Mapping weather risk – A multi-indicator analysis of satellite-based weather data for agricultural index insurance development in semi-arid and arid zones of Central Asia

Index insurance has been introduced as a solution to tackle several challenges that prevail in the agricultural insurance sector of developing countries. One of the main implementation challenges in these countries is the lack of reliable weather data for index development and implementation. The increasing availability of satellite data could ease the constraints of data access. Meanwhile, the suitability of various satellite products for yield estimation across world regions has to undergo a thorough assessment. This study contributes to the literature by systematically analyzing the accuracy of some globally available satellite data, namely the Global Satellite Mapping of Precipitation (GSMaP), Climate Hazards Group InfraRed Precipitation with Station (CHIRPS), and the Global Land Data Assimilation System (GLDAS) compared to ground-level weather information for 14 different indicators for the case of Uzbekistan. Our analysis indicates that those sources may provide the necessary data for an accessible and adequate climate service. However, a considerable risk of overestimation and underestimation depending on the source of satellite data may exist, especially for precipitation data in the conditions of Central Asia. Among the tested datasets, GSMaP showed a relatively better performance than CHIRPS in precipitation estimation for drought and flood detection. In order to reduce detection inaccuracy, the application of satellite weather products for index insurance is possible when temporal aggregation (e.g., monthly, seasonal) is considered. Globally available climate data could serve as a good source to establish index insurance products in Central Asia; however, a careful selection of source and index is required.