AGRICULTURE GLOBAL PRACTICE TECHNICAL ASSISTANCE PAPER INCREASING AGRICULTURAL PRODUCTION AND RESILIENCE THROUGH IMPROVED AGRO- METEOROLOGICAL SERVICES WORLD BANK GROUP REPORT NUMBER 94486-GLB MARCH 2015 AGRICULTURE GLOBAL PRACTICE TECHNICAL ASSISTANCE PAPER INCREASING AGRICULTURAL PRODUCTION AND RESILIENCE THROUGH IMPROVED AGRO- METEOROLOGICAL SERVICES © 2015 World Bank Group 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org Email: feedback@worldbank.org All rights reserved This volume is a product of the staff of the World Bank Group. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of World Bank Group or the governments they represent. The World Bank Group does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of World Bank Group concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. World Bank Group encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone: 978-750-8400, fax: 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank Group, 1818 H Street NW, Washington, DC 20433, USA, fax: 202-522-2422, e-mail: pubrights@worldbank.org. Cover Photo: Erick C. M. Fernandes. CONTENTS Preface v Acknowledgments vii Acronyms and Abbreviations ix Executive Summary xi Chapter One: Agro-Meteorological Observation Networks 1 IGAD Climate Prediction and Applications Centre 1 Kenya 2 Ethiopia 7 Chapter Two: Traditional and Non-Traditional Agro-Weather Observation Systems 17 Traditional Instrumentation 18 Nontraditional Instruments 20 Special Agro-Meteorology Related Sensors for Use with Automatic Weather Stations 21 Area Rainfall Estimates from Lightning Locating Systems 22 Discussion 24 Chapter Three: Adoption Barriers for Agro-Weather Observations 27 Cultural Barriers 27 Bureaucratic Barriers 28 Financial Barriers 30 Chapter Four: Cost-Effective Strategies for Upscaling Agro-Meteorological Observation Systems 33 Estimating a Cost-Benefit Ratio for Agro-Meteorological Services 33 Potential Impacts on Ethiopia and Kenya 35 Other Benefits 36 Cost-Effective Strategies for Improving Agro-Meteorological Observation Systems 37 Chapter Five: Models for Sustainability of Agro-Meteorological Observation Systems 41 The Spread of the Cellular Telephone Network Across Sub-Saharan Africa 42 Concerns Regarding Representativeness 43 Public-Private Partnerships 43 Chapter Six: Best Practices for Institutional Strengthening and Capacity Building 47 Strategic and Implementation Plans 48 Concept of Operations 49 Parallel, Coupled Development Efforts 50 Example of an Agro-Meteorological Network 52 Chapter Seven: Final Thoughts 53 References 55 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services iii Appendix A: Lessons to Be Learned from the Weather for All Initiative 57 Appendix B: KMS Weather and Crop Review 61 BOXES Box 1.1: Agro-Meteorological Stations 3 Box 1.2: Rwandan Farmers to Benefit from Insurance Agreement 6 Box 1.3: Improved Access to Weather Information Leads to Better Farming Outcomes 8 Box 5.1: Weather Philippines Foundation to Roll Out 1,000 Weather Instruments in 2014 46 Box 6.1: Defining Capacity Building under the PIC Directive 48 FIGURES Figure 1.1: Illustration in a KMS Review 3 Figure 1.2: Recent Floods and Impact in Kenya 4 Figure 1.3: Distribution of Agro-Meteorological Stations in Ethiopia 9 Figure 1.4: Rainfall Distribution (in Millimeters) for February 11–20, 2014 10 Figure 1.5: Observed versus Normal Rainfall (February 11–20, 2014) 10 Figure 2.1: Global Lightning Frequency 23 Figure B.1: Analysis of Rainfall, Temperature, and Vegetation Conditions 63 iv Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services PREFACE Informing farmers’ production decisions with timely weather forecasts and related climate information services is a fundamental means of improving agricultural pro- ductivity. In Sub-Saharan Africa, where agriculture is overwhelmingly rainfed and yield gaps are very high, access to these information sources will be instrumental in reducing the vulnerability of African smallholders by enabling them to manage risk more effectively. The information and communication technologies needed to improve and expand agro-meteorological observation systems are widely available and readily transferable from one setting to another. In addition to news and weather services that farmers can use in making immediate and short-term decisions, these channels are useful for longer-term, educational pur- poses as well—purposes that in more industrialized countries typically fall under the rubric of agricultural research and extension. In countries like Ethiopia and Kenya, which are the focus of this report, the generation of adaptive farming technologies and methods and their effective dissemination to producers who are prepared to use them will determine a great deal. The farmer who is comfortable referring to sources of information that he or she has come to trust can be an active agent in what Theodore Schultz once called “transforming traditional agriculture.” Farmers who remain unreached by these services are far less well positioned to do so. And we should remember just how much we are asking of them as we seek to “enlist” them as agents of change and even drivers of economic development. Schultz is more than a passing reference here. In 1964 he argued persuasively that farmers in low-income countries were rational actors who made efficient use of the resources available to them. He depicted them as rational but highly constrained in terms of access to resources and knowledge. This may seem intuitive now, but back then in the larger field of development economics, it was revolutionary—enough so to eventually earn him a Nobel Prize in Economics in 1979. Our language has changed somewhat, and Shultz’s use of the term “knowledge” clearly encompassed information available to the farmer and the farmer’s skill and know-how. And the issues touched upon in this document touch upon those as well. This is not only about the farmers themselves, however. It also relates to the institutions and policies of the country in which they live and operate. As its name clearly suggests, “agricultural meteorology” is very much an interdisciplinary field. This is more than a question of what academic discipline agro-meteorologists consider themselves primarily affiliated with, or whether they consider themselves agriculturalists or meteorologists. It also has much to do with how a national government organizes itself institutionally—in this case how jurisdiction over agricultural meteorology is divided between and assigned Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services v to national hydro-meteorological services and to ministries The public sector comes to mind first because so many of of agriculture and the extension services and national agri- the benefits generated by improved agro-meteorological cultural research systems they generally run. services are public goods. The public sector is, however, by no means the only prospective source of financial or tech- In this document, the author refers to the creation of nical support. This document also refers to the civil society financially sustainable weather observation networks in and nongovernmental organizations (NGOs) that operate Guinea and the Philippines that are likely applicable to within developing countries and to the pool of potentially many developing settings, including those of East Africa interested private investors, whether they are based in the and the Horn of Africa. The author also addresses a prac- country concerned or outside it. International develop- tical imperative that is familiar to virtually every initiative ment agencies like the World Bank are clearly interested that achieves a new level of investment and commitment parties as well. Everything we need is already available to by central governments allocating scarce public resources. us and should be used. This is the need to prove its worth quickly and decisively early on. Agricultural meteorology needs to generate and Juergen Voegele to demonstrate significant impacts to persuade higher lev- Senior Director els of government such as finance ministries and executive leadership that public financial commitments are good investments. vi Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services ACKNOWLEDGMENTS This report was commissioned by the World Bank Agriculture Global Practice under the Technical Assistance Project: Agroweather Tools for Adapting to Climate Change led by Ademola Braimoh. The report was written by John T. Snow with support from Nicole Robertson Van Every from the University of Oklahoma. We are grateful for the input of the “Agroweather Team”: Adul Aggrey, Tesfahiwot Dillnessa, Ijeoma Emenanjo, Peter Gitika, Fitsum Kidane, Francis Ngari, Carolyn Odicko, Idowu Oladele, Maria Ana de Rijk, and Simon Wambua. Jim Cantrell, Sarian A. Akibo- Betts, Beaulah Noble, Lansong Zhang, Lucy Kangaru’a, Sophie Nelly Rabuku, Hope Nanshemeza, Mistre Hailemariam, Mahlet Tsigie, and Elizabeth Karuoya provided valuable support to the team. The report benefited greatly from invaluable suggestions from peer reviewers. We would like to thank Peter Ambeje (Kenya Meteorological Services), Sofia Bettencourt, Vladimir Tsirkunov, Makoto Suwa, David Rogers, Lucy Hancock (Global Facility for Disaster Reduction and Recovery [GFDRR] Hydromet Team), Andrew Karanja, Teklu Tesfaye, Tim Robertson, Parmesh Shah (Agriculture Global Practice), Andualem Shimeles, Girma Mamo (Ethiopian Institute of Agricultural Research), and Robert Stefanski (World Meteorological Organization). We thank Diarietou Gaye, Guang Zhe Chen, Mark Cackler, Tijan Sallah, Severin Kodderitzsch, and Eija Pehu for assistance rendered at various stages of the project. We also thank Gunnar Larson for his untiring efforts in proofreading and editing the report. This activity was supported under the Bank Netherlands Partnership Program. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services vii ACRONYMS AND ABBREVIATIONS ATA Agricultural Transformation Agency JRC Joint Research Center of the European AWS Automated Weather Station Union CCAA Climate Change Adaptation in Africa JRC-IPSC JRC Institute for the Protection and Security CCM Certified Consulting Meteorologist of the Citizen CCU Climate Change Unit KARI Kenya Agricultural Research Institute CIMO Commission for Instruments and Methods of KMD Kenya Meteorological Department Observations KMS Kenya Meteorological Services CLPA Climate Prediction and Adaptation KRCS Kenya Red Cross Society CRDI International Development Research Centre MOU Memorandum of Understanding CSO Civil Society Organization NDVI Normalized Difference Vegetation Index DMCH Drought Monitoring Centre Harare NGO Nongovernmental Organization DMCN Drought Monitoring Centre with NHMS National Hydro-Meteorological Services headquarters in Nairobi NMA National Meteorological Agency EIAR Ethiopian Institute of Agricultural Research NOAA National Oceanic and Atmospheric FAO Food and Agriculture Organization (of the Administration United Nations) NWS National Weather Service FEWSNET Famine Early Warning Systems Network OU University of Oklahoma FIM Finnish Markka PAGASA Philippine Atmospheric, Geophysical, and FIR Far Infrared Astronomical Services Administration GDP Gross Domestic Product QC/QA Quality Control/Quality Assurance GIS Global Information System REST Relief Society of Tigray GMFS Global Monitoring for Food Security SMS Short Message Service GMT Greenwich Mean Time SRG Standard Rain Gauge GPS Global Positioning System TAHMO Trans-African Hydro-Meteorological HARITA Horn of Africa Risk Transfer for Adaptation Observatory ICPAC Climate Prediction and Applications Centre TMA Tanzania Meteorological Agency ICT Information Communication Technology US United States ICTP Inter-Centre Training Programme USAID U.S. Agency for International Development IDRC International Development Research Centre UV Ultraviolet IFC International Finance Corporation (of the VPI Vegetation Productivity Index World Bank Group) WBBGT Wet-Bulb, Black Globe Temperature IGAD Intergovernmental Authority on Development WFP World Food Programme ILRI International Livestock Research Institute WPF Weather Philippines Foundation IRI International Research Institute for Climate WMO World Meteorological Organization and Society Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services ix EXECUTIVE SUMMARY BACKGROUND Agriculture sustains humanity.1 The global population stands at some 7.3 billion as of 2014 (Population Institute 2014). Combined with rising incomes and changing consumption patterns and expectations about living standards typical of urbanizing populations throughout much of the developing world, along with population growth and rising demand for food and other agricultural products, agricultural production is severely challenged, locally, regionally, and globally. Further complicating this challenge are the approximately 860 million people living at or below the US$1.25 per day poverty line. As many as 75 percent of these individuals live in least-developed countries and work small plots for both subsistence and local markets. In these countries, agricultural production tends to be a proportionately sig- nificant driver of economic development. Most smallholder farms are less productive and less profitable than they could be, and the gaps between actual and potential yields are very large. Smallholders lack access to inputs such as modern seed and fertilizer and to services such as credit and information which are instrumental in enabling more modern farmers to manage risk and optimize production using more productive technologies and management practices. This lack of information and technical knowledge results in a skills gap that inhibits their adop- tion of these technologies and practices and reduces their efficiency when smallholders attempt to modernize their operations (World Bank 2007). Public extension programs tend to be underfunded and have limited contact with farmers in low-income coun- tries, and they are generally not effectively supported by agricultural research systems that generate applied knowledge. Supporting infrastructure is most often lacking as well, making coordination between the different actors in the supply chains linking farmers to consumers difficult and expensive, increasing costs and lowering revenues. 1 “Agriculture” is defined here as the art, practice, and science of farming, including cultivation of the soil for the growing of crops (food, animal feed, fiber, industrial feedstock, and fuels) and the rearing of livestock and poultry for meat, milk, eggs, hide/fur/leather, feathers, bone, sinew, wool, and dung. More broadly it also encompasses forestry, fisheries, and aquaculture. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services xi Agriculture is highly sensitive to weather and climate. In Applying meteorological information to enhance agricul- India, for example, annual productivity depends heav- tural productivity entails helping farmers to capitalize on ily on each year’s monsoon rainfall. A good monsoon opportunities presented by favorable weather and climatic year sees plentiful rainfall over wide areas and results in conditions, while working with them to mitigate, to the a bountiful harvest. A bad monsoon year, with limited extent practical, the impacts of adverse conditions. Agro- amounts of rain or rain limited to certain areas, results meteorological services are intended to enable producers in a poor harvest and often widespread food insecurity. to maximize their returns on investments in seed, seed- The great sensitivity of agriculture to weather in the short lings, animals, feed, fertilizers, and human effort while term and to climate in the medium to long term make it improving their ability to manage risk. the economic sector’s most vulnerable to climate variabil- ity and change.2 Agricultural meteorologists work with local extension agents, soil scientists, agronomists, experts in livestock Agricultural meteorology or “agro-meteorology” and poultry, hydrologists, and agricultural engineers to took shape as a field of applied science at the end of the monitor, interpret, and forecast the effects of weather nineteenth century, at the beginning of the global effort to and climate on plant distribution, crop yield, water-use apply scientific principles to improving agricultural pro- efficiency, phenology of plant and animal development, ductivity. Meteorological considerations enter into assess- the energy balance of managed and natural ecosystems, ing the performance of plants and animals whose growth and, conversely, the impact of vegetation on climate and is a result of the combined effect of genetic characteristics weather.3 Typical challenges include optimizing the tim- and their response to the local environment. The weather ing of planting and harvesting of various crops, properly is a major component of the local environment, which applying irrigation systems, and forecasting the emer- also includes soil properties, terrain, other plants, farming gence and onset of a wide range of plant and animal practices, and so on. For more than a century, agricultural pests and molds and fungi to ensure timely applications of meteorologists have worked to develop tools, techniques, insecticides and fungicides. Agro-meteorologists can also and practices that farmers can use to make agriculture provide forecasts in support of controlled burns and aid in more resilient. the management of wildland fires. On seasonal and interannual timescales, agricultural meteorologists assist farmers and government ministries 2 “Weather” refers to atmospheric phenomena that have timescales that of agriculture in forecasting the size and quality of yields range from hours to days to one or two months, whereas “climate” refers to atmospheric conditions that have timescales that range from a few months and in estimating the condition of winter crops. In recent to a season to a year to a decade or more, or even longer. In this sense, the years, supporting agricultural adaptations to seasonal and terms “weather” and “climate” identify regions along a continuous spectrum interannual variability—most often expressed in terms of of atmospheric conditions, weather describing rapidly changing events, and cli- warm or cool, wet or dry periods lasting through a season, mate describing slowly changing ones. Climate can be represented in terms of a normal, long-term average, and year-to-year fluctuations—the interannual to several years, to a decade—has also become an impor- variability—around which that average when viewed over a period of a few tant area of endeavor for agro-meteorologists. They seek hundreds of years, has fallen within a bounded “range” of values. Common to mitigate the negative impacts of seasonal and interan- drivers of climate variability are the oscillations that occur in Earth’s coupled nual weather risks by advising farmers on which crops to ocean-atmosphere system. An example is the El Niño and La Niña (ENSO) events, shifts of warm, tropical Pacific Ocean currents that can dramatically plant in coming seasons, the purchase of crop insurance, affect seasonal weather patterns around the world. Other drivers include vol- canic eruptions and solar phenomena. Sometimes climate varies in ways that suggest a component of randomness is inherent in Earth’s climate system. “Cli- 3 In the last 20 to 25 years, weather forecasts have come to routinely demon- mate change” is a long-term continuous change (either increase or decrease) strate high levels of accuracy in the 0- to 7-day range, and sometimes in as in a climate normal (for example, an increase in the long-term average tem- many as 10 to 12 days. Climate outlooks, such as seasonal forecasts, are far less perature) and/or the range of climate variability (for example, more frequent, accurate, but in many cases, such as when there is a strong planetary signal, more intense thunderstorms together with fewer small showers). As the range such as an El Nino or La Nina in the equatorial Pacific, are sufficiently reliable increases, the year-to-year variations in a variable such as temperature or pre- to be used in decision making and will generally show over a period of years a cipitation should be expected to be greater, and so new extreme values are likely. significant improvement in outcomes over chance. xii Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services changes in farming practices (for example, contour plow- vices to farmers and government officials will improve ing, building of dams), and so on.4 agricultural production and help build resilience in the face of increasing variability, from seasonal to decadal Much of the work of agricultural meteorologists is done timescales. While some advice is provided on how farm- with the same observations, numerical models, and anal- ers can use meteorological and climatological informa- ysis and forecast techniques used by general meteorolo- tion in their operations, this is not the main thrust of the gists in a national hydro-meteorological service. However, report. It focuses instead on the generation and provision many agro-meteorological services require special obser- of such information in a timely manner and in formats vations in the fields, crop and ecological models, and tai- that inform decision making by farmers and by ministe- lored techniques for purposes of analysis and forecasting. rial staff. This is intended to support the Agroweather Agro-meteorologists therefore require not only a sound Information project in increasing the adaptive capacity knowledge of applied meteorology and climatology, but of farming communities in Kenya and Ethiopia. The also an understanding of agronomy, plant physiology project seeks to improve farmers’ access to relevant infor- and plant and animal pathology, and common agricul- mation on weather and climate, and to develop farm tural practices. The World Meteorological Organization management capabilities in a context of climate change. (WMO) has recognized the need for specialized training The project will also raise awareness of the practical util- of agricultural meteorologists through the publication of ity of agro-weather information products and improve detailed guidelines for their training (WMO 2009). extension services. THIS REPORT For these purposes, six types of activities were undertaken: This study was undertaken in support of the World Bank 1. Assess the current level of agro-meteorological project, Agroweather Tools for Adapting to Climate observation networks and monitoring capacity vis- Change. The overall goal of this pilot project is to estab- à-vis international standards and suggest strategies lish community-based agro-weather risk management for improvement. tools. These tools are to be supported by a flow of weather 2. Review a number of traditional and nontradi- and climate information via information and communi- tional agro-weather observation systems and their cation technology (ICT) delivery systems. The project has relevance. short-, medium-, and long-term objectives. In the short 3. Identify adoption barriers associated with tradi- term, the project will raise awareness of the practical util- tional and nontraditional approaches and how ity of agro-weather information products and services and they may be overcome. of their direct application to farmers’ needs and concerns. 4. Quantify the costs and benefits of the traditional It will seek and incorporate feedback from partners as the and nontraditional approaches, and provide ex- specific products are developed. In the medium term, the pert advice on cost-effective strategies for upscal- project will build a community-based platform to facilitate ing agro-meteorological observation systems. better access to relevant agro-weather information and 5. Identify business models, including innovative decision support systems. In the long term, it will expedite public-private partnerships and key private sector the flow of products, services, and information to decision players, that can ensure the sustainability of the makers at national and local levels, and will ultimately be agro-meteorological observation systems. instrumental in increasing food supply, increasing income, 6. Make recommendations on global best practices and sustaining the health of the environment. for institutional strengthening and capacity build- ing to support development and testing of new This report is based on the premise that the provision principles and approaches. of meteorological and climatological products and ser- These activities cover a broad range of issues, some tech- 4 Here we take the word “mitigate” to mean “to lessen” or “to make less severe.” One can mitigate risks but never eliminate them. As opposed to “acceptance,” nical, others dealing with policy and procedure, and still “mitigation” is inherently proactive. others dealing with government organization. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services xiii While World Bank staff are the report’s primary intended The recommendations presented in the report are more audience and will determine its ultimate distribution, the indicative than definitive and are intended to provide document is likely to be of interest to people at the World general advice about guiding the transformation of agro- Meteorological Organization, national hydro-meteoro- meteorology in this part of Africa. More detailed on- logical services, and ministries of agriculture and environ- the-ground surveys and fact-checking will be required to ment (well beyond those in Ethiopia and Kenya), among inform decision making regarding policy, investment, and others. operations. xiv Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services CHAPTER ONE AGRO-METEOROLOGICAL OBSERVATION NETWORKS While the national agricultural meteorology programs in Ethiopia and Kenya are very limited, a number of international development agencies and NGOs are operating in both countries and have been implementing various agro-meteorological services as components of more general initiatives to improve agricultural productivity and resilience. Although a comprehensive review of all of the actors at work in the area of agro-meteorology and agro-climatology in the two countries was far beyond the remit or the resources of this study, a number of the more prominent agencies and organiza- tions at work in the field are described in this chapter. IGAD CLIMATE PREDICTION AND APPLICATIONS CENTRE The countries of East Africa and the Horn of Africa are prone to extreme weather and climate events such as floods and droughts which have severe impacts on agricul- tural production. In 1989, 24 countries in eastern and southern Africa established a Drought Monitoring Centre with headquarters in Nairobi (the DMCN) and a subcen- ter in Harare called the Drought Monitoring Centre Harare (DMCH) in response to devastating weather-related disasters. In October 2003, the heads of state and govern- ments of the Intergovernmental Authority on Development (IGAD) held their 10th Summit in Kampala, Uganda. The DMCN was adopted as a specialized IGAD insti- tution and renamed the IGAD Climate Prediction and Applications Centre (ICPAC) in order to more fully reflect all its mandates, mission, and objectives within the IGAD system. A protocol integrating the institution fully into IGAD was signed on April 13, 2007. Today, the center is responsible for 11 member countries: eight IGAD member countries, namely, Djibouti, Eritrea, Ethiopia, Kenya, Somalia, Sudan, Republic of South Sudan, and Uganda, as well as Burundi, Rwanda, and Tanzania. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 1 ICPAC intends to become a regional center of excellence accessible to a wide variety of users. The information in climate prediction, producing applications for climate presented goes into substantial detail regarding such mat- risk management, environmental management, and sus- ters as measurement data, observing equipment, and the tainable development. Its mission is to provide timely layout of meteorological observatories (box 1.1). Some of early warning weather information that enables its mem- the observing equipment is locally fabricated in the KMS ber countries to cope with various risks associated with calibration laboratory. climate variability and change. Its responsibilities extend to environmental management, sustainable development, From the reported data, the Agro-meteorological Section and poverty reduction. of the KMS produces periodic Weather and Crop Reviews, such as the Review for Dekad 7 (appendix B) (figure 1.1). ICPAC produces a variety of climate status reports, weather forecasts, and climate outlook products, including Many of the instruments and observing approaches used 10-day, monthly, and seasonal climate and weather bul- by KMS are traditional. Data are collected for 10-day letins, climate watch and El Niño updates, and annual cli- periods (dekads) and then reported to the Agro-meteoro- mate summaries. It also provides Normalized Difference logical Section of KMS. No direct or systematic interac- Vegetation Index (NDVI) data, using satellite imagery of tion with farmers is reported. The KMS website suggests land-cover vegetation from the National Oceanic and that the KMS is primarily in the data collection business, Atmospheric Administration (NOAA). ICPAC’s “custom- with little attention given to products and services for ers” are national hydro-meteorological services and vari- delivery to farmers.5 ous ministries of the governments in the region. The KMS also produces longer-range quarterly Outlooks containing agro-meteorological information. These Out- KENYA looks are disseminated by press release and then sent to The agriculture sector is the largest contributor to Kenya’s government ministries. Farmers are advised to remain in gross domestic product. In 2005, agriculture, includ- liaison with the Ministry of Agriculture for recommenda- ing forestry and fishing, accounted for about 24 percent tions about appropriate crops based on the most recent of the gross domestic product (GDP), as well as 18 percent information about upcoming rains. The Outlooks are posted of wage employment, and 50 percent of export revenue. to the KMS Facebook page where they are easily accessi- Kenya is a leading producer of tea and coffee, and the ble to farmers. However, the primary audience appears to third-leading exporter of fresh produce, such as cabbages, be the Ministry of Agriculture that uses the information to onions, and mangoes. Small farms grow most of the corn promote foot security among other purposes. and also produce potatoes, bananas, beans, and peas. An estimated 15 to 17 percent of Kenya’s land area has suf- Kenya swings between drought and flood virtually every ficient fertility and rainfall to be farmed, with 7 to 8 per- year. Because much agricultural land, along with towns cent high-quality farmland. In 2006, almost 75 percent and villages, are located along Kenya’s streams and riv- of working Kenyans made their living by farming. About ers, flooding can have major impacts both locally on indi- one-half of Kenya’s total agricultural output is nonmar- vidual farmers and nationally on the nation’s food supply. keted subsistence production. (Maps in figure 1.2 illustrate major floods in Kenya.) Heavy flooding in March and April of 2013 was the subject of At the top of its list of responsibilities, the Kenya Mete- orological Service (KMS) gives the “[p]rovision of mete- orological and climatological services to agriculture, 5 A reviewer reported that there is an ongoing devolution process in KMS forestry, water resources management, civil aviation and with respect to its agro-meteorology service. Reportedly, KMS has established the private sector including industry, commerce and pub- 47 county offices in this devolution process. These offices are currently staffed only with a head, but the plan is to dispatch one hydro-meteorologist and one lic utilities for the better exploitation and utilization of agro-meteorologist to work at each county office. The county office will work natural resources for national development.” The KMS with county government to provide more geographically tailored info. The aptly posts its information resources online, making them author was unable to find any further information on this effort. 2 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services BOX 1.1. AGRO-METEOROLOGICAL STATIONS In Kenya there are two types of stations: GRADE A and These observations are made in 32 stations spread through- GRADE B. Grade A stations are operated and manned by out the country. In addition to the above, the following crop Kenya Meteorological Department (KMS) staff, while Grade B data are obtained from the 13 agro-meteorological stations stations are run by other organizations, for example, Ministry of (Grade A and B): Agriculture, universities, and agricultural research institutions. • Variety of the grown crop • Stage of development attained by the crop Currently there are 13 agro-meteorological stations in the • General assessment of crop performance country (11 Grade A and 2 Grade B). • Damage by pests, diseases, and adverse weather Types of Observations • State of weeding in the farm In all of these stations, normal meteorological parameters • Plant density are measured on a daily basis and the data are conveyed to • Soil moisture the Agro-meteorological Section of KMS after every 10 days. Expected yield (assessed visually), is normally observed at the The data include the following: end of each 10 days and along with the meteorological data is • Air temperature in degrees Celsius (maximum, mini- communicated to the agro-meteorological section to facilitate mum, wet bulb, dry bulb, dew point) crop-weather impact analysis. • Soil temperature in degrees Celsius at 5-, 10-, 20-, 30-, 50-, 100-centimeter depths In order to obtain a general overview of crop performance in • Sunshine duration in hours the country, especially on the main staple crops in the country— • Radiation in megajoules per square meter namely, maize, beans, and wheat—all 32 stations report on the • Wind speed in meters per second at 2-meter height. stage of crop development, general assessment of crop perfor- • Calculated relative humidity (%) at 0900 hours and mance, and yield expected (visual) from the farmers’ farms on 1500 hours the basis of what they see from nearby farms and oral interviews • Pan evaporation in millimeters per day with farmers they come across from areas far from their reach. • Calculated potential evapotranspiration in millimeters per dekad (10-day period) • Rainfall in millimeters per day Source: Kenya Meteorological Department (KMD). Undated. FIGURE 1.1. ILLUSTRATION IN A KMS REVIEW Source: Dekad 7 (March 1–10) 2014 Rainfall Distribution. Extracted from Kenya Meteorological Service Dekadal Agro-Meteorological Bulletin Weather and Crop Review for Dekad 7. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 3 FIGURE 1.2. RECENT FLOODS AND IMPACT IN KENYA Source: United Nations Office for the Coordination of Humanitarian Affairs and Kenya Red Cross, January 11, 2013. an article by Thomson Reuters Foundation correspondent way toward this purpose. These include the continued Katy Migiro in which Kenya Red Cross Secretary Gen- development of RANET-Kenya for getting information eral Abbas Gullet was quoted as follows: “For the farming out to rural areas via radio and Internet. KMS has also communities that have lost their livelihoods, they will need started a pilot effort of roving agro-meteorological semi- assistance for the next three to six months. The pastoralist nars with support from the WMO and the U.S. Agency communities that have lost their animals also would need for International Development (USAID). While this will assistance, so another vicious circle” (http://www.trust.org cover only a limited number of communities during the /item/20130425095508-c7hi9?view=print). Gullet went on specified pilot period, the seminars are expanding aware- to lament the fact that virtually all of the rainwater from ness of the usefulness of agro-meteorological information. the floods poured into the Indian Ocean and Lake Victoria, Agro-meteorology is also represented in Care Internation- rather than being stored. During the following dry season, al’s Participatory Scenario Planning efforts and in ICPAC drought would set in and aid agencies would spend millions programs supported by the Rockefeller Foundation. of dollars trucking water to many of the same areas that experienced floodwaters. The Institute for Meteorological Training and Research (IMTR) offers a six-month course in Applied KMS recognizes the need to expand more into the area of Agricultural Meteorology covering bio-meteorological agro-meteorology, and a number of initiatives are under interrelationships, agro-meteorological observations and 4 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services measurements, soil science and agro-ecological zoning, KENYA—NONGOVERNMENTAL crop and animal production, agronomy, statistics in agro- ORGANIZATIONS AND OTHER PARTNER meteorology, ICT, and remote sensing. AGENCIES Outside the agencies of the Kenyan government, impor- The Kenya Agricultural Research Institute (KARI) tant work in agro-meteorology is being done by inter- is Kenya’s national institution for research and develop- national development agencies and nongovernmental ment in food crops, horticultural and industrial crops, organizations (NGOs), often in partnership with the livestock and range management, land and water man- KMS. Here we discuss a few of these. agement, and socioeconomics. KARI has been develop- ing weather- and climate-related technologies to help The Syngenta Foundation for farmers cope with the difficult conditions found in many Sustainable Agriculture parts of Kenya, conditions that are being exacerbated by The Syngenta Foundation for Sustainable Agriculture has increasing interannual variability. With support from the a number of regional divisions carrying out projects to Rockefeller Foundation, the KARI Climate Change Unit increase crop productivity, protect the environment, and (CCU) was established in 2010. The Unit facilitates col- improve health outcomes across East Africa. Its projects laborative work on regional climate change science and include teaching farmers how to conserve water through adaptation and mitigation mechanisms, and engages in minimum tillage and rainwater harvesting, and providing capacity building for researchers working on these issues. microfinance solutions to growers. Syngenta also provides Activities relate to science and theory, assessments, inte- thousands of Kenyan farmers with index-based weather grated soil fertility management, conservation tillage, insurance coverage through the Kilimo Salama (“Safe farm- rainwater harvesting, pest management, weed control, ing” in Kiswahili) program. The coverage targets two and carbon sequestration. CCU’s mandate also includes main types of farmers: larger-scale barley, wheat, coffee, developing or adopting technologies and methods that maize, export ornamentals, and vegetable growers; and can be applied to cope with climate change impacts on smallholder and subsistence farmers who grow mostly food insecurity, water scarcity, and the emergence of pests for local consumption. Kilimo Salama is a partnership and diseases. between Syngenta and UAP Insurance Kenya and tele- communication operator Safaricom. As a result of these initiatives by KARI, traditional farmers have been informed by timely information about the tim- Kilimo Salama uses automated, solar-powered weather ing of the two annual rainy seasons, helping them to effec- stations and a low-cost, mobile phone payment and data tively plan farming activities throughout the year. This has system to offer thousands of farmers in parts of western become more important in recent years, as droughts and and central Kenya affordable “pay-as-you-plant” crops. dry spells have become more frequent, from once every The weather stations were installed and maintained by ten years or so to as often as once every three years. While Syngenta, while being overseen by the Kenyan Minis- many Kenyan radio stations issue weather forecasts, only try of Agriculture, which served as the neutral party as a few provide practical information directed at farmers. required by reinsurers. KARI identified a need to provide combined information from both the meteorology department and agricultural When weather data transmitted over Safaricom’s 3G data experts. The strategy is to provide a detailed weather fore- network from a particular station indicates that drought, cast around one month before each planting season which excessive rains, or other extreme conditions have occurred tells farmers how much rain is predicted, when it will start, and likely crippled crops, all farmers registered with that and therefore what types of crops to plant. When rainfall station automatically receive payouts directly via Safari- predictions are good, they can invest in hybrid seeds and com’s M-PESA mobile money transfer service. The pay- fertilizer, and maximize their yields. When predictions outs compensate farmers for losses to crops like maize, are for a dry season, choosing drought-resistant crops is wheat, beans, and sorghum, with credits for fertilizer and a better option. seed to help them restart farming after a loss. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 5 “Extreme weather, particularly drought, traps many Afri- BOX 1.2. RWANDAN FARMERS TO BENEFIT can farmers in poverty because it robs them of the means FROM INSURANCE AGREEMENT to recover,” said Marco Ferroni, Executive Director of the Syngenta Foundation (http://www.eurekalert.org Rwandan small-scale farmers will be part of 1 million farmers in East Africa to insure their crops against unpre- /pub_releases/2010-03/bc-fmp030210.php). “We have in dictable weather and other related problems. Kilimo Salama a microinsurance strategy that will work. By utilizing state-of-the-art risk management tools, revolu- This follows the signing of two grants agreements of tionary mobile phone technologies, and the knowledge and $3.9 million (about Rwf 2.6 billion) signed on February 5, 2014, in Nairobi by International Finance Corporation (IFC) expertise of farmers and rural business men and women, and Syngenta Foundation for sustainable agriculture. The we have developed for the first time a model for providing agreement would enable expansion of index-based insurance farmers with reliable, low-cost cover from the vagaries of to small-scale farmers in Rwanda, Kenya, and Tanzania. extreme weather.” The grants were issued by the Global Index Insurance Facility, a multidonor trust fund financed by the European “We believe Kilimo Salama can revolutionize insurance Union, Japan, and the Netherlands and implemented by and make it accessible to farmers,” said James Wambugu, IFC and the World Bank. It is estimated that the project Executive Director of UAP Insurance Kenya. “By using will bring index insurance to approximately 1 million the weather stations to verify local weather conditions, we small-scale farmers within two years. are avoiding claims procedures that have created mistrust Marco Ferroni, the Executive Director of Syngenta Foun- and led people to avoid insurance. As such, this strategy dation for Sustainable Agriculture, said Syngenta has has the potential to make agricultural micro-insurance developed a product dubbed “Kilimo Salama,” being affordable and attractive for smallholder farmers and eco- administered by selected insurance firms in the region. nomically viable for insurance companies in developing In Rwanda, for instance, Syngenta Foundation is offering the countries that had previously written off the agricultural product in partnership with SORAS Insurance Company sector” (Red Orbit 2010). Originally established in Kenya and the international reinsurer, Swiss Re Corporate Solutions. and Tanzania, this index-based insurance was extended “Our aim through this initiative is geared towards enhanc- to farmers in Rwanda in February 2014 (box 1.2). ing food security as well as minimizing poverty among other sufferings to majority of farmers in the region,” said Fer- Discussion. While weather index agricultural insurance roni, noting that the insurance product is accessible to all has a number of attractive features, it can give rise to a categories of farmers—small-scale and large-scale farmers. number of issues as well, and some of these are present in In late 2012, the government of Rwanda in partnership the Syngenta effort in Kenya. The Syngenta Foundation with the Syngenta Foundation for Sustainable Agriculture is effectively the same organization as Syngenta, a major introduced Kilimo Salama, the scheme that uses satellite supplier of seed and agricultural inputs. Syngenta weather stations to monitor how rainfall variability and drought index insurance is sold as a part of the seed and agricul- affect crop production. tural inputs supply chain. Syngenta is also supplying the So far, many Rwandan farmers have gained from the crop meteorological observations in Kilimo Salama project, insurance product; with a recent example of 7,086 maize in which they only work with Safaricom, one of several farmers from Bugesera, Ngoma, and Kirehe districts who mobile operators in the region. Participating farmers are received their insurance payout totaling Rwf 42.1 million last Tuesday. therefore limited in their options to Syngenta and Safari- com products. This suggests the potential utility of a func- Source: Ministry of Rwanda Ministry of Agriculture and Animal Resources. tioning regulatory framework to foster a more competitive marketplace in which farmers have more alternatives avail- able to them in terms of product and service providers. livestock insurance program and the World Food Programme (WFP) R4 initiative, which will be discussed below. How these Other insurance-related programs at work in the region or Syngenta’s initiatives relate to the activities of the National include the International Livestock Research Institute’s (ILRI) Hydro Meteorological Service (NHMS) is unclear. 6 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services Weather index insurance programs appear to be a promis- ing area for public-private partnerships, given for instance ETHIOPIA the strong interests that private insurance companies have Agriculture accounts for over 40 percent of GDP and in high-quality forecasts and observations. 80 percent of exports, and engages 80 percent of the labor force in Ethiopia. The country has vast areas of fertile land, a diverse climate, generally adequate Trans-African Hydro-Meteorological rainfall, and a large labor pool. Around 75 percent of Ethiopia’s approximately 85 million people are Observatory dependent on subsistence agriculture, which is almost Another NGO promoting agro-meteorology in Kenya entirely rain fed and small scale. Both farmers and and elsewhere in Africa is the Trans-African Hydro- pastoralists are highly dependent on the weather for Meteorological Observatory (TAHMO), a collaboration their livelihoods. between Oregon State University and Delft University of Technology. The TAHMO seeks to build a dense network—a station every 30 kilometers—of hydro- Coffee, tea, and, more recently, flowers are the only major meteorological stations across Sub-Saharan Africa. This commercial crops. Ethiopia is Africa’s second biggest effort recognizes that accurate, high-resolution temporal maize producer. Its livestock population is believed to be and spatial hydro-meteorological data are essential to the largest in Africa, accounting for 10.6 percent of Ethi- supporting agriculture, combating diseases, and enabling opia’s export income in 2006/2007—with leather and weather prediction and climate modeling (box 1.3). leather products making up 7.5 percent and live animals 3.1 percent. Its first step has been an effort to catalog and make acces- sible all the currently collected weather and climate data for Sub-Saharan Africa to encourage climate analyses, ETHIOPIA—GOVERNMENT AGENCIES avoid redundancy in future TAHMO hydromet station The National Meteorological Agency installations, and prioritize the coming years of effort. Ethiopia’s National Meteorological Agency (NMA) oper- ates an extensive observing network, with four categories The long-term goal of a station every 30 kilometers of observing stations: entails the production and deployment of some 20,000 » First Class—22 Synoptic stations where such stations. It is TAHMO’s intent to utilize innovative meteorological observations are made for the sensors coupled with rapidly evolving information and purposes of synoptic meteorology. Observations communication technologies, for example, cell phones, so are taken of 18 atmospheric variables every hour, that each station costs no more than $200. The stations 24 hours a day. will be placed at schools and universities and integrated in » Second Class—>150 Principal or Indicative local educational programs. Data from the stations will be stations where meteorological observations are combined with those from models and satellite sensors to made for climatological purposes. Observations obtain detailed insight into the distribution of water and are taken of 13 atmospheric variables every three energy stocks and fluxes. hours at the following Greenwich Mean (GMT) times (0300, 0600, 0900, 1200, 1500 GMT). TAHMO has built and demonstrated a prototype acous- » Third Class—Ordinary stations where three tic disdrometer (rain gauge) that can be produced for €10, meteorological elements are observed: maximum less than 1 percent of the cost of a commercial equivalent air temperature in a day, minimum air temperature, with similar specifications (Hut 2010). The disdrometer and total rainfall. Observations are taken at 0600 and was developed in the Netherlands and tested in Tanza- 1500 GMT. nia for a total project cost of €5000. TAHMO has been » Fourth Class—Rainfall Recording stations seeking out other innovative ideas for instrumentation via where the total rainfall amount over a 24-hour period open design competitions in Kenya. is observed. Observations are taken at 0600 GMT. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 7 BOX 1.3. IMPROVED ACCESS TO WEATHER INFORMATION LEADS TO BETTER FARMING OUTCOMES “When every drop of rain counts: Managing climate and the University of Nairobi has made significant inroads in risks in the Greater Horn of Africa.” Ochieng Ogodo study sites in three Kitui county locations: Kaveta village in In her small vegetable plot in Kenya’s Kitui county, Mary Central district, Kyome village in Migwani district, and Kitoo Mueni Samson scrutinizes a rain gauge and carefully notes the village in Mutomo district. According to William Githungo of results. It is the middle of one of the worst droughts to hit East the Kenya Meteorological Department, the project has made Africa in recent decades. And though Kitui has been spared an important contribution by integrating climate informa- the worst of it, Samson’s modest harvest of green grams is due tion with knowledge on dry land agriculture techniques—and in no small measure to her close eye on the weather. helping farmers translate this knowledge into their decisions. In surveys conducted in 2010, over 80 percent of farmers “I read our rain gauge every day and keep records,” she says. involved in the study rated the improved information so useful “In the past, weather forecasting was a distant drum to many that they would be willing to pay for it. “Evaluation,” he says, of us, but it now plays a tune we understand well. It helps us “has shown that farming decisions based on forecast informa- know the amount of rains we get in a season and this helps tion may contribute substantially to reduce risk and improve us in planning.” productivity and profitability.” Samson’s skill and knowledge in using weather data to plan Adapting farming techniques in Kenya her activities is an important strategy for coping in this already Initial research stages with the Uuniko Farmers Field School dry region, where rains are less and less dependable. Along involved consultation between forecasters, researchers, and with many of her neighbors, she has learned to gather data farmers. The team then focused on testing farming approaches and apply forecasts to her farming decisions as a member of and technologies of greatest interest to local communities. Each the Uuniko Farmers Field School. season, they set up demonstration plots: one set served as an experimental control, and was cultivated using traditional farm- Erratic weather, little rain ing methods; in another, water and soil fertility management According to Edward Nduli, secretary to the field school, life techniques—such as terracing fields, using fertilizers, and cus- has been a struggle for decades in this part of Kenya, but tomizing the selection of crop and seed varieties—were chosen things were never as bad as they are today. in light of seasonal forecasts, and input from farmers. When much better yields resulted from enhanced farming practices “For a region where the majority of us are small-hold farm- tailored to climate data, the school began using these strategies ers who depend on crop production for our livelihoods, the as a group. Rainwater run-off that once went to waste, for exam- advent of unpredictable seasonal variations visited upon us ple, is now being diverted to gardens. The community is also serious challenges for survival,” he says. The last good rains constructing ground water tanks as a means to harvest water he can remember were back in 1997. But thereafter, he says, for dry spells. Households have learned effective ways to use “We saw the decline of rains with extreme erratic seasons, animal manure and crop residues to nourish their soils, saving and this has continued since. [Before,] we knew short rains valuable money on fertilizers. Farmers have seen a significant from October to December, while the long ones came March increase in harvests, even with little rainfall. With more water to May,” says Nduli. “But the years changed and yields came available from harvesting techniques, members of the group are down significantly… we started relying on relief food from practicing agroforestry with fruit trees and eucalyptus. Nutri- local NGOs and international organizations like Red Cross, tious grains like sorghum, green gram, lablab beans, and finger Panda, and (German aid agency) GTZ.” millet are moving to the center of farming activities because of their drought tolerance. As a result of these efforts, says Nduli Information worth paying for proudly, “We are no longer major receivers of relief food.” In Kenya, a partnership with the national meteorological department, the International Crops Research Institute for the Source: CCAA Human Impact Stories. IDRC/CRDI website. (Accessed Semi-Arid Tropics, the Kenya Agricultural Research Institute, November 14, 2014.) It is unclear how many total stations the NMA has in oper- Common comments about the data from this assembly of ation, but numbers between 400 and 1,200 are reported stations are that (1) while the number of stations is reason- in the literature (see figure 1.3). Most of these are likely ably good, coverage is uneven, and is particularly sparse in the Ordinary or the Rainfall Recording stations. lowlands, with greater concentration in the central parts; 8 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services FIGURE 1.3. DISTRIBUTION OF AGRO-METEOROLOGICAL STATIONS IN ETHIOPIA Source: National Meteorology Agency, Ethiopia. (2) stations are marked by data gaps; and (3) data from any support for agricultural research and development on some stations may not be easily accessed in part because its website, and linkages to the Ministry of Agriculture or records are a mixture of paper and digital materials. to NGOs working in Ethiopian agriculture appear to be limited. Dekadal, Monthly, and Seasonal Bulletins issued Data from this network are transmitted to the NMA’s central by Agormet are posted in highly accessible PDF format office for analysis. The information produced subsequently for easy dissemination. by NMA includes daily weather reports, 10-day weather summaries and 10-day forecasts, monthly weather sum- Extracts from the NMA dekadal report for February 11 maries and forecasts, and three seasonal predictions a year through 20, 2014, are presented in figures 1.4 and 1.5. covering the two rainy seasons and the dry season, known as Bega. The service also provides maps showing rainfall The Rockefeller Regional Project on Agro-met advisory received as a percentage of normal rainfall, vegetation con- to farmers—Training of Trainers on Weather and Cli- ditions, and impacts on crop and livestock production. mate Information and Products for Agricultural Exten- sion Services in Ethiopia, began in 2010. While NMA is Other items under “Data Service” include the “NMA Data the lead government organization, the Ministry of Agri- Policy,” “NMA Service Charges,” “Dataset and Informa- culture and its Rural Extension Service also play central tion,” and the “Data Request Form.” These describe the roles. The WMO’s Agricultural Meteorology Programme policies for those outside NMA who need specific data of the Climate Prediction and Adaptation (CLPA) Branch sets. Only a limited number of top government agencies of the Climate and Water Department has been a key and researchers collaborating with NMA can routinely supporting partner, responsible for the implementation of receive data for free. The NMA makes little mention of many aspects of this program. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 9 FIGURE 1.4. RAINFALL DISTRIBUTION (IN MILLIMETERS) FOR FEBRUARY 11–20, 2014 Source: National Meteorology Agency, Ethiopia. FIGURE 1.5. OBSERVED VERSUS NORMAL RAINFALL (FEBRUARY 11–20, 2014) Source: National Meteorological Agency, Ethiopia. Note: This plot shows the observed versus normal rainfall for the period: <50 percent, much below normal (gray); 50 to 75 percent, below normal (yellow); 75 to 125 percent, normal (light green); >125 percent, above normal (dark green). This Ethiopian program was based on the experience The goal is to exploit Ethiopia’s extensive observing net- and success of the WMO Roving Seminar Programme work in ways that provide better practical knowledge of in West Africa and other African countries, in particular agro-meteorological services and applications to farm- Mali, on the application of agro-meteorology to directly ers in order to improve farming practices and increase assist farmers in making operational decisions. or secure agricultural production. To meet this goal, the 10 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services Rockefeller Foundation supported NMA to engage with government agricultural research institutions (both fed- the Ministry of Agriculture and provide training to agri- eral and regional), operate in the field as well, in addition cultural extension agents and agricultural experts and to to universities and agricultural technical and vocational assist them in providing better practical knowledge of training colleges. Some, perhaps most, of these research agro-meteorological services and applications to farmers. and academic establishments have agro-met divisions or departments as well as agro-climatic research and studies. Based on these experiences and with feedback from the They also have meteorological stations, some of which are Ministry of Agriculture, the objectives of this project are under the direct control of NMA. twofold. One objective involves “training the trainers” in meteorological and agricultural agencies to familiarize Ideally, the assessment would have covered these research agricultural extension services with the use of weather and academic institutions, providing a more complete pic- and climate information that NMA provides in opera- ture of the current level of agro-meteorological observa- tional farm management. The other objective involves tion networks and monitoring capacity across the country.7 discussing with farmers the information they need to Unfortunately, locating these various institutions provided make decisions, which is the heart of the training and thus to be difficult as many have a limited Web presence and closes a feedback loop. Around 26,000 extension agents little of their work is documented in accessible English- work in some 600 administrative districts in Ethiopia—a language publications, so their (potentially significant) number well beyond the project’s ability to reach. Train- contributions remain obscure. ing of trainers therefore selectively targets 30 extension agents in each of the country’s 10 regions.6 The WMO seminars also increase the interactions between the local Ethiopian Institute of Agricultural agricultural extension services and the local NMA staff. Research (EIAR) This is crucial for NMA to provide better services for the The EIAR is a semiautonomous federal institution that agricultural community. in the last three years has taken the lead in reaching out to farmers regarding agro-weather and agro-climate research and extension. Its Division of Agrometeorology, As a result of these efforts, the Ethiopian Agricultural Biometrics, and GIS Research has been engaged in a wide Research Institute and Agricultural Transfor- range of climate research aimed at enhancing the applica- mation Agency have already put in place the use of tion of climate information by smallholder farmers. Since agro-meteorological advisories over selected districts in 2010, that division of EIAR has been piloting tailored Ethiopia. Several reports about this program note that decision support tools in the form of agro-weather and working closely with farmers is important to determine agro-climate advisories to farmers in the four major agri- and then service their agro-meteorology information cultural states of Ethiopia—Amhara, Tigray, Oromia, needs and to support adaptation to changing climate, and and the Southern region. so enhancing food security. This climate change adaptation initiative, Making Ethiopian The report focused almost exclusively on NMA, pre- Agriculture Climate Resilient: Towards Networking and Coordina- senting it as the governmental organization dealing tion to Mainstream Climate Change Adaptation into Food Security with agro-meteorology and agro-climatology. While and Sustainable Development, has been funded mostly by the NMA is nationally mandated to deal with meteorologi- Rockefeller Foundation, with a significant contribution cal issues and has a network of meteorological stations from the government. This project has different compo- dispersed throughout the country, a number of other nents such as climate modeling, impact and adaptation assessment, and climate mainstreaming. The component 6 A reviewer noted that these statistics are likely somewhat out of date; they are the sta- tistics cited in the documentation for the program. The reviewer indicates that the total 7 A reviewer noted that these diverse research and academic institutions have number of extension workers, often referred to as “development agents” (DA) is more significant agricultural monitoring capacity because they are better staffed than than 60,000, and the actual number of kebele administrations is more than 1,000. NMA. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 11 piloting tailored climate products to farmers was among selected districts in Ethiopia. The agency has found that the major ones. This was a phase one project and was working closely with farmers is important for determin- completed last year. ing and then servicing their agro-meteorology information needs and for supporting adaptation and food security. A second phase project with additional funds from the Rockefeller Foundation was launched in 2013, Reducing the NONGOVERNMENTAL ORGANIZATIONS Vulnerability of Smallholder Farmers in Ethiopia to Climate Variabil- AND OTHER DONOR AGENCIES IN ity and Change through Application of Climate Information and Best ETHIOPIA Bet Management Practices. This second phase has components In Ethiopia, at least five international NGOs are involved addressing different aspects of agro-advisories to farmers. in climate monitoring and yield estimation in Ethiopia.8 These include Global Monitoring for Food Security In addition, the EIAR has been working together with (GMFS), the Agriculture Unit of the Joint Research regional and international partners like the Association Center of the European Union, Institute for the Pro- for Strengthening Agricultural Research in Eastern and tection and Security of the Citizen (JRC-IPSC), the Central Africa (ASARECA), the International Maize and Famine Early Warning Systems Network (FEWSNET), Wheat Improvement Center (CIMMYT), the Interna- the World Food Programme (WFP), and the Food and tional Crops Research Institute for the Semi-Arid Trop- Agriculture Organization (of the United Nations) (FAO). ics (ICRISAT), the Inter-Centre Training Programme The GMFS, the JRC-IPSC, and the FEWSNET rely on (ICTP), and so on, on the issue of climate-induced remote-sensing imagery in their crop monitoring, using risks in agriculture. The EIAR also operates 30 agro- the analyses of crop and vegetation responses derived meteorological stations established in its research centers from NDVI or vegetation productivity index (VPI) imag- and subcenters. ery and water balance modeling based on rainfall esti- mates or similar products as well as outputs from global Ethiopian Ministry of Agriculture climate models. The Ethiopian Ministry of Agriculture has branch offices and agents posted in each administrative unit across the The FAO and the WFP are extensively involved in country, from the federal level down to peasant associa- Ethiopia’s crop and food supply assessment mission. tions. The presence throughout the lowest levels allows the The information generated is used by government and Ministry to collect data to estimate crop yield and produc- international agencies for the contingent planning in tion forecasts by direct observations of fields using teams humanitarian assistance, then in routine development of contract farmers. This information is rolled up through plans through technology extension services. It is only the national agricultural bureaucracy to the regional and WFP, in collaboration with the agriculture ministry and federal levels, where dekadal and monthly bulletins are NMA, that has developed a weather-based risk financing prepared and disseminated. These contain information tool known as Livelihood Early Assessment and Protec- on farming activities, rainfall, and crop status, and provide tion (LEAP), of which the Water Requirement Satisfac- an early warning of possible crop failures. tion Index is a part. A handful of Ethiopian universities have only recently Agricultural Transformation Agency come to include climate and climate change in their cur- The Agricultural Transformation Agency was recently ricula, including postgraduate programs in meteorology established as a federal government organization to sup- offered by Arba Minch, Mekele, and Haramaya Uni- port the transformation of agricultural practices on the versities. This leaves relatively few technically capable many smallholder farms in Ethiopia. The agency has ini- tiated the Climate and Environment for the Sustainable Development program. Under this program, it has already 8 “Support to Food Security Information Systems in Ethiopia.” Mid-Term Eval- put in place the use of agro-meteorological advisories over uation Report by Vianney Labé. 12 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services individuals who can translate weather and climate infor- attempts are made to move from pilot projects to large, mation into productive and profitable farming. regional-scale efforts. Even if efforts to deploy more sur- face stations are successful, the separation between sta- At the present time, the main effort to provide agricul- tions, together with large regions of complex terrain, tural meteorological services to farmers and pastoralists in suggests that scaling up such insurance programs may be Ethiopia appears to be put forth in four separate projects problematic in many areas. Obtaining the spatial infor- supported largely by the Rockefeller Foundation as part mation on precipitation distribution using satellite sensors of its global “Building Climate Change Resilience” initia- calibrated by the ground station data and sometimes sur- tive. These projects involve EIAR, NMA, Mekele Univer- face radars appears to be a viable approach. sity, and Oxfam America, as well. In the HARITA project, IRI Columbia University devel- Syngenta has not extended its crop insurance effort to oped an automated system that calculates insurance pre- Ethiopia, where it has focused instead on genetic studies miums and payouts from satellite-based rainfall estimates aimed at improving crops. This may have been to avoid (Osgood 2010). Further, the Enhancing National Climate duplication of efforts with Oxfam America and other Services (ENACTS) program, a joint development effort development agencies which have been encouraging such by Ethiopia National Meteorological Agency, the Inter- insurance efforts. Ethiopian insurance companies are also national Research Institute for Climate and Society, and offering what appear to be similar products. the University of Reading, demonstrated that combining satellite and gauge data overcomes, to a degree, the short- In the Horn of Africa Risk Transfer for Adapta- comings of the interpolated gauge and satellite estimates tion (HARITA) project, during the period from 2007 taken by themselves.10 Others, such as Rojas and Ahmed to 2012, Oxfam America, Swiss Re, and their partners (2012), have explored the possibilities of extending simi- developed and implemented a pilot integrated risk man- lar index-based insurance across all of drought-prone agement framework in Ethiopia’s drought-prone north- Sub-Saharan Africa using satellite data to compute the ernmost state of Tigray. This program enabled farmers required indices. to strengthen their food and income security through a combination of improved resource management (risk These satellite-based techniques for estimating rainfall use reduction), insurance (risk transfer), and microcredit (pru- both downward pointing radars (an active sensor) and radi- dent risk taking). The pilot project was funded by the ometers that are sensitive to the visual, infrared, and micro- Rockefeller Foundation and Swiss Re. As of June 2014, wave emissions from the precipitating cloud. By observing the techniques piloted in the HARITA project continue both visual and infrared emissions, clouds are located and as Oxfam’s Rural Resilience (R4) Initiative, implemented their height (related to the type of cloud) estimated. The in partnership with the World Food Program, and begin microwave signal provides information on the presence expansion in both Ethiopia and other African countries. of raindrops. Such schemes work well, but they require Companies like Nyala Insurance, which have been ground truth data from surface rain gauges for calibration. involved in this project, have experienced considerable success in designing innovative weather insurance prod- Like scans from surface-based radar, satellite-based tech- ucts that protect a range of farmers.9 niques provide remotely sensed area measurements. These are in contrast to the point measurements of tradi- The current approach of many groups offering index tional rain gauges and allow meteorologists to see in some insurance products is to use point observations to decide degree of detail what is happening over a wide area, filling payouts. This may create problems in the future when the gaps between surface-based sites. However, all schemes for remotely sensing area rainfall need surface measure- 9 An informative brief published by the International Food Policy Research Insti- ments for calibration. This is best accomplished on the fly tute in its series 2020 Vision for Food, Agriculture, and the Environment titled “Innovations in Insuring the Poor: Providing Weather Index and Indemnity Insur- ance in Ethiopia” by Eyob Meherette, December 2009. http://www.ifpri.org/. 10 Climate Services Partnership. http://www.climate-services.org/. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 13 in real time, rain event by rain event. This in turn requires cultural productivity (similar thoughts apply to NMA). near-real-time, high-quality data streams, produced via The reasons Laichena offers are numerous, including a reliable communications and high-speed quality control strong focus on aviation weather by KMS and the domi- and quality assurance (QC/QA)/archival systems. nance of conservative elements (“we have not done this before”) in both KMS and the Ministry of Agriculture, as well as tradition-bound farmers. Consequently, agri- Discussion and Recommendations cultural meteorology does not have a strong, well-placed In both Kenya and Ethiopia, weather and climate are rec- champion among the agricultural communities or among ognized as very important factors influencing agricultural the many competing interests within the government. production by the national hydro-meteorological services, KMS and NMA, respectively, together with the efforts Meteorology and climatology are areas of endeavor in from KARI (Kenyan Agricultural Research Institute) and which real-time environmental data streams are essen- EIAR, and their associated ministries of agriculture. tial. The same data that are used to prepare and verify daily forecasts, when accumulated over time, become the In both Ethiopia and Kenya, information on weather and basis for climatological analysis and prediction. KMS climate in farming and herding areas is being used by the and NMA are severely constrained by their lack of real- government to estimate the amount of food being pro- time environmental data on the meso- and micro-scales.11 duced. However, it was difficult to find from Web searches Without such data at the relevant temporal and spatial much evidence of agricultural weather services being scales, it is not possible to effectively forecast either day-to- provided to farmers to assist in increasing agricultural day weather or seasonal climate at the farm level, which productivity and improving resiliency. (Indeed, agricul- is, of course, the scale of most interest to farmers. tural meteorology has a low profile on the websites and in the official statements of both KMS and NMA.) The Weather forecasts (zero hours out to about one week) programs supported by the Syngenta Foundation (Kenya) improve only when there are data to verify forecasts and and the Rockefeller Foundation (Ethiopia) appear to be assess forecast skill. Meteorologists learn to identify key the best documented efforts. weather features and build their understanding of the unique aspects of a particular place or region only through An excellent example of regional-scale meteorological thorough analysis of the details of ever-evolving weather products in support of agriculture is the analysis and fore- patterns over that region. One common approach to fore- cast products and services provided by ICPAC. These are casting both weather and climate that can be used to some essential for an effective agricultural meteorology pro- extent to provide local information from larger-scale fore- gram. However, ICPAC is at a long administrative and casts is “downscaling.”12 However, even with downscaling, bureaucratic distance from the farmers who could use it is necessary to have good observational coverage of the this information. The challenge to nascent agricultural meteorologists in both Kenya and Ethiopia is to inter- 11 Mesoscale and microscale have a variety of definitions. Here we consider pret these (and other locally produced) products in light the mesoscale to be events with space scales of a few kilometers to a few tens of of local farmer needs, then put the resulting information kilometers and with timescales from 0 to 48 hours, with a strong focus on 0 to about into comprehensible messages that support appropriate 6 hours. We consider microscale to be events with space scales of few meters to a actions in local agricultural communities. That does not few kilometers and with timescales from 0 to no more than an hour or so. 12 In meteorology and climatology, downscaling, or “localization,” is the pro- appear to be happening, except in the most general way. cess of deriving finer-resolution information (for example, for a particular loca- tion) from coarser-resolution numerical model output. Most agricultural impacts Laichena (2010) of the Kenya Institute for Public Policy researchers feel that the horizontal resolution of most numerical models is gen- Research and Analysis (KIPPRA) has provided a reflec- erally too coarse to be used effectively by farmers in its original format. While tion on why neither KMS nor the Kenyan Ministry of much useful information can be derived from course definition numerical model output without the need for downscaling (such as national- and regional-scale Agriculture have developed a strong agricultural mete- estimates appropriate to the ministry level), to support decisions at the farm scale, orology program, given that they both recognize the it is necessary to try to add value to a forecast scenario by making it more appli- importance of weather and climate to their nation’s agri- cable for finer-resolution decisions such as the timing of planting and harvesting. 14 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services region to provide verification of the downscaled values. weather- and climate-based advice on what and when An effective verification process allows the downscaling to plant, irrigation and water management, what pes- technique to be fine-tuned to improve its performance. ticide and fertilizer to use at the correct timing, as well as other relevant agricultural support services. NGOs appear to be the most important players in agri- cultural weather and climate services in both Kenya and In the performance of such activities, the agricultural Ethiopia. The Syngenta Foundation for Sustainable Agri- meteorologist is part of a team. Agricultural extension culture reports note that the Ministry of Agriculture is personnel, who are usually deployed by the ministry of short on technical staff, but good at contracting. Working agriculture, play an important intermediary role between with the Ministry of Agriculture, they are operating an developers of agro-meteorological products (national agro-meteorological observing network, monitoring the hydro-meteorological services; ministries of agriculture, weather (especially precipitation), and providing an inno- the environment, and water; universities; research insti- vative insurance program founded on those observations. tutes) and farmers. Communications media, particularly Similar observations probably apply to the experience of radio and cellular telephones, are vital, making telecom- the Rockefeller Foundation in Ethiopia. munication companies an important part of the agro- meteorologist’s team. Finally, international development Both the International Development Research Cen- agencies and NGOs may be important team members tre (IDRC) and the TAHMO have innovative programs who provide resources and expertise not locally available. underway to collect and provide weather and climate information to farmers. These appear to be both com- A key reason for low-level use of agro-meteorological plementary and collaborative with the KMS. Both efforts services in African countries such as Kenya and Ethiopia emphasize the importance of grassroots efforts to reach is often the lack of effective liaison by extension agents the rural poor. In the case of Ethiopia, the bureaucratic between the institutions providing agro-met information culture in NMA seems to be more closed than that in and the farming community. This challenge is compounded Kenya, so it is not clear how these initiatives will fare when by insufficient education and training of the farmer-users they attempt to expand in that country. While there are no on how to effectively utilize such information, including international standards per se for agricultural meteorol- the farm advisory services that provide specific agricultural ogy, given the great diversity within agriculture around advice from general weather and climate information. the globe, a number of best practices can be inferred: Education and training for both groups are essential. » Understand the decisions made by farmers which can be influenced by agro-meteorological Given their different core missions, national hydro- information. meteorological services should for the most part play a » Express agricultural meteorological information supporting role, collecting agro-meteorological data, per- in terms that farmers can quickly understand and forming quality assurance and control, and then gener- apply in their decision making. ating products and services at the proper temporal and » Deploy observing systems and observational net- spatial scales, while the ministry of agriculture focuses works that are crop focused—that measure quan- on delivery of weather and climate products to farmers tities directly relevant to the weather and climate through extension programs. All this suggests that special experienced in the farm fields. efforts are needed to strengthen the weather and climate » Utilize communication means (radio, cell phone/ capabilities of both the NHMS and the extension pro- Short Message Service (SMS) messaging, printed grams within the ministries of agriculture. Success in materials) that are easily accessed and utilized by providing improved services requires close collaboration the farmers. between the two entities. » Communicate to farmers how proper utilization of weather and climate information can increase their Extension agents should be knowledgeable about farm- productivity while reducing risk. Provide practical, focused weather and climate information and be able Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 15 to communicate their knowledge clearly to farmers. » Continual improvement of the forecasts to make They should also be able to assist farmers in utilizing them more farm focused—that is, applicable to local that information in their decision making. Extension conditions—instead of being based on large areas agents also need to represent farmers’ needs to the (this likely will require the development of downscal- agro-meteorologists for additional weather and climate ing schemes as well as the establishment of meso- products and services. In the end, the whole national and micro-scale environment observing networks); agricultural meteorology program should be driven by » Strengthening of teamwork and coordination farmers’ needs. between meteorologists, agriculture personnel, and the communications services/media for expediting To this end, one might explore opportunities to leverage the delivery of tailored agro-meteorological exten- the activities already under way by other international sion services to farmers; and development agencies and NGOs, such as those being » Development of a system whereby real-time data operated by the Rockefeller Foundation (Ethiopia) and by at agricultural meteorological stations can be avail- IDRC and the Syngenta Foundation (Kenya, Tanzania, able locally to the farmers, especially pastoralists Rwanda, and perhaps elsewhere). Activities that could be with poultry, cattle, sheep, and goats, as requested. leveraged include the following: » Joint training of agricultural meteorologists, exten- In addition, better understanding is needed about the sion agents, agronomists, and farmers which dem- joint (multipartner) business practices and finances behind onstrates the importance of agro-meteorological the procurement, installation, and management of agro- information for agriculture should be continuous meteorology observing systems by entities other than and structured to facilitate the implementation of NHMS. It may be that both regulation to ensure a free agro-meteorological information in farmer deci- market and incentives to encourage following best prac- sion making; tices are required. 16 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services CHAPTER TWO TRADITIONAL AND NON-TRADITIONAL AGRO-WEATHER OBSERVATION SYSTEMS Observations and measurements of both atmospheric and biological variables are essential in any agro-meteorological or agro-climatological endeavor. In order for agro-meteorologists and agro-climatologists to assist farmers and others in the agricul- ture sector with planning, forecasting, and other services that can be essential to per- sonal survival and economic success, data, quantitative and qualitative, are required which describe the ever-evolving states of the atmosphere and the biosphere. Parallel agricultural data are also needed to assess the impacts of agricultural activities and processes on the environment and climate. The necessary types of observations and measurements and their spatial density and temporal frequency depend on the purposes for which the data will be used. Charac- terization of the agro-climate of a region, monitoring and prediction of the evolving climate, and management of natural resources requires local, regional, and national coverage over a period of a few decades of a number of atmospheric variables. To ensure that the observations are representative of an area, spatial density of observations depends on the topography of the region of interest, with complex, mountainous terrain requiring many more points of observation than open plains.13 For climatologic pur- poses, a temporal frequency of once or twice per day measurements is often satisfactory. To detect small climatological trends in data from a limited number of years of observa- tion requires careful selection of instruments and a good QC/QA process (see below). Long-term climatological information provides the background conditions within which day-to-day weather and farming operations are carried out. Such operations 13 Agriculture in mountainous regions is often highly dependent on small regions with favorable microclimates scat- tered within larger regions that are generally unfavorable to farming or herding. Even in open country, observations are likely representative of a region only 3 to 5 kilometers in radius around the observing point. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 17 include numerous weather-influenced decisions, such as in temperature data. This can easily be due to a change- when to plant, proactive responses to pests and diseases, out in the sensor making the measurement rather than management of irrigation, and when to harvest. Also to a natural event. Only by exploring the metadata can included in such daily operations is the early warning of the cause of such apparent warming be identified. Such the onset of natural hazards (for example, flash flood- factors must be taken into account when the goal is to esti- ing, wildfire, late or early frost, heat stress on livestock), mate changes of a few tenths of a degree in mean values together with monitoring of such events as they unfold. To over long periods of time. provide information in support of the many weather deci- sions necessary in day-to-day farming operations, more Commercial sources have developed and provide a very frequent and additional observations are needed. The wide range of equipment and systems to acquire, store, preparation of advisories and services on farming meth- transmit, and process data on the atmospheric and bio- ods, including irrigation, pest management, and other logical variables of interest to agricultural meteorologists types of microclimate manipulation, also require special- and climatologists. The next two sections review limited ized data. To support such standard agro-meteorological samples of these items, focusing on items of particular activities, hourly or even one- to five-minute observations interest to the agro-meteorologist. are often necessary. TRADITIONAL An important step in the observation and measurement process is QC/QA of the data to ensure its quality. This INSTRUMENTATION needs to be a rigorous process to ensure that failing sen- What we call “traditional instruments” for the most part sors and drifts in calibration are identified as early as pos- were devised in the 19th century and refined during the sible and that accuracy and precision of measurements first half of the 20th century. They are characterized by remain within the desired bounds. Even with simple the need for trained humans to read, record, and send in measurements such as temperature, the QC/QA process the collected data for others to scrutinize in a QC/QA has to be thorough to ensure that all reported values, par- process, and then analyze for operational use. In many ticularly those of extremes, are meaningful. Therefore, cases, instrument components are fragile and require QC/QA is particularly important where the goal is the careful handling. Often, routine resupply of expendables detection of trends that are small in magnitude from a is necessary. Most such instruments require regular main- limited number of years of data. All too often, instrument tenance and calibration. calibration drift or malfunction has been interpreted as an emerging trend. A selection of traditional instruments is presented and dis- cussed in the following paragraphs. These illustrate both Experience has shown that a combination of automatic the diversity of hardware and the range of measurements checking and review by trained human eyes is the best for necessary to support agricultural meteorological opera- such efforts. As a general rule in the QC/QA process, no tions. All of the systems presented remain widely used. data are ever discarded but rather are flagged to call the attention of users to the data being out of bounds. RAINFALL INSTRUMENTS The most basic of all agricultural measurements is that of As important as the data is the observing site metadata rainfall. Since the dawn of history, farmers have observed associated with them. Such metadata are essential in the amount of rainfall by collecting rainwater in various understanding the properties of the data streams. These devices. Many of these devices are simple containers of metadata need to be comprehensive, including not only various configurations that are read by eye. By the mid- such basic information as station location and observing 1950s, more elaborate recording mechanical devices had procedures, but also detailed data on the particular instru- become available. In recent years, not only the quantity of ments in use. An example of the importance of the latter rain that falls, but the rate at which it does so have become is the need to understand the cause of an abrupt warming of interest. 18 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services TEMPERATURE AND HUMIDITY This quantity is important to the agricultural meteor- INSTRUMENTS ologist because the temperature at the surface often Temperature measurements are fundamental to many varies substantially from the air temperature as meas- agricultural products and services. For example, soil tem- ured at 2 meters. Under dry conditions, during a hot perature determines optimal planting dates while growing summer day, the “grass temperature” might be much degree days allows farmers to follow the maturation of higher than the air temperature. Conversely it can be crops such as corn and plan for harvest dates. Most tra- much, much lower at night during dry spells under a ditional temperature sensors are some form of liquid-in- clear sky. In fact, the grass minimum temperature might glass thermometer or utilize the distortion of bimetal or be 0°C (32°F) or less, while the 2-meter air temperature liquid-filled metal devices. recorded simultaneously is 4°C to 5°C or higher. This can produce a condition known as ground frost and is The moisture content of the air is another atmospheric important information for farmers so that protective variable of great importance. Usually this is expressed measures can be taken. in the form of either “dewpoint” or “relative humidity” (though exactly what this last quantity means is often The temperatures at various depths in the soil impact poorly understood). Many other moisture variables are plant germination and how rapidly plants take up water used specifically in particular applications. and nutrients, and thus affect the rate of plant growth. Soil temperature affects the breakdown of parent material Data on moisture, together with corresponding tempera- and how fast microorganisms work. Both are important in ture data, allow agro-meteorologists to advise farmers on adding and returning nutrients to the soil. Further, certain heat stress on animals. Such data can also be accumulated plant diseases and pests thrive in certain soil temperatures, in ways that allow the prediction of the emergence or so knowledge of soil temperature is important for taking appearance of a range of molds, fungi, and insect pests appropriate measures. on crops, allowing farmers to minimize damage by tak- ing appropriate preventative measures. Humidity data are A variety of thermometers, both liquid-in-glass and dial- also essential to assess wildfire danger. type using bimetal strips, have been developed to measure soil temperatures at a variety of depths. Measurements A number of instruments can be used to provide a contin- are taken under grass cover or bare ground, depending uous record of temperature and moisture. Simple systems upon the intended purpose. using clockwork chart drives have been devised. The most common example is the hygrothermograph. There is no easy traditional way to measure the profile of soil moisture. This is unfortunate because soil moisture is often a quantity of great interest to farmers and agro- TRADITIONAL AGRO-METEOROLOGY meteorologists alike. The gravimetric method is the only INSTRUMENTS approach that was developed prior to the modern era Temperatures within crop canopies—such as in (and is still used for calibration purposes because it pro- orchards—can be of interest to assist in preventing frost vides a direct measure of the water content of the soil).14 damage and monitoring for the possible onset of plant In this method, a soil core down to the depth of interest diseases and various pests. Farmers and agricultural mete- is taken using a standard Kirg tube; the core is sliced orologists use any of several “orchard grove thermome- into segments, and each (moist) segment is weighed. The ters” to keep track of temperatures in orchards and fields. segments, one at a time, are dried in an oven to remove The grass minimum temperature is the minimum temper- ature measured (usually somewhat after sunrise) in open 14 The gravimetric method is described in detail in Johnson, A.I., 1962: Methods air in short turf, with the bulb of the thermometer just in of Measuring Soil Moisture in the Field—Evaluates methods for measuring soil mois- contact with the tips of the blades of grass (usually about ture and describes the equipment used. Contributions to the hydrology of the United 5 centimeters above the ground). States—Geological Survey Water-Supply Paper 1619-U. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 19 all moisture, and then weighed again. The difference in of sensor readings. Many of these systems can provide the before and after weights is a direct measure of the automated generation and transmission of meteorological soil moisture. This is clearly a time-consuming process reports, make special observations when preset thresh- that requires both extensive training and specialized olds are crossed, and provide alert messages when preset facilities. thresholds in key variables are exceeded. In arid environments, in addition to measuring the water Automatic weather stations are most powerful when oper- arriving via rainfall or the flow of streams and rivers ated in a network across a region. The data streams can be across an area, net evaporation from the surface land and brought to a central location for quality checking, analy- water bodies must also be determined. Such information sis, merger with other data such as terrain and soil maps is essential for planning and implementing irrigation pro- and information on when to plant, irrigate, fertilize, and grams, as well as estimating the return on investment to harvest. Individual stations will usually have some storage be received from implementing strategies to reduce water capability, so that when communications to the central loss from crops and the soil. collection point are interrupted, data continue to be col- lected. It is then forwarded to the central collection point Weighing lysimeters are used to directly measure the when communications are restored. The network of indi- loss of moisture from both plants and soil (Davie 2002). vidual station computers manages all the necessary com- These are complex devices to operate and maintain, and munications protocols. are usually found only in agricultural research institutes and universities. As a substitute, agro-meteorologists and Using a network of automated systems to provide weather agricultural engineers have devised what is termed (in the and climate observations has many advantages over tra- United States) the “Class A Evaporation Station.” ditional manual systems. These advantages can be sum- marized as follows: NONTRADITIONAL » Standardization of observations, both in time and INSTRUMENTS quality AUTOMATIC WEATHER STATIONS » Greater reliability—real-time continuous measur- Since the end of World War II, meteorological observing ing of parameters on a 24/7 basis and measuring equipment has evolved rapidly in paral- » Improved accuracy (eliminates reading errors, lel with the growth of modern technology. While numer- subjectivity) ous novel approaches to sensing temperature, humidity, » Collection of data in a greater volume, for exam- precipitation, and other atmospheric variables have been ple, one-minute or five-minute data as opposed to developed, perhaps the greatest advances have been in hourly or once per day the replacement of human observers and handwritten » Automatic adjustment of sampling intervals of dif- observing forms by electronic data logging and transmis- ferent parameters in response to changing weather sion systems. Within a few seconds of being made, obser- events vations from across a wide region can be collected at a » Automatic QC/QA during collection and report- central location, error-checked, and made available to ing stages, including automatic alerts to users and meteorologists and other users. maintenance personnel when errors are detected » Automatic message generation and transmission, Many of these technologies can be packaged to produce including alerts when critical thresholds are crossed “automatic weather stations,” some examples of which » Automatic data archiving are shown below. These stations can be configured with » Access to data, both real time and archived, locally a wide variety of sensors to provide data to meet specific or remotely operational requirements. An on-board computer or » Collection of data in remote, harsh, or dangerous “data logger” handles the collecting and initial processing climates 20 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services Automated observations also bring their own set of disad- Unfortunately, there are no standardized practices and vantages. Some of these are as follows: only limited guidance available on implementing and » There is a high initial cost of instrumentation maintaining observing systems in harsh climatic envi- and associated equipment and then ongoing costs ronments. Each such observing system must therefore be of operation, such as for maintenance, electrical designed as a one-off, often trial-and-error configuration. power, communications, security, and so on. Special instruments, hardened computers, and stronger » It is not possible to observe all desirable parameters towers, as well as more frequent routine maintenance, automatically; at key locations, it may be necessary will likely be required, all of which comes with significant to augment automatic observations with a human additional costs. observer to obtain information such as cloud cov- erage and cloud types. SPECIAL AGRO- » If solar panels are used to power a station, this may limit the amount and type of instrumentation, METEOROLOGY RELATED local computing, and telecommunication equip- SENSORS FOR USE WITH ment that can be used. AUTOMATIC WEATHER » Final quality control is best carried out by a staff of STATIONS trained operators working on a 24/7 basis. In this section, we show a few illustrative examples of » The high volume of data generated requires the instruments that have been developed to assist the agro- development of a data archival system that can be meteorologist. costly in its own right, which will require periodic forward migration as software changes. Water from rain or dew remaining for any extended » Routine preventative and as-required corrective period of time on the leaves of many different types of maintenance, together with periodic sensor cali- plants, particularly those with a dense leaf canopy, is a brations, require a staff of trained maintenance major contributing factor to the emergence of plant dis- technicians. ease. For example, limited air circulation and longer leaf wetness favor late blight development on tomatoes, as well Maintenance and calibration of the numerous devices as potatoes.15 An agro-meteorologist can contribute sig- and sensors comprising automatic weather stations nificantly to an integrated pest management program by require technicians with training and experience, spare providing data on leaf wetness. sensors, repair parts and expendables, and the availability of appropriate facilities. Most such stations are designed In addition to soil temperature and soil moisture content, to operate in mid-latitude conditions and in areas where soil salinity is a vitally important measure, particularly in significant supporting infrastructure is routinely available. irrigated areas. Most commercially available automatic weather stations In hot and humid environments, livestock health is a con- are designed for use in the moderate climate zones of the tinuing issue. One measure of animal comfort (and of northern hemisphere, although the meteorological and people as well) is the Wet-Bulb, Black Globe Temperature climatological requirements for observations are valid (WBBGT). Experience and field testing has shown that for all climatic zones. Special attention must be given heavy physical effort by livestock and humans should be to the design of automated systems intended for service avoided if WBBGT > 32.9°C. WBBGT is widely used in in the tropics, deserts, and mountainous regions, as well as climates susceptible to severe weather events such as hurricanes. Extremes in temperature, humidity, wind, 15 Late blight is a serious plant disease that mainly attacks potatoes and toma- toes, although it can sometimes be found on other crops, weeds, and ornamen- precipitation, and solar radiation together with sea salt– tals in the same botanical family. Late blight was a factor in the Irish potato enhanced corrosion, birds and insects, and lightning can famine in the 1850s, where entire potato crops rotted in the field or in storage all pose special challenges for automatic weather stations. because of late blight infection. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 21 the feedlot industry to mimic the effect of solar radiation Incoming long-wave Far-Infrared (FIR) radiation from on the comfort of cattle. It also can be used for comfort 4.5 to more than 40 μm is also absorbed by the surface, measurements of other animals and humans. which heats up and emits FIR back to the sky. The four components of net radiation are thus the incoming and Other animal comfort measures that can be derived by reflected solar radiation (from which the surface albedo data collected with an automatic weather station include may be calculated) and the downward and upward infra- the following: red radiation. Simple net radiometers combine two sen- » Temperature Humidity Index (THI) sors, one pointed toward the ground and one pointed » Heat Load Index (HLI) toward the sky. The sensors are wired together so that out- » Accumulated Heat Load Units (AHLU) put of the system is the sum (the net) of the four radiation » Current evaporation rate and total evaporation for components. Even simpler systems use only one system to the day view both the sky and the ground. Temperature Humidity Index, Heat Load Index, and Radiation instruments were historically quite expensive. Accumulated Heat Load Units are important measures In recent years, however, much less expensive instruments of animal comfort. As livestock starts to experience even have become available. While these lack the precision of mild heat stress, feed uptake drops, weight gain slows, the more expensive systems, they can produce data per- and fertility rates drop. In dairy cows the milk produc- fectly adequate for a wide range of agro-meteorological tion drops. As the heat stress worsens, animals can die purposes. unless appropriate steps are taken. Even one heat event managed correctly can pay for the cost of an automatic Measuring precipitation is a challenge with any automated weather station. In times when heat stress is not an issue, station. Most of the readily available systems have moving the agro-meteorologist can use the information to link components that can produce a variety of errors and have animal comfort to feed consumption and live-weight gain. ongoing maintenance requirements. Fortunately, various systems to measure rainfall without moving parts have The current evaporation rate is a measure of the rate come on the commercial market in recent years. These of loss of surface moisture. The evaporation rate can be include laser distrometers and small vertically pointing determined from a Class A pan by monitoring it continu- radars. Low-maintenance precipitation measurement sys- ously with an automatic weather service. These data can tems offer a major advantage over the common tipping be related to the rate of evaporation in animal pens, also bucket and weighing gauge processes. affecting in-pen humidity readings and thus, cattle com- fort. The current evaporation rate is also very useful in Systems such as those shown here are relatively expensive. irrigation scheduling. However, considering over the life of the instruments (10, 20, or more years) the potential savings from the reduced Another quantity of interest to the agro-meteorologist is maintenance requirements, such instruments become the net radiation received by the surface (usually covered more cost effective. with some type of plants). The main applications for net radiation data are in the study of evapotranspiration and AREA RAINFALL ESTIMATES in climatology, meteorology, and hydrology for the meas- urement of the overall radiation balance. FROM LIGHTNING LOCATING SYSTEMS Net radiation is the balance between incoming radiation The instruments considered up to this point provide sur- from the sun and sky and outgoing radiation from the face-based, point-location measurements. While such ground. Short-wave radiation of 0.3- to 3-μm wavelength measurements are essential to an agro-meteorologist, reaches the Earth’s surface, where some of the radiant they do not provide a full picture of what is happening energy is reflected and the rest is absorbed by the surface. weather-wise over a region. This is especially true with 22 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services FIGURE 2.1. GLOBAL LIGHTNING FREQUENCY Source: Image from National Aeronautic and Space Administration (NASA); data for 1995–2002. Note: Annual global lightning (flashes per square kilometer per year). As implied by the mapping of lightning frequency shown here, thunder- storms are concentrated in Sub-Saharan Africa and other tropical regions of the world, where many developing countries are located. The equatorial regions of Africa and South America can experience thunderstorms on more than 140 days per year with some areas having storms more than 200 days per year. Such storms account from much of the annual rainfall in these regions. regard to rainfall. The traditional tool for the meteo- that these lightning location technologies are much rologist in monitoring rainfall over an area has been less expensive to install and maintain, they may be an weather radar, an active sensing technology.16 However, attractive alternative to weather radar for many coun- weather radars are expensive to procure, install, oper- tries in Africa. ate, and maintain. Consequently, weather radars have proved impractical for many of the least-developed While a single lightning sensor can provide useful infor- countries in Africa. mation regarding lightning strikes within its detection range (a few hundred kilometers), operating a network A significant fraction of the total annual precipitation of such sensors spaced a few hundred kilometers apart in Sub-Saharan Africa is produced by thunderstorms. allows the monitoring of thunderstorms over or advanc- Thus it is very fortunate that in recent years, significant ing into a region in detail. Recent developments in light- advances have been made in lightning location tech- ning detection (both cloud-to-ground and in-cloud), nologies that now allow such passive systems to serve, location, telecommunication, and timing technologies in large part, as substitutes for weather radars.17 Given have greatly enhanced the capability to acquire and quickly process signals from a network of lightning sen- sors deployed across a region, resulting in useful data sets 16 Radar is an “active sensing technology” in the sense that electromagnetic of time and location of discharge occurrence, intensity pulses are generated by the radar and emitted into the atmosphere to produce and polarity, and other factors. These networks now also electromagnetic echoes. These echoes are then processed by the radar system to allow the near-real-time monitoring of the evolution of determine the location and movement of particles in the air. the occurrence of lightning flashes in a particular storm, 17 Lightning locating systems are “passive sensing technology” in that they do not emit electromagnetic radiation but rely on detecting electromagnetic signals from which much can be inferred regarding the state of produced by nature in the lightning discharge. the producing thunderstorm (figure 2.1). Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 23 Using correlations between spatial and temporal densi- als with only modest training, offering opportunities to ties of flashes and storm rainfall, maps of lightning flash involve local people in the observing process. Although rates can be presented in formats similar to what is done reporting can be slow, in the age of cell phones it can be for radar reflectivity. These pseudo-reflectivity values can made much faster than the traditional mail-at-the-end-of- be used to enhance thunderstorm prediction and could the-month approach. The primary disadvantages are that serve as a cost-effective radar alternative in underdevel- many of the instruments are fragile and easily broken, the oped regions of the world. By applying formulae similar readings are subject to a variety of errors, and the whole to those used with radar reflectively fields, the lightning- system is manually operated, limiting the data available based values can be used for precipitation estimates. for real-time operations and making the development of a When accumulated through time, just as is done with central archive a challenge. radar-based precipitation estimates, one can display accu- mulated (estimated) rainfall over periods ranging from one In many situations, traditional instrumentation has been hour to several days, weeks, or months. supplanted by automatic observing systems, which offer many opportunities for improving meteorological services The lightning detection and locating and associated sig- of all types. Automatic weather stations are most powerful nal processing systems marketed by Vaisala, Earth Net- when operated in a network where the data streams from works, the U.S. Precision Lightning Network (USPLN), the various observing stations are brought together in real and others, and the value-added resellers such as Weather time at a central collection facility. One can then follow Decision Technologies, Inc., illustrate what can be done in near real time the evolution of weather as it happens with modern detection techniques and advanced signal across a region. processing systems. Automatic weather stations bring their own challenges. For current purposes, lightning detection and locating sen- Many of these can be addressed through a careful system sors are well suited for installation on or proximate to cell design process. A backward planning process is often the towers (see discussion in Chapter 6). A network of such best design approach. In this case, the products and ser- detectors is very attractive because the sensors are passive vices to be supported immediately and in the future by with low power requirements while the telephone network the automated observing network are detailed first. These supporting local cell service provides the reliable, wide- products and services should be defined very carefully and bandwidth communications necessary to rapidly move in complete detail. User input is critical to this discussion. the lightning data back to a central server for processing. For example, a NHMS should undertake the establish- ment of an agro-meteorological network in close consulta- Total lightning and other lightning detection systems have tion with academic institutions, the ministry of agriculture, one shortcoming. They require that lightning occur before and agriculture extension programs in particular. they can be used to infer the presence of deep convection, precipitation, and other weather elements. With deep con- Site selection can be challenging because a number of cri- vection, this means the initial stages of the formation of a teria must be met. Access to power (unless a solar panel cloud cannot be detected. Further, much rainfall comes from is practical), telecoms (which may include access to a cell stratiform (layered) clouds that seldom produce lightning. network), security, lightning protection, and access for Lightning detection networks need to be complemented technicians are among these. Planners must also deter- with other observation systems to detect precipitation asso- mine how representative a given site is. ciated with forming thunderstorms and stratiform clouds. The standard WMO guidelines and recommendations DISCUSSION for site selection should be followed to the extent possi- Traditional instrumentation still has many applications, ble (WMO 2010). However, these recommendations are particularly with regard to agro-meteorology. Many largely for synoptic observing stations, often located in the of the instruments are suitable for reading by individu- large open spaces of airfields. Thus, they may not always 24 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services be appropriate for the meso- or microscale observations Forward migration of the archive as software and hard- needed by agro-meteorologists. The WMO Integrated ware changes occur needs to be considered—a data Global Observing System (WIGOS) program has begun archive should be considered a national resource to be to recognize the need for local networks. preserved in perpetuity. Maintenance, both preventive and corrective, and calibration are expensive and time The stations of the proposed network should be flexible consuming, but they are essential if a sufficient return- enough to absorb new hardware and software compo- on-investment is to be obtained from the network. The nents as technology changes and upgrades become avail- NHMS needs to allocate sufficient funding to support the able. Since the controller, data logger, or similar device is long-term life cycle of the network. the heart of any automated weather station, special atten- tion needs to be given to its specifications to ensure ease Finally, the human side of upgrading a network from of use and flexibility in swapping out existing sensors or manual to automatic needs to be carefully considered and installing additional ones. This often means initial over- factored in to plans. Human observers may be suspicious specification on the capabilities of the data logger and that deployment of automatic systems will mean the loss the available bandwidth on the communications system. of employment. They may also doubt that sufficient funds Good design favors the use of open-source software and will be forthcoming to properly maintain the automatic avoids use of proprietary software wherever possible. In equipment. addition to making changes to sampling rates, a proactive quality assurance program is necessary. One approach is to offer former observers the first oppor- tunities for training to become the operators, QC/QA The central data archive needs to be designed early in the specialists, and maintenance technicians for the upgraded process, when planners need to determine what data will network. This may represent a major investment to raise be stored and how metadata will be handled and made their skills to the levels needed. The training should there- available.18 Network metadata are an essential part of fore be started well in advance of system deployment the data archive. At a minimum, these metadata should as a central part of a transition plan. These individuals include a thorough history of the site and each instrument should be involved in the planning and implementation of and observing system used in the network. This technical upgrades and extensions of the network so they develop a history of the network is essential to climatological work. sense of ownership for the upgraded network. 18 Most archives are set up to receive all data collected. The QA process does not discard any data but provides flags on data that are questionable or missing. These flags become part of the data set. This allows future users to access the “raw data” and then provide any filtering or corrections as they see fit. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 25 CHAPTER THREE ADOPTION BARRIERS FOR AGRO-WEATHER OBSERVATIONS As may be inferred from previous comments, this report recommends each nation have a comprehensive agricultural meteorology program supported by a nationwide observing system. However, there are a number of significant adoption barriers to establishing and sustaining such national programs. These can be classified as cultural, bureaucratic, and financial.19 CULTURAL BARRIERS Cultural barriers can crop up in many guises, including the all too familiar, “we’ve never done that before and are not going to start now,” in bureaucracies. Here we comment on cultural barriers likely to be encountered in the field, where agro-meteorologists, working with and through agricultural extension agents, attempt to encourage small landholders to make better use of modern weather knowledge. Around the world, farmers tend to be conservative in their farming practices. This seems to be particu- larly true with small landholders living at or near the subsistence level and in areas where local village or tribal links remain strong. In many settings, local indigenous weather predictors rely on close observation of plants, animals, and other natural indi- cators. Many have long experience with farming and weather events in the area. Local lore, such as the phenology of plants and animals in the area, can provide insights into local, microscale weather and climate phenomena, as can careful study of the clouds that come and go overhead.20 19 Due to the lack of materials either published or posted to easily identified websites, this section is necessarily some- what broad and lacking in specificity. There may well be tighter connections between the NHMSs in Kenya and Ethio- pia, their respective Ministries of Agriculture and their research institutes, and farmers than what is suggested here. However, those connections are not readily apparent from a distance. The thoughts expressed here were formulated after a careful reading of several reports by nongovernmental organizations operating in the two countries and from what could be gleaned from the various websites of the organizations of interest, and after discussions with individuals who had worked in those countries. 20 Phenology is the study of periodic plant and animal life cycle events and how these are influenced by seasonal and interannual variations in climate, as well as habitat factors (such as elevation). Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 27 As an example of a local indigenous weather forecaster, This suggests that an agro-meteorologist coming into a consider 73-year-old Wilson Yoeze, resident of the Same new region should start by building positive relationships district, one of the seven districts of the Kilimanjaro with the local indigenous weather forecasters. Further, Region of Tanzania). He has been predicting local rain- the agro-meteorologist might consider recruiting local fall since 1965, looking to the skies and to local flower and indigenous predictors and farmer-leaders to become tree species for clues to the weather. “When we started,” he local observers. Such individuals can be provided with a says, “we were depending on observations of certain signs. set of simple meteorological instruments, and the neces- We looked at stars and clouds. . . . There are clouds that sary training to use them. It is now fairly easy to work only appear when the season is about to begin.” Accord- up a system whereby such individuals can complete sim- ing to Yoeze, if a single thick cloud leads the sky cover as ple reports on a cell phone and submit them to a central it moves from east to south, a good rainy season is on its number. Such local readings, properly quality assured and way. Another sign of long rains is the appearance of a controlled, can be an important meso- and microscale bright star from the east known as ngate kere. “When none supplement to data from standard observing sites. This of these appear,” he says, “there will be no rain at all involvement should build a local sense of ownership of and starvation will be great” (Ogodo 2012). Indigenous the agro-meteorological program in the area, so that it is weather forecasters often have significant influence on no longer perceived as being dictated from the national critical farming events such as dates of planting and har- capital. Quality assuring and controlling such data is a not vest, or the movement of livestock to winter or summer a trivial matter and requires significant investment of staff pastures. They also may predict rain or dry weather, and time, hardware, software, and funds. so strongly influence decisions about water management. To date, the methods of indigenous weather predictors have been poorly understood, but locally, their advice is BUREAUCRATIC BARRIERS actively sought out by many farmers. The agro-meteorologist and agro-climatologist are spe- cialists in the same way the aviation meteorologist, the Traditional forecasters have long guarded their knowl- hurricane forecaster, the fire weather meteorologist, and edge and shared their predictions only with the elders and the river hydrologist are specialists. Where the national farmers in their villages. In spite of their mysterious ways, agricultural weather program is situated within a govern- at the local level, their predictions can add a degree of ment varies by country, most often in either the agriculture precision that has been missing from the coarse regional ministry or the national hydro-meteorological service. outlooks typically provided by national hydro-meteoro- logical services. Given the strength of tribal organizations The national hydro-meteorological service has profes- and the tight knit populations in many villages, the best sional and technical personnel to support data collection, approach appears to be for agro-meteorologists to work access international model output, do the necessary anal- with and through local indigenous weather forecasters. ysis and synthesis of data and model output, and produce One such effort is found in the Tanzania Meteorologi- the general forecasts that agro-meteorologists need to pro- cal Agency (TMA). According to research team leader vide specialized products and services for the agricultural Henry Mahoo of Sokoine University of Agriculture, in community. Agricultural observing networks can provide Same district, the TMA and indigenous forecasters now the national hydro-meteorological service forecasters with work together on a consensus forecast that bridges the meso- or microscale data. Such fine-scale data can allow two knowledge systems. As a result, he says, “the image of the general forecasters to follow the evolution of weather indigenous forecasters is now positive to both farmers and systems in ways that might not otherwise be possible. Data district authorities.” Same District Council has since com- from the agricultural network also help further develop mitted funds to sustain these consensus forecasts. In gen- the nation’s general climatological database. The national eral, it may take special training by experts in indigenous hydro-meteorological service is clearly a professional knowledge and traditional lore of agro-meteorologists for home for the “meteorology” in “agricultural meteorology them to be effective in such an approach. program.” 28 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services On the other hand, the work of agro-meteorologists leadership is to establish clear policies and procedures supports and extends the work of agricultural extension that push the national hydro-meteorological service and agents, who are almost always employees of the ministry government ministries in the desired direction, and then of agriculture. Agro-meteorologists will often accompany to put in place an incentive structure that will make ensure extension agents into the field to meet with farmers and these policies and procedures are followed. discuss their weather and climate issues, and will work closely with farmers, agronomists, agricultural engineers, A memorandum of understanding between the national and extension agents to address a wide range of agricul- hydro-meteorological service and the ministry of agricul- tural problems. The agro-meteorologist needs the support ture is a highly useful way to arrive at a consensus about of extension agents in organizing local people to serve what the responsibilities and objectives of a national agri- as volunteer observers and to acquire land for automatic cultural meteorology program are. The memorandum weather stations. The senior leadership of an agricultural can serve as a reference for the development of a standard meteorology program needs frequent access to the senior operating procedure in which the specific agencies that leadership of the ministry of agriculture to ensure the agro-climatologists and agro-meteorologists are responsi- resources of the national agro-meteorology program are ble to is specified. It can also stipulate the responsibilities properly focused on the nation’s agricultural priority areas. of the concerned institutions and agencies and the indica- The ministry of agriculture is clearly the professional home tors that will be used to rate their performance, as well as for the “agro-” in “agro-meteorology program.” the financial commitments that will be required of them in covering salaries, facilities, and operating expenses. The agricultural meteorology program—and by exten- Nepal provides an excellent example of an agricultural sion, its staff—can easily find itself trying to work within information management system that provides weather the different rules, policies and procedures, and perfor- and climate information to government institutions and mance expectations of two different government agen- farmers alike.21 The national agricultural meteorology cies. That is not a good situation. program is a field-oriented organization position, with only limited desk time for most professional staff. Respon- Complicating the question of where an agricultural mete- sibilities for linking to regional programs such as Regional orology program would be most suitably located is the Center for the Training and Application of Operational deep institutional conservatism typical of so many gov- Agro-meteorology and Hydrology (AGRHYMET) and ernment bureaucracies. ICPAC, as well as international aid agencies should be specified. A joint annual evaluation process, of both pro- Complicating this decision is the fact that government gram and individuals, including feedback to the user com- bureaucracies tend to be very conservative and seldom munity, should be described in detail. welcome significant changes. This conservatism all too often leads to the “not invented here” syndrome or to Whether or not the ministry of agriculture is the adminis- competitive empire building by senior managers. In too trative home of the national agro-meteorology program, many cases, these last are accompanied by a reluctance to that program will necessarily be a bridging activity that openly share real-time and archived data or to adopt the links the national hydro-meteorological service to the open software systems that provide virtual sharing of such extension programs of the ministry of agriculture. information between the national hydro-meteorological service, relevant government ministries, academia, and the private sector. Much depends on the policies and procedures of the 21 An example of the development of such an agricultural management infor- mation system providing timely essential weather and climate information to country involved as well as institutional cultures of the government authorities and key user groups such as farming communities is national hydro-meteorological service and the minis- being developed for Nepal. For more details, see http://moad.gov.np/ppcr try of agriculture. The challenge to the nation’s senior /datadigization.pdf. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 29 an observing system established and made operational, FINANCIAL BARRIERS new agro-meteorological products and services designed Finances are always a particular challenge in small and implemented (with significant input from the poten- government organizations such as the national hydro- tial users), and farmers trained by agricultural extension meteorological service, and even for small programs within agents and agro-meteorologists in how to use the informa- large ministries, such as an agricultural extension service. tion in those products and services. Their budgets are almost always set by central government authorities, often through drawn out administrative and Owing to financial constraints, particularly in the least- legislative processes. Neither the hydro-meteorological developed countries, many national hydro-meteorological service nor the extension services are generally consid- services have very limited numbers of staff. Retaining ered high-profile agencies at higher levels of the national staff, particularly those with expertise in areas like climate power structure, and both tend to lack champions at those impacts, is difficult. Providing time and support for the levels. This lack of popularity has important ramifications professional development of staff, participation in interna- for a national agro-meteorology program. tional meetings, and undertaking new initiatives in small organizations places additional burdens on those who are Regardless of how well they are funded, the national left to attend to day-to-day work, decreasing morale and hydro-meteorological service and the ministry of agri- reducing support for new ventures. culture are the primary caretakers for their country’s “weather asset” and its “agricultural asset,” respectively. Changing this situation requires convincing the highest Any effort to establish an agro-weather or agro-climate levels of government that investment in NHMS fund- program, together with an appropriate observing network, ing and staff will have significant positive impacts on the must involve both—especially regarding budget matters. country. This is a challenging task for the leadership of an NHMS. The impacts may not be seen for three, five, A number of national hydro-meteorological services or more years after the investment is made, while govern- around the world are strongly focused on aviation mete- ment leaders tend to focus on much shorter-term issues. orology. In some countries, there is a separate agency In many respects, this is a circular problem in that to jus- with its own budget—often more generous than the one tify investment, performance has to be demonstrated, but for the NHMS—handling aviation meteorology at the the performance cannot be demonstrated until well after country’s internationally recognized airports. This has the investment is made. occurred in large part because of the visibility of avia- tion weather services to senior government leaders and It is during this initial three- to five-year period that NGOs high-ranking staff, as well as to top business personnel, and international aid agencies can have an impact, cover- all of whom fly frequently (and formal requirements ing many of these start-up costs and providing the nascent from international agencies such as the International agro-meteorology program the time needed to demon- Civil Aviation Organization [ICAO] to have airports so strate its value to the government. recognized). For this approach to result in a sustainable program, it may A comprehensive agro-meteorology program, imple- be necessary for the NGOs and international aid agen- mented in partnership between the NHMS and the min- cies to make parallel investments in the national hydro- istry of agriculture can attain prominent visibility with meteorological service and ministry of agriculture. (Some top government officials when the use of its information examples of this have been mentioned previously.) Such by farmers has a dramatic impact on agricultural pro- parallel investments may be necessary because few if any duction. This is more likely to happen once each of the of the national hydro-meteorological services and agricul- required elements is in place: trained professional staff, tural extension services in the least-developed countries and support staff, hardware and software, national observ- some of the more-developed ones appear to have staff with ing system, and so on. Trained staff must be recruited, the skill sets required. These include strategic planning; 30 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services end-to-end systems engineering leading to acquisition and partnership agreements in a tightly coordinated effort that installation of high-tech equipment; and new business provides initial support and capacity building, while leav- development, marketing, new product development, and ing the ownership of the program with the NHMS and design and implementation of delivery systems. responsible ministries. These are the same skills that are required to establish a To be sustained over the long term, a comprehensive comprehensive agricultural meteorology program and agricultural meteorology program must have sufficient agricultural weather observing network of the sort recom- impact on its country’s agricultural production to attract mended here. Expertise and experience on such matters budgetary support from the highest levels of national gov- must be brought in from NGOs, private sector consulting ernment. Alternatively, the program may find a way to companies, and universities with agricultural programs partner with interested private sector entities, though a in Africa, Europe, and North America. This should be combination of public and private funding sources may done through appropriate contractual, collaborative, or be preferred. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 31 CHAPTER FOUR COST-EFFECTIVE STRATEGIES FOR UPSCALING AGRO-METEOROLOGICAL OBSERVATION SYSTEMS Quantifying the costs and benefits associated with an agro-meteorology program is a challenging exercise. It is particularly challenging in developing countries such as Ethiopia and Kenya where agricultural production is based on a very large number of often very small farms that cultivate a wide variety of crops and raise a variety of different livestock animals. Because no comprehensive study of the costs and benefits of agro-meteorological ser- vices appears to have been carried out anywhere in East Africa, this analysis is limited to a review of the few relevant studies that have been undertaken, and to reporting preliminary results that are indicative in nature. ESTIMATING A COST-BENEFIT RATIO FOR AGRO-METEOROLOGICAL SERVICES Experience in other countries shows that agro-meteorological services help guarantee that farming is as efficient as possible (minimizing costs) and results in a good quan- tity of high-quality agricultural products (maximizing the return on farmers’ invest- ments). These benefits come through improving the bases on which farmers make decisions. Each such decision that is informed by enhanced meteorological or climato- logical information results in reduced costs for routine farming activities and enhanced income at the end of harvest. The benefits stemming from ordinary farming decisions made with enhanced weather and/or climate information have been the subject of a number of studies. The most relevant of these appears to be that by Ansalehto et al. (1985). This study was a joint project of the Finnish Meteorological Institute, the Agricultural Research Centre, and the Association of Agricultural Producers in Finland. The goal was to identify the requirements for further agrometeorological research and the need for daily services Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 33 to Finnish farmers, to estimate costs and benefits of such » Harvesting: ~FIM 44.5M/year services, and to make suggestions for the development of » Other actions: ~FIM 7.4M/ year an agrometeorological service for the country as whole. » Total: ~FIM 120M/year In three growing seasons from 1982 through 1984, the Following are a few comments on the results drawn from working group offered experimental services to 230 the Ansalehto et al. (1985) study: farmer-participants, all in a small area in southern Fin- » Across the 230 farmer-participants, each individ- land. These farmers were offered weather forecasts spe- ual farmer likely saw only a small net benefit or cifically tailored to agricultural purposes through an even a net cost, perhaps sufficiently small to be lost automatic answering machine, five days a week for each in normal year-to-year variations. But collectively, of the three growing seasons (May to September). Farm- averaged across the three seasons, the overall net ers were presumed to be able to make good use of the benefit was found to be significant. specialized meteorological information on their own; they » The biggest impacts were associated with the plant- were not given any training on how to use the information. ing and harvesting phases, so products and services supporting these activities should be a priority for To evaluate the economic impact of this prototype service, an agro-meteorology program. questionnaires were sent to the farmers. The responses » Recall that the farmer-participants received no to this survey revealed that it was very difficult for many training on how to best use the forecast products farmers to assess the amount of economic benefit they that were provided. One can easily imagine that had gained from the specialized meteorological services. with a program of aggressive initial training fol- However, based on the responses of the farmer-partici- lowed up with annual update and refresher train- pants taken as a whole and on other agro-economic data, ing, it is likely the cost-benefit ratio might change to the potential benefits and losses were considered and the 1/50 or even 1/100. cost-benefit ratio was estimated by agronomic experts to » For planning purposes, it seems reasonable to be about 1/30. assume that cost-benefit ratios in the range of 1/30 to 1/50 to 1/100 (➔ low/medium/high) For a similar service covering all the agricultural regions bracket the likely values for most agro-meteorology of Finland, the researchers estimated the cost to be about programs. FIM 4.2M a year, assuming strong leveraging of existing » The Finnish Meteorological Institute had a well- observing networks and other assets.22 Thus, the expected developed observing system installed across the yearly benefit for Finland as a whole resulting from the country at the time of the study. Thus, capital costs provision of specialized agro-weather services would be were taken to be very low, with most of the costs an increase of about FIM 120/year in national agricul- coming from refining and enhancing the proto- tural production. type service used in the study. Continuing annual costs were taken to come mainly from the addition Further, extrapolating from the study’s findings with 230 of necessary new staff. The cost-benefit ratio was farmer-participants, this net benefit would result from the based only on the increased continuing costs. following areas where farmers had to make decisions: » If there are capital costs and they are paid using » Sowing: ~FIM 44.5M/year local funds, then they should be amortized over » Spraying and protection: ~FIM 29.6M/year the life of the equipment and the facilities. This would increase the cost-benefit ratio (make it less 22 FIM = Finnish markka. All FIM amounts in the text are in 1984 values. Fin- attractive) to some extent. On the other hand, if land converted to the Euro (February 2002), however, pseudo exchange rates the capital costs are paid by an agency external to are still maintained for calculation purposes, for example, in 1984, US$1 = the nation, then they would not need to be amor- 6.0100 FIM ➔ FIM 4.2M ~ US$699,000, while FIM 120M ~ US$20 million. Taking into account inflation in the two currencies, in 2013 ➔ FIM 4.2M ~ tized. It is this last situation that is assumed in what US$938,000, while FIM 120 M ~ US$27 million. follows. 34 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services weather- and climate-related extension work; limited POTENTIAL IMPACTS ON real-time operations ETHIOPIA AND KENYA º Initial capital investment: US$20 million Utilizing the information from the report by Ansalehto et º Continuing annual cost: US$5 million al. (1985) as discussed above and further informed by the º C/B ratio: ~1/50 ➔ ~ US$250 million increase estimation process outlined in Hautala et al. (2008), we in agricultural production can make some rough estimates of the economic impacts Modern/Advanced—national network of automated of various types of agro-meteorology programs. systems supplemented by traditional instrumentation, professional staff now able to support full-scale real- Much depends on the type of agro-meteorology program time operations and fieldwork for training and exten- that is implemented. Three alternatives warrant note: Tra- sion support ditional/Minimal, Transitional, and Modern/Advanced. º Initial capital investment: US$50 million In each case, there is an initial capital cost for procurement º Continuing annual cost: US$10 million and deployment of equipment, routine maintenance of º C/B ratio: 1/100 ➔ US$1 billion facilities, and hiring and training of staff to use the new observations, new tools, and new techniques to provide Applying these results to Ethiopia and Kenya: more and better products and services. There is also an ongoing annual cost for staff, maintenance of equipment In Ethiopia, agriculture accounts for about 39.4 percent and facilities, and continued training. In light of the Finn- of the national GDP = US$41.72 billion (2012 estimate ish report, I assume that the cost-benefit ratio will improve or about US$16.44 billion).23 Thus a traditional, minimal significantly with both an increasing suite of meteorologi- agro-meteorology program would increase the agricul- cal and climatological products and services together with tural contribution to GDP by about 0.18 percent; a Tran- vigorous training of farmers on how to interpret the infor- sitional program, by about 1.5 percent; and a modern/ mation and apply it to everyday decisions. advanced program, by about 6.1 percent. Again, all the numbers here are only rough estimates but In Kenya, agriculture accounts for about 24 percent of should be of the right order of magnitude. the national GDP = US$40.70 billion (2012 estimate or US$10 billion).24 A traditional or minimal agro- Let us assume three types of agro-meteorology programs: meteorology program would increase the agricultural contribution to GDP by about 0.30 percent; a Transitional Traditional/Minimal—mostly traditional instrumenta- program, by about 2.5 percent; and modern/advanced tion, with automated systems at only a very limited program, by about 10 percent. number of locations; limited professional agromete- orological staff ➔ very limited field training or exten- Thus, the magnitude of the benefit to be gained by deliv- sion work; no real-time operations (more or less the ery of a full suite of agro-meteorological products and current status in both Ethiopia and Kenya) services, coupled with an aggressive training program º Initial capital investment in equipment, facilities, for farmers on how to use this information in their rou- training: US$3 million tine decisions, can encourage the nation’s leadership to º Continuing annual cost: US$1 million maintain and further develop the program delivering º C/B ratio: 1/30 ➔ ~ US$30 million increase in such services. agricultural production Transitional—still mostly traditional instrumentation, It is a combination of (1) the development and delivery but more automated systems in regional networks; of a suite of agrometeorological products and services, staff of agro-meteorologists and supporting techni- cians still modest in size but now sufficiently large 23 http://www.worldbank.org/en/country/ethiopia. to begin routine training of farmers and support 24 http://www.worldbank.org/en/country/kenya. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 35 coupled with (2) an aggressive training program for farm- One of the ways a national agro-meteorology program ers, which produces the maximum level of success. And can assist farmers in dealing with hazardous weather is by (3) access to risk management tools including insurance serving as a “neutral party” supporting weather indexed and microcredits. insurance. Here an automated observing network, over- » The above benefits will be realized only once seen by the agro-meteorology program, can play an the agro-meteorological program fully matures. important role in providing verification values for fields This will take several years following establish- near the station and ground truth/calibration for satellite ment or significant enhancement, so the above sensors used to cover a wider area. figures may be considered estimates for an objec- tive of increased economic benefits to be realized Weather indexed insurance is a risk management tool at the 5- to 10-year point. Of course, more mod- that is gaining wide acceptance as an alternative to a tra- erate benefits will be realized within a few years, ditional crop insurance program. As the name implies, with incremental amounts coming as the program weather indexed insurance is based on local weather indi- moves toward attaining the above objectives. ces. Payouts are triggered by the specified components of » A minimal or traditional program does not contrib- a weather index rather than crop yields. The components ute much to the national GDP. One has to decide of a weather index include measurable weather variables, if it is even worth doing, as the small amount of such as temperature or rainfall; a specified duration; and resources deployed might have a bigger impact if specific weather stations. Once the weather data have been used for other purposes. obtained from the agreed upon station(s), an index can be derived by reviewing how the weather variables have Again, the crops grown, agricultural practices followed, impacted past crop yields over time. In addition, a weather and environmental conditions found in Ethiopia and index will account for the impact of weather factors on Kenya are very different from those in the regions exam- crops during different stages of development. For exam- ined in the two studies adapted here. It would be worth- ple, insurance will payout if data from a nearby weather while to carry out a careful agro-economic study of the station shows too much rainfall or not enough by a certain benefits likely to accrue from an agro-meteorological pro- key date in the growing season. The payout amount will gram in Ethiopia and Kenya, if only to have more confi- be indexed to historical accumulated rainfall for the area. dence in the findings. Multiperils insurance, based on a sequence of several indices covering the growing season, can encompass the OTHER BENEFITS full range of hazardous weather events, giving farmers There are also significant additional economic benefits comprehensive coverage. obtained through mitigation of and recovery from the impacts of hazardous weather and climate events, which Weather crop index insurance is being currently used in may require extraordinary decisions by farmers. Enhanced Malawi, Tanzania, Rwanda, India, and the Philippines. agro-weather and agro-climate forecasts and warnings The World Bank is reportedly working with Jamaica on can prevent, at least partially, some of the damage caused weather insurance for their coffee crop. A weather crop by drought and floods. A big element of this mitigation is index insurance program was launched in Kenya in 2010 the removal of the “surprise factor” wherein farmers are to insure tea, coffee, and corn producers against drought. caught unaware of impending hazardous weather. There is anecdotal evidence that this is helping to sustain the livelihood of some small-scale Kenyan coffee produc- Early warning of climate-scale events such as droughts or ers (Njagi 2014). flooding allows farmers to shift seed types to ones more tolerant of the anticipated conditions. Livestock can be Both the insurance company and the farmers rely on a moved or sold in anticipation of loss of pasturage. Aid neutral party, such as a national agricultural meteorology agencies can also ready emergency supplies if famine program, to ensure that the observing network is operat- begins to develop. ing continuously and producing good data. 36 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services well as be mailed to a central location at the end of a COST-EFFECTIVE specified time period. The cell call goes to an automated STRATEGIES FOR IMPROVING receiving system that records the data, does simple error AGRO-METEOROLOGICAL checking, and provides a first-order analysis. OBSERVATION SYSTEMS Such a cell-phone-based approach can provide critical In practice, many countries appear to have at least two temperature and precipitation data on a once-per-day agro-meteorological observation systems. One is based on basis. This is often a good first step toward operational the traditional instruments previously described and uti- agro-weather forecasting focused on providing farmers lizes observers or local volunteers with only modest train- with daily weather and climate updates and advisories ing. The other is a national network of automatic weather (perhaps through the same cell phone used to send the stations. When managed properly, the two systems com- data, but also through radio and other media). plement one another. Such upgrades to a traditional network cost a few hundred ENHANCED TRADITIONAL NETWORKS U.S. dollars per station, but US$50,000 to US$75,000 to As its simplest, an observing site in a traditional network install the central data-collection computer. There will be has only a small rain gauge. The several hundred station deployment and training costs to get the new equipment network reported for Ethiopia appears to be of this type. into the field and to show the observers or volunteers how The observations from such networks are usually recorded to operate it. There will be continuing charges for the cell by hand and often mailed into a central office. services and maintenance of the central computer. It may be desirable to train the operators at the central computer Such observations, when passed by a rigorous qual- to a higher skill level in programming and system admin- ity assurance process, provide invaluable data for both istration. Additional personnel should include a small long-term climate studies and monitoring of how given team of three or so quality assurance meteorologists who growing seasons are evolving. Again, Ethiopia appears to review the raw data in parallel with the automated proce- provide a good example. dures, flagging questionable data. The attractiveness of a traditional network is that it is (rel- MODERN AND ADVANCED NETWORKS atively speaking) inexpensive to establish and maintain. In There are several different costs involved in developing a some cases, the equipment, such as rain gauges, can be national agro-meteorological observation system. As sug- locally fabricated. The primary disadvantages are the lim- gested in the previous section, these can be categorized as ited amount of data provided and the slowness with which either capital or one-time costs or continuing costs. the data are collected. Examples of capital or one-time costs for each site. Such traditional networks can be enhanced by providing º Purchase or long-term lease of the site additional instruments together with the necessary train- º Civil works and site preparations as needed (access ing for properly reading them and recording the observa- road, security fence, electrical power if available tions. For example, many traditional stations of the U.S. or else a solar panel/battery system—this last is Cooperative Observer Network also include minimum a good back-up against outages even if electrical and maximum temperature data. To minimize costs, power is available) these can be obtained from instruments such as the Grove º Tower or mast Thermometer (which gives minimum and current tem- º Instrumentation peratures) or the Six’s Thermometer in (which gives mini- º Data logger and software mum, maximum, and current temperatures). º Telecom interface (could be microwave radio, cell phone connection, hard wire to Internet, and so A simple data sheet can be used to record the data by hand on; can be two-way to allow remote interrogation in a format that then can be transmitted by cell phone as of the data logger) Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 37 º Cables and related hardware will require close interaction with the technical staff º Labor for site improvements, installation, and so on that performs maintenance of the automatic weather º Miscellaneous costs, such as shipping costs related stations and the observers in the enhanced traditional to getting equipment on site network Site costs can run between US$10,000 and US$100,000 Software developers—responsible for the software in per site, depending on the number and types of in- the data loggers in the field and the central computer struments installed. A good mean cost would be Technical maintenance/calibration staff—responsible US$30,000. These costs do not include site acquisition for routine maintenance and calibration of equip- or preparation costs or shipping costs for equipment. ment, and for corrective maintenance as required Central data collection facility Staff costs are difficult to address because much of º Facility preparation (space modifications, provi- the relevant information, such as typical salary data, sion of power, provision for air conditioning, are not available. Further, the number of people re- dehumidification, heating as necessary) quired is not known. But if we assume 8 to 10 people º Telecom interface for receiving incoming data of the above types are assigned to operate, maintain, streams (and transmitting commands to data log- and further develop the hardware and software at the gers in the field) central control facility, and an average compensa- º Computer(s) for receiving, decoding, quality check- tion (salary + benefits) of $15,000/year, then we get ing, and basic analysis (this could be the same sys- $120,000/year to $150,000/year in staff tem that receives data from the traditional network) º Display system Operating costs Telecom charges Central data collection facility costs can run between Repair parts and spare instruments US$50,000 and US$350,000. These costs do not in- Expendable supplies for operating and maintain- clude facility preparation costs. ing the data collection system Vehicles, fuel, and related (a small group of vehi- Central maintenance facility cles are needed for all agro-meteorologists to go º Facility preparation (space modifications, provi- out in the field to study local conditions and to sion of power, provision for air conditioning, support technical maintenance/calibration staff dehumidification, heating as necessary) making routine site visits) º Facility furnishings, to include workspace for main- tenance and calibration, and storage space for Operating costs can run between US$25,000 per year spare instruments, repair parts, and expendables and US$250,000 per year. º Tools and calibration equipment º Miscellaneous costs, such as shipping costs In light of the above cost analysis, the following strategy should be cost-effective. Central maintenance facility costs can run between US$50 and US$250,000. These costs do not include facility Step 1a: In parallel, build up and enhance the exist- preparation costs or shipping costs for equipment. ing traditional network of rain and temperature ob- servations taken by observers and volunteers. Where Examples of continuing costs the cell network allows, provide observers with cell Staff—compensation, benefits, and training/profes- phones to call in their observations once per day. sional development activities. All staff should have ap- Step 1b: In parallel, establish an automated dial-in data propriate university degrees in areas related to mete- delivery system so that people can begin sending infor- orological observations or agro-meteorology mation. This computer should be of sufficient capa- Data analysts—responsible for all aspects of receiv- bility to service both the dial-in delivery of data from ing, decoding, and quality checking of received data; network of traditional sites, data from traditional sites 38 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services received via the mail, and the acquisition of data from Synoptic observing sites operated by the national hydro- the modern or advanced sites. Note that this central meteorological service data collection goes in at this early stage, so that it is Airfields ready to receive data from the automatic weather sta- University and high school grounds tions (as well as from the enhanced traditional network). Step 3. Begin with a small number of automatic Step 2. Select sites for the installation of the automatic weather stations deployed to key locations. As the sta- weather stations supporting agro-meteorology. Bear tions come online, evaluate how well the central data in mind that this is an agro-meteorology network, so collection computer is working. the stations will be concentrated in agricultural re- Step 4. Add more stations to lower-priority sites as gions with sites selected to be representative of local funds permit. This process could take several years. farming operations. Ideally sites would be within, but The staff at the central data collection facility and the distinct from, fields. However, this is not always prac- agro-meteorologists they support gradually come to tical, so sites at the edge of fields may be utilized. understand how to utilize the growing data streams to Some candidate sites, in descending priority, are as produce new products and services. follows: The data stream from both the enhanced traditional net- Agricultural farms and test plots maintained by the work and the growing automatic weather station network ministry of agriculture may have significant value to others in the government, Plots of land on large-size farms. Farmers with large the agri-business community, the regional water manage- landholdings are more likely than small landhold- ment/hydro-electric community, and so on. It is strongly ers to let one have a small plot in a field for this recommended that the raw data from both systems be purpose. Small landholders need to grow crops archived, together with relevant metadata and data qual- on every square foot they have. ity control flags where erroneous data are suspected. Fur- Other government facilities such as airports, military train- ther, the archived data should be accessible to the public ing areas, public utilities as quickly as is practical. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 39 CHAPTER FIVE MODELS FOR SUSTAINABILITY OF AGRO-METEOROLOGICAL OBSERVATION SYSTEMS National hydro-meteorological services face a wide variety of challenges. These include all or some combination of severely constrained budgets; limited infrastructure, with little or no technical support, internal or external; noncompetitive salaries, so limited ability to train and then to retain talented staff; lack of access to training or the latest information on weather and climate monitoring and forecasting in their region. Thus setting up, operating, and sustaining a nationwide (or at least agricultural region-wide) agro-meteorological observing network that collects data centrally in real time may seem to be an impossibility. To overcome the various obstacles to setting up and sustaining an agricultural net- work, the leadership of the national hydro-meteorological service must look outward, first within the government, then to international aid organizations, and finally to the private sector. The ministry of agriculture is an obvious partner, but the ministries of energy, transport, natural resources, environment, and emergency services should also be considered as potential partners. A business case specific to each agency can be developed showing the value of real-time data and the tailored products and services that flow from such data to the accomplishment of that agency’s mission. While not all agencies can provide financial support, access to land, facilities, professional and techni- cal staff, and morale support in budget requests can be in-kind outcomes. International development agencies can provide capital funds for initial equipment purchases and training of staff, as well as access to technical expertise not otherwise available. How- ever, most such aid programs are for a specific period of time, usually just a few years. Perhaps the greatest obstacle to operating and sustaining an agricultural network that provides data in real time to a central office is the cost of required telecommunica- tions network. This cost is one that few national hydro-meteorological services can cover. Unless the national government operates a nationwide telecommunication net- work (and some do, for law enforcement, military coordination, and the like) which Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 41 can be accessed at little or no charge by the NHMS, then address deployment, telecommunications, and mainte- it is necessary for the NHMS to consider nontraditional nance issues. Such leveraging can provide long-term sus- approaches. The following paragraphs discuss one such tainability for an observing network that otherwise might possibility: working with local telecommunications/cell be impossible. phone companies through a public-private partnership. Leveraging can occur through the following: One note of caution: public-private partnerships have » Utilize inexpensive feature phones provided to vol- been tried in many contexts. While some have worked unteer observers to send data from enhanced tra- well, others have not. It is important to recognize at the ditional observing sets and to return weather and outset that while the NHMS has a primary interest in climate information to users. providing improved service to the nation, the private » Make cellular telephone network towers (hereafter sector partner(s) are interested only in increased prof- “cell towers”) the locations of choice for the net- itability. Public-private partnerships must be carefully work of automatic weather observing stations. Ide- constructed so as to produce results that satisfy all the ally, instruments can be either on the tower or in parties involved. the security area at its base. » Utilize the communications network that supports Further, as an alternative to a partnership approach, the the cellular telephone network as the means to col- national authority regulating or licensing telecommunica- lect observations. tions/cell phone companies could require such compa- º If this is not feasible to put instruments on the nies provide such connection and transmission services to tower or in the security area, then a nearby NHMS and other government agencies at little or no cost observing station still can be linked to the through the regulatory process. While an attractive shortcut telecommunications network via a short-haul, to establishing a partnership arrangement, it likely would low-power microwave transmission system. create an adversarial relationship with the telecommunica- » Utilize the technical staff of the telephone company tions company, forestalling access to sites, technical staff, to install and maintain the observing equipment. and support within the government. In any event, while such low- or no-cost connections might be justifiable in the For the above to occur, business models are needed that case of true emergencies, such as flash flooding or wildfires, provide a “win” for the national telecommunications com- they are harder to justify for routine day-to-day services. pany and the national hydro-meteorological service. It is likely that to make such a partnership viable, the business arrangements will also involve some combination of pri- A Nontraditional but Viable Approach vate instrument manufacturing companies, international A nontraditional approach to meeting some of these chal- weather data aggregators/distributors, and international lenges to establishing and sustaining weather observing aid organizations. networks is through a partnership with a public or private utility company or service that operates a telecommuni- cations network. An obvious first choice is the cellular THE SPREAD OF THE telephone network, but one should not overlook the pos- CELLULAR TELEPHONE sibilities offered by railways, electrical utilities, aviation NETWORK ACROSS SUB- route control networks, law enforcement networks, and so on. Here we focus on the cellular telephone network, but SAHARAN AFRICA the principles would be more or less the same with other The leveraging of local cellular telephone networks is utilities, public or private. viable in large part due to the rapid rate at which such networks are spreading across the globe, including all of Leveraging cellular telephone networks as the foundation Sub-Saharan Africa. In developing nations, the cellular for an agro-meteorological observing network offers the phone has become a “leapfrog” technology. Such nations national hydro-meteorological service opportunities to will almost certainly never have a hardwired telephone 42 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services network. Everywhere there are people, cell towers are “non-standard” in this case is [sic] not correspond to becoming ubiquitous. Even rural areas are rapidly receiv- the sensors—all of them are standard, but it is cor- ing cellular telephone coverage. Populated areas often responding to the installation site. The point is that have several different telephone service providers. While the CIMO guide exactly describes all the require- there will likely remain areas without cellular telephone ments to the place, surroundings and the heights of coverage for some time to come, these are almost certainly installation of each individual sensor. The cell tower remote areas with few inhabitants. meteorological station is not satisfying any of these requirements. Where they exist as for-profit ventures, the telephone companies have strong incentives for maintaining the To address this criticism, the authors instrumented cell functionality of the equipment supporting the services towers and performed comparison studies of observations they provide. They have the cash flow to provide security, from two cell tower stations with those from correspond- reliable electrical service, and wide bandwidth connectiv- ing “nearby” (7. 2 kilometers and 5.4 kilometers distant) ity at each cell tower. The companies also have trained observing sites that meet the requirements of the WMO’s electronics and mechanical staff to properly install and Commission for Instruments and Methods of Observa- maintain the equipment at each cell tower site. The com- tions (CIMO) guide. panies are always looking for more paying customers and new services they can sell. Weather information, par- In their study, Bakhtin and his colleagues found surpris- ticularly agro-weather information, is likely an attractive ingly high correlations between observations of tempera- product for cell companies to market. ture, humidity, pressure, and wind speed at a cell tower site and a nearby standard station. Only the wind direc- In nations where the telephone company exists as a state- tion measurements exhibited a loss of correlation when owned or state-controlled enterprise, similar arguments the instrument was in the “wind shadow” of the cell apply, except that the company then has a stronger incen- tower. This situation can be monitored through a quality tive to provide important public services at very low cost assurance process because orientation of the wind sensor or even for free (though publically owned companies often relative to the cell tower is known from site metadata. An still have to recover all or at least a significant fraction of effective solution is to utilize two wind direction sensors, their operating costs). one on each side of the tower. The findings indicate that for many purposes meteorological stations mounted on cell towers are able to provide usable data for the main CONCERNS REGARDING meteorological parameters. REPRESENTATIVENESS Observations at cell towers have been challenged on the PUBLIC-PRIVATE grounds of failing to meet the WMO observing standards PARTNERSHIPS for representativeness. As an example of this common In the absence of a direct budget increase from the national criticism, Bakhtin et al. (2012) commented: government, a sustaining arrangement with the private sector—broadly termed a “public-private partnership”— [S]ome scientists, and decision makers have ex- is a potentially viable approach to improving the capacity pressed a lot of doubt (the discussion was initiated of the national hydro-meteorological service for imple- by WMO secretariat on CIMO25-XV meeting in menting and sustaining an agro-meteorology program Helsinki at 2010) regarding quality of the data ob- and supporting observation network. Based on the litera- tained by such “non-standard” stations. The terms ture on development in Africa, it appears that Uganda has already adapted national policies in this direction and is enjoying some success, including, for example, the Mobile 25 CIMO is the WMO’s Commission for Instruments and Methods of Observa- tions. This is the group within the WMO that sets the standards for observations Weather Alert forecasts for fishermen on Lake Victoria and measurements of the atmosphere. (WMO 2012). Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 43 The foundation for such partnerships is the fact that weather and others (roughly analogous to the early morning farm data and products developed using that data have value. report, which the radio stations broadcast for free because While the value for the data from a single location is usu- they can sell commercial time at the beginning and the ally minimal, it increases rapidly as the area covered and the end of the report). density of coverage increases. Specialty data, such as might be produced by an agro-meteorological observing network, In reality, a public-private partnership can be quite also have a relative high value to certain business sectors. complex and have nuances that are not always appar- ent at the beginning to the staff of the national hydro- The situation is complicated by the fact that a national meteorological service. As an example, consider that a hydro-meteorological service owes its customers and users seemingly simple business arrangement for taking advan- a certain mandated level of service provided essentially free tage of the cell phone opportunity might entail a national of charge. However, by partnering with private sector enti- hydro-meteorological service partnering with the cellular ties, such as data aggregators and distributors, the national telephone company and a private sector weather company hydro-meteorological service can market its data and per- to establish and sustain a weather observing network. The haps some of its products both locally and around the world. resulting data are then provided to the national hydro- meteorological service at little or no charge for use within It may also be there are local and regional private sector the country while the private sector weather company and customers that the national hydro-meteorological service is the cell company cover their costs and make a profit by under no obligation to support anything more than the stand- marketing the data worldwide. However, such a simplis- ard weather and climate information provided to all its cus- tic arrangement is not likely to be attractive to either the tomers or users. Partnering with an appropriate private sector cellular telephone company or the private sector weather weather forecast company that will use the local data to pro- company, because it puts all the risk on these two entities. duce tailored products (for an appropriate fee) is another way of realizing some of the value inherent in the data. A solution is for the national hydro-meteorological ser- vice, with initial support from a development agency, to A telephone company may be interested in working take on some of the responsibility and some of the risk with the national hydro-meteorological service on pro- for sustainability through commercialization both within viding telecommunications and other services to sustain the country and perhaps a defined regional market- the observing network for a variety of reasons. A simple place. The role of the development agency is to cover the but important reason is to be seen as a good member of capital costs, as is the case today, while ensuring that the the community, providing service to the national hydro- NHMS staff receives the necessary training and profes- meteorological service on a pro bono (no cost) basis as the sional development required to be a full partner in such actual costs to the cell company are likely small. a multiplayer effort. The national hydro-meteorological service would remain responsible for the provision of Another reason is motivated by the fact that accessing those services mandated by its nation’s government and weather information is one of the major uses of cell phones. the sponsoring development agency, and will have to Partnering with the national hydro-meteorological service learn how to balance that role with internal and regional to bring in local data is a way of fostering the development commercialization efforts. Success by the national hydro- of applications for the dissemination of local weather meteorological service, demonstrated by continuing information via cell phones. This in turn fosters the sale operation of the observing network at five years after of more phones and more minutes of usage, increas- commissioning, could be rewarded by further investment ing revenue to the cellular company. It could be that the by the development agency. phone company would provide telecommunications and other support if the national hydro-meteorological service As a side note on this, the development agency can makes it the primary means by which the resulting deci- play important roles in fostering regional collaboration. sion support information is disseminated to the farmers This not only increases the size of the potential business 44 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services opportunity for the private sector companies but also is the country’s population against these hazardous events, essential to having an effective early warning network. the Direction Nationale de la Météorologique (National Meteorological Directorate) identified the need for an Unfortunately, few of the African national hydro- efficient weather information system that operated in real meteorological services have staff with the skill sets time to support storm monitoring, nowcasting, and the required (strategic planning; end-to-end systems engineer- production of storm warnings when appropriate. ing leading to acquisition and installation of high-tech equipment; new business development, marketing, sales; The Directorate partnered with Earth Networks, Inc. new product development and delivery) to fully play the and Cellcom (the national telephone company) to design, roles necessary to acquire, install, maintain, and sustain a install, and operate an early warning system. The resulting modern weather observing network. History suggests that system of lightning locating stations went online in August this is also true for most nongovernmental organizations 2013. A total of 12 such stations were installed at cell tower or government aid agencies. sites across Guinea. Each station consists of a total light- ning sensor combine with an automatic weather station. One option is carefully crafting public-private partner- ships with private sector entities that have these skills. The data from this system are used to produce localized These partnerships should stipulate that the private sector warnings on severe storms; radar-like, real-time displays partner will assist designated individuals in the national of storms over Guinea; rainfall estimates for various accu- hydro-meteorological service in acquiring such skills. The mulation times; and two-week hydrologic forecasts for international development community should encourage national and hydrological services and other government and foster such public-private partnerships by requiring a and industrial users. realistic sustainability plan as a condition for funding. Earth Networks has many of the upfront costs associated The private sector entities can be either local, such as the with installing and operating this network. That company cellular network company, the local electrical utility, or the is doing so as it views the Guinea network as a demonstra- national airline, or international. Examples of the latter are tion that this technology is mature and can work well in Earth Networks and Accuweather from the United States the difficult environments of Africa. and MeteoGroup from the United Kingdom; there are also entities that might be considered “semi-private,” such as MeteoFrance International, which could play useful roles. PHILIPPINES The Philippines are a subtropical nation of thousands of Finally, it is unfortunate, but we have an example of how islands spread over 300,000 square kilometers. Conse- not to attempt to build a weather network based on cell quently, the Philippine Atmospheric, Geophysical, and phone towers in Africa. This is provided by the “Weather Astronomical Services Administration (PAGASA), the Information for All Initiative.” This was a multipartner national hydro-meteorological service, faces daunting geo- effort launched with great fanfare in 2008 that collapsed graphic and logistic challenges in developing a national within about 18 months. These should be borne in mind hydro-meteorological observing network. PAGASA must when entering into a public-private partnership. also deal with several typhoons per year as the Philippines lie on the main tropical cyclone track in the Western Pacific. Two success stories suggest themselves as examples of how novel approaches can produce observing networks Recently, PAGASA, through its parent organization, the that are sustainable. Department of Science and Technology, received a spe- cial government appropriation to deploy about 85 auto- GUINEA matic weather stations and automatic gauging stations. The nation of Guinea is prone to severe local storms While these stations will be helpful when installed, this still during portions of each year. In order to better protect left many data void regions. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 45 BOX 5.1. WEATHER PHILIPPINES FOUNDATION TO ROLL OUT 1,000 WEATHER INSTRUMENTS IN 2014 To help provide accurate localized weather forecasts and mit- during severe weather situations,” said WPF General Man- igate risks during times of natural calamities, Weather Phil- ager Celso C. Caballero III. ippines Foundation (WPF) is completing its rollout of about To complement its initiative, the foundation trained about 1,000 automated weather instruments to various cities and 358 city and provincial risk reduction personnel, public infor- provincial capitals nationwide in 2014. mation officers and administrators nationwide to maximize Most of the instruments consist of automated weather sta- use of tools on its website www.weather.com.ph. tions (AWS) which are capable of determining temperature, WPF reported installation of 205 AWS in all cities and pro- pressure, humidity, wind speed and direction, solar radiation, vincial capitals of the League of Cities of the Philippines and and rain locally. Localized weather data is then processed to League of Provinces of the Philippines last year. provide the localized weather forecasts. The rollout of another 197 AWS were made possible by WPF completed 402 AWS in 2013 and conducted train- private donors and site partners such as mall chain SM ing on the use of this hardware as well as related software and Pilmico Foods Corp. Rappler, Globe Telecoms, Primer found in weather.com.ph. The provinces, cities and various Group, iAcademy, Geiser-Maclang and Cebu Pacific. media as well as private individuals relied on the AWS and its WPF is a non-profit organization founded by Aboitiz Founda- website for the latest forecasts during the onslaught of super tion Inc., Unionbank and Europe’s leading weather service typhoon Yolanda. provider mminternational (formerly Meteomedia). WPF aims to deliver critical and accurate localized weather infor- “This is in line with the foundation’s thrust to contribute in mation readily available through the WPF website, www disaster risk reduction and climate change adaptation. Tools .weather.com.ph. in the website, for instance, assist and provide complimen- tary information to LGUs in planning and decision-making Source: Extracted from Aboitiz Foundation, 2014. To address the need for a dense national hydro- In 2012, the foundation deployed 82 weather stations. By meteorological observing network, in 2012, Aboitiz the end of 2013, there were 402 stations deployed. The Foundation, UnionBank, and a European weather ser- goal is about 1,000 such stations across all the inhabited vice provider MMinternational (formerly Meteomedia) islands (box 5.1). The foundation’s automatic weather sta- founded the Weather Philippines Foundation, Inc. The tions are considered a supplement to the official stations intent of the foundation is to improve understanding operated by PAGASA. The data from PAGASA’s stations of the importance of nationwide disaster preparedness and MMinternational forecasts are made readily avail- and promote timely responses to threatening weather able to the general public through the WPF website (http: conditions. The first steps the foundation has made //www.weather.com.ph). have been to establish a weather portal and to begin deploying automatic weather stations. Since its found- It appears that this is a private venture, with the intent to ing, many local, provincial, and national organizations complement and augment PAGASA’s observing system. as well as large industrial and financial concerns have The data from the stations are used by MMinternational joined with the foundation to support its deployment of in its commercial weather forecasting operations around weather stations. the world. 46 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services CHAPTER SIX BEST PRACTICES FOR INSTITUTIONAL STRENGTHENING AND CAPACITY BUILDING For an agro-meteorological program to have a significant impact on overall national agricultural production, the organizations involved—primarily the national hydro- meteorological service, but also the extension program of the ministry of agriculture— must change to faster, more focused ways of providing decision support services to farmers. The comments that follow focus mainly on the national hydro-meteorological service, but parallel remarks likely apply to the extension program as well. Currently, the national hydro-meteorological services in many African countries base their forecasts almost completely on numerical model output from various meteoro- logical centers in the developed world provided at no cost through the World Mete- orological Organization.26 This model output is coarse in spatial resolution, with new output usually available every three, six, and twelve hours each day. Such output is suitable for forecasting on the synoptic (large) scale and, to some extent, on the sub- synoptic scale, but not for forecasting on the meso- and micro- (or “farm”) scale, the scales on which the weather events that impact people, livestock, and crops occur. In the current forecast process, only limited use is made of local data in part because of the small amount that are available. A consequence of this situation is that farmers do not receive timely agro-weather information at temporal and spatial scales of real use to them in planning and conducting day-to-day activities. Once the enhanced traditional observing network is in place and providing high spatial density, once-per-day observations of a few key variables, and the network of automatic weather stations starts to extend across key parts of the nation (specifically, the agricultural regions), forecasters at the national hydro-meteorological service will start to have timely access to much more local data. Applied meteorological studies utilizing these new data streams will produce new analysis and forecast products tai- lored to support farmers’ decisions. 26 These centers include, among others, those of the U.S. National Oceanic and Atmospheric Administration (NOAA), the United Kingdom Met Office, the European Center for Medium Range Weather Forecasting, Météo France, and the Japanese Meteorological Agency. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 47 To get the full value from these new data sources, refo- BOX 6.1. DEFINING CAPACITY BUILDING cusing and perhaps redeployment of staff and resources UNDER THE PIC DIRECTIVE will be necessary. Near-real-time quality assurance of the incoming data, even when prescreened by numerical sys- Capacity building can be defined as an evidence-driven process of strengthening the abilities of individuals, organi- tems, still requires careful examination by trained eyes. zations, and systems to perform core functions sustainably, Forecasters will need to consider meso- and microscale and to continue to improve and develop over time. Com- meteorological events, preparing nowcasts and short- munity capacity building often refers to strengthening the range, more detailed forecasts for the period 0 to 48 hours skills, competencies and abilities of people and communi- out. They will now be able to watch weather evolving and ties in developing societies so they can overcome the causes so develop capabilities to issue warnings of the imminent of their exclusion and suffering. According to the United onset of hazardous conditions. The agro-meteorologist Nations Development Programme: will participate in this short-range forecasting, producing In the global context, “capacity” refers to the abil- tailored products for the agricultural community. Since ity of individuals and institutions to make and the national network of automatic weather stations will implement decisions and perform functions in an build out gradually over a period of years, these changes effective, efficient and sustainable manner. At the individual level, capacity building refers to the will be evolutionary rather than revolutionary. process of changing attitudes and behaviours— imparting knowledge and developing skills while STRATEGIC AND maximizing the benefits of participation, knowl- edge exchange and ownership. At the institutional IMPLEMENTATION PLANS level it focuses on the overall organisational perfor- The evolving situation described in the preceding para- mance and functioning capabilities, as well as the graphs is a classic example of a capacity building effort ability of an organization to adapt to change. that involves individuals, organizational structures and Institutional Capacity Building addresses Capacity operating procedures, and an infusion of new technology. Building beyond the provision of education and train- Developing the ability to plan and then execute that plan ing of professionals. It aims to enhance the capacity as described in this section will result in greatly increased of governments, business, non-governmental groups managerial capacity and technical capabilities within the and communities to plan and manage resources efficiently and effectively. It also aims to improve national hydro-meteorological service. institutional arrangements for sustainable develop- ment. This implies addressing Capacity Building on To ensure that the end result is the one desired—an a long-term, strategic level. Concepts such as leader- effective agro-meteorology program for the nation— ship, awareness, and constituency building are part a comprehensive strategic plan together with a detailed and parcel of institution building. implementation plan are essential. The strategic plan should be short, providing a vision, clear statements of Source: Excerpt from Promoting International Collaboration (PIC). Undated. high-level goals and objectives, and target dates for meet- ing those goals. The implementation plan should be com- prehensive and detailed. It should describe the strategic operations, design studies, and requirements documents and tactical actions necessary to attain the goals, and lay that are hard to obtain any other way. out a detailed timeline for completing them. These plans must reflect a systems perspective. They are probably best In the case of establishing and sustaining an agro- developed using a backward planning approach. meteorology program, this means one must examine the “system” that extends from the observations taken in Planning with a systems perspective means accounting the field to delivery of products that influence decision for the behavior over time of a system as a whole in the making by farmers. In considering this chain of sequen- context of its environment. Systems concepts and lan- tial activities, one must avoid linear, reductionist thinking guage bring a precision to developments of concepts of that focuses too narrowly on isolated parts of this system. 48 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services Rather one must consider both the parts to the system and formance through “implementation” at the bottom (via especially the interconnections between them, and on how interconnections on the forward path) and through exten- the various parts interact through those interconnections sive and comprehensive “Verification and Validation” to produce the desired results. Some of these interconnec- processes (establishing a set of nested feedback loops). tions will be the forward path along which observations These last are all too often overlooked or glossed over in are made, data are collected and quality assured, analy- all too short “acceptance testing.”27 ses are made and forecasts produced and both delivered to uses. Other paths in the system may double back to It may be advantageous for present purposes to think in give rise to feedback loops. For example, the data quality terms of multiple but strongly coupled planning efforts car- assurance effort will result in communications to observ- ried out by working groups. The coupling comes through ers in the enhanced traditional network and to technical the overall goals and objectives for the agro-meteorology staff responsible for the corrective maintenance program, program and the associated concept of operations. so that data quality remains high. The “environment” includes the farming community, other government agen- CONCEPT OF OPERATIONS cies, the universities, private sector entities, and interna- Once the high-level goals and objectives have been estab- tional and nongovernmental organizations. lished, the first and in many ways most important element in the system design process is the development of a con- In general, backward planning means focusing first on the cept of operations for attaining those goals.28 The concept overall goals and objectives for the program as a whole, then of operations provides a conceptual view of how the agro- working backward, determining system requirements—that meteorology program will meet the needs of the various is, stating what is necessary in terms of staff skills and exper- users of its products and services without being overly tise, organizational procedures and policies, and technologi- technical or formal. The users may be farmers, colleagues cal capabilities—to attain the goals. For this to happen, as in the ministry of agriculture’s extension service, senior “Requirements and Architecture” and “Detailed Design” management in the ministry of agriculture and other min- are developed, they will include specific performance met- istries, and so on. The government or international agency rics that will form the basis for the testing and evaluation to funding the program during establishment and then dur- be carried out under “Integration, Test, and Verification” ing operations should be included as a “user.” The concept and “System Verification and Validation.” of operations should address in conceptual or scenario fashion how the needs of each of these users will be met. There are a variety of backward planning approaches that can help guide planning and ensure that all essential points are covered. One that is relevant to the develop- 27 Validation can be expressed by the query “Are you building the right thing?” ment of an agro-meteorology program is the V-model. Verification can be expressed by “Are you building it right?” Another way of The V-model emphasizes early identification of goals and expressing this difference is to note that verification is always done against the requirements (technical terms) and validation is always done against the real objectives, formulation of a concept of operations that world or the user needs as articulated in the concept of operations. reflects both user needs and the operating environment, 28 A useful overview of the concept of operations and where it fits within a sys- thorough and testable system requirements, detailed tems development process is available at the Mitre systems engineering guide design, implementation, rigorous acceptance testing of (http://www.mitre.org/publications/systems-engineering-guide/se-lifecycle- building-blocks/concept-development/concept-of-operations) as well as in IEEE the implemented system to ensure it meets the stated Computer Society, March 19, 1998, IEEE Guide for Information Technology—System requirements (system verification), measuring its effective- Definition—Concept of Operations (ConOps) Document (IEEE Std 1362-1998). This ness in addressing goals (system validation), and ongoing guide was originally intended for software development projects but is easily operation and maintenance. While rigid in process, the adapted to other types of projects. It does not specify the exact techniques to be used in developing the ConOps document, rather, it provide approaches that might V-model is flexible in application. be used. This document states that the main objective of a concept of operations is to “communicate with the end user of the system during the early specification The strength of the V-model is that the requirements stages to assure the operational needs are clearly understood and incorporated into and design phases are linked to the system’s actual per- the design decisions for later inclusion in the system and segment specifications.” Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 49 The relevant portions should be iterated with the users to specific design responsibilities to working groups. Senior ensure all concerned understand what is planned. managers must insist that the working groups coordinate with one another as they carry out their assigned tasks. The intent of the concept of operations is to facilitate a Each working group may also use the V-model to orga- common understanding of ideas, challenges, and issues nize its work. Regular systematic reports to senior man- related to establishing a national agricultural meteorology agement are essential for coordination. program between the national hydro-meteorological ser- vice and those it will support with this program. It should Possible working groups might include the following29: include the full range of factors that are needed to sup- » Agro-meteorological products and ser- port the establishment of the program (that is, policy and vices: What are the products and services that will procedures, organizational structures, leadership, person- be delivered by the program? How and by whom nel, training, materiel, facilities, and resources). It is also will these products and services be produced? Are critical that the concept of operations addresses annual new analysis tools or forecast techniques required? operating costs to be incurred once the program is estab- If so, how will these be developed and tested? What lished (that is, operation and maintenance of the observ- meteorological data streams are required by those ing system, applied research and development [R&D] to producing the new products and services? What develop new tools and techniques using the data streams, agricultural data are needed? Can these be sup- continued training and professional development for staff) plied by the ministry of agriculture? How will rou- and describes how these costs will be met. As the name tine validation and verification of these products suggests, the concept of operations should be stated at a and services be accomplished? What training will conceptual level, without much in the way of details or analysts and forecasters require? Outside the “reg- selection of a specific solution. ular” customers or users, how will data be shared? » Organization and staffing of the agro- As the national agricultural meteorology program comes meteorological program. How will the new into existence, the concept of operations helps sustain a agro-meteorological program fit into the existing common vision of the program. In accordance with the organization structure? How will it be connected verification and validation feedback loops in the V-model, to the extension program of the ministry of agri- the original concept of operations developed at the culture? A distributed approach, with most of the beginning of system acquisition should be updated after human resources well out in the field, may be opti- developmental and operational testing to convey how the mal, but how does that fit with overall ministry agricultural meteorology program will actually work. This and governmental policy? How many professional update is needed since the program may come to include (forecasters and analysts), technical (systems admin- some additional capabilities not originally envisioned at istration, software support, data quality assurance), the beginning, and may not include some capabilities that and support, maintenance, and calibration staff are were omitted for financial or other reasons. required? What are the required qualifications for each position? What are the initial performance For those individuals charged with developing the pro- expectations for each position? Who will evaluate gram, the concept of operations is a first step toward the performance of each individual in the program? developing system requirements. From where will these individuals come (internal transfers, new hires, or elsewhere)? What training will they require? Where will they be located? PARALLEL, COUPLED » Customer or user relations for the agro- DEVELOPMENT EFFORTS meteorological program. Who are the Once the initial concept of operations for the nascent agro- meteorology has been developed, senior management of 29 The various working groups listed here should be taken as illustrative, not the national hydro-meteorological service can then assign definitive, as should the questions that follow each one. 50 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services customers or users for agro-meteorological deci- to develop for calibration? How many spare sen- sion-support information? What information does sors, tools and equipment, and repair parts do we each one require? Does this change through the need initially? Is there adequate space somewhere course of the year? How will the products and for a maintenance and calibration and equipment services produced by this program be delivered storage facility? Should all this be kept centrally or to users? How will the users be trained to prop- should some of it be at regional locations? Who do erly apply the new products and services in their we want to be able to program the data loggers? decision making? What are the roles for the agro- How will the maintenance staff contribute to the meteorologist and the extension agents of the min- metadata file for each site? How many maintenance istry of agriculture in this training effort? What people are needed, and where will they be stationed communications medium is appropriate to each around the country? How many vehicles do they customer or user? Does the mass media—radio, need? What training does the maintenance and cal- TV, Internet—have a role in dissemination? How ibration staff requires and how will it be provided? will the information be packaged (tabulations, » Central data collection facility. Where will graphics, downloads, and so on)? How does each the central facility be located? What are the space, group of customers or users provide feedback on environmental conditions, power, telecommuni- the information they receive, and who is respon- cations, and security requirements for the central sible for acting on that feedback? facility and data archive? What size of computer, » Enhanced traditional network design. What capacity of storage system, emergency power sys- additional observations can reasonably be asked of tem, and what size of technical and maintenance local volunteers? What additional instruments and staff will be required to maintain everything in the related items are required? How many spares and facility? Who will be responsible for keeping the repair parts are likely to be needed? Who will be central collection computing systems up to date responsible for installation and maintenance? What is and functioning properly? Who will provide system the best approach to once-a-day collection of the vol- administration for the central collection facility unteer’s observations—should they use a cell phone? systems? Who will develop the software for screen- Should mail-in of observation forms continue to be ing the data as it arrives? What tests, by data item, required? How will quality assurance be performed should this initial screen provide? How many ana- on the incoming data? How will the volunteers in the lysts per shift will be required to ensure continuous field receive feedback on the quality of the observa- quality assurance of the incoming data, both from tions they provide? What training do these people the enhanced traditional network and the agro- require? Who should deliver that training? meteorological network? How will the computer » New agro-meteorological network design. screen and the analysts deal with errors discovered Where in the nation are measurements needed in data? What elements of metadata will be main- immediately? Where in the nation are measurements tained, and who is responsible for maintaining needed over the next three years? What quantities each element? What are the desired policies and need to be measured at each location? Longer term, procedures for handling missing or erroneous data what additional locations and measurements may as it goes into the data archive? be required? How do we provide for access, power, telecommunications, security, and maintenance at Note how often “training” is mentioned in the working each location? What capabilities do we want in the group descriptions. Making training and professional data logger? What command-and-control capa- development of operational staff has to be a top priority. bilities do we want from the central data collec- Similarly, it is essential that staff members who have been tion facility? What policies and procedures do we trained are subsequently properly utilized and retained. need to develop for routine and corrective mainte- Where deployment of new equipment is involved, it nance? What policies and procedures do we need is important to ensure that training is accomplished in Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 51 parallel with the deployment of the equipment so that and implemented by scientists, engineers, and students those trained can immediately make use of their new of the University of Oklahoma (OU) and at Oklahoma knowledge and skills. State University (OSU). At each site, the environment is measured by a set of instruments located on or near a EXAMPLE OF AN AGRO- 10-meter-tall tower. The measurements are packaged into “observations” every 5 minutes; the observations are then METEOROLOGICAL transmitted to a central facility every 5 minutes, 24 hours NETWORK per day, year-round. The Oklahoma Climatological Sur- A good example of a very effective agro-meteorology pro- vey (OCS) at OU receives the observations, verifies the gram serving a diverse group of farmers and others in the quality of the data and provides the data to Mesonet cus- agricultural industry, including state and federal agencies, tomers. It only takes 5 to 10 minutes from the time the is provided by the Oklahoma Mesonet. The Oklahoma measurements are acquired until they become available Mesonet is a network of 120 environmental monitoring to the public. For more detail on the Oklahoma Mesonet, stations covering the state of Oklahoma. It was designed see http://www.mesonet.org/index.php. 52 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services CHAPTER SEVEN FINAL THOUGHTS Ethiopia, Kenya, and nearby countries in Africa urgently need to find more effective adaptation strategies to manage growing weather and climate risks. This is particularly the case in regard to their agricultural base since so many of their citizens are small- holder farmers working at or just above subsistence levels. These weather and climate risks present such a complex set of challenges that government institutions, NGOs, and donor organizations cannot address them one at a time, in isolation from the others. A comprehensive systems approach—in the broadest sense of that phrase— appears to be essential. Enhancing and strengthening the capabilities and capacities of key government agencies, more specifically the national hydro-meteorological services together with the ministries of agriculture, water, energy, and emergency management, will be essential to build levels of resilience to the environmental changes anticipated to unfold in the next few decades. Given the limited resources available, it is likely that everyone involved in agriculture in this region, from the farm to the national ministry, will need to adopt novel collaborative approaches and become much more flexible in terms of plants sown and agricultural techniques used if farmers are going to continue to produce the food needed by the people of these nations. This will include utilizing new types of partnerships and collaborations, in particular ones involving the private sector. There are numerous issues that must be addressed to provide the range of agro- meteorological services that are almost certainly going to be essential in the medium future. This report has discussed some of them, focusing somewhat nar- rowly on possibilities for improvements to surface observations and measurements, given that the resulting data are the foundation for producing many types of prod- ucts and services, and particularly so for those supporting agricultural decisions at all levels. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 53 The coming changes will also provide unique opportu- aid agencies in agro-meteorology and agro-climatology nities. The NHMSs in this region have historically been provides a unique opportunity for the NHMSs to demon- starved for funds, resulting in minimal staffing, inability strate their value to their nation, build stronger relation- to retain key staff, deferred maintenance or replacements, ships with natural allies like the ministry of agriculture and a strong dependence on financial assistance from aid and influential private sector entities, and so become more and development agencies to carry out new initiatives. The competitive for increased government funding. recent renewed interest by several different NGOs and 54 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services REFERENCES Aboitiz Foundation. 2014. “Weather Philippines Foundation to Roll Out 1,000 Weather Instruments in 2014.” January 29. http://aboitizfoundation.org/AF /index.php?p=2502&type=2&sec=114&aid=11439&pg=. Ansalehto, A., E. Elomaa, M. Esala, and A. Nordlund. 1985. “The Development of Agrometeorological Services in Finland.” Technical Report Number 31, Finnish Meteorological Institute, Helsinki, Finland. Bakhtin, Roman, Arkadiy Koldaev, Yuri Lanin, and Sergey Sarychev. 2012. “Cell Phone Tower Mounted Meteostation and Standard Meteostation Data Four Seasons Inter Comparisons. RPA Typhoon.” Roshydromet. https://www.wmo.int/pages/prog /www/IMOP/publications/IOM-109_TECO-2012/Session1/P1_14 _Koldaev_Cell_phone_tower_meteostation.pdf. Davie, Timothy. 2002. Fundamentals of Hydrology. London: Routledge. Ericsson. 2013. Mobile Weather Alert. May 15. http://www.ericsson.com/the- company/sustainability_corporateresponsibility/technology_for_good/mobile _weather_alert. Global Humanitarian Forum, 2008. Weather Info for All. Geneva, Switzerland. http://publicintelligence.net/weather-info-for-all-initiative-2008-2012/. Hautala, R., P. Leviäkangas, J. Räsänen, R. Öörni, S. Sonninen, P. Vahanne, M. Hek- kanen, M. Ohlström, B. Tammelin, S. Saku, S. Seppo, and A. Venäläinen. 2008. “Benefits of Meteorological Services in South Eastern Europe—An Assessment of Potential Benefits in Albania, Bosnia-Herzegovina, FYR Macedonia, Moldova and Montenegro.” Project Number 16669. VTT Working Papers. VTT Technical Research Centre of Finland, Vuorimiehentie, Finland. Hut, Rolf. 2010. “Trans-African Hydro-Meteorological Observatory.” June 14. https: //www.youtube.com/watch?feature=player_embedded&v=Nh7GDD3Ssr8. Kenya Meteorological Department (KMD). Undated. “Table of Agro-Met Stations.” http://www.meteo.go.ke/obsv/agro.html. The Kofi Annan Foundation. 2013. The “Weather Info For All” Initiative. May 15. http://kofiannanfoundation.org/%E2%80%9Cweather-info-all%E2%80%9D- initiative. Laichena, J. 2010. “Mainstreaming Agrometeorological Advisory Services into the National Agricultural Policy.” http://www.future-agricultures.org/publications /search-publications/event-publications/fac-roundtable-on-climate- change/1049-mainstreaming-agrometeorological-advisory-services-into-the- national-agricultural-policy/file. Migiro, Katy. 2013. “With 62 Dead, Kenya Must End Drought-Flood Cycle— Red Cross.” Thomson Reuters Foundation. April 25. http://www.trust.org /item/20130425095508-c7hi9/?source=spotlight. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 55 Njagi, Kagondu. 2014. “Coffee Insurance Cuts Bitterness of Extreme Weather for Kenyan Farmers.” Thomson Reuters Foundation. July 18. http://www.trust.org /item/20140718062615-809c2/. Ogodo, Ochieng. 2012.“When Every Drop of Rain Counts: Managing Climate Risks in the Greater Horn of Africa.” CCAA Human Impact Stories. IDRC /CRDI. http://www.idrc.ca/EN/Resources/Publications/Pages/ArticleDetails. aspx?PublicationID=992. Osgood, Daniel E. 2010. “TR10–08 HARITA IRI Report to Oxfam America—Final Report for IRI MIEL Planning and Technical Support for HARITA Micro-Insur- ance Pilot USA.” http://iri.columbia.edu/publications/id=1002. Partanen, P. 2013. Personal communication. Population Institute. 2014. “Welcome to the Population Institute.” http://www .populationinstitute.org/?gclid=CNfBnPObmr0CFRQS7AodoBIA8w. Promoting International Collaboration (PIC). Undated. “Collaborating to Accomplish Change.” http://www.promoting-intl-collaboration.com/WhatWeDoCont.html. Red Orbit. 2010. “Mobile Phones, Weather Stations to Shield Kenya’s Farmers.” March 5. http://www.redorbit.com/news/international/1832510/mobile _phones_weather_stations_to_shield_kenyas_farmers/. Rojas, Oscar, and Shukri Ahmed. 2012. “Feasibility of Using the FAO-Agricultural Stress Index System (ASIS) as a Remote Sensing Based Index for Crop Insur- ance.” Forum for Agricultural Risk Management in Development. https://www .agriskmanagementforum.org/content/feasibility-using-fao-agricultural-stress- index-system-asis-remote-sensing-based-index-crop-. U.S. Department of Transportation, Federal Highway Administration. 2006. “Coor- dinated Freeway and Arterial Operations Handbook. Chapter 3. A Framework for Coordinated Operations in a Corridor.” Publication FHWA-HRT-06-095. World Bank. 2007. Agriculture for Development: 2008 World Development Report. Washing- ton, DC: World Bank. ———. 2014b. “Kenya: Country at a Glance.” http://www.worldbank.org/en /country/kenya. ———. 2014a. “Ethiopia: Country at a Glance.” http://www.worldbank.org/en /country/ethiopia. World Meteorological Organization (WMO). 2009. Guidelines for the Education and Train- ing of Personnel in Meteorology and Operational Hydrology—Volume 1/Supplement No. 2— Guidelines for Curricula in Agricultural Meteorology (WMO 258/Supplement 2). Geneva, Switzerland: World Meteorological Organization. ———. 2010. Guide to Meteorological Instruments and Methods of Observation (WMO-No. 8) (2008 edition, Updated in 2010). Geneva, Switzerland: World Meteorological Organization. ———. 2012. “Mobile Weather Service Improves Safety of Fishermen in Uganda.” Press Release. May 10. http://www.wmo.int/pages/mediacentre/press_releases /documents/lakevictoriafinal10may.pdf. World Meteorological Organization. 2013. Mobile Weather Alert: Communicating weather warnings via mobile communications. May 16. http://www.wmo.int /gsearch/gresults_en.html?q=mobile+weather+alaert&submit.x=0&submit.y=0. 56 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services APPENDIX A LESSONS TO BE LEARNED FROM THE WEATHER FOR ALL INITIATIVE “The world’s poorest are also the world’s most vulnerable when it comes to the impact of climate change, and the least equipped to deal with its consequences. Today you find cell phone towers in almost every part of Africa. We have never been able to establish weather moni- toring on that scale, until now. By bringing together the expertise and resources of different public and private actors, this project may help to save lives and improve the livelihoods of communities in Africa living on the frontlines of climate change.” Kofi Annan, former UN Secretary General, announcing the Weather for All Initiative Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 57 Unfortunately, to the best that can be determined, this THE WEATHER INFORMATION effort soon failed. Between November 2009 and May FOR ALL INITIATIVE—A FAILED 2011, contact was lost with all 19 weather stations.30 EFFORT Launched with great fanfare in 2008 by Kofi Annan and LAKE VICTORIA PROJECT(S) the Global Humanitarian Forum (GHF), the Weather After the initial launch of the WIFA initiative, a subproj- Information for All (WIFA) initiative was a public- ect within the initiative was started by the Kofi Annan private partnership formed to increase access to reliable Foundation. This was intended to improve the weather weather and climate information throughout Africa. As and climate information received by fishermen and farm- announced, the GHF, WMO, Earth Institute at Colum- ers around Lake Victoria (Kofi Annan Foundation 2013). bia University, mobile telecommunication companies The project was based in Rarieda Constituency in Saiya Ericsson and Zain, and NHMSs and governments of County, Kenya, and was supported by the Svenska Post- participating countries were to work together to install kodStiftelsen, Health and Climate foundation, Aga Khan 5,000 new automatic weather stations at new and exist- University and African Centre of Meteorological Appli- ing mobile network sites across the continent by 2013 cations for Development (ACMAD). (GHF 2008). As a first step toward building a lake weather observ- The project aimed to use the mobile telecommunica- ing network that would meet real needs, surveys were tions network to improve both the continent’s weather designed and conducted to determine how the communi- observing network and the availability of weather infor- ties around Lake Victoria used the weather information mation through the dissemination of forecasts and early they received and to determine what additional services warnings through mobile short message service (SMS; could meet their needs. Ericsson 2013). Like the WIFA initiative, this subproject also ended when This increased access to weather information was GHF ceased activities in 2010. It does not appear that any intended to help those rural communities most affected weather observing equipment was deployed. by and vulnerable to hazardous weather events and vari- ations in seasonal to interannual climate. For example, Less than a year later, management of the WIFA initiative seasonal outlooks could help farmers decide what and was passed to the ACMAD, where a new project known when to plant, likely increasing their crop yields and as Mobile Weather Alert was created to build off of the incomes. Communities, health agencies, and govern- work previously done under the WIFA initiative (WMO ments could also use early warning of weather patterns 2013).30 The objective of this new project was again to use to take preventative action to limit the spread of climate- existing mobile phone technology, infrastructure, services, sensitive diseases like cholera and malaria. Early warn- and applications to develop and demonstrate a sustain- ings could also be used to alert people on Lake Victoria, able warning service in the region around Lake Victoria. where an estimated 5,000 fishermen die each year due to storms and accidents (GHF 2008). Of the three planned Ericsson, WMO, MTN Uganda, the Uganda Depart- phases of the WIFA initiative, only the first was com- ment of Meteorology, the National Lake Rescue Institute, pleted before the GHF ceased all activities in 2010 due to and the Grameen Foundation worked together to test and lack of funds (WMO 2013). deliver a range of mobile communication options for pre- disaster weather alerts for fishermen. The pilot study was In this first phase, 19 automatic weather stations fabri- launched in May 2011 in Uganda at a district along Lake cated by Fairmount Weather Systems Ltd. were installed Victoria called Kalangala. A survey of 1,000 fishermen on existing Zain and Ericsson mobile network sites around Lake Victoria, in the nations of Kenya, Tanzania, and Uganda. 30 For details, see http://betastations.fairmountweather.com/all_locations_page.php. 58 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services found that they valued the possibility of receiving accu- º Despite good intent and high-level political rate and specific weather information to their mobile support, the organizers did not seem to have a phones in addition to the general weather forecasts on the business plan focused on funding for long-term radio (Ericsson 2012). sustainability; there was no obvious path to data or content commercialization with the insurance No information about the Mobile Weather Alert project industry, the energy industry, the media, and can be found after May 2011. consumers. Lack of such a sustainability plan likely deterred many of the potential investors WEAKNESSES LEADING one would like to attract to such an effort. » Technology and operations TO THE COLLAPSE OF º Technology focused on lowest-cost sensors and THE WIFA INITIATIVE related field equipment rather than robust, It appears that the WIFA initiative failed within two years environmentally appropriate, low-maintenance of being announced. Further, it appears to have accom- hardware. plished little beyond having introduced to the region the º There was minimal innovation or recognition of idea that cell towers might open opportunities not previ- opportunities offered by leapfrog technologies. ously considered. » Perhaps most importantly, lack of buy-in by the local national hydro-meteorological services in the The available reports suggest the following weaknesses in region the WIFA initiative led to its demise: º The limited involvement of, and hence limited » Lack of sufficient capital at start-up and subse- sense of ownership of the program by the services quent inability to attract funds for the realization in the participating countries (P. Partanen 2013, of the announced plan personal communication) almost certainly doomed º Simply put, broad ambition was not matched this program from the start. However, it is unclear with sufficient capital. Consequently, the pro- that the services, even if they had been involved, gram’s vision had to be continuously scaled back had the required expertise or resources to play the into various “pilot zones,” “concept testing,” and roles necessary for the success of the program. “community based impacts.” It finally reached º Siting concerns by the services over representa- the point where the broad national and interna- tiveness were not properly addressed. tional developmental impact had been designed º Public-private partnerships between the NHMS out of the initiative. In summary, WIFA prom- and the providers of the systems were not prop- ised too much for the available resources, and erly addressed. delivered too little to have a real impact. » Lack of application development º Unable to synchronize funding, an implement- º Impactful humanitarian applications of the data ing agency, measurement system manufactur- were never able to get off the ground; data col- ing, and deployment cycles, it failed to produce lection was seen as an end in itself. a coherent, sustainable program. WIFA’s failure º Little attention was given to converting data to became another example of the very different actionable information for commercial applica- timescales on which the public and private sec- tions such as energy or utility, insurance, TV, tors operate. and mobile media. Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 59 APPENDIX B KMS WEATHER AND CROP REVIEW KENYA METEOROLOGICAL SERVICE DEKADAL AGROMETEOROLOGICAL BULLETIN WEATHER AND CROP REVIEW FOR DEKAD 7, 2014 1–10 MARCH, 2014 HIGHLIGHTS ON RAINFALL AND TEMPERATURE Rainfall activities in the country tended for increment as compared to the previous dekad. Eastern Region received the highest amount of rainfall countrywide with Mwea station recording 47.7mm as compared to 53.7mm recorded in Kisii station in Lake region in the previous dekad. In Western and North Eastern regions, there was no rainfall within the dekad. In Rift Valley region, Narok station recorded the highest amount of Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 61 rainfall of 12.4mm. In the Coastal region, Mtwapa sta- Eldoret-Kapsoya tion received the highest amount of rainfall with 37.1mm. The station received no rainfall during the dekad. The In Nairobi Dagoretti received the highest amount of rain- average air temperature and pan evaporation reported fall of 6.6mm and in Lake region, Suba station recorded were 18.9°C and 82 mm respectively. There was no report 23.1mm. In Central region, Thika station recorded the on sunshine duration. highest amount of rainfall of 5.2mm No phenological report. There was a general increase in the maximum temperature country wide. Lodwar station recorded the highest maxi- CENTRAL KENYA HIGHLANDS mum temperature of 38.4°C compared to 37.7°C recorded AND NAIROBI AREA REGION in the Wajir station in the previous dekad. The minimum Nyeri temperature increased significantly in most stations in the The station received rainfall amount of 1.8mm over the country with Nyahururu station in Central region record- dekad. The average air temperature was 19.4°C. There ing the lowest temperature of 6.9°C compared to 7.5°C was no report on pan evaporation and sunshine param- reported in the same station in the previous dekad. eters. Maize crop was at the wax ripeness stage and in failure state due to insufficient rainfall with below normal yield being expected. For a more comprehensive summary of rainfall and other meteorological parameters. Kabete The station recorded rainfall amount of 4.2 mm over the CROP AND WEATHER REVIEW dekad. The average air temperature recorded was 19.6°C. FOR DEKAD 7; 1–10 MARCH There was no report on sunshine duration and the Pan 2014 Evaporation was 61mm. Maize was at flowering stage and in poor state due to insufficient rainfall. Below normal yield is expected. NYANZA AND WESTERN REGIONS Kakamega The station did not record any rainfall. The mean air Thika temperature was 23.4°C. There were no records on sun- The station recorded rainfall amount of 5.2 mm. The shine duration and Evaporation Pan. mean air temperature and pan evaporation recorded were 22.3°C and 57.8 mm respectively. There was no report No phenological report on sunshine duration. Maize was at harvest stage and in poor state as it had been adversely affected by insufficient rainfall. Below normal yield is expected for the crop. Kisii The station recorded rainfall amount of 8.1mm. The mean air temperature and Pan Nyahururu The station recorded rainfall amount of 0.2 mm. The Evaporation were 23.0°C and 57.6 mm respectively. No mean air temperature and pan evaporation recorded were records on Sunshine duration. 15.6°C and 71.7 mm respectively. There was no report on No phenological report sunshine duration. No phenological report. RIFT VALLEY REGION Dagoretti Kitale The station reported rainfall amount of 6.6 mm during The station did not record any rainfall. There were no the dekad. The average air temperature and Pan Evapo- records of average air temperature and sunshine dura- ration were 20.6°C and 62.6 mm respectively. No report tion. The Van Evaporation record was 61.5mm. on sunshine duration. Maize had been disposed at their wax No phenological report. ripeness stage. 62 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services FIGURE B.1. ANALYSIS OF RAINFALL, TEMPERATURE, AND VEGETATION CONDITIONS EASTERN KENYA REGION Maize and was at maturity stage and in fair state with below nor- mal yield being expected. Meru The station recorded no rainfall. The average air tem- Embu perature and Pan Evaporation recorded were 18.9°C and The station reported rainfall amount of 24.9 mm. There 57.5mm respectively. No report on Sunshine duration. was no report on Temperature, Van Evaporation and Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 63 Sunshine duration respectively. Maize was at maturity stage » Over the Northwestern counties (Tur- and in fair state with below normal yield being expected. kana, West Pokot and Samburu), Rains over few places in the morning in the first few Katumani (Machakos) days and sunny intervals for the rest of forecast The station recorded rainfall amount of 16.4 mm. The period. In the afternoons, expect mainly sunny average air temperature reported was 22.1°C. There was intervals with showers and thunderstorms over no report on pan evaporation and Sunshine duration. few places throughout the forecast period. The Maize was at harvesting stage and in poor state due to adverse effect wet condition is expected to resurrect then pasture and veg- by too much sun. Below normal yield is expected for the crop. etation in this region. » The Central highlands including Nai- robi area (counties of Meru, Murang’a, COASTAL REGION Kiambu, Nyeri, Nairobi, Embu, Nyanda- Msabaha rua, Tharaka and Kirinyaga), are expected The station received rainfall amount of 1.3 mm. The aver- to experience mainly cloudy early morning with age air temperature and Pan Evaporation recorded was rains over few places giving way to sunny intervals 28.8°C and 55.3 mm respectively. There was no report throughout the forecast period. In the afternoons, on sunshine duration. Mangoes were at 100 percent fruit setting sunny intervals, showers and thunderstorms are stage and in good state. expected over several places throughout the fore- cast period. The expected showers will continue to enhance Mtwapa the state of the crops and vegetation in the region. The station received rainfall amount of 37.1 mm. The » Northeastern counties (counties of Mars- average air temperature recorded was 28.1°C. There was abit, Mandera, Wajir, Garissa and Isiolo), no report on sunshine duration and Pan Evaporation. are expected to experience rains over few places coupled with sunny intervals throughout the day Mangoes were at flowering stage and in good state. during the forecast period. This dry condition will con- tinue to worsen the state of pasture and vegetation in this EXPECTED WEATHER AND region. CROP CONDITIONS DURING » Southeastern lowlands (counties of Taita THE NEXT 10 DAYS; 11–20 Taveta, Makueni, Machakos and Kitui), are MARCH 2014 expected to experience cloudy mornings breaking into sunny intervals in the morning session over » Counties within the Lake Victoria Basin, the forecast period and during the afternoon ses- Highlands west of the Rift Valley, Nyamira, sion showers over few places giving way to sunny Kericho, Bomet, Uasin-Gishu, Nakuru, intervals throughout the forecast period. Narok, Trans Nzoia, Elgeyo Marakwet, » The wet conditions forecasted will regenerate the crops, pas- Nandi, Laikipia, Kajiado, Vihiga and ture and vegetation in the region. Busia), are expected to experience rains in few » In the Coastal strip (counties of Mom- places in the morning during the first few days and basa, Malindi, Kilifi, Lamu, Kwale, etc.), the rest other days would be sunny interval. In the is expected to experience mainly showers over few afternoons, there would be showers and thunder- places, giving way to sunny intervals during the storms interchanging with sunny intervals through- morning session and sunny intervals in the after- out the forecast period. The afternoon showers will noon throughout the forecast period. continue to improve the state of the crops and the vegetation, » The showers expected will have a refreshing impact on the which are still in farms in the region. crops in this region. 64 Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services For feedback or further guidance, Contact: Director, Kenya Meteorological Services, Agro-meteorological Advisory Services Division, Dagoretti Corner, Ngong Road, P.O. Box 30259, 00100 GPO, Nairobi Tel: +254 (0)20 3867880-7/3876957/3873682; Fax: +254 (0)20 3876955 E-mail: agromet@meteo.go.ke; Website: www.meteo.go.ke ©2014 Kenya Meteorological Service Increasing Agricultural Production and Resilience through Improved Agro-Meteorological Services 65 A G R I C U LT U R E G L O B A L P R A C T I C E T E C H N I C A L A S S I S TA N C E P A P E R W O R L D B A N K G R O U P R E P O R T N U M B E R 94486-GLB 1818 H Street, NW Washington, D.C. 20433 USA Telephone: 202-473-1000 Internet: www.worldbank.org/agriculture