Spatially Smoothed Crop Yield Density Estimation: Physical Distance vs Climate Similarity

Many crop insurance studies have pointed out that considering spatial yield similarity can help provide more precise premium rating. We use Bayesian Kriging for spatial smoothing to consider such similarities when estimating crop yield densities. This article’s innovation is that the spatial smoothing is based on climate space, which is composed of climatological measures. We compare the climate-space smoothing with a general physical space (longitude–latitude space) smoothing. The test results are favorable to the proposed climate-smoothing method. Climate smoothing performs particularly well in states that have many missing counties and varied climate due to varying topography.