A model for improving climate services in Africa

In developed countries, we are accustomed to having access to long and detailed records on weather and climate conditions, demographics, disease incidence and many other types of data. Decision makers use this information for a variety of societal benefits: they spot trends, fine-tune public health systems and optimize crop yields, for example. Researchers use it to test hypotheses, make forecasts and tweak projections from computer models. What's more, much of these data are just a mouse click away, for anyone to access for free. [Examples for climate and health]

Across much of Africa, however, it's a different story. By most measures, Africa is the most "data poor" region in the world. Wars and revolutions, natural and manmade disasters, extreme poverty and unmaintained infrastructure, have left massive gaps in socioeconomic and environmental data sets. Reliable records of temperature, rainfall and other climate variables are scarce or nonexistent. If they do exist, they're usually deemed as proprietary and users must pay to get access. This is not an inconsequential matter. Without readily available, reliable data, policy makers' ability to make smart, well-informed decisions is hobbled.

The problem of data access persisted even in Ethiopia, regarded as having one of the better meteorological services on the continent. Thanks to the recent efforts of Tufa Dinku, a climate scientist at the International Research Institute for Climate and Society, the situation has improved considerably.

"It used to be that in order to get data for a given place, you'd have to submit a written request to the National Meteorology Agency and then pay according to how much you needed. The process would take at least three days," says Dinku, who used to work at the agency. "Now it takes three seconds."

Tufa and members of IRI's Data Library team worked with his former colleagues at the NMA to develop a groundbreaking set of online climate data maps. The new maps allow any one to access 30 years of temperature and rainfall data for any spot in the entire country, down to a 10-kilometer square. This kind of coverage has never existed before in Ethiopia.

“ It used to be that in order to get data, you'd have to submit a written request to the National Meteorological Agency, and then pay according to how much you needed. The process would take at least three days. Now, it takes three seconds.”

"The information could be used in a number of applications, including long term investment and developmental studies, tourist information, disaster reduction and preparedness planning and more," said Dula Shanko, NMA's Deputy Director-General, during the official public launch of the maps in December 2011. "With this new service, NMA is shifting from one-desk service to one-click service," he said, referring to how easy it will be for anyone with an internet connection to access the data.

The project was funded by the U.S. National Oceanic and Atmospheric Administration and in large part by Google.org, the philanthropic arm of the technology company, which has been interested in improving the prediction and prevention of infectious-disease outbreaks in East Africa.

"The outputs of this project are a practical means of maximizing data to produce useful routine information," says Stephen Connor, the project's principle investigator, who is currently a fellow at the University of Liverpool. "While this effort was motivated by a health project, the resulting information products are of critical value to agriculture, livestock, water and other climate-sensitive development sectors in Ethiopia and elsewhere."

For example, IRI's Dan Osgood uses the hybrid data to develop index insurance contracts that help Ethiopian farmers lower their vulnerability to crop loss from droughts. "Perhaps most groundbreaking here is that NMA and IRI figured out a new way for climate data to be shared so that it does the greatest public good. This is something that sectors can actually use, not just about producing more data and improving it. It's starting off from the needs of the sectors."

In a paper published in the World Meteorological Organization's WMO Bulletin, Dinku and his coauthors detail the process that led to the collaboration with NMA and to the development of the new maps. He wrote the paper with NMA's Kinfe Hilemariam, the late David Grimes of the University of Reading's TAMSAT research group, and Connor.

The authors contend that by overcoming the problems of data availability, access and use they have been able to further the "effective and efficient use of climate information in Ethiopia." What's more, they argue that their experience in Ethiopia is a model which can be used to improve climate services throughout Africa.

The online map rooms are the culmination of years of work by Dinku to combine satellite rainfall and temperature estimates with on-the-ground station data. The hybrid data sets allowed Dinku to generate historical data for the whole country, not just where rain and temperature stations happened to be. Given that most are near highways and urban areas, rural areas stand to benefit greatly from Dinku's work.

"This alleviates the inadequacy of climate data particularly for rural Ethiopia, where such data is most needed and people are most vulnerable to climate variability," Dinku says.

Members of the IRI's Data Library group visited the NMA offices numerous times over the course of the project to help implement the new products. Rémi Cousin developed the maprooms and then trained the NMA's staff to use IRI's technology to disseminate the agency's climate information in a way that best met users' needs. John del Corral helped to port the existing mapping and visualization platform used by the Data Library to the agency's servers, and later helped design the new web site to host the climate maps.

The successes in Ethiopia have helped Tufa and his IRI colleagues start to make inroads in other countries. "We're exploring the possibility of doing something similar in Madagascar and Tanzania," Dinku says. "In many ways, the hard work has already been done. The raw satellite data we processed for Ethiopia is available for all of Africa, and the methodologies and computer coding we developed for the NMA can easily be adapted to another location."