Bayesian Approach for Midseason Crop Yield Forecast and Crop Insurance Premium Calculation

Abstract

Both the indemnity and index-based have limitations. In addition, the lack of quality, consistent and timely data hinders the smooth functioning of risk assessment and claim settlement, which has an impact on the accurate pricing of insurance premia and farmers' satisfaction. Considering the growing challenges of climate change and the increasing vulnerability of farmers, we aim to develop a risk assessment model. Our work takes a digital data approach we integrate high-quality and consistent remote sensing data in the yield estimation. We further illustrate to forecast through Bayesian linear regression capturing the relation between the yield and regressors. Our findings suggest the use of the Bayesian approach over the traditional approach for crop insurance pricing. The proposed approach has the ability to capture the uncertainty in both the data and model parameters.