Farmer Perception, Recollection, and Remote Sensing in Weather Index Insurance for Agriculture in the Developing World: an Ethiopia Case Study
Abstract
New insurance products use remote sensing to trigger weather insurance payments for growing numbers of low-income farmers in developing countries. Because projects typically target data-poor regions with little validation information available, and are built upon participatory farmer approaches, data collected from farmers are increasingly relied upon to design accurate products. However, there are well known limitations in farmer recollections that could lead to problematic insurance if used inappropriately. We ask two questions: 1) is there evidence that farmer data has any information about actual drought events and 2) is there value in addressing recollection issues when using farmer data? We investigate issues concerning farmer perceptions and remote sensing for risk protection through one of the largest participatory insurance projects in the developing world, the R4 Rural Resilience Initiative of the World Food Programme (WFP) and Oxfam America in Ethiopia. Our approach is to test the cross-consistency of farmer-reported seasonal vulnerabilities against the years reported as droughts in independent satellite data sources. We use logistic regressions to test the extent to which remote sensing estimates can predict farmer's recollection of historical drought years in 86 communities in Ethiopia (gathered through farmer interviews) and rainfall deficits observed via satellite. This strategy compares prediction quality for alternate processing strategies of farmer recollection data-if there is evidence of a drought in biophysically based remote sensing estimates and that event is independently identified through farmer recollections, there is evidence that the event reported is not spurious. We compliment these with regressions and Heidke Skill score tables to map the seasonal timing of historical drought events. We find evidence that farmer reported events are reflected in multiple remote sensing datasets, and that utilizing strategies of repeated interviews over time, and to some extent, aggregating independent village reports over space lead to improved predictions. These findings are important to understanding the quality of and strategies for utilizing this information, and for verifying the appropriate remote sensing approaches as index insurance continues to scale.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFMPA22A..03P
- Keywords:
-
- 9810 New fields (not classifiable under other headings);
- GENERAL OR MISCELLANEOUSDE: 4328 Risk;
- NATURAL HAZARDSDE: 4337 Remote sensing and disasters;
- NATURAL HAZARDSDE: 6309 Decision making under uncertainty;
- POLICY SCIENCES