Winter wheat yield index insurance design based on multi-source remote sensing and meteorological observation data

Abstract This study utilizes Sentinel-2 satellite data, remote sensing time series data MOD13Q2, and daily meteorological data from the meteorological station in Weinan City, Shaanxi Province. The aim is to accurately extract winter wheat planting areas and construct a remote sensing index system that is highly correlated with winter wheat yield. The index system covers wheat in Weinan City and includes meteorological data throughout the growing period. A multi-source fusion data prediction model for winter wheat yield is established using multiple linear regression methods to explore the relationship between multi-source data and winter wheat yield. Additionally, a comprehensive model is designed to calculate the claim thresholds and actuarial pure rates for index insurance products with an accuracy of 10 meters. The output prediction error rate of the final model is within 11%, indicating that this product can effectively reduce the basis risk of index insurance, improve the spatial accuracy of rate determination, and provide valuable insights for future index product research.