Verifying Satellite Data with Farmer Reports

Elisabeth Gawthrop
Inside Index Insurance
2 min readDec 30, 2016

Index insurance uses satellite-based rainfall estimates to calculate how much money a farmer will receive in a drought. For index insurance to expand and reach more farmers, these estimates must be consistent with actual rainfall received by farmers. In 2016, rainfall reports from farmers in Ethiopia agreed pretty well with the index insurance satellite rainfall payouts. In this new animation, farmer rainfall measurements in Ethiopia are blue and satellite-based estimates are in yellow. Around 100 villages participated in this as part of the R4 Rural Resilience Initiative, which is led by the World Food Programme and Oxfam America.

“Check out how coherently you can see the main rainfall patterns sweep across the regions in the animation, evident for both the farmer and satellite — that’s what we need for index insurance,” said Daniel Edward Osgood, who leads IRI’s financial instruments sector team.

“The key with a good index is that it captures the main thing, e.g. was there a major seasonal drought? To work well, an index has to filter out the noise of the day-to-day small scale rainfall differences,” said Osgood.

But comparing satellite information and farmer reports isn’t as simple as colorful animation; Osgood’s team must pay attention to detail when analyzing the data or the insurance product won’t work. For example, said Osgood, satellites often see tiny rainfall amounts that are not visible in a rain gauge. Another issue is that sometimes rain gauges are visited every few days, leading to a delayed rainfall value that has a couple of days totaled.

Osgood added that this year, farmers and an Ethiopian-based field team from the REST organization were able to gather a lot of local farm data using cell phones. Osgood’s team is testing new phone-based prototypes with projects including CropText with Michigan Tech and an Earth Institute grant on Scaling Index Insurance through Mobile Networks. These prototypes are currently being tested in Ethiopia and the Dominican Republic, with hopes to expand.

The key to scaling this to reach more people is that farmers drive the design and validation for insurance demand and its impacts, said Osgood. That can’t be done at large scales unless we harness technology in new ways, but these innovative, experimental methods are difficult to fund in their trial period. If you are looking for an end-of-year gift opportunity, consider this effort, or email Osgood for more information.

--

--