Increasing the flow of data

GIS improves rainfall data collection and information services in West Africa

Aondover Tarhule
Zakari Saley-Bana
Peter J. Lamb

Climate researchers have developed a system that uses GIS, computers, and the internet to improve rainfall data management and information delivery to farmers in West Africa.

Sub-Saharan Africa is highly dependent on rainfall. More than 90% of the land is used for farming, very little of which is irrigated. Despite this reliance on rainfall, there are relatively few monitoring stations in the region that gather the data that farmers need to plan their seasonal cultivation processes. Even in areas where rainfall data are collected, several weeks can pass before the information is processed and made available in a form that is useful to farmers.

To speed up and simplify the data collection and management procedures, a team of researchers from the University of Oklahoma has developed a geographic information system (GIS) that monitors rainfall and its seasonal patterns. Known as Rainwatch, the system can also automatically generate visual representations of the data that can be easily interpreted by interested parties, including farmers.

The team has initially tested Rainwatch in Niger, where the Direction de la Météorologie Nationale du Niger (DMNN) is responsible for monitoring weather and climate. The country suffered a severe drought in 2009, followed by its wettest year in a generation in 2010, and then a return to severe rainfall deficiencies in 2011. Although there are more than 200 stations in Niger’s rainfall monitoring network, most are ‘rain gauge only’ sites maintained by volunteer observers. They report rainfall data to DMNN’s operations office in Niamey once a day by telephone or radio. Only 14 stations transmit data on an hourly basis throughout the year, using telex and phone.

From these data, DMNN compiles rainfall reports that are broadcast on national and local radio and on national television – although TV reception is limited to the major urban areas. DMNN also publishes regular bulletins for the country’s eight provinces, and shares data with policy makers and the national committee for early warning and disaster management systems (Comité Nationale du Systeme d’Alerte Précoce et de Gestion des Catastrophes).

Although rainfall levels are broadcast on radio daily, it can take up to two weeks before DMNN releases data that have undergone any kind of analysis. Users of rainfall data outside the research community are not interested in exact rainfall statistics. Most farmers and other groups who depend on rainfall prefer qualitative information relating to previous seasonal patterns. Farmers, for example, simply want to know if the weather is dry, wet, or normal for the time of year. A long delay in delivering processed data means they cannot rely on the information, and cannot plan ahead.

Customised

Rainwatch was developed to alleviate such limitations, and improve the way rainfall data is collected, managed and disseminated throughout West Africa. The system consists of a database and a program that customises several functions of ArcGIS and MapObjects software. The database is linked to a graphics feature, which automatically updates the related charts and graphs as new data are added. The software adaptations make it easy for the user to process and view the data, and prepare it for publication and distribution.

When users log on to Rainwatch, they see a map showing the geographic locations of rainfall monitoring stations throughout the country. Users can click on the relevant icon to view the rainfall data for a particular station, then choose to compare the figures for a particular period of time against the median or with other years or even with the results from other stations. The user can then use the program to produce a variety of graphics to illustrate the data.

The number of sites and/or years that can be seen simultaneously is limited only by the amount of information on the database. Users can also request further analyses of the data to show the frequency and intensity of rainfall in certain areas, or view the occurrence of dry spells; information that is especially useful to farmers.

The results are, of course, only as good as the data. Rainwatch works best and provides the most accurate analyses when an optimum number of observers regularly contribute data to the system. By simplifying the data management processes, Rainwatch could be the catalyst needed for many countries to improve their rainfall monitoring procedures. The researchers hope the system will be adopted more widely throughout West Africa where other more complicated rainfall data dissemination systems have had limited success.

Increased availability
Rainwatch uses self-explanatory symbols and easy-to-understand terminology. When the system was tested, new users quickly became comfortable and could navigate their way through its processes within ten minutes. The test users also found it easy to follow the system’s logic, and fully understand and interpret the graphics they produced.

Based on feedback from users so far, the research team is developing an updated version of the program. Users suggested including a feature that would trigger an early warning system once rainfall, or lack of it, reached a certain threshold. Users also wanted to be able to export data to spreadsheets easily for further analysis. The upgraded Rainwatch will also include other climatic variables, such as temperature, streamflow, and soil moisture – which are linked to activities like irrigation scheduling.

Another important change will be to make Rainwatch compatible with free GIS software. This will make it available to users who cannot afford the product licence for the ArcGIS program, something that has been required up to now. A lower cost version of the system would make it possible to train more observers and equip observation stations throughout the region, which could act as local weather information centres. Providing more localised services could lead to farmers having a greater awareness of rainfall data, and a higher likelihood of being able to use the information. With all these improvements, the researchers hope that Rainwatch will become the African counterpart to Australia’s Rainman rainfall monitoring software.

The current version of the system is mainly suited for use by national meteorological services, and organisations such as the African Centre of Meteorological Applications for Development (ACMAD) based in Niamey, Niger. These centres could rapidly process and share their analyses with the media, researchers, educational institutions and agricultural extension services.

There are already plans to expand the use of Rainwatch beyond Niger. Starting in 2010, the long-term plan is to develop it into a web-based application that would be available to anyone with internet access. It could then be used alongside other climate information initiatives, using radio or cell phones, for example, to deliver weather details to people in rural communities.

The researchers believe that Rainwatch can especially benefit national meteorological services by improving the automation of rainfall data collection and database management. The ability to produce easily interpreted charts and graphics increases the likelihood that the information is distributed to more people. These are critical features in reducing the time lag between collecting the data and delivering it to farmers, and providing it in a format to help them adapt to a changing climate.

-------------------------------------------------------------------------------------------------------------------------------------------------------

Aondover Tarhule is chair of the Department of Geography and Environmental Sustainability at the Universi....
Zakari Saley-Bana is a former graduate student of the Department of Geography and Environmental Sustainability, the University of Oklahoma.
Peter J. Lamb is director of the Cooperative Institute for Mesoscale Meteorological Studies at OU.

-------------------------------------------------------------------------------------------------------------------------------------------------------

Related links

Direction de la Météorologie Nationale du Niger

African Centre of Meteorological Applications for Development (ACMAD)
Centre Africain des Applications de la Météorologie pour le Développement

RAINMAN

National Oceanic and Atmospheric Administration

Climate Knowledge for Action

29 November 2011

Copyright © 2014, CTA. Technical Centre for Agricultural and Rural Cooperation (ACP-EU)