, verification, and model post-processing and calibration con- it may be more cost effective and could ensure continuity of sidered and operationalized. data services. Use of LAMs. In most instances, running a limited area Data sharing. As much of the available data as possible will modelling (LAM) is not the most optimal use of resources need to be made available to the global prediction centers to quickly improve forecast capabilities, which should be via the WIS/GTS or in some instances directly, if data restric- one of the goals of most NMHSs. The level of skill and the tions prevent open access. Many countries share only a resolution of global mesoscale7 models are now so good that See WWRP/WGNE Joint Working Group on Forecast Verification Research. 6 Mesoscale—in meteorology, it is the study of weather systems of horizontal resolutions generally ranging from around 5 km to several hundred kilometers, 7 and temporal resolutions typically ranging from 1 hour to 1 day (WMO 1992). Examples of mesoscale weather systems are: sea breezes, squall lines, and large convective cells. Vertical velocity often equals or exceeds horizontal velocities in mesoscale weather systems due to non-hydrostatic processes such as buoyant 28 Weathering the Change: How to Improve Hydromet Services in Developing Countries? FIGURE 2.6 Modelling Systems A lack of understanding of how to make full use of global and regional model guidance leads to an unwarranted use of LAMs. One argument given by NMHSs in LMICs for the local Hydrological Modelling Global NWP development of NWP is the cost of access to the full suite of Systems Systems digital products from GPCs, or lack of open access to these digital data. However, this is not a good reason for operat- ing an inferior system. It is far more effective for NMHSs to Modelling make full use of the numerical guidance from GPCs. Where Systems higher resolution is needed (such as hazardous-prone areas or over complex orographic regions), it is preferable to use the best available LAMs at the regional level (supported by Limited Area Regional NWP WMO Regional Specialized Meteorological Centers (RSMCs) Modelling Systems Systems or by consortia (WMO 2017). This should be followed by forecast verification and feedback, model post-processing Note: Central to forecast production is the numerical modelling system, which and calibration, model output interpretation, and delivery comprises three main geographically dispersed elements—the global, regional, of services at the national level—the so-called “Cascading and local systems. Increasing the resolution of the global systems is reducing the need for separate regional and local modelling systems to provide basic mete- Forecasting Process.” This approach has been introduced orological predictions. The latter are used where they either provide very high in many regions as a part of the WMO Severe Weather resolution or unique data, which are assimilated at these scales. Hydrological modelling is highly dependent on local data and is almost always operated Forecasting Demonstration Project (SWFDP). It strength- nationally with initial and boundary conditions provided by local observations ens the links between the global NWP production centers, and global or regional NWP. There is a growing preference for all models to be run as ensemble prediction systems, which enable risk to be better predicted. regional centers that interpret information from the global centers, and NMHSs (which are responsible for issuing alerts, advisories, and severe weather warnings). The SWFDP has these models can only be challenged by NMHSs with very been a particularly successful program, leading to the full high levels of investment and highly qualified experts. For operationalization of the concept in southern Africa and the example, the European Centre for Medium-Range Weather expectation that the SWFDP in other regions will become Forecasts (ECMWF) is currently running a global 9 km fully operational once the skills are in place. deterministic model and an ensemble prediction system at 18 km. And it aims to run a global high-resolution ensemble When would a LAM be appropriate? An NMHS should con- prediction system up to two weeks ahead with a horizontal sider running very high-resolution (~1 km or less), convec- resolution of about 5 km by 2025 (ECMWF 2016). While tion-permitting numerical models over hazardous-prone noting that lateral boundary conditions can be sourced areas or over complex orographic regions. But this should from GPCs, the inherent errors introduced in passing data occur only if: through a model boundary mean that a significant benefit ■■ Robust and reliable telecommunications infrastructure (to is achieved only when the regional and local model can be import the required volumes of data) and supercomputers run at substantially higher resolution than the global model. are affordable to (i) support models of competitive resolu- Initial conditions can also be sourced from GPCs for a pure tion with optimal parameterizations of the physical process- downscaling approach, but the benefits will be limited with- es; and (ii) address high-quality processing and full data as- out assimilation of high-quality local observations at high similation, spin-up, and cycling issues. resolution. In addition, effective (operationally) running, maintaining, and updating locally of these LAMs requires a ■■ A large and expert staff of scientists and computer techni- long-term commitment to continuing scientific and technical cians are available to develop and maintain such a system expertise—along with a complex infrastructure of high-level 24 hours per day, 365 days per year, in an operational envi- computing and telecommunications—both of which are ronment. Identification of high-risk areas is a prerequisite, major challenges for LMICs. which is not always available in many client countries. acceleration of a rising thermal or acceleration through a narrow mountain pass. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 29 In developed countries, many NMHSs are phasing out their (iv) short-range forecasts; (v) medium-range forecasts; and local LAMs. Generally, the current freely available model (vi) long-range forecasts (Figure 2.7). As the name implies, codes have been conceived for training and research the severe hazard forecasting subsystem focuses on the purposes only, lacking the optimal parameterizations of extreme events within the NMHS’s responsibility—such as the physical processes, the configurations tuned for the tropical cyclones, thunderstorms, storm surges, high winds, regions of interest, and the developments required to keep extreme heat and cold, and rainfall and flooding. Nowcasting them up-to-date in terms of science and technology. Thus, focuses on the immediate weather (nominally 0 to 6 hours); these codes are not well suited for operational use. The it is highly dependent on rapidly updated observations, notion that lower-resolution, locally run models can be usually from radar and satellites. Several advanced NMHSs equivalent, or better, than the global mesoscale models is have pioneered the development of these systems and made misguided. them available to the WMO community. Very short- and For example, in the summary of the recent annual reports on short-range forecasts cover up to 3 days ahead. Medium the application and verification of ECMWF’s forecast prod- range extends the forecast up to 15 days, and long-range ucts (ECMWF 2017), most ECMWF Member and Co-operating forecasts up to seasonal timescales. States highlighted that overall the ECMWF deterministic Wherever possible, the forecasts should be based on ensem- model performance is as good as their LAMs’ performance. ble predictions, which provide valuable information about For that reason, a number of ECMWF Member States (such the likelihood of an event, and coupled with information on as Switzerland, Norway, and Portugal) are phasing out their the potential impacts, provide users with warnings of the local LAM configurations at 7 to 12 km resolution, retaining severity risks (Fleming et al. 2015, Rogers et al. 2018). only those LAM configurations at much higher resolutions (such as 2.5 km and less) with data assimilation. This illus- Impact-based forecasts emphasize what the weather will trates that the current ECMWF model resolution has provided do rather than what the weather will be. Impact-based fore- a direct cost savings for them. It also highlights that simply casting and warning services focus on translating weather running a model without data assimilation serves little or no operational purpose. FIGURE 2.7 Forecasting and Warning Systems Use of “ensemble prediction systems” (EPS). An import- ant and continuing development for operational weather Severe hazard forecasting forecasting is the use of EPS, which are capable of providing system uncertainty information, associated with NWP results. EPS Long-range Nowcasting system/Flash are a powerful tool in predicting and early detection of forecasting flood system severe weather events. For impact-based forecasting, the guidance EPS may be used to help estimate the probability of weather Objective and impact hazards for use in the estimation of Risk (= Probability x forecasting Impact). The combined use of very high-resolution limited and warning area deterministic NWP (and increasingly limited area systems ensemble systems) is relevant for short-range forecasting Medium-range Very short-range forecasting forecasting (typically up to 48–72 hours ahead)—in conjunction with system system global and regional guidance—which helps NMHSs at an Short-range early stage in severe weather forecasting. forecasting system Forecasting and Warning Systems Objective and impact forecasting and warning systems cover Note: Forecast and warning systems cover all timescales from the immediate all timescales from the immediate to seasonal. This compo- (nowcasting) to climate timescales. Each range is often a specialization of a fore- caster. For example, severe hazard forecasting may be the responsibility of flood nent includes: (i) meteorological and hydrological severe forecasters, tropical cyclone forecasters, drought forecasters, or a combination of hazards; (ii) nowcasting; (iii) very short-range forecasts; these and other expertise. 30 Weathering the Change: How to Improve Hydromet Services in Developing Countries? TECHNICAL INSIGHT 2.2 Putting the Focus on Weather Impacts The WMO puts a high priority on impact-based forecasts and Impact-based forecasts warning services as a way to increase NMHSs relevance and NMHSs must have access to the best available numerical guid- utility. What are impact-based forecasts? They focus on what ance to generate timely, accurate, and specific meteorological the weather will do rather than what the weather will be—and and hydrological forecasts. This is often the most challenging by doing so, translate weather hazards into sector-specific part, but increasingly possible, as higher resolution numerical impacts. The hope is that as a result, those exposed to a haz- weather predictions become available from the WMO global ard will have a better understanding of the risk and will more production centers (GPCs) and WMO Regional Meteorological likely take appropriate action (Fleming et al. 2015). Specialized Centers (RSMCs). This can be taken a step further But using impact-based forecasts to boost the NMHSs output nationally if the country has the capacity and capability to run requires that NMHSs put a greater emphasis on service deliv- very high-resolution models (better than 2 km resolution), ery. Already, WB- and GFDRR-supported modernization efforts which incorporate assimilation of local data from radar and emphasize service delivery. Moving beyond weather fore- other observing systems. If this is not possible, reliance on casting requires effective partnerships with many different the global and regional centers may be sufficient. government agencies. This is perhaps one of the most difficult Establishing impact-based forecast and warning services things to achieve and where the WB has a role through its can be done in two phases: the first focusing on developing convening power to bring together many of the actors and a qualitative or subjective forecast that relies mostly on the stakeholders to help NMHSs create the necessary partner- experience of stakeholders to establish warning thresholds; ships and data sharing arrangements, as well as encouraging and the second focusing on quantitative estimates of vulnera- other development partners to support this approach. bility and the use of tools to calculate impacts. hazards into sector-specific impacts, and developing sectoral This tool can be used to display meteorological warnings responses to mitigate those impacts (Technical Insight 2.2). and sector-specific impact warnings. It also highlights the By focusing on impacts, it is expected that those exposed importance of common color-coding for specific levels of to a particular hazard will have a better understanding of risk regardless of the hazard, impact, or sector. This way a the risk and will more likely take appropriate action. While better “feeling” for risk is established across all sectors. not essential for the successful implementation of impact- based forecast and warning services, it is highly desirable Modelling and forecasting progress model. The strategy to develop a map-based warning system, which divides the for improving the use of high-resolution NWP and LAM, country into a convenient grid and uses colors to represent as a part of the overall forecasting processes, is based on warning levels within each of the grid boxes. This approach the WMO Global Data-processing and Forecasting System was initially developed by MétéoFrance and has been (GDPFS) guidelines (WMO 2016a). Mapping current capabil- adopted Europe-wide through the MeteoAlarm system (see, ities as a first step to identify weaknesses should be followed for example Stepek et al. 2012). Other countries are follow- by developing and implementing an action plan to improve ing this example, which enables at-a-glance stakeholders the overall forecasting system (Annex 2). to visualize the geographical extent and type of warning. Updated frequently, warnings evolve during the course of the event and in response to actions taken to mitigate risks. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 31 Category 3: Support Systems nationally recognized and respected. It is the only ISO stan- dard against which organizations can be certified or regis- The third category of the system of systems—the tered through a third-party audit process. support systems—underpins the value chain ■■ ISO 9004:2009, Managing for the sustained success of an organization—a quality management approach. It focuses on of meteorological and hydrological services. It achieving sustainable success by meeting the needs and ex- includes: (i) quality management systems; (ii) ICT pectations of customers and other stakeholders. It promotes systems; (iii) technology infusion systems; and self-assessment as an important tool, which enables ongoing review of the level of maturity attained by the QMS. Howev- (iv) capacity building. er self-assessment is not a substitute for a third-party audit process. Corporate governance. Governance relates to the processes Quality Management Systems and structures that ensure an organization is directed, con- Quality management systems (QMSs) are an essential part trolled, and held to account. It focuses on how an organiza- of the normal operation of organizations—possibly mak- tion is managed, how risks are monitored, and how value is ing the difference in saving lives and property. Aviation added for the community, government, and other stakehold- requirements have forced most NMHSs to adopt ISO 9000 ers. The main components of a sound corporate governance standards, but they have not been universally applied. framework are: (i) promoting and ensuring adherence to As of November 2012, the International Civil Aviation a code of conduct and values; (ii) risk management; (iii) Organization (ICAO) has required quality management sys- continuity of service; (iv) occupational health and safety; (v) tems for meteorology as a standard. However, few NMHSs ongoing development of staff competencies; (vi) providing have proper management systems to regulate their internal timely and accurate reports to senior/executive manage- decision processes (Figure 2.8). ment; (vii) a published Service Charter that sets out the standards of service to the community; (viii) contributing to The overall management of the NMHS may be discussed in the organization’s annual reports; and (ix) contributing to the context of a QMS, leading to an assessment of whether the strategic and operational planning process. the current organizational structure of the institution is fit for purpose. ISO 9001 provides a rigorous management An ISO 9001 QMS is a management tool to measure the framework to enable NMHSs to identify and meet the ongoing performance of the corporate governance activities requirements of their customers, monitor and measure their of an organization. The WMO Guide to the implementation own performance, and identify opportunities to continually improve service delivery (WMO 2013). The adoption of a FIGURE 2.8 Quality Management Systems QMS is a strategic decision. If well planned, appropriately resourced, and efficiently implemented, it provides a cost-ef- Institutional fective management system. Management Processes and The ISO 9000 series consists of three standards on good Systems quality management practices: Quality Management ■■ ISO 9000:2005, Quality management systems—fundamen- Systems tals and vocabulary. It describes the fundamentals of QMSs Operational and the terminology used in the ISO 9000 family. Management Systems ■■ ISO 9001:2008 (updated to ISO 9001:2015), Quality man- agement systems—requirements. It can be applied to all types of organizations (public and private), regardless of Note: Quality management systems, which are at the heart of well-run organiza- size or industry group—enabling both product and service tions, apply equally to operation of the various service delivery and production organizations to achieve standards of quality that are inter- systems and the overall management of the organization. 32 Weathering the Change: How to Improve Hydromet Services in Developing Countries? of a QMS for NMHSs 8 is an essential starting point for intro- FIGURE 2.9 Information and Communication Technology ducing a QMS (WMO 2013). Systems Information and Communication Technology Systems External Cloud ICT systems are fundamental to the operation of a modern Computing Systems NMHS. They include data, computing hardware communica- tion, and increasingly, managing access to cloud computing (Figure 2.9). An efficient organization has a central repos- Data Communication Systems itory for data and information—including all observational Information and data and numerical predictions (regardless of source), Communication post-processing, and applications. This enables new services Technology to be generated easily, and O&M of the ICT systems to be Systems Computer Hardware optimized. The current practice of each new observing or and Software forecasting system setting up stand-alone servers should be Systems discouraged. A key requirement is broad bandwidth access to the Internet to access very large data volumes on a daily basis Telecommunication from the global prediction centers, satellite- and ground-based Systems remote sensing, and non-traditional national sources (like the Internet of Things (IoT) and crowdsourcing). This is an area where staff retention in an NMHS is often difficult because of highly paid opportunities in the private sector—raising the Note: ICT systems, which are central to a well-functioning NMHs, are increasingly a combination of internally supported systems and externally sourced services possibility of external national contracting. (such as Infrastructure as a Service (IaaS)), provided under contract by specialist IT firms. Big data is motivating radical changes in the way ICT systems are being used in meteorological and hydrological services. Digital, gridded meteorological and hydrological data are national circumstances, but the transition from traditional increasingly being combined with data in multiple formats databases to distributed file systems is inevitable to match from other sources—like transportation networks, social the need to maintain large amounts of useful data available media, and health services—in integrated maps to provide online. A major benefit is less need to maintain many local impact forecasts and targeted warnings. This requires the operational systems in favor of a few online platforms. ability to gather, store, and manipulate data from these Artificial Intelligence (AI). Another major ICT advance is multiple, disparate sources, which, in turn, requires a cen- tralized, flexible, extensible ICT architecture. the role of AI in forecasting. This is critical to mining big data and providing customized meteorological and hydrological Cloud computing. As NMHSs access higher temporal and services. This transition to so-called smart meteorological spatial resolution data from global centers, and integrate and hydrological services allows forecasts to be individual- information from multiple national sources, higher perfor- ized and targeted to situations such as flooding, landslides, mance will be required from computing systems, resource and storm surges. This hyperlocal service can be realized management, and networking (see for example Jiao 2016). through mobile Internet by building user profiles based on A cloud-based approach is essential to achieve this. Infrastructure as a service (IaaS) is discussed in Chapter 3. mobility, demographics, and understanding specific real- While such a platform can be developed by an NMHS, it is time user scenarios. Cooperation with the private sector more likely that smaller services will acquire this capability may be the best way to realize this potential. through commercial service providers. The balance between internal and external cloud services will depend on specific https://library.wmo.int/pmb_ged/wmo_1100_en.pdf 8 Weathering the Change: How to Improve Hydromet Services in Developing Countries? 33 Technology Infusion Systems FIGURE 2.10 Technology Infusion Systems In most of the NMHSs in LMICs, the ability to further their own development is limited by the absence of a culture of research and development (including external research and External Research development, internal research and development, and the and Development Systems transition of research to operations) (Figure 2.10). It also limits their participation in external research and devel- opment initiatives. The result is that the more advanced Commercial off-the-shelve NMHSs and the private sector control the direction of the Systems field. Thus, greater effort is needed to help NMHSs develop skills in verification methods, timeseries analysis, and Technology downscaling techniques. Twinning arrangements with more Infusion Systems Transition of advanced NMHSs and technical centers can go some way Research and Development into to addressing the lack of research and development in Operational many low-income country NMHSs. Similarly, participation Systems in WMO-led research initiatives and development projects Internal Research (such as the SWFDP) can help build expertise. Traditional and Development instruments used in the WB-funded projects are not well Systems suited for providing support for research—underscoring the need for special efforts like targeted grant funded programs and twinning arrangements. Note: Technology infusion systems refers to the combination of research and Adopting and implementing new ICT strategies is a key development, performed in-house or externally, that can be transitioned into operations—such as hardware or software or new forecasting techniques. area for technology infusion. Exploiting AI, big data, and cloud computing technologies are critical to the future of any NMHS. This requires NMHSs to build or access the skills statistics, and machine learning (Staudinger et al. 2018). needed to implement these strategies. These create greater challenges for LMICs, especially where training and sustaining a workforce is already difficult. A big problem is that public sector staff are often lost to private Capacity Building sector activities outside of the GWE, where remuneration for Strengthening human capacity covers meteorological and technical skills is often significantly higher. This is a chal- hydrological training, business development, stakeholder lenging area in WB implementation projects that requires training, and end-user training and outreach (Figure 2.11). special attention. Management skills are often overlooked This is an area where significant investment needs to occur and should be a focus within WB modernization projects. to strengthen the human capacity within the NMHS. Based on feedback from many training events and NMHSs’ man- Core disciplines. The traditional skills in meteorological agement, it is clear that training should be concentrated on and hydrological sciences, which build on a foundation in on-the-job training with external training focused mostly on mathematics and physics, continue to be the foundation of building a cadre of trainers to conduct the national training. meteorological and hydrological services. However, these The piecemeal approach, focusing on one or two individu- need to be expanded to include new skills without necessar- als, does not always work well unless their responsibility is ily reducing the core disciplines. to teach others. ICT skills. Software engineers and meteorologists and Keeping up with rapidly evolving GWE skills. Training hydrologists with programming and ICT skills are a crit- of NMHSs’ management and technical staff is an integral ical component of modern services. The ability to utilize and essential part of keeping pace with the evolving GWE. observations and prediction products requires the ability to There are significant changes in education and training with manipulate and interpret big data. Communication skills are greater emphasis on communication programming, big data also essential. The ability to interpret scientific information 34 Weathering the Change: How to Improve Hydromet Services in Developing Countries? FIGURE 2.11 Capacity Building ■■ Provide on-the-job training in the use and interpretation of model products from the GPCs, and introduce a fore- casting process. ■■ Train forecasters in the interpretation of probabilistic NMHS Institutional Education Training forecasts into useful information for decision making by users. ■■ Train public weather service advisors (or staff tasked with public weather service duties) in developing and main- taining close relations with key stakeholders (notably, di- Capacity Stakeholder saster managers and civil protection authorities), as well Building Institutional Training as assisting them to understand and make better use of probabilistic forecasts and warnings. Typically, this training would involve several experts for about two weeks each, but due to the fast pace of devel- End-user Training opment in forecasting and communication, it should be and Outreach repeated at regular intervals. Such training took place at the Pakistan Meteorological Department during preparation of a modernization project (Technical Insight 2.3). Note: Capacity building underpins the human resources within the NMHS and Training trainers. Specialized international training at WMO within the main users and stakeholders. regional training centers (RTCs), for example, should be carefully selected for staff to be qualified as trainers of the and communicate it to the public, government agencies, and other staff. These training opportunities should have specific businesses is critical. objectives that can be tested and evaluated against staff per- formance. So-called south-south training can be particularly Social communication skills. The growing use of social effective where staff from different LMICs agencies share their media to exchange information requires its own particular knowledge and experiences. It is also helpful to establish a set of skills—including writing, content design, communica- formal process to measure the training’s usefulness. tion, customer service, digital marketing, and analytics. The Twinning with advanced NMHSs. Well-defined standard ability to measure the performance of social media posts, operational procedures, including a checklist of activities for for example, is increasingly critical to understand how peo- weather forecasting, are also required for improved quality ple respond to forecasts and warnings. forecasts that meet stakeholder and other end-user require- Business practices. Whether in a NMHS or in the private ments and expectations. To address this, we recommend sector, knowledge of business practices is essential—includ- that twinning arrangements be made between the devel- ing project management, business modelling, business oping country NMHS and other NMHSs that can help pro- planning, marketing, financial management, procurement, vide guidance on forecasting techniques—including model and risk management. The assumption by many NMHSs that post-processing and calibration, effective use of ensembles, business practices are the purview of only the private sector and operational forecast guidance. This approach is prac- needs to be challenged. ticed by the Met Office, the National Weather Service, and Météo-France, which provide twinning support to a number On-the-job training. The most effective, cost-effective, and of developing countries. But it could also be provided by sustainable training is conducted on-site by experts using more advanced NMHSs within a region to share expertise as the systems and equipment available to the staff. This is part of so-called south-south cooperation. Where there is a especially useful for departments with many staff, such as need for flood forecasting, the licensing of digital products forecasters and public weather service advisors. Individual from GPCs is recommended to provide quantitative precipi- trainers might be expected to do the following: tation forecasts as an input to locally executed flood models. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 35 University education. Countries with university programs rel- to the NMHS. The ability to manipulate and analyze data evant to staff of NMHSs and related agencies usually have fewer from multiple sources and provide hyper-local forecasts and problems recruiting staff than those countries that do not. warnings tailored to individuals’ needs is at the heart of this service. The basic building blocks remain the same—deliver Summary of Services and Systems services and adjust those services based on feedback from The main purpose of NMHSs is the provision of services users by: (i) monitoring and observing the natural and human that meet the needs and requirements of users. This cannot environment, and (ii) modelling and predicting changes that be overemphasized. The era of data collection as the pri- impact the lives and livelihoods of people. The underpinning mary goal is past, and NMHSs need to focus on developing technologies and skills are challenging for even the most cost-effective and high-quality services for users. Further, advanced NMHSs. given their limited budgets, the emphasis needs to shift from a heavy infrastructure-oriented approach to one that In the next chapter, we focus on key concepts for moderniz- takes advantage of the range of available services external ing NMHSs. TECHNICAL INSIGHT 2.3 Training Program to Strength Forecasting and Public Weather Services Training is an integral part of developing and maintaining Public weather services operational skills. Training can be conducted on-site by 1. How to meet the key stakeholders’ specific requirements experts, usually in two week-long courses for each topic. The for hydromet information (including accuracy, timeliness, following is an example from the Pakistan Meteorological the use of nontechnical language, and agreed formats); Department, addressing the following subjects: how to handle uncertainty and how to get the message across about uncertainty and probabilities. Numerical weather prediction and forecasting 2. Guidance on the effective delivery and communication of forecasts and warnings to the end-users (like the general 1. Understanding the meaning of the Ensemble Prediction public, farmers, fishers, etc.), including the use of nontech- System (EPS) and the potential behind that technique, nical language. including in support of decision-making processes. 3. Guidance on preparing key messages for senior PMD manage- 2. Guidance on the procedures to follow for a continuous ment or government officials during severe weather events. and consistent operational weather forecasting process— 4. Guidance on working with the media (radio, TV, print) to including analysis, verification, use of deterministic versus deliver and communicate (both written and oral as applica- probabilistic forecasts, and use of ensembles to estimate ble) forecasts and warnings in agreed formats and times. uncertainty. 5. How to interpret probabilistic forecasts and turn them into 3. Demonstration of capabilities of European Centre for messages understandable to all the stakeholders, in sup- Medium-Range Weather Forecasts (ECMWF) ecCharts, and port of decision making. how to build and use a Dashboard aligned with the opera- tional weather forecasting process. 6. Simple call-to-action messages for end-users (especially the public), as part of forecasts and warnings of severe 4. Forecast exercises, using case studies (monsoon and tropi- weather events (should be prepared in advance in consul- cal cyclones applications). tation with the National Disaster Management Authorities 5. On-the-job practical training on the above points, using (NDMAs) and Provincial Disaster Management Authorities (live) performance of the system (ecCharts) for weather (PDMAs)). forecasting. 7. How meteorological services can contribute to the mitiga- tion the impact of natural hazards. 8. On-the-job practical training on all of the above points, using the daily forecasts (and warnings if applicable). Modernized automatic weather station in Tajikistan. Source: GFDRR, 2018 Weathering the Change: How to Improve Hydromet Services in Developing Countries? 37 Chapter 3 Key Considerations for Creating Effective NMHSs > Many NMHSs in low- and middle-income countries (LMICs) continue to follow a traditional business model—rooted in a culture of manual meteorological and hydrological observations—to support general forecasts and outlooks about a day ahead. > The problem is that these NMHSs, which tend to be static, are often at a level that is too low to operate and maintain more than a basic level of services—on top of not paying much attention to their users. > To be relevant, NMHSs, like all service organizations, should provide the services their users and stakeholders need and want. These needs far exceed the capability of most NMHSs in LMICs that are operating within a traditional business model. > Thus, there is an urgent need for different business models—ones that can help improve the efficiency and effectiveness of NMHSs. Good planning and strategies are needed, along with the tools to design and test operational concepts prior to designing and implementing solutions. 38 Weathering the Change: How to Improve Hydromet Services in Developing Countries? Introduction this as expansion financing. A different approach would be required if the plan has a different purpose. The traditional business model of a national meteorological Market, customers, clients, and users. The NMHS must and hydrological service (NMHS)—relying on manual obser- understand the overall market and its structure. Vital ques- vations and trying to accomplish everything alone—needs tions are: How do people use information? Is government to be urgently overhauled to deliver the services that users funding sufficient to discharge the public service respon- and stakeholders now need and want. What type of model sibilities of the NMHS adequately? Is there a willingness would help transform weak agencies, especially in low- and to pay for services within the country’s economic sectors? middle-income countries (LMICs), into more robust ones? What are these sectors? Are sectors aware of how to use This chapter explores the key considerations for creating weather, climate, and hydrological information in their deci- effective NMHSs. It begins with the importance of business sions? How do these sectors currently get information? What plans and business models, then looks at a growing recogni- are the gaps that the NMHS can fill? How will new products tion that more public-private sector engagement could help and services be introduced? What are the areas for potential strengthen meteorological and hydrological services—before growth? Who else is likely to operate in this market? What detailing the tools available to design and test operational is the NMHS’ share of the total market? And what are the concepts prior to designing and implementing solutions. success factors and future developments for NMHS? Competition. Inevitably, in a market that is growing and Business Plan becoming increasingly important to economic sectors, there will be competition between the public and private sectors Why a business plan? The business plan of any organization, and among private sector providers of meteorological and be it for profit, non-profit or government, describes what hydrological services. Rather than being defensive, the NMHS the organization does, its operational and financial goals, needs to look at opportunities and consider its relationship and how it intends to meet them. Many NMHSs consider with commercial providers. A first step is recognizing distinct business planning to be the domain of the private sector. roles and responsibilities for the public and private sectors. A But in practice, any organization that needs to demonstrate second step is articulating the competitive advantage of the its value—ministries responsible for meteorological and NMHS: (i) knowledge of meteorological and hydrological con- hydrological services, ministries of planning, ministries of ditions of the country; (ii) close cooperation with government finance, and tax payers—must be able to articulate its busi- departments and communities; and (iii) its potential position ness plan. The plan provides structure to the organization as the national authority for meteorology and hydrology. By by explaining why it does what it does and what makes it doing so, this provides the NMHS with a unique position as a successful. Without a business plan, it is difficult to assess partner for joint ventures with the private sector. Alternatively, the funding requirements for the organization and whether it may choose to compete directly for commercial services. further investment is good value for money. In the case of However, care should be taken to avoid distorting the market. an NMHS, it is an opportunity to capitalize on the govern- Restricting private sector access to data, for example, limits ment’s investment in modernization to provide sustained, competition and eventually reduces the interest of economic high-quality, value-for-money services. sectors in paying the market value for services—possibly resulting in an underdeveloped market, dominated by the Expansion financing. Business plans are usually targeted NMHS but with a limited revenue stream. to a specific need and audience. In the case of an NMHS, which is going through a process of modernization, the plan Marketing. In countries where the NMHS is the dominant explains: (i) what the improvements in capabilities will pro- provider of services, there is little incentive to market its vide; (ii) its strategy to sustain these services; and (iii) the services. However, this often leads to a lack of awareness by nature of the products and services. It can be used to help the public and government of NMHS capabilities. Explaining convince the government to increase its budget to cover what the NMHS has to offer is marketing. NMHSs may think the additional costs of operations and maintenance (O&M) they are in the public eye because of the kind of services associated with the new higher level of services. We refer to they provide. However, they are often only visible to the Weathering the Change: How to Improve Hydromet Services in Developing Countries? 39 public and the government during extreme events—and most often, when there is a failure in the system. Thus, a proactive Business Models approach is needed to maintain better awareness of NMHS The business model is central to the business plan, focus- capabilities. To use a marketing expression, the NMHS is the ing on how the NMHS operates. It is a conceptual structure “brand” leader in the country. that supports the viability of a product or organization and includes the purpose and goals of the organization and how “In Belarus the public obtains forecasts from all sorts of differ- it intends to achieve them. It covers all the business pro- ent apps and social media sites, generally never caring about cesses and policies that an organization adopts and follows. the sources of the forecasts. But when these forecasts are And it is supposed to answer who is your consumer, what wrong, they blame Belhydromet. So we have had to establish a value you can create or add for the consumer, and how you presence on all social media channels to deliver good informa- can do that at reasonable costs. Like most businesses, an tion and protect our reputation.” —Mr. Roman Labaznov, Head NMHS has both external and internal consumers. of Belhydromet Cost centers. Each of the “systems” (or in some cases the Production and delivery. The NMHS should provide an “subsystems”) comprising an NMHS can be treated as an overview of its modernized systems from production individual cost center. How each is operated can be analyzed through service delivery (the value chain). How the NMHS to determine the most cost-effective and efficient way to goes about its business is important from the perspective of understanding its efficiency and effectiveness. provide the services required of, or by, these systems. Each of the systems is a production center with a well-defined set Research and development (R&D). Infusing new technol- of consumers. For example, the monitoring and observing ogies is a critical element of maintaining the relevance of system’s outputs are consumed by: (i) the modelling system, the NMHS in an ever-changing economy (see chapter 2). But and (ii) the objective and impact forecasting and warning in developing economies, R&D is often not part of NMHSs. system (Figure 2.6). However, this can be addressed by working with regional partners and participating in WMO research activities, Staff limitations often compromise the effectiveness of each thereby showing that the NMHS is actively benefitting from of the systems—particularly in the observing networks, developments within the worldwide meteorological and modelling, ICT, and service delivery systems—and absent hydrological community. access to external research, the ability to infuse new tech- nologies is limited or nonexistent. Few LMICs address staff Management and organization. Change management is a shortages by outsourcing. In many instances (especially critical component of the modernization process to ensure with ICT and service delivery), the private sector has the that not only is the technology fit for purpose but also the capacity to provide contract services—such as maintaining staff are adapting to their evolving roles. The business plan computing systems, hosting servers, staffing, web-based should describe the skills of the NMHS workforce and pro- services, and mobile applications. In the Middle East, it is vide confidence that there is a capable management team. common practice to hire contract staff. Currently, the Kuwait Risk analysis. It is critical that NMHSs identify, assess, and Meteorological Service has four Kuwaiti nationals on its staff manage operational and business risks. Doing this effec- and about 150 contract staff, mostly from Egypt. These are tively increases confidence—including by the responsible hired individually, but it would also be possible to contract government ministry—that the NMHS is well managed and a firm to provide these staff—as is done by the U.S. National can anticipate and deal with institutional problems. Weather Service to provide technical support staff to the National Centers for Environmental Prediction (NCEP). In Finances. Financial planning should focus on worst-case the latter case, this is a business decision based on govern- and best-case scenarios—for example, a worst-case scenario ment rules that limit the hiring of full-time civil servants. would be what will happen in five years if there is no budget uplift to support O&M. It should anticipate information fail- Traditionally, as government departments, NMHSs in LMICs ures that may contribute to turning a hazard into a disaster are obliged to follow a simplistic and often inefficient approach and use a variety of economic tools to determine the NMHS’ to budgeting. A fixed amount of money is spent mostly on financing needs. salaries of civil servants. A limited O&M budget supports the 40 Weathering the Change: How to Improve Weather Services in Developing Countries? observing networks and the rest of the organization’s infra- budgets or development partners. The NMHS maintains structure. Thus, investment depends largely (and sometimes, a large field staff. The advantage of this approach is the completely) on external support from development partners. certainty that the data are restricted to the NMHS, which Increasingly, however, governments are moving organiza- has control over how they are used, sold, or distributed. tions, such as NMHSs, from a departmental to an agency Also, some degree of continuity is assumed, providing the operating model in some regions like Africa (Rogers and equipment is properly maintained. But the disadvantage Tsirkunov 2013a). This switch forces the NMHSs to look more is that it requires significant capital investment to renew carefully at efficiency and effectiveness to meet the demands the system periodically. Rarely is a capital expenditure of the customers and justify additional investment. It is often budget available for the routine upgrade of equipment accompanied by an expectation on the part of ministries of and most operate until they fail. Over time, the quality of finance or treasuries that there will be a return on capital the network may decline as equipment ages. employed. As a result, many NMHSs are forced to focus heav- 2) Observational networks as a service: In this model, the ily on selling services, mostly data services. But because they NMHS contracts with another party to meet its obser- often lack the capability to improve these services, they resort vational requirements. It may involve outsourcing the to protectionism through legislation to maintain their monop- entire network with the contractor providing all of the olies. This is an increasingly untenable position as more users equipment, maintaining it, and providing the NMHS with of meteorological and hydrological services have access to quality-controlled data that is ready for use in analysis the Internet and multiple sources of information. and forecasting, and may be sold to third parties. An Thus, NMHSs operating with an agency model must look advantage of this model is that there should be a fixed at their overall business model and restructure themselves cost, or at least a known annual cost, for service provi- to remain relevant. The various models used to deliver the sion. This would mean no significant capital expenditure overall services were discussed by Rogers and Tsirkunov during the lifetime of the contract, and the contractor (2013a) and are the subject of ongoing studies supported by would be expected to meet NMHS requirements for GFDRR. One of these studies is looking at different models of quality and quantity of data supplied by the network. public and private engagement in the hydrometeorological Depending on the contract, the NMHS may retain some sector and comparing performance against good public-pri- right to the equipment installed on its property—partially vate practices in other sectors. mitigating the risk of the contractor failing to perform to the requirement. In effect, the contractor is pledging the The main areas where different models can be effectively equipment as collateral against performance. Having employed are in support of the service delivery systems; the different contractors provide different aspects of the monitoring and observing systems; ICT systems (like data ser- observing network might be an advantage. A short-term vices and cloud computing); modelling systems; and capacity but important downside of this model may be the need to building (like training). The outsourcing or privatization of an redeploy or make redundant significant staff; in general, entire NMHS could be a future option, but it is not considered redeployment should be an option, but there is some here. National observations are a core responsibility of all operational risk. Another issue may be variations in con- NMHSs—and are often a key component of efforts to improve tractual arrangements. NMHSs, because existing systems have become obsolete or Mexico tried this model in 2018 when it planned to award nonfunctional, owing to lack of investment. They also absorb a contract for a public-private partnership to provide a considerable number of staff, which are themselves part data collection, transmission, processing, and delivery of the observing network. There are a number of potential services as part of modernizing its national weather options for observation and data services: service. This contract included providing: (i) automatic 1) Own and operate: In this model—which is almost weather stations, radiosondes, radar, hydrometeoro- exclusively used in NMHSs of developing countries—the logical stations, and lightning detection; (ii) collection, NMHS owns, maintains, and operates its own equipment. integration, distribution, and storage of observation Equipment is often operated well beyond its depreciated network information; and (iii) O&M of the equipment. life and replaced when capital is available through national The period of the contract was 20 years, with transfer Weathering the Change: How to Improve Hydromet Services in Developing Countries? 41 of intellectual property for any products and information (higher resolution than global and regional models); (ii) produced. But the intended outsourcing faced practical data not available to the global and regional models logistical and implementation challenges, resulting in are assimilated; and (iii) the LAM performs better than the cancellation of the initial bid. Despite this setback, the global model and others. But justification is often the market for such services is growing, encouraging anecdotal and not backed by rigorous analysis of model competition among potential suppliers of observational performance—leading to unsubstantiated, exaggerated networks as a service. claims of the benefits. The costs of maintaining the computational infrastructure and model codes are signif- 3) Data as a service: In this model, the observational net- icant, with impacts on other cost centers. However, for work, or part of it, is maintained and operated by a third hydrological forecasting, where the models are specific party, which provides a data service to the NMHS. This to a river basin, these models must be run locally. model differs from the previous one, in that it is assumed that the vendor retains its rights to resell the data. It may 2) Outsource NWP services. Reliance on regional and take the form of a public-private partnership, where the global centers for NWP guidance is a form of outsourcing. NMHS and the vendor cooperate and play complementary Some centers provide these data without charge, while roles. For example, the NMHS may restrict itself to specific others require a license fee to access their full set of uses that would not impinge on the vendor’s business. digital products. The European model is a good example Similar arrangements already exist, such as lightning of countries collaborating to provide a very high level of detection, which has high commercial value, and are the NWP capability by pooling resources in a single center. subject of discussion for emerging global data sets. The Another solution is bilateral arrangements between right to reuse these data beyond the immediate needs countries with advanced NWP and neighboring countries of the NMHSs would depend on the type of contract. In that lack this capacity. practice this should be a low-cost option, but with some Choosing between options 1 and 2 is a matter of cost, qual- significant disadvantages—notably, the NMHS would ity, and reliability. Note these options are not necessarily have restricted data rights, which may limit its options to mutually exclusive to each other as locally owned and develop its own external business prospects. Alternatively, operated LAMs still require global NWP outputs as boundary the NMHS may partner with the vendor, or others, or both, conditions. to develop new business opportunities, where additional value is added by other NMHS activities. Contractually, if Outsource ICT functions. The ICT system, which underpins part of the data set were considered essential (as defined all of the functions of a modern NMHS, lends itself most by WMO resolution 40), they would be made available obviously to outsourcing. After all, most NMHSs cannot through the WMO Information System without restriction. hire or retain sufficiently qualified staff in this area, where there are competing (and often more attractive) demands In practice, some variant on each of these models is likely for this expertise, and where the technology is changing so to create the optimal solution—depending on the specific rapidly. Not only do many developing countries want to run circumstances of each country. their own numerical models but they also want to own and Numerical prediction. There are serious questions about operate high-performance computing (HPC) systems. At the whether NMHSs should run a local limited area model same time, more advanced services are pooling their numer- (LAM)—as opposed to drawing on global or regional models ical prediction expertise in regional centers, or outsourcing (see Chapter 2)—pointing to the need for a business case to their computing needs,1 or both. justify which path should be taken. The options, which are The European Centre for Medium-Range Weather Forecasts not necessarily exclusive to each other, include: (ECMWF) is one of the best global modelling centers, and 1) Own and operate LAMs locally. This option would be its computing requirements are substantial. However, its justified if: (i) the model’s resolution is sufficiently high high-performance computing facility is operated under https://www.meteoswiss.admin.ch/home/latest-news/news.subpage.html/en/data/news/2015/9/meteoswiss-and-cscs-pave-the-way-for-more-detailed-weather- 1 forecasts.html 42 Weathering the Change: How to Improve Weather Services in Developing Countries? a service level contract to a high-performance computer In China, the China Meteorological Administration’s (CMA’s) vendor with the requirement to provide a certain level of operational hybrid cloud (Figure 3.1) highlights how inter- processing capability, memory, and storage. In effect, they nal cloud and public cloud systems are used to host and do not own the computers on which their models run. deliver meteorological services and host NWP products and Amendments to the contract allow upgrades to the main observational data. The public cloud has broad bandwidth systems and extend the contract support period. Most of the and flexible resource allocation to support a diverse data- supporting technicians are provided by the supplier, with base and network protocols. CMA has demonstrated that only a few staff employed by ECMWF. This approach ensures cloud-based operations, based on a big data platform, save ECMWF has access to the newest technologies within a fixed energy, provide faster data sharing, quicker deployment of annual operating budget. new systems, easier maintenance, and better support for scientific experiments (Jiao 2016). This type of platform Infrastructure as a service (IaaS). This platform is a form of may be developed internally by the NMHS or accessed com- cloud computing that provides virtual computing resources mercially through commodity cloud services. Although not over the Internet. It is one of three main categories of cloud readily applicable to NWP, commodity cloud services are computing services; the others being software as a service increasingly being used by advanced centers, such as the and platform as a service. IaaS is a way to integrate and cen- U.S. National Weather Service and the UK Met Office (Dodds tralize data and information. The goal is to improve process- et al. 2017) for data management and access. ing and delivery of data and remove the need for developers to move data between environments, and with the use of Application of ICT. Rapid adoption of Application advanced APIs, data can be better shared with consumers Programming Interface (API) technology and IaaS by private (Jiao 2016). This approach is also readily adapted to new sector weather services means that these services are readily sources of weather observation data from nontraditional accessible. The widespread use of smartphones in developed data sources. In addition, data can be readily shared with and developing countries alike enables citizens to find and the NMHS from external sources, which improves the quality use weather data from a variety of private and public sources, of impact-based forecasts that rely heavily on vulnerability, not necessarily their own national providers, which makes the exposure, and the response of those at risk. monopolistic tendencies of some NMHSs untenable. Common FIGURE 3.1 China Meteorological Administration (CMA) Switches Operations to a Hybrid Model From 2015, CMA started to build an internal cloud computing platform (IaaS, Infrastructure as a Service)—which consists of 1 national center and 31 provincial nodes to integrate meteorological data and application for centralized monitoring and management. A public resource is also employed in a CMA cloud, not only to deliver meteorological service but also to host numerical weather prediction products with real-time verification and NWP interpretations, as well as the satellite data of long time-series analysis. Source: Jiao 2016. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 43 business models of NMHSs include: (i) government to govern- the value chain—from the delivery of observations to the ment (G2G), where meteorological and hydrological services provision of services to end-user customers and clients. are provided to other government departments (such as disas- Working with partners, the World Bank Group (WBG) is ter management, water resources, and agriculture), either helping LMICs expand their range of options and solutions as part of the agencies funded responsibilities or through to sustainably grow their economies, reduce poverty, and intra-governmental financial transfers; (ii) government to expand opportunities. While countries have unique needs, public (G2P), where the NMHS provides public weather and the WBG is helping find the right mix of public and private public hydrological services as part of its funded mandate; funding to meet their objectives.2 and (iii) government to business (G2B), where the NMHS pro- The 17th World Meteorological Congress in 2015 gave a vides a fee for service to private clients. new perspective to partnerships by acknowledging the Governments have often perceived their NMHS as a cost growing involvement of entities belonging to the “private recovery vehicle, which can boost its operating budget by sector” (private companies, citizen’s associations, bloggers, selling services to the private sector and directly charging etc.) in weather, climate, water, and related environmental other government departments for data and services. While matters). Since then the WMO Members highlighted the this may appear attractive, it tends to incentivize behaviors, different, and at times, complementary roles and respon- which have an overall detrimental effect on the quality of sibilities of NMHSs, academic institutions, research and services available from the NMHS. It should also be noted technological agencies, and the private sector. They felt that that cost recovery of many NMHS in developing countries is closer interactions between the public and private sectors marginal at present due to the limitations of the market and would stimulate innovation and facilitate cross-fertilization, challenges related to the quality of service being offered. benefitting the society. But they also recognized that such Governments need to accept that the value of the NMHS activities could lead to the proliferation of weather and cli- is in the provision of public goods services, especially to mate information of various nature and quality—which could support warnings to protect life, property, and livelihoods. challenge the NMHSs mandate to disseminate authoritative The fee-for-service model inevitably puts the NMHS in direct weather information and warnings to the public and disaster competition with the private sector in areas where the pri- management authorities as well as the rest of the enterprise vate sector is often the more skilled. in delivering on their missions.3 Switch to service-oriented models. The move from infra- The World Bank’s support to the hydromet sector fully sup- structure to services is a challenge for development partners ported the efforts of the WMO to expand cooperation between and some client countries that are not well aligned with the various stakeholders participating in the Global Weather these business and technical shifts, and who need to adapt Enterprise (GWE) in line with the UN Sustainable Development their procurement practices and operational models. There Agenda 2020, which emphasizes the need for more partner- is a tendency to reinforce the developing countries’ “own ships between public and private actors as a necessity for and operate model” which, rather than sustaining devel- achieving the ambitious Sustainable Development Goals opment, may contribute to expanding the technological (SDGs). SDG 17 calls for strengthening the means of implemen- gaps that are already present between the more and less tation and revitalizing the Global Partnership for Sustainable advanced NMHSs and the private sector. In practice, this Development. In 2018, the WMO adopted a Policy Framework switch is not an option, but essential if NMHSs are to remain on Public-Private Engagement (Box 3.1). The topic of the viable services. public and private engagement in weather, climate and water related services is present in the program of the 18th World Public and private sector engagement. The nature of Meteorological Congress (June 2019). the relationship between the public and private sectors is changing and will inevitably affect the public sector’s business models. Any effort to improve public services must explicitly consider the role of the private sector throughout http://www.worldbank.org/en/about/partners/maximizing-finance-for-development 2 EC 70 Doc 12.2 Public-Private Engagement. 3 44 Weathering the Change: How to Improve Hydromet Services in Developing Countries? Box 3.1 WMO and the Private Sector The WMO’s Policy Framework on Public-Private Engagement their engagement in the GWE. Both the private and public sec- is intended to guide global, regional, and national action by tors recognize the opportunities for innovation and growth, the WMO to promote active engagement between the public, based on science and technological advances. Shared value is private, and academic sectors, and with other stakeholders created by promoting data sharing, accelerating the uptake of to manage and participate in the GWE. An all-inclusive research and technology developments into operations, and weather-enterprise approach aims at maximizing the bene- stimulating the creation of new services. The sustainability of fits to citizens. It aims to build understanding and enhance the GWE is a matter for joint action and can be best achieved cooperation, strengthen opportunities for WMO Members, by assuming complementary roles, which minimize overlap their NMHSs, and the private sector, on the basis of ethical and unnecessary competition, and where the strengthens of behavior to ensure a level playing field, enable efficiency and each sector are reinforced. innovation, and utilize an inclusive approach to funding basic Standards and recommended practices. The WMO facilitates infrastructure and research. There is considerable opportu- worldwide cooperation in meteorology and hydrology. It sets nity to improve the efficacy and reach of warnings, forecasts, standards and provides recommended practices to enable a and other services through more cooperative relationships unified global exchange of data in weather, climate, hydrology, among the sectors. and related fields, a harmonized data processing and forecast- People-first. WMO has adopted a “people-first” approach to ing system, and that services are provided with an acceptable public-private engagement, which is widely accepted as a level of quality and standard to specific economic sectors and vehicle to achieve the UN SDGs. This implies that the focus the public. It is important that both the public and private sec- of GWE should be on improving public safety and the quality tor meteorological and hydrological services adhere to these of life. Partnerships should, therefore, help increase access standards, particularly in the areas of observations and fore- to essential, affordable, and fit-for-purpose products and casting, to ensure the public and economic sectors’ confidence services, which eliminate vulnerabilities and exposure to in the GWE. Quality assurance is an ongoing issue for both weather, climate, and hydrological impacts. public and private sectors with greater emphasis being placed on verification and validation of all aspects of the value chain. Shared Values. It should be recognized that the public, aca- demic, and private sectors coexist and benefit mutually from Strategic Planning informed by the NMHS. These external stakeholders are not always considered at the outset of the strategic planning Given the rapidly changing landscape of the GWE, strategic process, but it is essential to understand their current and planning is an essential building block for the effective future requirements of the NMHS and to develop the plan- development of the NMHS. A relatively simple but effective ning process around these requirements. approach is promoted by the WMO,4,5 and it is recommended The South African Weather Service’s vision is an exemplar of that this be followed at the earliest stage of any effort to a user-oriented organization—“The South African Weather modernize or improve an NMHS. Service (SAWS) . . . has a vision of “A weather SMART Strategic goals. In many instances, there is a rush to invest Nation”—where citizens and institutions are able to use in an NMHS without fully understanding the strategic goals quality and reliable weather- and climate-related data pro- of the organization, which may be poorly articulated or non- vided by the organization to enhance the quality of their lives existent. For a public facing body, the NMHS must ensure and build resilience to extreme weather events and mitigate that it is set up to meet the needs of those it serves—its users, climate change impacts. “. . . SAWS will achieve this vision customers, or clients (the exact nomenclature depends on by evolving into a Weather and Climate Centre of Excellence the specific country circumstances), which are the external that provides innovative solutions to achieve a Weather- end-user of the services whose decisions and actions are smart (WeatherSMART) region, sustainable development World Meteorological Organization—Strategic Plan Template. 4 WMO Integrated Strategic Planning Guide. 5 Weathering the Change: How to Improve Hydromet Services in Developing Countries? 45 and economic growth through leadership in meteorological, ■■ Monitor and modify based on changes in the external envi- climatological and other related sciences, the development ronment of the organization. of relevant and innovative applications and products utiliz- Theory of change. Variations on this model include issue- ing cutting-edge technology and establishing and leveraging based planning, alignment, scenario planning, self-orga- collaborative partnerships.” nizing planning, real-time planning, theory of change, and Planning process. The strategic planning process helps to balanced scorecard (WMO 2016a). Each of these models is align the organization’s goals to fulfil its mission and achieve adapted to specific needs, but in essence all are intended its vision. The main output of the process is a plan, which is to achieve the goals of organization, efficiently and effec- used to communicate with internal and external stakehold- tively. WMO favors the theory of change model because of ers the goals and actions required to achieve those goals. its emphasis on results-based management. This model The basic rationales are to: first identifies the desired long-term goals and then works back to identify all of the conditions that must be in place, ■■ Agree on a common purpose. and how these are casually related, for the goals to occur. ■■ Build consensus around goals, objectives, and priorities. These are all mapped in an outcomes framework or logic ■■ Increase effectiveness. model that provides the basis for identifying what type of activity or intervention will lead to the outcomes identified ■■ Increase efficiency. as preconditions for achieving the long-term goal. This way ■■ Provide the basis for resource allocations and ensure value the link between activities and the achievement of the long- for money. term goals is more fully understood.7 ■■ Understand, predict, and adjust to changing circumstances. ■■ Improve decision making. Concept of Operations (CONOPS) ■■ Improve organizational abilities. The services provided by NMHSs can be considered a com- ■■ Improve communication. plex, dynamic “system of systems” that deliver a spectrum ■■ Assess performance. of meteorological and hydrological services to a variety of users—from public guidance and warnings to specialized There are several different approaches to strategic planning. services for other government agencies. They may also play The World Bank’s ten-step guide is to:6 a role in the provision of bespoke, commercially oriented ■■ Agree on the process. services. The purpose of the CONOPS is to ensure that: (i) ■■ Carry out an environmental scan—how the organization re- there is a consensus among all stakeholders so that every lates to its external environment. partner understands and supports the proposed operational system; (ii) risks are reduced by ensuring every aspect of the ■■ Identify key issues, questions, and choices. system is determined before it is procured or implemented; ■■ Define or review and update the organization’s values, mis- and (iii) quality improvement is built in by taking every sion, and vision. opportunity to leverage existing and new infrastructure to ■■ Develop a shared vision—where the organization wants to increase system performance. be in five years. The CONOPS is a document that describes the scope and ■■ Develop a series of goals. characteristics of the proposed system and the way the sys- ■■ Agree on key strategies to achieve goals. tem (or system of systems) will be used (Technical Insight 3.1). It must consider all stakeholders, ensuring that the ■■ Develop an action plan (Concept of Operations) that ad- CONOPS is readable and relevant to high-level decision dresses goals, objectives, and work on an annual basis. makers and systems operators (UCAR 2010). And it is built ■■ Write the strategic plan. on the strategic plan, business modelling, and stakeholders’ https://siteresources.worldbank.org/INTAFRREGTOPTEIA/Resources/mosaica_10_steps.pdf 6 Center for Theory of Change. 7 46 Weathering the Change: How to Improve Hydromet Services in Developing Countries? TECHNICAL INSIGHT 3.1 A Sample CONOPS The CONOPS may include the following: How and how well the needed capability is currently being met. Why the system is needed and an overview of the system itself—for example, a severe weather forecasting How the system will be used, including operations, platform. maintenance, and support. The full system life cycle from deployment to Scenarios illustrating specific operational activities decommissioning. involving the use of the system. Different aspects of the system—operations, maintenance, The CONOPS should answer the following questions: support, and decommissioning. Who are the stakeholders involved with the system? The different groups of users—for example, end-users of the system, forecasters, technicians, public weather When will the activities be performed? advisers, and their different skills and limitations. Where are the geographical and physical locations of the The physical environment and locations in which the system? system is used and supported. Why does the NMHS require this system? The relationship between the system and other systems. What are the known elements and the high-level capabili- When the system will be used, and under what ties of the system? circumstances. How will the system be resourced, designed, built, and maintained? It is recommended that the CONOPS addresses the following issues 1. Scope 4. System overview a. Vision for the system a. Specific goals and objectives that are measurable and b. Outline of the contents of the document time bound c. Purpose for implementing the system b. Interdependencies between subsystems d. Intended audience/beneficiaries c. Confirmation that the system’s capabilities will satisfy its mission e. Limitations of content covered 5. Operational and support environments 2. Knowledge references a. Facilities a. Discussions with stakeholders and experts b. Equipment b. Studies of systems from other countries c. Hardware c. Analysis of mission requirements and operational needs d. Software d. Recommendations offered by vendors and product e. Personnel manuals f. Operational procedures g. Maintenance, training, and support requirements 3. Operational description a. Summary of each user’s role and activities 6. Operational scenarios b. Clarification of the order of user operations a. A range of stakeholders’ perspectives c. Summary of the operational process procedures b. A range of stress/failure scenarios (both typical and d. Description and flow diagrams associated with organi- extreme circumstances) zational decision-making and management structures Note: A sample Concept of Operations for modernizing and operating an NMHS is based mostly on the approach taken by the COMET program of the University Corporation for Atmospheric Research in developing a flash flood early warning guidance system. Source: UCAR 2010. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 47 requirements. The system of systems approach shown in ■■ Section 6 describes operational scenarios. Figure 2.1 provides the basic building blocks of the CONOPS. ■■ Section 7 summarizes operational, organizational, and other It can be used to identify the current status of any NMHS impacts during development. and to visualize investments, component by component, in each of the systems and subsystems to achieve a particular ■■ Section 8 analyzes the proposed system. level of improvement. The complexity of each system and its ■■ Section 9 includes abbreviations and acronyms. subsystems varies depending on the size, level of develop- ■■ Section 10 includes annexes. ment, and resources of an individual NMHS. A pre-requisite An annotated table of contents of a CONOPS is shown in of each of the individual systems is access to sufficient staff Annex 4. This can be used as a template for preparing a with the capacity to understand and operate a particular CONOPS document. Because this follows an IEEE standard, system. At the same time, access to various systems within it is important to adhere to the structure of the document the observing systems block is essential for the rest of the as closely as possible for the document to be considered a systems to function and produce various products for deliv- CONOPS. The rest of this chapter looks at Sections 1–8, and ery to the users—and eventual establishment of a monitoring Box 3.2 cites some common mistakes to avoid in drafting and feedback loop to the users. All associated equipment, the CONOPS. facilities, material, software, hardware, policy and technical documentations, services, training, and personnel required for operations and support of the NMHS are included as part of each subsystem. How each system is supported depends Section 1: Scope on the choice of business model. Section 1 describes the approach used to develop the CONOPS. Given that the primary mission of NMHSs is to The CONOPS should evolve with the system. It is a vehicle to serve the needs of the public and government, the CONOPS communicate high-level quantitative and qualitative char- should be based on an assessment of user requirements, acteristics of the system to the user, developer, operator, which should be derived from a series of stakeholder work- and other stakeholders. The CONOPS addresses the chal- shops—including an element of user education, since many lenges involved in ensuring the NMHS is fit for purpose. It is users do not understand what can be expected from an “ide- intended to be a structured document with an architecture ally” functioning NMHS. defined by the 1362-1998 IEEE guide for information tech- nology system definition Concept of Operations (CONOPS) The initial approach to preparing the CONOPS may use con- document.8 This formality services several purposes: a cept analysis—which is the process of analyzing a problem common understanding of the systems requirements, a domain and an operational environment—for specifying the common format to test and exchange information among characteristics of the proposed system from the users’ per- projects within the NMHS community and with others, and spective. This approach helps to clarify and resolve vague proven approach to development of operational systems. and sometimes conflicting needs, wants, and opinions by The CONOPS has 10 sections. reconciling divergent views. The goal is to minimize the potential for designing a system in which each individual ■■ Section 1 describes the approach used to develop the CON- function meets its specifications, but the system, as a whole, OPS. fails to meet the users’ needs. ■■ Section 2 provides a list of reference documentation used in the creation of the CONOPS. ■■ Section 3 describes the current NMHS systems. Section 2: Referenced ■■ Section 4 discusses the justification and nature of changes Documents based on the most current information. Section 2 summarizes the documents used in the prepara- ■■ Section 5 provides information on proposed system con- tion of the CONOPS. It should reference the following WMO cepts. standard and guidelines, at a minimum: http://sse.tno.nl/IEEE_STDS/SESC/1362-1998.pdf 8 48 Weathering the Change: How to Improve Hydromet Services in Developing Countries? ■■ WMO, 2014. The WMO Strategy for Service Delivery, WMO- ■■ http://www.wmo.int/pages/prog/www/WIS/manuals_ No. 1129. Geneva, Switzerland. http://www.wmo.int/pages/ guides_techregulations_en.html (the link to all Technical prog/amp/pwsp/documents/WMO-SSD-1129_en.pdf Regulations above). ■■ WMO, 2016. WMO Technical Regulations: General Meteoro- These should be supplemented by certain NMHS documents, logical Standards and Recommended Practices. WMO-No. including: 49, Volume I. Geneva, Switzerland.  ■■ The statutory authority of the NMHS, which states its man- ■■ WMO, 2016. WMO Technical Regulations Meteorological dates and legal responsibilities. Services for Air Navigation. WMO-No. 49, Volume II. Gene- ■■ User requirements—these would be acquired during any va, Switzerland.  strategic planning process, but, in any case should be fre- ■■ WMO, 2006. WMO Technical Regulations: Hydrology. WMO- quently updated and will drive continuous improvements in No. 49, Volume III. Geneva, Switzerland.  the NMHS to remain responsive to these requirements. ■■ WMO, 2013. WMO Technical Regulations: Quality Manage- ■■ The Strategic Plan of the NMHS. ment. WMO-No. 49, Volume IV. Geneva, Switzerland.  ■■ An analysis of alternative approaches, which could be em- ■■ WMO, 2011. Manual on Flood Forecasting and Warning. ployed to meet the strategic objectives of the NMHS. WMO-No. 1072. ■■ Any other documents, which may include project appraisal ■■ WMO, 2006. Guidelines on the Role, Operation and Manage- documents, project descriptions, government data sharing ment of National Hydrological Services, WMO-No. 1003. policies, and any other information that may influence the operations of the NMHS. ■■ WMO, 2017. Manual on the Global Data-processing and Fore- casting System (GDPFS). WMO-No. 485, 2017 edition, Gene- va, Switzerland. BOX 3.2 Avoiding Common Mistakes Several common mistakes can be made when developing CONOPS (UCAR 2010). These include: 1. The NMHS expecting its system vendors, contractors, or 4. Assigning unqualified staff for CONOPS development. other external partners to develop the CONOPS, as a part Adequate representation by personnel with experience of the other deliverables. in the organization’s strategic, operational, technical, The NMHS must ensure that its staff is responsible for the administrative, financial, and communication programs development of its CONOPS, albeit with external expert is essential. assistance. 5. Adopting another organization’s CONOPS. 2. Postponing CONOPS development until after the system Borrowing another organization’s plan is likely to ensure has been designed and delivered. that mistakes of others will be repeated. Worse, A CONOPS is a document that must be drafted before a stakeholders will have a much weaker commitment system design is finalized and kept continuously up-to- than if they are given the opportunity to contribute to date as systems requirements change. the development of their own CONOPS. 3. Allocating inadequate staff resources for CONOPS 6. Neglecting to update a CONOPS while the new system is development. being implemented and once it becomes operations. The process of conceptualizing NMHSs’ systems is The CONOPS should reflect the actual system design, complex, tedious, and time consuming. It requires mission requirements, and operational vision. It should study tours to learn about current practices in CONOPS be reviewed and updated regularly to remain relevant. development from other leading NMHSs. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 49 Section 3: Current System of the high volume of observations, owing to their lack of the ability to assimilate these data in nowcasting systems Section 3 of the CONOPS describes and assesses the current and very short-range forecasting models. Also, automation operational system. This includes: (i) background, objec- does not necessarily mean instruments should be left unat- tives, and scope of the current system and its subsystems; tended, given that in extreme environmental conditions, (ii) operational policies or constraints; (iii) the details of the instruments need to be inspected. current system and subsystems; (iv) modes of operation for Automation versus manual systems. The transition from the current system; (v) users of the services provided by the entirely manual to even quasi-automation is a difficult, NMHS; and (vi) staff and the support environment. relatively long-term (two to four years), and costly step in Possible approaches. Based on the current status, the anal- all countries—especially countries that rely on unskilled or ysis should propose different approaches to developing the semiskilled labor and that might be uncertain about future NMHS. This might include: (i) maintenance of the current employment (Lynch and Allsopp 2008). NMHS management business model, (ii) privatization of some or all functions, (iii) often overlooks the fact that introducing automatic systems partnerships with external public and private entities, and requires fundamentally changing data collection routines (iv) outsourcing of some functions. The analysis should also and operating procedures. This process is costly because consider the activities of other development partners, which it requires extensive staff training, significant technical may be actively investing or proposing to invest in the NMHS. support, and parallel manual and automatic observations Frequently these efforts are limited in scope, but they can for at least one year at all climatic and other stations with have a significant bearing on modernizing the NMHS—if those significant historical records. efforts encompass key components of the system or require a Much more effort is needed to exploit the benefits of an particular business model to be sustainable. It is not uncom- automated network than is done in most modernization mon that one project results in unplanned delays in another, projects. In many instances, automatic stations are viewed as a result of incomplete exchange of technical information. as supplemental to manual stations, and no effort is made Very often the proposed investment is made because the to transition the network to fully integrated automatic sta- NMHS’s process for monitoring and observing, modelling, tions—which is why it is important to engage all staff mem- forecasting, and service delivery rely largely on outmoded bers, including field observers, so that they fully understand systems. A motivation for modernization includes: (i) the the benefits. Vandalism is also a generic problem in most acquisition of new observational capabilities (such as fully developing countries, especially where solar panels are exploiting new radar technologies required for nowcasting used. Experience suggests that where the local community and very short-range forecasting); and (ii) access to, and is engaged and can access the data directly from the station, best use of, high resolution global and regional model out- vandalism is less of a problem. Over time, the number of puts—for generating tailored products for users and operat- observers at the automatic weather station sites should be ing the latest technologies for dissemination and delivery of gradually reduced, and their duties should be transformed forecasts and warnings. from pure observers to technicians, station guards, or com- munity climate extension workers.9 NMHSs tend to pay more attention to modernizing monitor- ing networks and instruments, usually the most expensive Upper-air measurements. These are particularly important part of the program, which has a high risk of being unsus- for forecasts, but many of the poorest countries have stopped tainable. Modern observation methods are typically based taking such measurements because of the cost of the expend- on as much automation as possible, because automation ables. Continuing to take upper-air measurements requires the increases the frequency and spatial coverage of the obser- government to commit to the additional O&M costs that accom- vations. While the advantage of the greater density of the pany the new capabilities. In the case of upper-air stations network and frequency of observation is apparent, most that are critical for global forecasts, development agencies or developing countries’ NMHSs are not able to take advantage the WMO might want to allocate special long-term funding. Creating a cadre of climate extension workers from existing observers could help improve the community’s response to hazardous weather and increase local 9 knowledge on climate resilience. 50 Weathering the Change: How to Improve Hydromet Services in Developing Countries? The rapid development of Global Positioning System–Radio the stakeholders and end-users—by building on technolog- Occultation (GPS-RO) technology means that atmospheric pro- ical advances to reduce the economic impact of disasters files of temperature and humidity may be derived from satellite of meteorological and hydrological origin, and to improve observations at virtually any point. This means that models climate resilience. can be initialized with data at every grid point. However, this What are the attributes of the system? From the overall sys- does not eliminate the need for radiosondes to provide direct tem perspective, the proposed modernization of an NMHS measurements of temperature, humidity, and winds aloft, and should possess the following attributes: as a reference network for calibration. ■■ Infrastructure independence: an architecture that allows the Overall system goals. In order to meet the NMHS’s strate- NMHS to operate under all circumstances, including during gic goals, the proposed system typically should be able to extreme hazards. accomplish the following: ■■ Scalability: the capability to accommodate growth and ■■ Affordable and easy to maintain observing networks. evolving technologies without requiring a complete replace- ■■ Effective and efficient data transmission, archiving, and ment of the systems. management, including rigorous quality assurance/quality ■■ Flexibility: the ability to tailor services to users’ future re- control (QA/QC) of data. quirements. The system should be flexible enough to inter- ■■ Access to and effective use of NWP ensembles. face with other agencies’ evolving systems. ■■ Reliable forecasting system, including impact-based. ■■ Comprehensiveness: the system should be able to meet all anticipated needs. ■■ Robust dissemination and communication. ■■ Extensibility: the ability to handle new types of observa- tions, numerical models, and forecasting systems. Section 4: Justification for What should an NMHS be able to do? To meet strategic and Nature of Changes objectives, the NMHS needs to integrate its systems and subsystems, along with incorporating legacy systems and Section 4 describes the rational for the proposed system. new systems provided by other development partners. Most NMHSs have a long history of providing meteorologi- Increasingly, parts of the system may be provided by others cal and hydrological services to their countries. Continuing as a contract or within a partnership with shared responsi- growth in the weather- and climate-dependent economic bility. In any case, the NMHS, or its contractors/partners, or sectors—and in the increasing exposure of a growing popu- both should be able to: lation to weather and climate extremes—makes it an imper- ative to improve the quality and level of services provided ■■ Receive observational data from a variety of international by the NMHSs and other service providers to protect lives, and national sources and use these data in the forecasting livelihoods, and economic development. system. One of the purposes of a modernization program is to ■■ Apply quality control to observational data. increase the number of services that can be provided by the ■■ Ingest, preserve, and provide access to observational data. NMHS to meet the growing needs of their users. There is ■■ Store observational data in a manner that is independent of an increasing demand from users for more accurate, timely, any particular hardware and software over long periods. and useful forecasts and warnings. Advances in knowledge and technology have increased the accuracy of meteoro- ■■ Retrieve NWP/EPS data from global and regional centers logical predictions that help save lives, protect livelihoods, and use these data in the forecasting and service delivery and improve economic performance. Most modernization systems. efforts are initiated because the NMHS has not kept pace ■■ Carry out verification of NWP/EPS forecasts and apply with these requirements or developments. It is critical to post-processing techniques (that is, statistic downscaling), improve the quality of weather, climate, and hydrological such as Model Output Statistics (MOS), for calibrating mod- information and services—and their accessibility and use by els throughout the country. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 51 ■■ Apply and use dynamic downscaling for improved under- services. Upper-air stations and radar are costly, but ben- standing of regional and local weather phenomena and eficial. Upper-air stations are particularly useful to improve fine-tuning of forecasts. the local and downstream forecasts, if these data are assim- ■■ Produce user-tailored numerical products for specific users. ilated in the global models. Radar would improve thunder- storm detection and quantitative precipitation estimates, in ■■ Use EPS techniques for impact-based forecasting. turn improving nowcasting and very short-range forecasts. ■■ Scale up in order to store and preserve observational and A significant investment in data servers would be required. forecasting data. The modelling and forecasting system would provide a data ■■ Code and decode observational and forecasting data using service to the modelling systems, the objective and impact international standards. forecasting and warning systems, and to end-users that require validated observational data. ■■ Organize and maintain an adequate metadata repository and asset management—including inventory register of the Modelling systems. All NMHSs should be able to exploit the contents of the systems, their conditions, maintenance, and available global and regional prediction centers’ products at replacement/calibration/upgrade schedules. the highest resolution available. The architecture of the mod- elling system, therefore, should be based on access to digital ■■ Provide electronic access to weather, climate, and hydrolog- data from these centers, including the purchase of licences, ical observational and forecasting data and information for if required. The choice of the NWP system should be in close all users based on established user rights and privileges to consultation with the relevant staff and management to ensure ensure that users are able to access all of those electronic due consideration of the continuity in usage of known mod- data and information to which they are entitled. els—preferably those with higher resolutions and established ■■ Provide access to weather, climate, and hydrological obser- mechanisms to provide direct technical support from global vational and forecasting data and information in a manner and regional center(s) running such models. Model selection consistent with current technology and the changing expec- also requires knowledge of how to apply each model to partic- tations of diverse users. ular physical situations, which requires an extensive objective ■■ Adapt to changing technology in order to provide the level of verification program. The use of EPS digital data would enable services desired by the users. the application of impact-based forecasting techniques, as well ■■ Exploit current knowledge and technology to provide the op- as the provision of the forecast uncertainty for decision making. timum level of quality services expected by the users. Statistic downscaling (such as Model Output Statistics) should be implemented for the calibration of both deterministic and probabilistic forecasts from ensembles. Dynamic downscaling Section 5: Concepts of should be considered for hazardous-prone areas to fine-tune and refine forecasts. To support application areas, including the Proposed System hydrology and agrometeorology, manipulation of digital Section 5 describes the proposed system. The proposed data (that is, coding and decoding) should be implemented. system would be derived from the analysis of the existing Introducing LAM should be done with caution, taking into con- system of systems, new and evolving user requirements, sideration the points made earlier. and any constraints imposed by the government and finan- Hydrological modelling systems should be flexible to allow cial instruments. Ideally, the proposed system follows the a wide range of forecasting models and should provide an system of systems architecture proposed above. interface with meteorological models to use quantitative Monitoring and Observing Systems. More frequent, precipitation forecast data. It is also important to consider high-quality meteorological and hydrological data would the affordability of the models, if they are not open source. improve short-range forecasts. A detailed network design Model selection requires knowledge of how to apply each would be required to determine optimum impact. If the model to a particular physical situation. This necessitates existing synoptic network is not dense enough, this should not only an understanding of the model and its use but also be expanded using automated stations where feasible—for the physical setting—including a detailed understanding example, augmentation would allow better agricultural of the underlying vegetation, topography, soils, geology, 52 Weathering the Change: How to Improve Hydromet Services in Developing Countries? and climate. This system can be used to specify the type of ■■ An integrated modelling environment within a single fore- hydrological models that are most favorable for forecasting cast system. purposes and appropriate to the available database and ■■ Significantly enhanced data and hydrologic modelling visu- spatial and temporal boundary conditions of modelling. The alization. system should provide for ensemble-based flood forecasting to add a probabilistic component by using inputs from multi- ■■ A framework to expand modelling and forecast capabilities ple NWP models—which, in turn, would require a data ingest from major rivers to minor tributaries, and flash flood–prone process. For example, ECMWF ensemble NWP fields of QPF areas. and QTF and an efficient hydrological observation network ■■ Extended forecast lead times to several days and hydrolog- would be required to automatically generate hydrological ical outlooks to weekly, monthly, and seasonally (based on forecast ensembles at all forecast locations. Hydrological remote sensing and snow modelling). forecasting based on an ensemble of NWP models would ■■ The capability for both short lead time and extended (sea- permit the modelling of the complete river basins (includ- sonal) probabilistic hydrologic forecasts, using ensemble ing snow modelling and reservoir simulation modelling) to NWP outputs (primarily, precipitation, and temperature produce monthly to seasonal hydrologic outlooks for water fields) as hydrological model inputs. resources planning and management purposes. ■■ Forecast verification, which would aid forecasters to identify The primary user of the modelling systems products and strengths and weaknesses in the forecast process to guide services would be the forecasters running the objective and model and forecast process improvements. impact forecasting and warnings systems, and other agencies. The hydrological forecasting system should provide: (i) access If hydrological modelling is done elsewhere, the modelling to global hydrological products; (ii) implementation of real- system would be expected to provide a feed of model data. time forecast evaluation, model verification, calibration, and Objective and impact forecasting and warning systems. post-processing systems; and (iii) visualization tools to allow The architecture of a modern forecasting system is based data from rain gauges, radars, and satellites to be overlaid on the implementation of real-time forecasting processes— seamlessly into single multisensory fields on an hourly basis. which consist of (i) quality-controlled data services from the Service delivery systems. The service delivery system monitoring and observing system; and (ii) verified, post-pro- should fully engage the users, and it should include the cessed, and calibrated NWP data and statistic and dynamic introduction of public weather and hydrological services downscaling from the modelling system. The forecasting for major sectors (such as DRM, water resources, agricul- systems would include: (i) interpretation and communica- ture, civil aviation, transport, energy, and health). The tion of products; (ii) visualization and processing tools to system should support: (i) developing and enhancing new enable integration, overlaying, and manipulation of obser- and existing user-tailored products and services; (ii) the vational and NWP/EPS data for nowcasting and very short- Common Alerting Protocol (CAP)10 capability; (iii) improving range forecasting, short- to medium-range forecasting, and dissemination mechanisms to all communities (particularly seasonal forecasting; (iii) real-time forecast monitoring and remote areas)—including mobile applications (for warnings, verification; and (iv) impact-based forecasting. It would also food security), FM radios, SMS, and web-based services, on require a quality management system and standard opera- top of traditional media; and (iv) developing mechanisms tional procedures. Probabilistic forecasting is an important for evaluation of forecast utility and user satisfaction. tool that should be implemented as a part of any modern forecasting system. Capacity building of staff would be criti- The system should specifically address developing and oper- cal to understand and use these new products. ationalizing impact-based forecast and warning services, and include a post-event review and assessment process. Modernization of hydrological forecasting should be In addition, the system could include an Agriculture and expected to provide: Climate Advisory Service (ACAS) (including drought moni- toring) and satellite assessment of crop condition. A National http://www.wmo.int/pages/prog/amp/pwsp/CommonAlertingProtocol_en.html 10 Weathering the Change: How to Improve Hydromet Services in Developing Countries? 53 Framework of Climate Services (NFCS) is also recommended ■■ Project management. to be included in the system—and as part of climate service ■■ Management training. delivery or the latter, it should contribute to the NCFS if this is developed outside of the NMHS. Also, a digital library of ■■ Financial management. climate-relevant information should be developed to digi- ■■ Quality management and QMS. tize original data (data rescue), quality control the historical ■■ Technical skills to support meteorological and hydrological data, and create a centralized and standardized database observing networks. (including metadata). ■■ Instruments and detectors for maintenance and engineering. ICT systems. A modern data management, communication, ■■ Enhanced skills in weather forecasting using numerical mod- and ICT system should be part of any NMHS. The moderniza- els on all timescales from nowcasting to long-range forecast- tion and expansion of an NMHS’s observation network system ing. and improvement in forecasting and service delivery systems ■■ Enhanced skills in weather forecasting based on remote require comparable improvements in ICT infrastructure and sensing. data integration capacity. The systems should be fully redun- dant, enabling continuous operations during extreme events. ■■ Enhanced skills in hydrological forecasting using numerical A critical, and sometimes overlooked, component is access to models. high speed Internet—which, if not available, should be intro- ■■ Understating of the end-to-end early warning production duced as soon as possible in the project. and delivery, including coordination with disaster manage- ment authorities. Capacity building. For a strong and effective NMHS it is essential to have continuous access to new skills for new and ■■ Impact-based forecasting and warning services. existing staff, provision of short- and long-term courses, and ■■ Mesoscale meteorology. on-the-job training. Some of these skills could be provided ■■ Verification and statistics methods, including big data meth- by universities and technical institutes. Generally, a steady ods. supply of meteorologists, hydrologists, and related specialists (at the B.Sc., M.Sc., and Ph.D. levels) are required to replace ■■ Database management. retiring staff. These new staff would need to be equipped with ■■ IT management skills. the skills and abilities to perform at the required skill levels, ■■ Skills in Public Weather Services (PWS) and service deliv- in line with the proposed advances in forecasting and observ- ery—including user/stakeholder consultation, communica- ing systems, and service delivery practices. Most NMSs and tion, negotiation, and feedback gathering. NHSs develop cooperative relations with local universities, so that the latter can provide qualified staff to the operational ■■ Knowledge of social, environmental, and economic sectors services that match current needs. sufficient to provide consulting services to their users. ■■ Enhanced skill in climate prediction using numerical meth- Technical- and service-related expertise will change as new ods. meteorological and hydrological products become available. Such changes would require review and upgrading of skills ■■ Public education and outreach. and responsibilities of staff and a long-term plan for staff- Leadership and management training should be conducted ing, to be defined in the context of the technical upgrades as a priority to facilitate the rapid development of high-qual- and service delivery strategy. Technical training should be ity management processes near the start of the project. offered to main stakeholders to facilitate the smooth opera- Team building workshops to build strong team performance tional processes to the benefit of all concerned. in the NMHS should also be conducted (for example, through leadership coaching). The capacity building system should Training requirements for capacity building. The NMHS support retraining of current technical staff, and recruitment should consider developing a plan (initial and ongoing), of new staff, as required, to suit the modernized environ- which addresses all capacity building needs prior to starting ment. An important component of capacity building relates any new operations and activities. Areas where training is to training support for the main stakeholders and end-users usually required include: to understand and be able to apply the NMHS’ information 54 Weathering the Change: How to Improve Hydromet Services in Developing Countries? and services in their decision-making processes. It would also be important to conduct outreach and public education Section 7: Summary of Impacts activities for communities. Service monitoring and feedback Section 7 describes the operational, organizational, and systems should be established to gather crowdsourced data implementation impacts of the modernization program. from trained volunteers, as well as evaluation of user satis- These should be documented and addressed, including: faction. Relations and collaboration with local universities ■■ Changes in the operational budget. should be established or strengthened. ■■ Changes in the operational risks. Changes in the maintenance budget. Section 6: Operational Scenarios ■■ ■■ Changes in the maintenance risks. Section 6 of the CONOPS document expresses what users ■■ New modes of operations, based on new forecasting tech- want and envision. Scenarios convey these needs in a non- technical language. Overlap exists between different scenar- niques. ios as a result of interaction between different users or due to ■■ Changes in the quantity, type, and timing of data to be input similarity between different activities. For example, the docu- into the system. ment might include scenarios for disaster managers, agricul- ■■ Uses of new data sources. ture extension workers, farmers, and the general public. ■■ Changes in procedures. A scenario is a step-by-step description of how the NMHS should operate and interact with both its users and external ■■ Numbers and skill levels of personnel needed for contingen- interfaces under a given set of circumstances. Scenarios cy operation. should enable readers to walk through them and gain an ■■ Changes in the number of personnel, skill levels, position understanding of how all principal parts of an NMHS func- identifiers, or location of personnel. tions and interacts. ■■ Revision of position descriptions to reflect changes in busi- For example, the following scenario for DRM highlights how ness practices. an NMHS would support emergency operations prior to, during, and after a major event: ■■ Commitment of resources to the new system. ■■ The need for cross-functional, interdisciplinary staff teams. ■■ The NMHS provides long lead time outlooks ahead of the monsoon season, based on numerical model guidance, with ■■ The development of education and increased training for regular updates (such as monthly). both NMHS staff and users. ■■ The NMHS provides long lead time outlooks on developing ■■ Improved opportunities for career development for NMHS adverse weather situation, up to 10 days ahead, based on staff. numerical model guidance. ■■ Relationship between the NMHS and external partners. ■■ The NMHS enables emergency operations to begin initial ■■ Articulation of business rules, templates, and other controls preparation. needed for operational implementation. ■■ Guidance provided by the NMHS is updated daily—and clos- ■■ Development of training for staff. er to the severe weather event and associated impacts and ■■ Consideration of parallel operations of the new and existing briefings are given on a frequent basis. systems. ■■ Severe weather forecasters must relocate to emergency op- ■■ Operational impacts during system testing and transition of erations center. the proposed system. ■■ Warnings are issued via the emergency operations center, and directly through the media and other communication channels. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 55 Section 8: Analysis of the Summary of Changing Roles Proposed System NMHSs and their private sector counterparts can comple- ment each other in their efforts to provide levels of service Section 8 considers various improvements, disadvantages to protect lives, livelihoods, and property in an era of and limitations, and alternatives and trade-offs. Some of the increasing meteorological and hydrological hazards. NMHSs benefits would likely include: need to evolve quickly, not because of the threat of com- ■■ More accurate and timely forecasts, based on full use of the petition from the private sector, but rather because society high-resolution ensembles provided by advanced numerical needs to be able to make informed decisions. The changing prediction centers, such as ECMWF. requirements of users, in response to growing threats, is the main driver for change. Services must be and must remain ■■ More confidence in warnings issued with greater under- fit-for-purpose. Therefore, they need to evolve as the infor- standing of risk and the ability to improve decisions in low mation requirements of their users change. probability, high impact extreme weather events. The traditional model of an NMHS doing everything alone ■■ More detailed shorter range forecasts. is no longer tenable from either an economic or technical ■■ Nowcasts and very short-range weather forecasts. viewpoint. New business models are needed that can help the public sector deliver what is required. Few of these mod- ■■ Dynamic, ensemble-based flood forecasts. els have been tested, and efforts are under way to determine ■■ Impact-based forecasts and warnings. the most effective ones under a variety of national condi- Some of the potential disadvantages include: tions. Static organizations cannot meet the changing needs of their users and customers. Strategies must change and ■■ High development costs. tools to evaluate and test new directions are needed. The ■■ Staff anxiety brought about by new responsibilities. Concept of Operations (CONOPS) is a powerful tool to help design and evolve an operational system. To be useful, it ■■ Poor staff morale without proactive change management. must be kept up-to-date as users’ requirements and techni- ■■ Higher operating and maintenance costs. cal opportunities change. ■■ Unrealistic expectations of users and stakeholders. This chapter has emphasized the role of public and private By adhering to this approach, it is possible to assess the sectors. There is growing recognition that creating resilient societies depends on harnessing the capabilities of both— impact on the operational system of any change. It should be something that needs to be kept in mind in any effort to help updated frequently and whenever there is a material change a country strengthen its meteorological and hydrological in the system, which may affect any aspect of the operations services. from operational costs to the delivery of services. In the next chapter, the focus is on practical approaches to modernizing NMHSs. 56 Weathering the Change: How to Improve Hydromet Services in Developing Countries? Specialists from MeteoRwanda checking soil temperature in Kigali. Source: D. Kull, World Bank Group, 2015. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 57 Chapter 4 Recommendations on Modernizing NMHSs > Although there is no definitive approach to designing NMHS modernization projects in low-and middle-income countries, because of many various country specific elements—such as institutional setups, NMHS capacity, natural conditions, cultural context, and project objectives—a lot has been learned over the past decade that can contribute to more successful outcomes. > A key problem is that the goals of the NMHSs are often not well articulated and not supported by an effective and evolving planning process, making it difficult for proposed investments to respond to these goals. > Thus, the starting point for a modernization project should be a strategic plan, which lays out the NMHSs’ goals and the means to achieve them. It should include the adoption of appropriate business plans and models and the implementation of the strategy within an operational plan—that is, a concept of operations or CONOPS. > The aim of the CONOPS is to explore the entire operational system from the perspec- tive of users and stakeholders—enabling various options to be tested theoretically to determine what is critical to achieving the goals. 58 Weathering the Change: How to Improve Hydromet Services in Developing Countries? Introduction observing systems, modelling systems, forecasting systems, and ICT systems; and (4) improving all aspects of project Despite the evident benefits of strengthening weather, management, which includes support for the project man- climate, and water information systems and services, and agement unit, and potentially international advisors and a modernizing NMHSs, considerable efforts are needed just systems integrator, if they have responsibilities spanning to convince governments and World Bank (WB) teams to more than one project component. proceed with project preparation and commit resources for preparing and implementing complex, and often relatively Component 1: Institutional Strengthening small, hydromet modernization activities. And once the This objective of Component 1 is to: (a) strengthen the preparation is done, there is implementation—and impor- legal and regulatory framework of agencies involved in the tantly, steps to make the improved services sustainable, modernization project (including the NMHS, disaster man- given the many instances of improved services in low-and agement, water resources, and agriculture); (b) improve the middle-income countries (LMICs) not holding up over time. NMHS’s institutional performance as the main provider of What is the best approach to NMHS modernization? Although weather, climate, and hydrological information; (c) build there is no definitive approach to designing these projects, a capacity of its personnel and management; (d) ensure oper- lot has been learned over the past decade that can contrib- ability of future networks and systems; and (e) support proj- ute to more successful outcomes. This chapter begins with ect implementation. This approach applies to the NMHSs and recommendations on the preparatory phase before moving may extend to other agencies. Where the NMSs and NHSs are on to those for implementation and sustainability. A key separate organizations, ensuring close collaboration is vital to enable flood forecasting and management. Some of the message is a need for NMHSs to articulate their goals and subcomponents—such as strategic planning, business mod- lay out a strategy early on for how to meet them. elling, and the preliminary CONOPS—help ready the project plan for implementation. The main subcomponents are: How to Carry Out NMHS Road mapping. This includes: (a) mapping all the relevant Modernization Programs paths that can be taken to meet end-users’ needs (route, actions, and milestones); and (b) holding workshops with The first step in the preparation process is for the project major stakeholders and users of the NMHS’s products and team to concentrate on matching the needs for moderniz- services (including development partners and ministries ing the NMHS with opportunities. This can take the form of responsible for planning and budgeting). identifying champions in the country and within the WB; identifying potential public and private sector partners; and Strategic planning. This includes: (a) holding joint work- identifying funding commensurable to the task (see Box shops with the major users of the NMHS’s products and 4.1). A key element is developing the concept of operations services (such as agriculture, emergency, health, water (CONOPS), which explores the entire operational system resource management, energy, and transportation); (b) holding internal workshops with agency staff; and (c) writ- from the perspective of users and stakeholders—enabling ing the strategic plan. various options to be tested theoretically to determine what is critical to achieving the goal. Business modelling. This includes assessing the applicabil- ity of different business models to achieve strategic objec- As for implementation, so far in this report, we have dis- tives, which would feed into developing the CONOPS. cussed the operation of an NMHS within the framework of a system of systems (see Chapter 2). But now it is useful Development of a concept of operations (CONOPS). This to map the systems into four larger components: (1) institu- includes: (a) holding workshops with development partners tional strengthening, which comprises the quality manage- to incorporate their activities; (b) conducting a review of ment systems, capacity building, and technology infusion institutions’ operations and management (such as observ- systems; (2) improving service delivery, which comprises ing and forecast system requirements; and (c) drafting of service delivery systems and feedback systems; (3) mod- CONOPS version used as the guide for designing the mod- ernizing infrastructure, which includes monitoring and ernized systems. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 59 INVESTING IN HUMAN CAPITAL REQUIRES A LONG-TERM APPROACH W ith financial support from the World Bank’s Central Kyrgyzhydromet is able to offer can strain staff and the families Asia Hydrometeorology Modernization Project they support. As NMHSs around the world continue to struggle (CAHMP) and the World Meteorological Organization to hire, train, and retain high-quality staff, modernization proj- (WMO), Kyrgyzhydromet sent five promising students to Russia ects need to include long-term human resource development to study hydrometeorology. Having completed their studies strategies at the Russian State Hydrometeorological University in Saint Petersburg and the Moscow Hydrometeorological College, all returned to join Kyrgzhydromet as staff, with some receiving additional operational training under CAHMP in Germany, Russia, and Uzbekistan. One young expert, who at the age of 26 attained a PhD in atmospheric physics from the Russian State Hydrometeorological University, is now working in the forecasting department, integrating modern technologies and approaches, and accelerating the institution’s transformation into a modern hydrometeorological service. His skills and therefore Kyrgyzhydromet are increasingly recognized region- ally and globally for progressive expertise in forecasting; the expert has even been accepted as a hydrodynamic model devel- oper for the Weather Research and Forecasting (WRF) Model by the US National Center for Atmospheric Research (NCAR). As part of the agreement to sponsor his studies, the expert com- mitted to working for Kyrgyzhydromet following attainment of his degree. The expert is also personally committed to contrib- uting to the development of his country and the modernization of Kyrgyzhydromet, further serving on the Council for Science, Innovation and New Technologies under the prime minister of the Kyrgyz Republic. While his contributions are and will continue to be impressive and significant, their sustainability may however be challenged. Although personally dedicated to Kyrgyzhydromet specialist maintaining manual equipment under extreme supporting Kyrgyzhydromet, the limited employment packages climate conditions. Source: U. Torobekov, 2016. 60 Weathering the Change: How to Improve Hydromet Services in Developing Countries? BOX 4.1 Initial Steps to Match NMHS Modernization Needs with Opportunities Seeking government and NMHSs commitment. The commit- Identifying potential opportunities. The focus should be on ment to modernize NMHSs should include socioeconomic engagement with private sector service providers and under- assessments (WMO 2015a) and building partnerships with standing broader developments in the country (such as in tele- project beneficiaries (such as disaster risk management, agri- communications, social media, and information technology). culture, and water resource management). Motivating staff ownership. This is important (especially for Introducing the modernization agenda. This agenda should often disincentivized NMHS staff) for tasks such as developing appear in country assistance strategies and country partner- the concept of operations (CONOPS), which will determine the ship strategies, usually through disaster risk management, scope and direction of the modernization project, along with a climate adaptation, food security, water resource manage- sense of pride in achieving success. ment, and other significant sectoral programs and projects. Scoping modernization activities. This should be based on: Identifying funding sources. The focus should be on finan- (a) natural risks and vulnerability assessment (types of nat- cial sources for preparation (such as the Global Facility for ural hazards, frequency, and exposure); (b) weather depen- Disaster Relief and Recovery (GFDRR) and Climate Risk Early dence of economy and user needs assessment; (c) evaluation Warning Systems (CREWS)) and for implementation (such as of NMHSs status and high-priority modernization needs; the International Bank for Reconstruction and Development (d) cost-benefit analysis of the potential modernization sce- (IBRD), International Development Association (IDA), and the nario; and (e) government commitment to sustain proposed Green Climate Fund (GCF)). These development partners will modernization. have to coordinate to extend the lifetime of projects from the Ensuring a substantial preparatory phase. This would include, usual five years to a more realistic timeline of seven to eight besides the above scoping activities, developing a strategy/ years. GFDRR is proving to be instrumental in supporting roadmap and an initial CONOPS—and if time and funds allow, these initial stages with technical assistance and economic identifying key procurement packages (systems integrator, and sector work, using grant and trust fund resources. The integrated ICT, design, and build or DBO/T). GCF also supports modernization efforts where there is a clear and strong link to climate adaptation. Developing a business plan. This should be based on the social and economic benefits of the new services to attract Crafting a strategy. The focus should be on encouraging the government financing for the additional operations and main- NMHS management to formulate and articulate a clear role tenance costs. for the future of their services (10–20 years ahead) and a clear strategy on how to get there as part of the moderniza- tion plan—with input from major stakeholders and service beneficiaries. Institutional bolstering. This includes: (a) developing training and educational institutions. a legal and regulatory framework for the NMS, NHS, and Design and implementation support. This includes: (a) disaster management operations (such as assessing new providing a detailed systems design for the NMHS and other business models and enhancing public-private engage- agencies, based on the CONOPS (for example, observation ment); (b) twinning support between the NMHS and one or and monitoring networks, and forecast and service delivery more advanced NMHSs for operational forecast and service systems) and implementation support through a systems guidance; and (c) developing and operationalizing standard integrator; and (b) procuring and integrating systems. operating procedures. Engagement with the private sector. This includes explor- Capacity building. This includes: (a) conducting NMHS ing potential business relations with the private sector staff training, retraining, and professional development; (b) as part of a pilot project. In most instances, it would be supporting the professional orientation of the NMHS senior difficult to make firm partnership agreements for private management; (c) providing educational support for staff; (d) sector engagement with the public sector within the value offering training for key users and stakeholders (may include chain of meteorological and hydrological services without media and communities); and (e) strengthening national some no-regrets evaluation of the potential benefits. The Weathering the Change: How to Improve Hydromet Services in Developing Countries? 61 CREATING AN ENABLING ENVIRONMENT FOR HYDROMETEOROLOGICAL MODERNIZATION PROJECTS F rom 2010–2015, the Slovenian Environment Agency (ARSO) implemented the national modernization project Better Observation for Better Environmental Response (BOBER), financed 85 percent by the EU Cohesion Fund and 15 percent by the Slovenian government from the national budget. Costing EUR 33 million, BOBER upgraded the meteorological (including radar), avalanche, and hydrological monitoring networks; established flood forecasting, drought monitoring, water quality monitoring and marine monitoring and forecast- ing systems; upgraded ICT systems including a new high-per- formance computer; and strengthened calibration facilities. While the idea for the project started already in 2005, activi- Commissioning of a oceanographic monitoring buoy in the Gulf of Trieste, Adriatic Sea. Source: L. Ravnik, ARSO, 2014. ties to create an enabling environment were needed to get it off the ground. For example, during preparation of the project, and human resources for maintenance of the upgraded obser- the Slovenian government recognized its importance and vation system were needed. In addition, project implementa- included it in the Operational Programme for Environmental tion was performed almost entirely utilizing existing human and Transport Infrastructure Development 2007–2013 under resources. More than 50 ARSO employees were involved, some the priority guideline Reduction of damage caused by water. of them full time, and only a few additional staff were hired. This program was also adopted by the European Union in 2007. In the middle of the project in 2013, the government supported Further, ARSO discussed in detail the content and priorities of a proposed amendment to the National Regulation on the clas- project outputs with internal and external users, including the relevant civil protection and rescue agencies, national authori- sification of construction with regards to complexity, thereby ties for water and environmental management, and the private minimizing the required paperwork for construction at obser- sector, such as the hydropower industry. This led to a project vation sites and enabling ARSO to successfully complete the design aiming to deliver comprehensive solutions suitable for project according to plan. most users, thereby building support. Before and during BOBER, ARSO and its supporters in the Floods in September 2007 in northern and western Slovenia Slovenian government made sure that there was an enabling caused significant economic damage and the loss of several environment at both the policy and operational levels to ensure lives, particularly in the town of Železniki, which experienced the project could be successfully financed and implemented. flooding beyond the 100-year return period. Following this While the requirements and possibilities for pursuing supporting disaster, the government recognized BOBER as a very important factors will vary from context to context, without the relevant non-structural flood prevention measure and fully backed it, policy and financial support mechanisms in place, as well as further stressing that the upgraded information systems would plans for sustainable operation of modernized systems, projects also inform the design of better performing future structural are destined for failure. And like the 2007 floods in Slovenia, flood control measures. hydrometeorological disasters often provide windows of oppor- tunity to help establish such an enabling environment, at least in Recognizing the inherent challenges of maintaining a modern terms of political will. observation network, ARSO developed its own distributed real- time monitoring system. Utilizing inexpensive bulk-produced Acknowledgments electronic parts and Linux run by embedded computers, every sensor at every station can be accessed and controlled from the Gregor Sluga, Acting Director General and project leader, central office, with software upgrades also performed centrally. ARSO and Klemen Bergant, Director of Meteorology and This approach ensured that no significant increases in financial Hydrology Office, ARSO. 62 Weathering the Change: How to Improve Hydromet Services in Developing Countries? framework for such engagement would be developed as a ■■ Installing computer visualization systems. Such systems part of this component and implemented within Components should be installed for user-defined locations (like civil 2 and 3, if the right conditions exist. protection offices and NMHS regional offices). Each system should be tailored to the stakeholder’s specific Business planning. This includes developing a sound finan- requirements. cial plan for the NMHS’s future operation based on all of the ■■ Implementing early-warning system pilots. Such pilots forgoing activities. It could be part of the results framework are particularly important if the NMHS has little or as a business plan to support higher operations and main- no experience with developing and using warning tenance costs. systems. Greater use of mobile telephone–based applications. Component 2: Enhancement of the Service Although such applications are optional, it is becoming Delivery System increasingly important to exploit the advances in mobile The objective of Component 2 is to enhance the service deliv- technology in even the poorest and most vulnerable commu- ery system by creating or strengthening the public weather nities to reach the population during severe weather events. service—including developing new information products for vulnerable communities and the main weather-dependent Creating a national framework for climate services. This sectors of the economy. The main subcomponents are: framework is intended to transform the traditional climato- logical role of an NMHS and other climate service providers Introducing or strengthening the public weather service to a full user-oriented service and to increase opportuni- (PWS).1 This covers services related to disaster risk manage- ties for NMHSs to provide relevant climate information to ment, agriculture, the media, civil aviation, health, energy, government decision makers and the World Meteorological and water resources—which are essential for delivering the Organization. Tasks may include: (a) application of a digital benefits of the modernized program. It includes: library of all climate-relevant information from all sectors; ■■ Developing a service delivery strategy. The PWS will and (b) application of a national framework for climate ser- function as the principal interface between the vices (linked to the Global Framework for Climate Services) technical provider of products and the users. It should by engaging all climate-sensitive sectors. be responsible for developing and implementing standard operating procedures with authorities (such Component 3: Modernization of Observation and as civil protection), as part of disaster risk management Telecommunication Infrastructures, Modelling, and service quality management. and Forecasting Systems ■■ Developing a PWS platform to provide impact-based The objective of Component 3 is to: (a) modernize the obser- forecast and warning services. Such a platform provides vation network systems, communications system, and ICT forecasts and warnings of the weather’s impact, system of the NMHSs and possibly related organizations based on information available from the forecasters, (such as disaster management agencies), based on the combined with geographic information system (GIS) business models and CONOPS developed in Component 1; vulnerability and exposure data to provide risk (b) improve the meteorological and hydrological forecasting assessments. It requires forecaster workstations for system; and (c) refurbish offices and facilities. The main producing decision support information tailored to subcomponents are: each of the PWS sectors. These systems can be turnkey or built to suit. Technical modernization of observation networks. This ■■ Installing media equipment. Such media equipment includes: (a) rehabilitating and reequipping meteorological, enables the NMHS to create broadcast-quality bulletins, hydrological, and other networks as required;2 (b) introducing as well as being used for conducting interviews, and ground-based remote sensing systems for nowcasting and education and public outreach activities. very short-range weather forecasting (for example, radar 1 The PWS provides a model for the delivery of all services, including climate services. Its principal function is to focus on (a) translating and interpreting meteorological and hydrometeorological forecasts into impact-based forecasts and information; and (b) communicating this information to all sectors, including the public. 2 This task may also support aviation, agriculture, or other sector-specific networks, depending on the NMHS’s responsibility. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 63 DELIVERING LOCAL BENEFITS YIELDS PUBLIC SUPPORT D uring the partial modernization of Kyrgyzhydromet’s Following rehabilitation of the station, Kyrgyzhydromet was observation network under the World Bank’s Central able to deliver higher quality and more timely river stage Asia Hydrometeorology Modernization Project (CAHMP), and flow information and forecasts to the community. Local a hydrological post on the Djety-Oguz River (among many residents now appreciate the services that the hydrological others) was rehabilitated in 2016. At first, however, there was post helps deliver, and actively support its operation and local opposition to this investment. Having been dysfunctional maintenance by ensuring it is not vandalized. It is clear that for many years, neighboring residents were resentful that improving service delivery incentivizes support and resources Kyrgyzhydromet had for years kept prime riverfront land with- far more effectively than simply asking for support. out delivering benefits to the community. weather surveillance equipment, wind profilers, and light- ■■ Developing archiving, database management, and ning detection networks);3 (c) upper-air measurements using digitizing capabilities. radiosondes, although temperature soundings may also be ■■ Providing computers and software to access and use available from commercial aircraft as a part of the AMDAR numerical weather predictions (NWPs) and climate (Aircraft Meteorological Data Relay) system; and (d) strength- models.4 ening quality control by setting up calibration facilities. ■■ Consider exploring the potential role of public-private Access to GIS data. This includes: (a) digital elevation partnerships or other alternatives to own and operate data (like for flood modelling); (b) demographic data for within the NMHS, if the capacity to develop these skills internally is very weak. vulnerability mapping; and (c) exposure data for critical infrastructure. Improvement of the hydrometeorological forecasting system. This includes: (a) introducing modern computer Modernization of the NMHS’s communication and ICT equipment for processing observational data from in situ systems. This includes: surface networks, satellites, upper-air stations, radar, and ■■ Consider creating a data center, which would support other in situ remote sensing systems (for example, radar, all information needs and simplify ICT procurement. wind profilers, and lightning networks); and (b) forecaster ■■ Increasing access to sufficient Internet bandwidth workstations to integrate observations and numerical to access global and regional data and products. weather prediction guidance to prepare forecasts and warn- This should be addressed early on, preferably during ings. A significant amount of effort should be spent on NWP preparation, using grant funds to enable NMHS access model verification, calibration, and post-processing, as well to the country’s telecommunication infrastructure. as on how to effectively use ensemble prediction systems Experience suggests that in most countries broadband (like probabilistic forecasting techniques). communication exists, but it is not always exploited effectively by the public-sector services. Refurbishment of NMHS’s offices and facilities. This includes updating office spaces and, in some instances, ■■ Introducing new communication equipment that meets new buildings to accommodate new ICT systems and pro- WMO Information System (WIS) standards. vide a modern working environment for operational staff. 3 Generally, all these ground-based remote sensing technologies are complex. If the NMHS staff has little or no experience with these technologies, they should be introduced with sufficient long-term technical support. It is also important to weigh alternative technologies that can partly fill the role of these in situ systems (such as satellite-based imagery that can be blended with rapid refresh systems for nowcasting), which are freely available to WMO members. Here outsourcing, or data as a service, may be a viable alternative to owning and operating the system entirely within the NMHS. 4 The CONOPS should, among other things, clearly determine the type of NWP support required. It should be based on a cascading forecasting process with reliance on the global mesoscale NWP and climate models’ data and products, followed by regional products; and finally, model verification, calibration, and post-processing at the national level. Only if clearly justified, national investment in dynamic downscaling of NWP and climate models should be considered. The latter is a last step, not the first one. 64 Weathering the Change: How to Improve Hydromet Services in Developing Countries? Experience shows that facilities that have deteriorated and International Advisors. The “soft” functions—such as are in a poor state of repair contribute to poor staff working strategic planning, assessment of business models, devel- conditions and motivation—besides being unable to house opment of the CONOPS, design of training programs, and modern equipment. stakeholder workshops—can be done through separate contracts to individual highly qualified international Component 4: Project Management experts that have the client’s trust and can take the time The objective of Component 4 is to support the NMHS’ tech- to advise the client on transforming services. Critical to nical team to fully engage on project management, devel- the success of this approach is (a) the availability of a opment, oversight, and implementation—with assistance pool of qualified experts with experience in integrated from national and international advisors and a systems inte- hydrometeorological modernization in developing coun- grator. Another critical element is a highly qualified project ties; (b) the experts’ level of expertise; (c) the amount of management unit, adequately staffed with procurement and time they can devote to the client; (d) excellent commu- other specialists knowledgeable about the complexity of nication skills; and (e) support from the NMHS manage- NMHS modernization projects. ment and PMU. There should also be enough flexibility for them to call on other experts if new needs arise. Project Managers. An essential element for success of a modernization program is a robust project management The drawback is that this approach increases the and implementation support team, which should include administrative workload and requires a qualified team an NMHS project management team—comprising a to coordinate input and monitor performance of multi- Project Management Unit (PMU), a specialized technical ple international advisors (IAs). Thus, there is still the support team, and a qualified World Bank supervision advantage of enabling projects to start up quickly and and implementation support team. make sufficient progress prior to contracting the system integrator. In practice, the service delivery strategy and Systems Integrator. This is a firm, typically a consor- the CONOPS would be developed through workshops tium of key technical providers, including experts from with stakeholders and the client’s staff (with the inter- advanced meteorological and hydrological services. national advisors’ help), and the system integrator would Their tasks include designing the new system, develop- develop the design and implementation plan to achieve ing implementation plans, supporting procurement, and the CONOPS. Another approach would be to start hiring making sure the procured systems work together to sup- the system integrator earlier and include the interna- port the optimal operation of the NMHS. When it was first tional advisors at the project preparation phase, using envisaged, the system integrator was also responsible financing available for technical assistance. for developing a training program, organizing workshops with stakeholders to develop the initial strategic and operational concepts, developing new tools, and provid- How to Make Modernization ing ongoing operational guidance. Programs Sustainable But the contracts for large-scale modernization are It is preferable to develop modernization projects in two typically multi-million dollars and take much longer to phases: (a) design, which focuses on service delivery stra- procure than is desirable for projects that typically last tegic planning, business modelling, and the CONOPS; and five years (the system integrator is normally hired at the (b) implementation, which focuses on acquiring assets and implementation stage of the project, although selection introducing new operational capabilities and services. can be done earlier). In some instances, because high- level approvals within government are required, the sys- The design phase can be developed through technical tem integrator hiring process can take up to 18 months. assistance projects and/or the hiring of individual advisors. This invariably means that the project’s start-up is quite It typically lasts at least 12 months, which ideally should slow and some of the tasks (like procurement) are started be initiated in parallel with the project preparation and without system integrator support, which may lead to funded by grant resources. Bank projects are usually five substandard technical design of the systems. years, but increasingly this is recognized as too short for Weathering the Change: How to Improve Hydromet Services in Developing Countries? 65 complex projects that involve strengthening weak institu- ■■ Building government (ministries of finance, economy, and tions. Obviously, the overall budget for the project needs planning) understanding of the NMHS’s importance through to be established and robust estimates of costs made. business planning, in hopes of a legally binding commitment However, this does not preclude a range of solutions, fixed in credit or a grant agreement to increase budget sup- including engagement with the private sector, which by its port and allocations for O&M costs. nature should be in the form of a partnership and thus a ■■ Developing a CONOPS and project design that is more likely joint financial venture. In 2018, the WB and WMO signed to be affordable and implementable by involving NMHS staff an agreement that permits governments to use WB financing in the process. to obtain WMO technical assistance5—which is particularly useful for basic training, setting up twinning arrangements, ■■ Introducing a quality management system in the entire and accessing WMO regional technical centers. NMHS and developing standard operating systems to pro- vide a uniform structure and process across the whole or- Operations and Maintenance (O&M). One of the key factors ganization. in defining the project’s affordability is assessing the O&M ■■ Preparing national strategic plans. costs of the future system, which should be done as part of preparing the CONOPS. In developing countries with very ■■ Mounting training programs on service provision. limited funding for NMHSs, any modernization effort will ■■ Organizing media workshops. generally increase the annual O&M costs. Thus, the govern- ■■ Building capacity and retaining qualified staff members, in- ment must agree up front to the incremental cost of running cluding developing additional incentives. the NMHS, or the project will be difficult, if not impossible, to sustain beyond the lifetime of the project investment. ■■ Testing new business models to strengthen sustainability. Experience suggests that the incremental O&M costs are ■■ Building partnerships with national and international stake- equivalent to about 10 to 15 percent of the investment holders, such as twinning arrangements for more advanced budget. Many donor-driven projects, which have focused on and developing NMHSs (see Technical Insight 4.1). providing observation equipment, have failed because the ■■ Setting up user and policy committees. NMS or NHS did not have the staff or resources to maintain the new observation capability. ■■ Building the membership of WMO technical commissions. ■■ Creating a modernization leadership team composed of Staffing. Most developed countries’ modernization efforts NMHS management, the project management unit, and a have also included costs for staff reductions because employ- special modernization team (including international advi- ees are often the largest operational expense—which raises sors and system integrators). the possibility of introducing automation as a cost-reducing strategy. However, this strategy is ineffective in most LMICs, ■■ Carefully assessing absorption capacity and ensuring that where staffing costs are relatively low. In those countries, the investment is scaled accordingly. relatively few qualified technical personnel are available, ■■ Ensuring close supervision, especially during the initial and more are usually needed to maintain modern observa- phase of project implementation; also often helpful is mobi- tion, telecommunication, and forecasting systems. lizing additional Bank-executed trust funds. Mitigating Risks. NMHS modernization projects are usu- Procurement issues. Procurement is complicated, owing to ally high-risk, high-reward efforts because they require the need to buy and install numerous and relatively small significant cultural changes within institutions—and intro- packages of various specialized interrelated hardware and duce technically complex information and communication software products—often high-performance computers and technologies—but have the potential to provide substantial communication equipment—and to arrange delivery and benefits to weather- and climate-sensitive sectors of society. installation of equipment in multiple and often remote loca- The risks of problems can be mitigated by a set of interlinked tions. A detailed design of networks and systems is expensive activities, including the following: and time consuming, and it is often unavailable at the time of appraisal. Consequently, many projects have shifted the Standard Form of Agreement for use by World Bank Borrowers. Provision of Technical Assistance by WMO under Bank-Financed Projects V.1 (October 12, 2018). 5 66 Weathering the Change: How to Improve Hydromet Services in Developing Countries? WORLD BANK HYDROMET INVESTMENTS RESULTS FRAMEWORK AND INDICATORS –RECURRENT CHALLENGES F rom 2005 to the present, the Bank has undertaken ■■ When user satisfaction is used as an indicator, the results hydromet modernization investments in about 30 coun- framework often promises annual updates, but in many tries. Their success was measured according to results cases only one survey is undertaken during implementation frameworks developed individually for each project. The expe- – the baseline survey that tends to occur somewhere around rience that has accumulated for these past fifteen years allows the mid-term review (MTR). While a survey at the project end for analysis of some recurrent issues regarding how results is planned, it often does not occur. frameworks are setup, as well as to identify approaches that ■■ Measures of the technical quality of forecasting (accuracy, allows for more effective use of results frameworks as tools for skill and timeliness) are commonly used as indicators, but project management. are often not described precisely, which leads to differences Analysis revealed that altogether around 150 different indica- in interpretation and monitoring approaches. For such tors were used in the results frameworks of hydromet projects. indicators to be useful, the NMHS needs to have a reliable They can be roughly grouped by objective/outcome as follows: verification system in place, and the parameter and lead- (i) Institutional strengthening; (ii) Strengthening weather fore- time of the forecast to be monitored needs to be clearly casting: and (iii) Tailoring user services. defined. ■■ In some cases, when accuracy of the forecast is used as By assessing project implementation reports including prog- an indicator, it rises and then falls and then rises again, ress on indicators and achieving targets, the following recur- suggesting that the chosen indicator does well or poorly by rent issues were identified: chance and does not measure underlying capacity growth. ■■ Most indicators show no change from the baseline until It must be recognized that depending on the instability mid-year 3 or year 4 of project implementation. Progress is and/or complexity of the weather occurring in a particular then achieved in a rush in the last year or two of the project reporting period, forecast accuracy can also display some lifecycle. This pattern is seen as a weakness in the design inconsistency; changes in forecast accuracy and skill should of the results framework as project indicators are intended therefore be viewed as trends rather than individual points. to support project monitoring, and to guide task teams to ■■ For the projects that involve system integrators, indicators identify implementation weaknesses. are worded so that progress does not formally exist until the ■■ Baselines are rarely prepared during project preparation. system integrator has devised a plan that is approved so that Sometimes baselines are not set in place for several years the progress can be assured to be aligned with that plan. during project implementation. Often, the lack of the crucial ■■ Some indicators simply do not measure what they say baseline is represented by a “zero”, which leads to incorrect they do. I.e., reliability of climate projections may not reflections of project progress (i.e. when the baseline is be measured by their downscaling. The existence of an eventually prepared, and it is some number greater than authoritative public weather service may not be measured “zero”, it appears significant results have been delivered, by a user satisfaction survey. which may or may not be the case.) Weathering the Change: How to Improve Hydromet Services in Developing Countries? 67 detailed design of networks to the project implementation Setting project performance indicators. Realistically, it phase and included them as the first phase of the integrator is hard to determine indicators for the results effective assignment. The integrator assignment also includes devel- framework, given that hydromet modernization projects are oping most procurement packages (technical specifications technically complex and their results are broad and multi- and bid documents) and implementing support to NMSs, sectoral. That said, we believe that the indicators need to NHSs, and project implementation units. Hiring an integra- combine quantitative measures related to forecast accuracy tor is a high priority and selecting the consultant through and qualitative measures related to services and benefits, quality- and cost-based selection (QCBS) should be initiated which adhere closely to the service delivery progress model well in advance, given that the project’s implementation (Annex 1), observation, telecommunication and forecasting will strongly depend on the consultant’s success. Thus, it is progress model (Annex 2), and, and the climate services important to develop a well thought through CONOPS before progress model (Annex 3). Measuring performance during entering the design and implementation phase. the execution of a project is essential to ensuring that a project achieves its objectives. Metrics must be realistic, Turnkey systems. An alternative approach involves using a measurable, and useful—enabling project managers to make turnkey contract for modernization, under which most tasks corrections during project implementation. Ideally the would be outsourced to a consulting firm to design-build-op- project metrics should align with or be identical to the key erate or design-build-operate-transfer. But such contracts in performance metrics of the organization. the hydromet sector have never been arranged in the WB, and rarely by other development partners. Particular attention Public facing entities, such as NMHSs, should focus on satis- has to be paid in the design-build to include legacy systems fying their users’ needs. Any service has two components— and those supported by other development partners. Thus, one focused on user satisfaction and another on technical it will take some time and significant efforts to properly aspects. For example, a warning service is most useful if it structure, test, and define safe areas of application for such is in time, at the right location and intensity. Figure 4.1. innovative contracting. There are an increasing number of through 4.8 show the scoring used by the UK Met Office firms and joint ventures with the ability to provide turnkey derived from first responders to a situation as well as from solutions. the public affected. TECHNICAL INSIGHT 4.1 Types of Twinning Arrangements It is difficult to realize the full benefit of a modernization hosted and operated by the South African Weather Service. program within an institution that does not have a full Arrangements for pairing NMHSs are best managed through complement of staff or that lacks training and experience in the WMO, which has the capacity to monitor these activities modern forecasting techniques. National meteorological and and, where necessary, can facilitate financial support. Ideally, hydrological services (NMHSs), however, do not operate in the WMO regional centers would play a leading role in this isolation and can benefit from the capacity and capabilities of process, because most (but not all) are located within the more more advanced services. advanced NMHSs. The pairing arrangement would provide Most WMO regions contain both advanced and developing operational backup for forecasters, enabling them to receive NMHSs. More advantage could be taken of the capacity of the guidance on complex weather situations. This guidance could most advanced NMHSs to provide operational support to the be in the form of bulletins or, in the case of life-threatening weakest. This assistance is provided in some regions, but it situations, through direct video or voice contact—much the is not universal, and it cannot be readily sustained without way that a central forecasting office works with regional financial support. offices within large countries. A useful example is the Cascading Forecasting Process imple- Frequent training is essential to increase and maintain the mented in southern Africa. It provides global and regional skills of an NMHS’s staff. Such training should be viewed as a NWP guidance to 16 countries through the WMO Regional continuous process in which all staff members are involved in Specialized Meteorological Centre (RSMC) Pretoria, which is a long-term program to improve their skills. 68 Weathering the Change: How to Improve Hydromet Services in Developing Countries? FIGURE 4.1 FIGURE 4.3 Range of Satisfactory Survey questions Purpose Purpose Parameter scores score Did you see or hear This helps to assess the effective- Measure the Impact level 0–3 More than 72% anything about this severe ness of the communication systems accuracy of Impact timing 0–3 of warnings weather warning? used to deliver the warning. public warnings score > 6 Impact location 0–3 Generally speaking how This would assess people’s accurate do you think most perceptions of the accuracy of the weather/flood forecasts information received. A low score, The “satisfactory score” can be used to measure year-on-year and warnings are? despite significant quantitative progress, increasing from an agreed baseline to something improvement in the forecasts, would in the range of 72 percent. The relatively large uncertainty indicate the need to improve public reflects the nature of predicting risk. awareness. Overall how useful would This would assess whether people The aim is to target only those at risk, rather than the entire you say the weather/flood are able to understand the informa- population. Making warnings available to everyone results forecasts and warnings are tion. these days? in over-warning, which should be avoided. The desired outcome is to persuade people at risk of a high impact of an event to take action on the relatively low probability The survey questions and the measurement of the effec- of occurrence. The purpose of impact-based forecasting is tiveness of warnings can be disaggregated by gender and the ability to forecast the probability of an event happen- other demographic indicators. They can also be turned into ing, based on a forecast of a hazard and the vulnerability a satisfaction index. Again, it would be reasonable to expect of those at risk to the impact of that hazard. The public is more than 75 percent of the population affected being sat- encouraged to understand their exposure to risk and to isfied with the services provided by the end of the project. act accordingly. However, the response depends on peo- An initial baseline survey would be indicative of the gap ple’s appetite for risk; it is therefore impossible to expect between the current and proposed services. that everyone deemed at risk will take appropriate action (Figure 4.2). From the perspective of disaster management, Frequent false positive (false alarms) and false negative the more people taking appropriate action in response to a (missed) warnings are the primary cause of lack of compli- warning is a useful metric. ance to subsequent warnings. Consequently, an important aspect of a forecast improvement project is the reduction on both false positive and false negative warnings and the FIGURE 4.2 use of probabilistic forecasts. It has been demonstrated that Satisfactory the use of probability estimates increases compliance, trust, Purpose Indicator score and decision quality (LeClerc and Joslyn 2015) and is more Measure the effec- Number of people who take important than reducing false alarms, because the cost of tiveness of public action as a percentage of the a bad decision is usually much greater than that of a false >75% warnings population at risk receiving a warning alarm. Hence the focus is on the use and improvement of impact-based forecasts and warnings. The introduction of Measurement of the public’s satisfaction with forecasts and an impact-based forecast and warning system is a binary warnings is also an approach used by many NMHSs to assess measure of project improvement (Figure 4.4). year-on-year changes in the quality of their services. Since not all people will take action, it is useful to measure peo- FIGURE 4.4 ple’s perception of a severe weather warning by sampling Activity Purpose a statistically significant number of those expected to have received it. Questionnaires have been developed by some Operational use of an Probabilistic forecasts and impact-based forecast and warnings, which reflect the likely NMHSs and can be adapted to most situations. Questions warning system impact of a severe weather event, might include those shown in Figure 4.3. are issued. Aligned with CONOPS Weathering the Change: How to Improve Hydromet Services in Developing Countries? 69 Subsequent improvement in the system depends on tech- Activity Purpose Indicator nical improvements in meteorological and hydrological Data shared with Overall forecast Percentage of national forecasts and access to vulnerability and other data, which the international improvement will data shared via WIS/ can be measured independently. community occur if data are GTS with the global shared with global modelling centers. From the perspective of the forecasting process, timely modelling centers Ideally, this number access to high-quality observations from various sources— in a timely manner. should include all Withholding should basic meteorological international and national—is important. International be considered and hydrological observations are available via the GTS or WIS, and not adversely parameters, but not much can be done to improve on the observations except to necessarily very high- volume data. ensure high reliability of the communication links. National observing networks are most useful if most of these data are shared within the global meteorological and hydrological In addition to the external measures of the forecast and community since the local application of a global model warning systems, the basic forecasting processes, can be is improved by any data assimilated from that location. assessed as in Figure 4.6. Metrics for national observations may include any or all of those shown in Figure 4.5. FIGURE 4.6 Activity Purpose Indicator FIGURE 4.5 Forecast of basic Demonstrate year- At a set of locations for meteorological on-year improve- which measurements Activity Purpose Indicator and hydrological ment in the basic are available: Number of A measure of the Number, aligned with parameters forecasting system, meteorological procurement and CONOPS Daily maximum applying standard and hydrological deployment of temperature methods used by stations installed equipment such as the internation- Daily minimum and operating automated weather al community. temperature to international stations, river Regional variation standards gauging stations, will indicate where Daily accumulated or standard upper-air stations, additional effort is rainfall operating and radar needed to improve Daily water levels procedures forecasts, e.g., urban areas, extra Each compared with Amount of data A measure of how Gbytes of data urban, mountain- the forecasts at the available to the well the individual available to the ous terrain, coastal same locations for days forecasters via a observations are forecasters expressed areas 1, 2, and 3 central database integrated into as a percentage of the overall data the total amount of to generate a matrix of management system data that should be skill scores (0–1) and and how well each available from all of monthly composite of the individual the national networks. scores for temperature, sensors operate The amount should be rainfall, and river levels well over 90%. Numerical Evaluate the The same as above, but Utilization of This measures Percentage of new weather quality of local nu- comparing raw model the observations how much of the data used in the prediction merical predictions products with these in the analysis available data is forecasting process (without forecaster data and with model and forecasting actually useful in intervention). The data from other centers process the forecasting comparison would for the same locations process. It reflects be against local for days 1, 2, and 3 on the quality of observations and the data as well as NWP from other the maturity of the centers. forecasting system 70 Weathering the Change: How to Improve Hydromet Services in Developing Countries? These are relatively straightforward verification methods Activity Purpose Indicator and appropriate for performance measures that can be Leadership Provide all managers with Number of shared with the public. One limitation is the need for about courses for the tools to manage their managers. three years of meteorological and hydrological data to create all senior staff through a change The goal is to managers process. The aim is to educate and the baseline. Since this may not be available, these metrics minimize resistance to train all. should not be given too much prominence in the early stages change and encourage the of the project but can become key performance metrics for full participation of all staff the organization in the latter part and beyond. More complex Align Ensure all departments are An organogram verification is appropriate for forecast departments to gauge departments fit for purpose within the aligned with with new future organization strategy and improvements in different aspects of the system. functions CONOPS A critical aspect of modernization is the introduction of new Technical Provide staff with skills Number of staff, ICT capabilities, which are required to be fully operational training required to carry out new aligned with functions in all departments CONOPS during the first part of the project to support other compo- nents. It is useful for project implementation, therefore, to monitor progress in the design, procurement, installation Internal development of, or external access to, production and operation of the ICT system (Figure 4.7). systems. The overall expectation should be that the modern- ized NMHS can provide a full level of services and that these FIGURE 4.7 are sufficient to meet the population’s needs. In the past we Activity Purpose Indicator would have expected that the NMHS would develop its own Installation and Measure progress in Number of production systems to attain a sufficient level of capability to operation of design, procurement, subcomponents be able to deliver this high level of public services, based on ICT system installation, and operation completed and building and operating its own systems. However, it is more of the ICT system. Identify aligned with realistic to support the notion that the NMHS should have problems that could slow CONOPS the implementation of access to sufficiently high-level production systems—meaning the overall modernization that they could achieve this level by accessing the required program and may require systems through twinning with other NMHSs, engaging with revision of the CONOPS the private sector, or some combination of internal develop- ment and external opportunities. Thus, they should either have, or have access to, production systems that underpin the Modernization brings with it the need for significant orga- required public services. Either way, the results framework is nizational change—new working methods, new skills, and not altered since the NMHS is ultimately responsible for the some restructuring of departments to align with the intro- outputs of the production system, which are measured as duction of new services. Change management is often over- performance indicators. looked—a measurable facet of the modernization process. It is suggested to consider the activities shown in Figure 4.8. Safeguards. Environmental and social risks of NMHS modernization projects are typically small. Moreover, the FIGURE 4.8 hydromet projects usually provide environmental benefits, given that they support lowering risks associated with Activity Purpose Indicator floods, drought and fire, winds, extreme weather events, Number of staff Provide all staff with the Number of staff. and even industrial accidents. Project activities are usually (at all levels) tools to cope with the The goal is to implemented within available hydromet sites and involve a participating necessary adjustments to educate and in change meet the objectives of the train all. minor installation of observation equipment with no or min- management future organization imal environmental disturbance. Specific attention should seminars and be given to the safe handling and disposal of equipment courses containing mercury (such as thermometers). Procuring larger equipment (such as Doppler radar) or constructing Weathering the Change: How to Improve Hydromet Services in Developing Countries? 71 MULTIFACETED CAPACITY SUPPORT TO ADOPT NEW TECHNOLOGIES Training of radar specialists and forecasters of the State Hydrometeorological Service of Moldova. Source: R. Keene, Enterprise Electronics Corporation (EEC), 2018. T he Moldova State Hydrometeorological Services (SHS) ■■ Development of standard operating procedures on effective received significant investments under the World Bank operation and use of the radar; Disaster and Climate Risk Management Project (DCRMP), ■■ Optimal utilization of radar products by forecasters; including the procurement and installation of a modern ■■ Development of a methodology for estimating the budget Doppler weather radar at the Chisinau International Airport needs for delivering radar products; and in 2014. The radar is jointly operated under a Memorandum ■■ Integration of the SHS radar data/products into the common of Understanding (MOU) between the SHS and Moldovan Air radar system of the NMA. Traffic Services Authority (MoldATSA). This MOU provides, inter alia, that while the SHS owns the radar, MoldATSA is Romania and Moldova share a common language, and the NMA tasked with its operation and contribution to maintenance. experts were already supporting SHS to strengthen their fore- This approach was pursued to help support the sustainability casting verification and Quality Management System (QMS), so of radar operations, where both agencies benefit from the real- this twinning approach was welcomed by both sides. time data it produces. The experience clearly illustrates the need to ensure that Unfortunately, by 2017 the radar had stopped functioning prop- NMHSs adopting new technologies (this is the first and only erly, the warranty had expired (due to financial constraints, a weather radar in Moldova) have sufficient capacity to fully very limited warranty had been purchased), and the Moldovan leverage and sustain them. While the original DCRMP project authorities were not able to fix the radar on their own. Despite should have better prepared the SHS to receive the radar, the the MOU, SHS and MoldATSA could also not agree on the best follow-on technical assistance project was able to rectify any course of action. It was therefore decided to contract the radar shortcomings. manufacturer to fix the radar under the World Bank/GFDRR In addition, financed by UNDP and GFDRR, the Austrian Reinforcing Weather and Climate Services in Moldova technical Central Institute for Meteorology and Geodynamics (ZAMG) assistance project. During their work, the manufacturer also helped SHS qualify for and join certain operational programs identified several shortcomings in how the radar was being of the European Meteorological Services Network (EUMETNET) maintained and operated. including MeteoAlarm, aiming to enhance the quality of mete- To further support the full leveraging and sustainability of the orological information through the existing European meteo- radar, under the same project, the radar manufacturer was rological infrastructure. This included integration and use of further contracted to provide significant technical training on radar products. Strong partnerships that provide complemen- radar operations and maintenance. In addition, experts from tary technical assistance and support should be leveraged to the Romania National Meteorological Administration (NMA) strengthen the sustainability of resilience investments. were tasked to provide coaching to SHS on the: 72 Weathering the Change: How to Improve Hydromet Services in Developing Countries? buildings will require a more detailed evaluation of the pub- with the WB playing an implementation support role, along lic health and environmental safety aspects, which should with ensuring compliance with agreed provisions of grants be carefully documented in the environmental and social or credits. However, many other partners do not actively management framework. A potential staff reduction during encourage the client to participate in the implementation or a large-scale replacement of manual observations by auto- supervision of a project. Consequently, there is a tendency matic instruments should be carefully evaluated within the to hand-over a finished product to the client, who may have framework’s scope and will require extensive consultations little incentive to continue its operation. Conversely, the with key stakeholders. That said, so far no hydromet mod- NMHSs, which have little experience in managing their own ernization projects in developing countries have triggered projects, expect others to deliver the solutions with little any incremental staff reductions. engagement on their part. An attitude of “we pay them, so they should do the all of the work”, misses the point that the Coordination with development partners. Nowadays, there NMHS is the beneficiary and thus should retain the respon- is a crowded and competitive space among development sibility for delivering the investment’s benefits. partners—underscoring the need for coordinating activities to ensure that the NMHSs really benefit and reduce the risk of International initiatives. A key one is the Regional overlapping and potentially conflicted interests. That is why Framework Program to Improve Hydrometeorological WB/GFDRR and WMO have encouraged cooperation among Services in Sub-Saharan Africa (Africa Hydromet)—a joint partners through engagement at both the international and effort of the African Development Bank (AFDB), WMO, national levels. Progress has been made through the First United Nations Development Programme (UNDP), World and Second Global Development Partners Conferences6,7, the Food Programme (WFP), French Development Agency (AFD), emerging Alliance for Hydromet Development8, and national WBG, and GFDRR. It aims to address the deficiencies high- NMHSs’ efforts to create more complementary projects. In lighted above by supporting the improvement of hydromet recent years, several international development partners services on the national, regional, and continental levels. At conferences have been organized by WMO, WB/GFDRR, the national level, the program seeks to modernize or build and other partners. And it is becoming common practice to infrastructure and strengthen service delivery. Sub-regional encourage clients to organize national events, especially efforts include standardizing procedures to promote when a large WB supported investment is being proposed trans-boundary collaboration, while Africa-wide efforts or is under way. will ensure that hydromet services across the continent are linked to regional and global centers, thereby improving One example of the effective coordination of development data availability and sharing and promoting partnerships. partners’ support to an NMHS is Myanmar’s Department of Meteorology and Hydrology (DMH). At the national level, led The Climate Risk and Early-Warning Systems (CREWS) by DMH with support from WMO, the focus is on technical Initiative is another example of development partners issues and firming up commitments from partners9 to carry collaborating toward building stronger people-centered, out specific activities and to identify any financial and tech- multi-hazard, early warning systems. It was launched in nical gaps. Priorities include introducing flash flood guid- 2015 as part of the COP21 Solutions Agenda. It aims to raise ance, sharing radar data, and reconciling existing activities US$100 million by 2020. The WBG serves as the trustee (it with new initiatives. is a Financial Intermediary Fund (FIF)), and there are three Implementing Partners—the World Bank/Global Facility But the different approaches of development partners can be for Disaster Reduction and Recovery (GFDRR), the World confusing for clients. For the WB, there is an expectation that Meteorological Organization (WMO), and the UN Office for the client would take full responsibility for the investment, Disaster Risk Reduction (UNISDR). 6 https://www.gfdrr.org/en/publication/meeting-summary-development-partners-roundtable 7 https://www.gfdrr.org/en/event/second-hydromet-development-partners-conference-beyond-business-usual-closing-capacity-gap 8 https://public.wmo.int/en/media/news/world-bank-and-wmo-announce-hydromet-development-alliance-commitment 9 The partners include WMO; WB; Japan Meteorological Agency (JMA); Japan International Cooperation Agency (JICA); Korean Meteorological Administration and its partners; China Meteorological Administration; USAID; Met Norway; Met Office; Asian Disaster Preparation Centre (ADPC); the Regional Integrated Multi-Hazard Early Warning System for Africa and Asia (RIMES); Bureau of Meteorology (BoM), Australia; Hong Kong Observatory (HKO); and Indian Meteorological Department (IMD), among numerous other international and national actors. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 73 Small-scale investments and technical assistance proj- ects. Despite the recommendation that modernization Conclusion projects should be transformative, large-scale investments When it comes to developing a modernization project, it may not be practical. In this case, the emphasis should be is essential to hire a skilled team to assist the client in the on improving the delivery of services by increasing the effi- initial planning stages—which include strategic planning, ciency and efficacy of the existing systems. business modelling, and a preliminary operational concept (CONOPS). Moreover, implementing these activities early In Armenia, for example, the first step was to develop a has the added advantage of building knowledge and trust roadmap with three scenarios of increasing complexity and with the client. It is important that the project is developed cost—offering a phased approach based on the availability of by them and not for them, although care must be exercised resources10 (Kootval et al. 2018). In the low-cost, technical to avoid infrastructure-heavy projects. In this way, owner- assistance scenario, grant funds of about US$350,000 are ship is achieved at the outset. This is important, given that proposed to: (a) improve short-range weather forecasting by most development projects are implemented by the devel- learning to utilize fully NWP products from GPCs; (b) improve opment partner, often with only perfunctory engagement of monthly and seasonal forecasts through training to apply the client. While expedient, it does not impart any skills in regional climate downscaling methods; and (c) improve project management to the client. hydrological forecasting by training staff on data manage- ment, use of statistics, hydrological modelling, and the visu- In addition, engaging with other partners is critical, and alization of geospatial data sets—and purchasing low-cost, projects need to be sufficiently flexible so that they can high-priority communication and computer equipment and adapt to the changing circumstances, which will include software. The expectation is that this limited intervention new donor-assisted activities. The CONOPS is designed to should produce better weather and hydrological forecasts, take these factors into account. By maintaining its currency, improve access to forecasts by the public and government it provides the NMHS with a tool to assess the impact of any stakeholders, and better support the early warning system. changes that will affect the NMHS’ operations. http://documents.worldbank.org/curated/en/684751548347371395/Modernizing-Weather-Climate-and-Hydrological-Services-A-Road-Map-for-Armenia 10 74 Weathering the Change: How to Improve Hydromet Services in Developing Countries? ANNEX 1 Progress Model for Service Delivery The Service Delivery Progress Model is adapted from WMO No. 1129—The WMO Strategy for Service Delivery and its Implementation Plan. The model can be used as a tool for assessing the level of development of NMHSs and creating an action plan to improve service delivery. Full details can be found at http://www.wmo.int/pages/prog/amp/pwsp/documents/WMO-SSD-1129_en.pdf. STRATEGY ELEMENT 1 Evaluate user needs and decisions Undeveloped Development initiated Development in progress Developed Advanced The users and their Users are known, but Users are able to contact NMHSs seek input on an An ongoing dialogue is requirements for products no process for user NMHSs and their feedback ad hoc basis from users to maintained with users or services are not known. engagement exists. is recorded. facilitate the development regarding their needs and User requirements for There are some formal of services. the services they receive. service delivery are not processes for integrating Requirements are defined Requirements are defined well defined. the feedback received in documents agreed upon in documents agreed upon into the development of with the customer but are with the customer and services. not routinely updated. routinely updated using User requirements are feedback from users. defined with limited documentation. STRATEGY ELEMENT 2 Link service development and delivery to user needs Undeveloped Development initiated Development in progress Developed Advanced No concept of service Services do not adapt to Services are developed User feedback is used Users are consulted to exists; products are simply changing user needs and and changed as to inform management facilitate development of issued. new technology. technology allows, but of changes and products and services. Products are documented engagement with users is developments to services. The service defined in the with limited descriptive ad hoc. Products and services are SLA is agreed upon with information. Products and services consistently documented. the customer based on are documented, and the SLAs are defined. user consultation. information is used to inform management of changes. STRATEGY ELEMENT 3 Evaluate and monitor service performance and outcomes Undeveloped Development initiated Development in progress Developed Advanced No measures are in place Some measures of Measures of verification User requirements Measures of performance for assessing performance, development are in place. and service delivery are in are used as data for are based on user needs, either in terms of accuracy The verification of place but are not informed performance measures. which are regularly or service delivery. accuracy and/or service by user requirements. Findings are used reported and consistently delivery takes place, to identify areas for used to inform decisions but no systematic improvement. on improvements. process exists to use this Subsequent actions are information to improve taken in an ad hoc manner. the service. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 75 STRATEGY ELEMENT 4 Sustain improved service delivery Undeveloped Development initiated Development in progress Developed Advanced No concept exists of The concept of service An action plan has been The action plan is being The status of service service delivery principles. delivery has been created to improve the implemented to improve delivery is reviewed on a introduced, and an current level of service service delivery; the regular basis. assessment of current delivery, and resources outcomes are being The action plan evolves in status has been have been identified to monitored. response to the outcome undertaken. implement it. of the reviews. STRATEGY ELEMENT 5 Develop skills needed to sustain service delivery Undeveloped Development initiated Development in progress Developed Advanced No concept or No formal training Most members of All members of staff are There is a culture of communication of service in service delivery is NMHSs are aware of the fully aware. providing best possible delivery principles exist. provided, though service importance of service Formal training is service delivery. delivery principles are delivery. provided. There is an Innovative ideas are Informally communicated. Some formal training is ad hoc process for routinely integrated into provided. staff to offer ideas for the continual service improvements to service improvement process. delivery. STRATEGY ELEMENT 6 Share best practices and knowledge Undeveloped Development initiated Development in progress Developed Advanced NMHSs are encouraged to share best practices in service delivery through formal training, twinning, mentoring, and other methods. 76 Weathering the Change: How to Improve Hydromet Services in Developing Countries? ANNEX 2 Progress Model for Observations, Telecommunications, and Forecasting OBSERVATIONS AND TELECOMMUNICATIONS Undeveloped Development initiated Development in progress Developed Advanced The NMHS has very few The NMHS has the Automation of observing Observations extend to The NMHS conducts manual synoptic stations capacity to support a networks with quality smaller scales and include research, introducing new and hydrological stations. synoptic meteorological control is routine. ground-based remote observational technologies It does not share these network1 and hydrological The NHMS accesses sensing techniques, and techniques as needed. data on the Global network. It shares these satellite data with, e.g., such as radar. The NMHS The observing network Telecommunication data on the GTS. The the capacity to derive may be able to take and is comprehensive, System (GTS). NMHS has sufficient staff precipitation estimates. integrate observations is sufficient to meet to maintain its observing The observing network from other parties. main user needs, and networks. is sustainable with incorporates external a sufficient budget It may access observations observations from other for operations and by outsourcing its suppliers, for example, maintenance. The observing requirements. agro-meteorological vertical structure of networks operated by a the atmosphere may be Ministry of Agriculture routinely measured. or hydrological network operated by a Ministry of Energy or Water Resources. Synoptic meteorological network—a network of stations at which surface and upper-air observations (at locations that give meteorological data representation of 1 the area in which they are situated, that could range many hundreds of kilometers) are made at standard times (i.e., main synoptic times: 0000, 0600, 1200, and 1800 UTC (Universal Time Coordinated); and intermediate synoptic hours: 0300, 0900, 1500, and 2100 UTC) for the purpose of presenting a comprehensive and nearly instantaneous picture of the state of the atmosphere (WMO, 2015c). Weathering the Change: How to Improve Hydromet Services in Developing Countries? 77 FORECASTING SYSTEMS Undeveloped Development initiated Development in progress Developed Advanced The NMHS provides up to The NMHS can provide The NMHS can provide LAM systems are available The NMHS has an two-days deterministic at least three-days 0 to 5 days forecasts locally or through extensive research forecast based on deterministic forecasts using global and regional regional centers. Using program and introduces graphical forecast based on access to global deterministic NWP and local data assimilation, new forecasting products retrieved from and regional numerical EPS data and products high-resolution short- technologies and different web sources. weather prediction data from GPCs; issues time scale forecasts are techniques; has the There is no verification of and products available on nowcasts and very produced with emphasis capacity to support forecasts. The NMHS does the GTS and/or graphical short-range forecasts on 0–6 hours for extreme requirements of other not operate forecasting on products available from up to 12 hours based on events. The forecasting NMHSs; is able to run a 24-hour, seven-days-a- WMO RSMCs. The NMHS extrapolating NWP and system extends from 0 to global, regional, and week basis, and warnings monitors the current blending remote-sensing at least 7 days based on national NWP and EPS are not issued. weather and hydrological observations; is able to a combination of global, systems. Forecasts of system. The NMHS has monitor major rivers and regional, and national weather and hydrological basic data-processing generate short-term flow deterministic NWP and impacts on specific sectors and archiving systems. and flood forecasts; has EPS data and products. are routine and generally It carries out subjective protocols for emergencies, The NMHS has the developed with users of forecast verification. backup of data and capacity to manipulate these forecasts. The NMHS There is no research products, and offsite digital data and to tailor has a well-developed and development, and storage facilities; carries forecasts to specific users. education and training the quality management out verification, and post- A multi-hazard warning unit. system is rudimentary. The processing; has some R&D system exists. It is able NMHS may not operate and a QMS. The NMHS to generate seasonal forecasting on a 24-hour, operates forecasting on stream flow outlooks, and seven-days-a-week basis. a 24-hour, seven-days-a- specialized hydrology Warnings are limited. week basis. products; and has full R&D capability. There are well- established relationships with partner agencies. 78 Weathering the Change: How to Improve Hydromet Services in Developing Countries? ANNEX 3 Progress Model for Climate Services CLIMATE SERVICES Undeveloped Development initiated Development in progress Developed Advanced The NMHS may operate a The NMHS designs, The NMHS has the The NMHS generates The NMHS has research limited national climate operates, and maintains capacity to develop and/ sub-seasonal to seasonal capacities and runs global observing system; collect national climate observing or provide monthly and forecast products, and regional climate data in paper form; systems; manages longer climate predictions, develops specialized models (sub-seasonal retrieve climate data data including QA/QC; including seasonal climate climate products; to decadal and longer); from different sources to develops and maintains outlooks, both statistical downscales long-term and works with sector- generate national climate data archives; monitors and model-based; conduct climate projections as based research teams products; participate in climate; oversees climate or participate in regional well as interprets annual and develops application regional climate outlooks; standards; performs and national climate to decadal climate models and develops and have very limited or climate diagnostics, outlook forums; interact predictions; covers all software and products no interaction with users. climate analysis, and with users in various the elements of climate suites for customized Typically, the NMHS does climate assessment; sectors; add value from risk management, from climate products. Staff not have staff dedicated to disseminates climate national perspectives on risk identification, risk have multi-disciplinary carry out climate services. products; participates in the products received assessment, planning modelling and statistical regional climate outlooks; from RCCs and in some and prevention, services expertise and can and interacts with users. cases GPCs for long- for response and downscale/calibrate The NMHS performs the range forecasts; conduct recovery from hazards, global scale information functions of a national climate watch programs; information relevant to to regional and national climate center providing and disseminate early climate variability and levels. The NMHS is able basic climate services. warnings. Staff are change, and information to receive and respond to Staff are proficient proficient in developing and advice related user requirements for new in climate statistics, and interpreting climate to adaptation; builds products. homogeneity testing prediction products, and societal awareness to techniques, and quality in assisting users in the climate change issues assurance techniques. uptake of these products. and provides information relevant to policy development and a national action plan. Staff have knowledge in climate modelling and methods for downscaling/calibration, risk and risk management and financial tools for risk transfer. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 79 ANNEX 4 Progress Model for Hydrological Services HYDROLOGICAL SERVICES Undeveloped Development initiated Development in progress Developed Advanced The NHS may operate Functions of NHS may The NHS is able to The NHS operates and In addition to the and maintain a very small include operation and operate and maintain a maintain a comprehensive foregoing capabilities, hydrological observation maintenance of a small hydrological observational hydrological observational the NHS has an extensive network; collect data hydrological observation network to monitor network to monitor major research and development in paper format; and network; hydrological major rivers, and take and some smaller rivers, program; and strong have very limited or no data management, with and integrate some and takes and integrates relationships with interaction with users. basic hydrological data- hydrological observations most of the hydrological partner agencies, taking a Typically, staff of NHSs processing, archiving from other parties. observations from leading role in the advice in this category are not and communication The NHS operates an other parties. The NHS and decision support. trained in hydrology. system; little or no interoperable hydrological operates a well-developed NHSs have the ability back-up / offsite storage; data management system; interoperable hydrological to generate customized and some interaction and have well-established data management hydrological products, and with users of hydrology protocols for emergencies, system; and has well- to develop hydrological data and products. backup of hydrological established protocols for application tools. There is no research data and minimum offsite emergencies, backup of and development, and facilities. The NHS carries hydrological data and rudimentary quality out water level and flow offsite facilities. The NHS management system. monitoring, and is able carries out water level and There are no relationships to generate short-term flow monitoring, and is with partner agencies. flow forecasts (low flows), able to generate short- flood forecasting, and term flow forecasts (low hydrological data products flows), flood forecasting, for design and operation and hydrological data of water supply structures. products for design and There is a small research operation of water supply and development unit; structures. The NHS is and a quality management also able to generate system. There are some seasonal stream flow relationships with partner outlooks, and specialized agencies. hydrology products. There is a research and development unit; and a well-established quality management system. There are well-established relationships with partner agencies. 80 Weathering the Change: How to Improve Hydromet Services in Developing Countries? ANNEX 5 Annotated Concept of Operations Concept of Operations (CONOPS) (Document #) for the {Name of the National Meteorological or Hydrological Service (NMHS} {Date} Prepared by: {Preparer} {Contact information} Weathering the Change: How to Improve Hydromet Services in Developing Countries? 81 Concept of Operations (CONOPS) Signature Page 82 Weathering the Change: How to Improve Hydromet Services in Developing Countries? Document Change Control Sheet Document Title: Concept of Operations (CONOPS) Date File name/version # Author Revision description Weathering the Change: How to Improve Hydromet Services in Developing Countries? 83 Table of Contents 1.0 Scope 1.1. Identification 1.2. Document Overview 1.2.1. Approach 1.2.2. IEEE Standard 1.3. System Overview 2.0 Referenced Documents 2.1. Standards and Guidelines 2.2. NMHS Documentation 2.3. Other Documentation 3.0 Current System 3.1. Background, Objectives, and Scope 3.1.1. Analysis of the Current Systems/Subsystems 3.1.1.1. First Analysis of Alternative Findings 3.1.1.2. Second Analysis of Alternative Findings 3.1.1.3. Legacy Evaluation Findings 3.1.2. Motivation for a New System/Subsystems 3.1.3. Modes of Operation of the Current System 3.1.4. Classes of Users 3.2. Operational Policies and Constraints 3.3. Description of the Current System/Subsystems 3.3.1. Description and Capabilities of Existing Systems/Subsystems 3.3.1.1. Monitoring and observing systems 3.3.1.2. Modelling systems 3.3.1.3. Objective and impact forecasting and warning systems 3.3.1.4. Service delivery systems 3.3.1.5. Actions, service monitoring, and feedback systems 3.3.1.6. Quality management systems 3.3.1.7. ITC systems 3.3.1.8. Technology infusion systems 3.3.1.9. Capacity building 3.4. Modes of Operations for Current System 3.5. User Classes and Other Involved Personnel 3.5.1. Organizational Structure 3.5.2. Profiles of User Classes 3.5.3. Interactions among User Classes 3.5.4. Other Involved Personnel 3.6. Support Environment 4.0 Justification for and Nature of Changes 4.1. Justification of Changes 4.2. Description of Desired Changes 4.2.1. Proposed System Attributes 4.2.2. Proposed System Capabilities 4.2.3. Proposed System Interfaces 84 Weathering the Change: How to Improve Hydromet Services in Developing Countries? 4.3. Priorities among Changes 4.4. Changes Considered but Not Included 4.5. Assumptions and Constraints 4.5.1. Assumptions 4.5.2. Constraints 4.6. Adverse Effects 5.0 Concepts of the Proposed System 5.1. Background, Objectives, and Scope 5.2. Operational Policies and Constraints 5.3. Description of Proposed System 5.3.1. Major System Components 5.3.2. Interface to External Systems and Data Flow 5.3.3. Capabilities or Functions of the Proposed System 5.3.4. Continuity of Operations 5.4. Modes of Operation 5.5. User Classes and Other Involved Personnel 5.5.1. Organizational Structure 5.5.2. Profiles of Users Classes 5.5.3. Users’ Capabilities 5.5.4. Interactions among Users and Stakeholders 5.5.5. Other Involved Personnel 5.6. Support Environment 6.0 Operational Scenarios 6.1. Disaster Management Scenario 6.2. Water Resources Management Scenario 6.3. Public Weather Services Scenario 6.4. Farming Community Scenario 6.5. Facilities 7.0 Summary of Impacts 7.1. Operational Impacts 7.2. Organizational Impacts 7.3. Impacts during Development 8.0 Analysis of the Proposed System 8.1. Summary of Improvements 8.2. Disadvantages and Limitations 8.3. Alternatives and Trade-offs Considered 9.0 Notes 10.0 Appendices 11.0 Glossary List of Figures List of Tables Weathering the Change: How to Improve Hydromet Services in Developing Countries? 85 PREFACE The NMHS {provide mission statement} {Statement about what will be possible after the completion of the modernization project} The CONOPS document provides a conceptual overview of the proposed system and sub-systems. The CONOPS is intended to support the evolution of a fully integrated, modernized and functional NMHS, which provides the level of services required by its users and stakeholders. The CONOPS is a living document and will be coordinated in a collaborative manner with users and stakeholders to ensure the viability of the concepts presented. 86 Weathering the Change: How to Improve Hydromet Services in Developing Countries? CONCEPT OF OPERATIONS (CONOPS) 1.0 Scope 1.2.1 Approach The initial approach taken by the NMHS technical team (TT) The CONOPS document describes the desired characteris- to develop this document used concept analysis, the process tics of the NMHS’s system from the users’ viewpoint. The of analyzing a problem domain and an operational environ- sections below identify the proposed system, provide a ment for the purpose of specifying the characteristics of the document overview and the approach used to generate the proposed system from the perspective of users. This helped document, and provide a brief overview of the system. to clarify and resolve vague and sometimes conflicting needs, wants, and opinions by reconciling divergent views. Using this 1.1 Identification approach, the potential for designing a system in which each The proposed system will include all of the associated sub- individual function meets its specifications, but the system as systems (Figure 1-1, Components of the Proposed NMHS a whole fails to meet the users’ needs, was minimized. System) that comprise the NMHS, including all associated This version of the CONOPS document is the initial version equipment, facilities, material, software, hardware, policy, based on extensive consultation with users, staff of the technical documentation, services, training and person- NMHS, and external technical experts. nel required for operations and support of the NMHS. The CONOPS is built on a user requirements document, which In order to meet users’ requirements, a complete reexamina- was prepared separately and included as an annex to the tion of all subsystems was required. CONOPS document in Section 10. 1.2.2 IEEE Standard 1.2 Document Overview {If appropriate cite IEEE Std. 1362-1998 guide for informa- tion technology system definition—Concept of Operations The CONOPS is a vehicle to communicate high-level quan- (CONOPS) document} titative and qualitative characteristics of the system to the user, developer, operator, and other stakeholders. The CONOPS addresses the challenges involved in ensuring the 1.3 System Overview NMHS is fit for purpose. {Describe the current structure of the NMHS; how it operates, etc. This should be based on a separate assessment of the orga- ■■ Section 1 describes the approach used to develop the nization, which should be referenced.} CONOPS; Section 2 provides a list of reference documentation used ■■ in the creation of the CONOPs; 2.0 Referenced Documents ■■ Section 3 describes the current NMHS systems; The standards, guidelines, and NMHS documentation used to develop the CONOPS are described in the sections that follow. ■■ Section 4 discusses the justification and nature of changes based on the most current information; 2.1 Standards and Guidelines ■■ Section 5 provides information on proposed system concepts; The standards and guidelines used in preparation of this document are listed below. ■■ Section 6 describes operational scenarios; ■■ Section 7 summarizes operational, organizational, and 2.2 NMHS Documentation other impacts during development; The following NMHS document was used to support the gen- ■■ Section 8 analyzes the proposed system; eration of this document ■■ Section 9 provides additional information; ■■ {statutory authority of NMHS} ■■ Section 10 is for annexes; and ■■ Review of NMHS ■■ Section 11 provides a glossary of terms. ■■ …. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 87 NMHS Project Management Documentation 3.1.1.1 First Analysis of Alternative Findings {If there are alternatives for the modernization of the NMHS’ ■■ {Results of stakeholder workshop} systems and subsystems, they should be described here in ■■ {User requirements document} terms of whether those alternatives met the requirements of the ■■ {Early version of CONOPS, if any} users. Scoring the current system against a complete or partial ■■ {Analysis of alternatives} modernization with respect to mission tasks accompanied by a technical and cost analysis is used to justify the new system.} ■■ {Project management plan} ■■ {Project appraisal document} In the initial analysis, several alternatives for the devel- opment of the NMHS were identified by the Analysis of ■■ {Project description document} Alternatives Team and examined {for example, the privatiza- tion of the NMHS; creating a commercial arm of the NMHS; 2.3 Other Documentation partnerships with the private sector; and focusing the NMHS Other documentation used to support the generation of this exclusively on public sector services, etc.}. The Analysis of document. {May include government directives on data shar- Alternatives Team rated the alternatives and selected the ing, for example.} alternative that would best meet the goals and objections of the NMHS mission. 3.0 Current System In the first step, the Analysis of Alternatives Team identified a The following subsections describe the background, objec- number of basic criteria from existing NMHS needs and laws tives, and scope of the current system and its subsystems; that each alternative would have to meet. Through their anal- operational policies or constraints; the current system and ysis, the Analysis of Alternatives Team eliminated the alter- subsystems; modes of operation for the current system; native of privatizing the NMHS, for example, as this alternative users of the services provided by the NMHS; and staff and would be unable to comply with current laws and regulations. the support environment. In the second step, the remaining alternatives were scored against the mission tasks, measures of effectiveness, and 3.1 Background, Objectives, and Scope the requirements of the NMHS. {Each of these needs to be {Describe how well or otherwise the current system meets the identified enabling a numerical score to be created. Measures mission needs.} of effectiveness might include proficiency in providing warn- ings to disaster managers, for example.} Currently, the NMHS’s infrastructure consists of manual observing systems, daily forecasts, etc., that do not meet the Lastly a technical and cost analysis with respect to the mission needs, or effectively address the emerging require- development of a new system/subsystems was performed. ments of the NMHS’s key users. {Describe briefly the history {This should guide and support the alternative selected for of the organization.} {Weather and climate dependence of all the new system and provide a strong argument based on its sectors of the economy are increasing, and the risk to lives and cost-effectiveness to deliver the mission tasks.} livelihood of citizens is growing as more people are exposed to 3.1.1.2 Second Analysis of Alternative Findings severe weather and flooding.} {If additional analysis is done it would be described here. This might include a separate risk analysis. It might also be used to 3.1.1 Analysis of the Current Systems/ evaluate new offers as alternatives to the preferred approach, Subsystems for example, from a new vendor of turnkey systems that had The NMHS’s systems are nearly obsolete, incapable of pro- not been considered during the first analysis of alternatives.} viding the breadth and depth of services that the proposed system will provide. Currently no single system or group of 3.1.1.3 Legacy Evaluation Findings systems exists {i.e., no turnkey solution} that will provide the {This would include an evaluation of systems that have been capabilities envisioned for the proposed modernized NMHS provided by other development partners and need to be system. To support this assertion analyses were performed included in the final modernized NMHS. These legacy require- on alternatives. ments might be included as a table.} 88 Weathering the Change: How to Improve Hydromet Services in Developing Countries? 3.1.2 Motivation for a New System/ 3.3 Description of the Current System/ Subsystems Subsystems {Describe the factors that motivate the need for a new system. This will include new observational technologies that will The NMHS’s current operating environment consists of a improve nowcasting, very short and short-range weather fore- number of subsystems that do not adequately fulfil its mis- casting and service delivery, etc.} sion needs. {This section should provide an overview of the whole system and how it performs currently. The details of The NMHS’s process for monitoring and observing, analyzing each subsystem would be discussed in sub-sections.} and forecasting, and dissemination rely largely on manual systems. The NMHS needs a system capable of meeting the needs of users, which requires the ability to provide timely, 3.3.1 Description and Capabilities of accurate, and relevant impact-based forecasts and warnings Existing Systems/Subsystems tailored to the specific needs of different groups. The following subsections provide a functional description of each subsystem and include any known limitations. A {A motivation for modernization might include the acquisition of summarization of the high-level functionality of these sys- new observational capabilities; for example, to exploit fully new tems is also provided. radar technologies requires nowcasting and very short-range forecasting capability, access to high resolution models, etc.} 3.3.1.1 Monitoring and observing systems In order to meet the NMHS’s strategic goals, the proposed {Function and current limitations} system should be able to accomplish the following goals: 3.3.1.2 Modelling systems ■■ {Service-related goals} {Function and current limitations} ■■ {Reliability of forecasting system} … 3.3.1.3 Objective and impact forecasting and ■■ {Access and effective use of NWP ensembles} … warning systems {Function and current limitations} ■■ {Affordable and easy to maintain observing network} ■■ {Etc.} 3.3.1.4 Service delivery systems {Function and current limitations} 3.1.3 Modes of Operation of the Current System 3.3.1.5 Actions, service monitoring, and feedback sys- Modes of operation of the current systems are provided in tems Section 3.4, Modes of Operation for Current System. {Function and current limitations} 3.1.4 Classes of Users 3.3.1.6 Quality management systems Classes of users for the current systems are identified in {Function and current limitations} Section 3.5, User Classes and Other Involved Personnel. 3.3.1.7 ITC systems {Function and current limitations} 3.2 Operational Policies and Constraints 3.3.1.8 Technology infusion systems {This might include interagency agreements on data sharing, {Function and current limitations} commercial sales of data, etc.} 3.3.1.9 Capacity building There are no other constraints beyond those mentioned {Function and current limitations} in previous sections and in Section 4.0, Justification for and Nature of Changes. Limitations of a number of the A summary of the high-level functionality/capability of each subsystems that comprise the current environment are dis- subsystem is provided in Table 3.1. cussed in Section 3.3, Description of the Current System/ Subsystems, and appropriate subsections. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 89 TABLE 3.1 Systems/Subsystems High-Level Functionality Summary System/Subsystem Functional area/high-level functionality Point of contact 3.4 Modes of Operations for 3.5 User Classes and Other Involved Current System Personnel {Describe why the current system is inadequate.} {Summarize how people interact with the current system. A later section (5.5) will describe this in detail.} The current modes of operation provide forecasting capabil- ities limited to {describe briefly}. Users, stakeholders, and other involved personnel are pro- vided in Table 3.2. TABLE 3.2: System Users System Number and Type of Users/Providers Services (actions, etc.) There are a total of {XX} user groups comprising public sector, businesses, and government stakeholders. Service delivery systems Staff: Observing systems The observing system is supported by {XX} field technicians. Internal users comprise the modelling and forecasting staff. Modelling systems The modelling system is supported by {XX} technical staff. Internal users comprise {XX} forecasters. Forecasting systems QMS Staff: ITC Staff: Technology infusion systems Staff: Capacity building 90 Weathering the Change: How to Improve Hydromet Services in Developing Countries? 3.5.1 Organizational Structure desired changes, priorities among changes, changes consid- {Organograms of how the NMHSs serve its public users and ered but not included, and assumptions and constraints of government stakeholders, and how the NMHS is organized building the new system. internally.} 3.5.2 Profiles of User Classes 4.1 Justification of Changes Profiles of the user classes for the current situation identified Advances in knowledge and technology have increased the above are the same as those addressed in Section 5.5.2, accuracy of meteorological predictions that help save lives Profiles of User Classes. Please refer to this section for and improve economic performance. The NMHS has not kept detailed information. pace with these developments. 3.5.3 Interactions among User Classes Interaction among the user classes for the current situation 4.2 Description of Desired Changes identified above are the same as those addressed in Section The following subsections list the proposed system attri- 5.5.3, Interactions among User Classes. Please refer to butes, capabilities, and interfaces. this section for detailed information. 3.5.4 Other Involved Personnel 4.2.1 Proposed System Attributes Responsible points of contact for the systems/subsystems From the overall system perspective, the proposed modern- are provided in Table 3.1, Systems/subsystems High-Level ization of the NMHS should possess the following attributes: Functionality Summary. ■■ Infrastructure independence: an architecture that allows the NMHS to operate under all circumstances including 3.6 Support Environment during extreme hazards {How are the various systems/subsystems supported and ■■ Scalability: the capability to accommodate growth and maintained?} evolving technologies without requiring a complete re- The current systems are supported entirely by in-house gov- placement of the systems ernment support. ■■ Flexibility: the ability to tailor services to users’ future requirements. The system should be flexible enough to 4.0 Justification for and Nature interface with other agencies’ evolving systems of Changes ■■ Comprehensiveness: the system should be able to meet all anticipated needs {General description of the need for change, followed by brief description of subsections that follow.} ■■ Extensibility: the ability to handle new type of observa- tions, numerical models, and forecasting systems Over the last {XX} years, the NMHS has been responsible for providing meteorological and or hydrological services to the country. Continuing growth in the weather- and climate-de- 4.2.2 Proposed System Capabilities pendent economic sectors and in the increasing exposure To meet strategic objectives, the NMHS must integrate its of a growing population to weather and climate extremes systems and subsystems and incorporate legacy systems make it an imperative to improve the quality and level of and new systems provide by other development partners. To services provided by the NMHS to protect lives, livelihoods, meet the challenges of today and the future, the proposed and economic development. NMHS system should provide the following capabilities. Currently the NMHS can provide only a limited number of ser- ■■ Capability to receive data from a variety of international vices that do not meet the growing needs of users adequately. and national sources and use these data in the forecasting system The following subsections describe the proposed system in terms of the justification of changes, a description of the ■■ Capability to ingest, preserve, and provide access to data Weathering the Change: How to Improve Hydromet Services in Developing Countries? 91 ■■ Capability to store data in a manner that is independent 4.4 Changes Considered but Not Included of any particular hardware and software over long periods There are no changes to the proposed NMHS system that ■■ Capability to scale in order to store and preserve data were considered but not included in the proposed list of ■■ Capability to provide electronic access to data and infor- NMHS system attributes, capabilities, or system interfaces mation for all users based on established user rights and identified in Sections 4.2.1 through and including Section privileges to ensure that its users are able to access all 4.2.3. Changes to the items provided in those sections will of the electronic data and information to which they are not be known until completion of the systems analysis and entitled design phase of the NMHS modernization program. Moreover, changes that may result in newly identified requirements ■■ Capability to provide access to data and information in and/or changes to requirements cannot be considered until a manner consistent with current technology and the changing expectations of its diverse users cost estimations have been provided. ■■ Capability to adapt to changing technology in order to provide the level of services desired by the users 4.5 Assumptions and Constraints ■■ Capability to fully exploit current knowledge and tech- This section identifies assumptions and constraints that may nology to provide the optimum level of quality services impact the system architecture or specific components of expected by the users the proposed system. 4.2.3 Proposed System Interfaces 4.5.1 Assumptions The proposed system will operate within the framework of The proposed NMHS system relies on a set of assumptions the national government, and be capable of interfacing with that is derived from the NMHS’s operational policies or other government systems, enabling the seamless transfer are inherent in its operational environment. The proposed of data and information. The volume and diversity of input NMHS system assumes that: and output data, and the expected heavy use of the system, ■■ For the next several years, data inputs will be in standard will have considerable impact on the NMHS computing and familiar format enabling full integration into the environment. system without requiring any reengineering; Interfaces with other government agency systems will be ■■ Expect the volume, variety, and complexity of data will accommodated by the proposed NMHS system. Other sys- continue to grow throughout the period of the NMHS tems the proposed NMHS system will interface with are modernization; provided below with additional detail in Section 5.3.2, ■■ Expect the numerical weather prediction technology Interfaces to External Systems or Producers. available from the international community to evolve ■■ WMO information System significantly throughout the period of the NMHS modernization and beyond; and ■■ ECMWF ■■ Expect the operations and maintenance budget for the ■■ WMO Regional Specialized Meteorological Center NMHS to keep pace with the technological improvements ■■ NCEP throughout the period of NMHS modernization and beyond. ■■ Etc. 4.5.2 Constraints 4.3 Priorities among Changes Constraints that may impact the system architecture or This section describes the functionality by increment that specific components of the proposed system are provided in the proposed NMHS system is expected to satisfy. {This may Section 5.2, Operational Policies and Constraints. be developed in a separate document if necessary.} 92 Weathering the Change: How to Improve Hydromet Services in Developing Countries? 4.6 Adverse Effects Constraints as currently known that may impact system architecture or major system components are provided The risk of not proceeding with the development of the pro- below. posed system are many and include the following: ■■ World Bank procurement rules; ■■ The NMHS will not be able to achieve its mission ■■ NMHS system design and implementation is flexible ■■ The public will not receive information that could protect and adaptable to changes in hardware, software, their lives and livelihoods communication technology, policy, personnel, locations, ■■ The weather- and climate-dependent sectors will not etc.; receive information relevant to protecting their business ■■ The NMHS system design is a balance of Commercial Off activities The Shelf (COTS) and developed software; and ■■ Emergency management of weather-related hazards will ■■ The NMHS system implementation is within the specified be compromised budget and time frame. ■■ The NMHS will not be positioned to provide adequate guidance, assistance, or services to agencies to manage their weather and climate risks 5.3 Description of Proposed System The proposed NMHS system described herein incorporates ■■ The NMHS’s role as the single authoritative voice for the information from user surveys, user workshops, require- weather warnings would be increasingly reduced ments documents, NMHS reviews, and recommendations ■■ The nation’s long-term climate records would be adversely from other reviewers on necessary and desired character- affected if these data will not be stored properly istics, as well as those learned from ongoing activities and other development partners’ activities. The design and development of the proposed NMHS system currently does 5.0 Concepts of the Proposed System not exist. The following subsections describe the concepts of the pro- posed system with respect to the background, objectives, Section 4.6 provides a bulleted list of risks in the event the and scope of the proposed system; operational policies and NMHS system is not designed and developed. constraints that apply to the new system; description of the proposed system; users, stakeholders, and other involved 5.3.1 Major System Components personnel; and the support environment. Proposed architectural characteristics/concepts are pro- vided in {this might include review of NMHS with specific recommendations; workshops on NWP and forecasting, etc.}. 5.1 Background, Objectives, and Scope Additionally, details on major components are presented in A high-level system overview has been provided in Section the Section 6.0, Operational Scenarios. 1.3, System Overview. Goals of the proposed NMHS system and the motivation for the new system are discussed in 5.3.2 Interface to External Systems and Data Section 3.1.2, Motivation for a New System/Subsystems. Flow {Here follows a detailed description of the proposed system.} The external interfaces include the observations ingested from the WMO information system; access to proprietary and nonproprietary numerical weather prediction data; 5.2 Operational Policies and Constraints and non-NMHS national and local sources of observations The proposed NMHS system is required to adhere to gov- and non-weather data. There are also external interfaces ernment policies for the provision and exchange of data. It with users, which can provide feedback on the utility of is mandated by {here cite any legal arrangements that govern the information they receive. This is also a crowdsourcing the operation of the NMHS and its relations with users, private interface to improve impact-based forecasts and warnings. sector providers, etc.} Figure 5.1, External Interfaces Context Diagram depicts Weathering the Change: How to Improve Hydromet Services in Developing Countries? 93 external interfaces along with associated data flows. The Standard operating procedures should be established for data flows are high-level, i.e., not broken down into individ- the different modes of operations and the expected level of ual elements. services that can be provided in the different modes. 5.3.3 Capabilities or Functions of the Proposed 5.5 User Classes and Other Involved System Personnel Capabilities of the proposed NMHS system are listed in The following sections describe the organizational structure Section 4.2 {they may also be expounded on in additional and classes of users, including user capabilities that are documentation}. NMHS capabilities by user class are pro- associated with the proposed NMHS system. vided in Section 5.5.3., Users’ Capabilities. 5.3.4 Continuity of Operations 5.5.1 Organizational Structure The following organizational chart, Figure 5.2, Organizational The proposed NMHS system is intended to be able to survive Interfaces, portrays the structure of the NMHS. and operate during hazardous events. Therefore, the ability to operate critical infrastructure and services away from the main operations center needs to be considered. Different 5.5.2 Profiles of Users Classes approaches will vary in terms of complexity and cost. A user is anyone who uses the services of the NMHS. A user class is determined by the way in which the user interacts with the NMHS system. The major user classes identified for 5.4 Modes of Operation the proposed NMHS system include the following: The modes of operation for the proposed system as currently ■■ End-user—receives forecasts and warnings and acts on known are: those warnings. An end-user is generally the public or a ■■ Nominal community. An end-user may also apply to a business that does not receive specially tailored products and services ■■ Degraded from the NMHS. ■■ Maintenance ■■ Government stakeholders—receive forecasts and warnings • Remedial maintenance from the NMHS usually in advance of any release to the • Preventative maintenance public. This information may be automatically transferred ■■ Upgrades from the NMHS and may be incorporated in decision- support tools; for example, for emergency operations. For ■■ Alternative site non-collocated hydrological services, this information would Nominal mode of operation describes the system when enable flood forecasts to be computed at a remote location. working at the optimum, i.e., the system is operational and ■■ Customers—these are users or stakeholders that receive working as intended. forecasts, warnings, guidance, etc., as a fee-for-service. Generally, this would apply to commercial users, who by Degraded mode and maintenance mode of operation agreement may repackage information for their clients describe operations in times when the system or subsys- or use the information in commercial decision-making tems are working using a reduced strength of operations. systems. For example, the system is in maintenance mode if the radar system is off-line. Forecasters may need additional backup ■■ Etc. from remote centers or use satellite data under these circumstances. 5.5.3 Users’ Capabilities ■■ The NMHS services correspond to users’ capabilities, Alternative site mode of operation can be described as needs, and desires for the proposed system. These occurring when the main operational center is out of action capabilities are organized according to user class, but due to the Impact of a hazard requiring operations to be some capabilities cross user class boundaries and may be conducted from an alternate site. employed by users in more than one user class. 094 Weathering the Change: How to Improve Hydromet Services in Developing Countries? End-user 5.5.5 Other Involved Personnel Figure 5.4, NMHS Modernization Project Oversight ■■ Receives forecasts and warnings through traditional or Structure depicts the organizational structure that is over- social media seeing the overall execution of the project. This organiza- tional structure includes the following elements: ■■ NMHS may provide apps to assist the end-user access information through smartphones and the Internet ■■ The Director General of the NMHS is the decision authority ■■ Provides NMHS with feedback through crowdsourcing ■■ {Can be populated by the Project Appraisal Document or tools where available, which provides feedback on the Terms of Reference of System Integrator (SI)} quality of the service and the actions taken by end-users to minimize their exposure to hazards following receipt ■■ Etc. of a warning ■■ Etc. 5.6 Support Environment The support environment will not be determined until the conclusion of the system analysis and design phase of the Government stakeholders modernization program. ■■ Emergency managers may be responsible for transmitting warnings to the public and others 6.0 Operational Scenarios ■■ Emergency operations centers may collocate the NMHS The CONOPS document expresses what users want and extreme weather forecast desk to enable forecasters to envision. Scenarios convey these needs in nontechnical lan- interact directly with emergency managers guage. Overlap exists between different scenarios as a result ■■ Hydrological service may collocate with the NMHS of interaction between different users or due to similarity forecasting office to enable flood forecasts to interact between different activities. For example, the document directly with weather forecasters might include scenarios for disaster managers, agriculture ■■ Water resource managers and dam operators access extension workers, farmers, and the general public. critical weather and hydrological information to manage A scenario is a step-by-step description of how the NMHS water levels should operate and interact with both its users and external ■■ Etc. interfaces under a given set of circumstances. Scenarios should enable readers to walk through them and gain an understanding of how all principal parts of the NMHS func- Customers tion and interact. ■■ May have varying degrees of capability depending on their operations, ranging from guidance based on forecasts to 6.1 Disaster Management Scenario ingesting data into their own decision support systems or combining the NMHS products with their own to generate This scenario represents one example of how the NMHS will ensembles and new forecast products for their clients. support the emergency operations prior to, during, and after a major event. ■■ Etc. ■■ NMHS provides a long lead-time outlook on a developing adverse weather situation, up to 10 days ahead based on 5.5.4 Interactions among Users and Stakeholders numerical model guidance; The proposed NMHS system described here is an overall conceptual workflow model that depicts where user classes ■■ Enables emergency operations to begin initial preparation; should interact within the system and with each other. ■■ Guidance is updated daily, and closer to the event horizon Figure 5.3 NMHS User classes, illustrates this conceptual briefings are given on a frequent basis; model. Weathering the Change: How to Improve Hydromet Services in Developing Countries? 095 ■■ Severe weather forecasters may relocate to emergency 7.0 Summary of Impacts operations centers; and Warnings are issued via the emergency operations center. ■■ 7.2 Operational Impacts ■■ Etc. Until the modernization program undergoes systems anal- ysis and design, operational impacts of the proposed mod- 6.2 Water Resources Management ernization are not known; therefore, impacts to the following Scenario have been omitted: This scenario represents one example of how the NMHS will ■■ Changes in the operational budget support water resources management. ■■ Changes in the operational risks ■■ NMHS provides seasonal and longer lead-time forecasts ■■ Changes in the maintenance budget to hydrologists to predict long-term availability of water and the risk of meteorological drought. ■■ Changes in the maintenance risks ■■ Etc. ■■ New modes of operations based on new forecasting tech- niques Changes in the quantity, type, and timing of data to be 6.3 Public Weather Services Scenario ■■ input into the system This scenario represents one example of how the NMHS will support public weather services. ■■ Uses of new data sources ■■ NMHS provides daily forecasts up to 10 days ahead to the ■■ Changes in procedures general public via traditional and social media; However, it is anticipated that the NMHS will have to imple- ■■ NMHS provides public weather information via smart- ment changes to the ways it conducts business in order to phone apps; achieve the agency’s mission, goals, and objectives. The proposed NMHS system will facilitate this endeavor. ■■ Warnings of severe weather are issued via the emergency operations center; and ■■ Impact-based forecasts are provided routinely. 7.2 Organizational Impacts The NMHS is examining current policies and business prac- ■■ Etc. tices and may have to develop and/or modify policies and business practices as necessary. The depth and breadth 6.4 Farming Community Scenario of the organizational impact is unknown at this time. This scenario represents one example of how the NMHS will Information with respect to the following has not been pro- support farmers. vided for this reason. ■■ In addition to the routine public weather services, ■■ Numbers and skill levels of personnel needed for contin- farmers receive special bulletins related to planting, gency operation harvesting, and protection of crops and livestock. These ■■ Changes in the number of personnel, skill levels, position impact-based forecasts enable farmers to manage inputs identifiers, or location of personnel and maximize yields. ■■ Revision of position descriptions to reflect changes in ■■ Etc. business practices ■■ Commitment of resources to the new forecasting system 6.5 Facilities ■■ The need for cross-functional, interdisciplinary staff teams Facilities’ requirements will be based on the future design ■■ The development of education and increased training for of the system. both NMHS staff and users 096 Weathering the Change: How to Improve Hydromet Services in Developing Countries? ■■ Improved opportunities for career development for NMHS ■■ Nowcasts and very short-range weather forecasts; staff ■■ Dynamic, ensemble-based flood forecasts; and ■■ Relationship between the NMHS and external partners ■■ Impact-based forecasts and warnings. 7.3 Impacts during Development 8.2 Disadvantages and Limitations The full extent of impacts during development will not be Potential disadvantages or limitations of the proposed NMHS known until completion of the system analysis and design system include: phase and as such has not been provided; however, impacts considered thus far including the following: ■■ High development costs; ■■ Articulation of business rules, templates, and other con- ■■ Staff anxiety brought about by new responsibilities; trols needed for operational implementation; ■■ Poor staff morale with proactive change management; ■■ Development of training for staff; ■■ Higher operating and maintenance costs; and ■■ Consideration of parallel operation of the new and exist- ■■ Unrealistic expectations of users and stakeholders. ing systems; and Operational impacts during system testing of the pro- ■■ Alternatives and Trade-offs Considered posed system. {May be documented elsewhere in support of the analysis of alternatives.} 8.0 Analysis of the Proposed System Various improvements, disadvantages and limitations, 9.0 Notes alternatives and trade-offs considered are covered in this section. The technical terms used in this document are defined below. Table 9.1 provides a list of acronyms used herein. 8.1 Summary of Improvements The proposed NMHS system, when implemented, will sub- TABLE 9.1: Acronyms sume and enhance the existing functionality provided by the Acronym Definition current system identified in Section 3.3.1; however, the pro- CONOPS Concept of Operations posed NMHS system will provide new sets of capabilities as NMHS National Meteorological and Hydrological Service offered in Section 4.2.2. The full extent of the capabilities to be provided by the proposed NMHS system will not be known until completion of the systems analysis and design phase of the program and will be addressed at that time as required. It is anticipated, however, that the proposed NMHS system will offer numerous benefits to users and may WMO World Meteorological Organization include the following: ■■ More accurate and timely forecasts based on full use of 10.0 Appendices the high-resolution ensembles provided by advanced numerical prediction centers, such as ECMWF; 11.0 Glossary ■■ More confidence in warnings issued with greater understanding of risk and the ability to improve decisions in low probability, high impact extreme weather events; List of Figures ■■ More detailed shorter range forecasts; List of Tables Weathering the Change: Climate change is impacting the lives and livelihoods of people everywhere. More How to Improve Hydromet people are vulnerable and exposed to the effects of extreme weather, especially through Services in Developing the impact of floods, landslides, drought, and winds. One way to cope is by reducing Countries? exposure to harm by providing more accurate, reliable, and actionable weather, climate, and hydrological information. This is amply demonstrated in high-income countries, where the impact of extreme weather events is mitigated by their ability to take early action based on meteorological and hydrological warnings. This is possible because these countries have invested in their publicly financed National Meteorological and Hydrological Services (NMHSs), encouraged the development of complementary private services, and invested heavily in research and development. Global Facility for Disaster Reduction and Recovery The Global Facility for Disaster Reduction and Recovery (GFDRR) is a global partnership www.gfdrr.org that helps developing countries better understand and reduce their vulnerabilities to natural hazards and adapt to climate change. Working with over 400 local, national, regional, and international partners, GFDRR provides grant financing, technical assistance, training, and knowledge sharing activities to mainstream disaster and climate risk management in policies and strategies. Managed by the World Bank, GFDRR is supported by 37 countries and 11 international organizations.