Spatiotemporal data fusion for precipitation nowcasting

Vladimir Ivashkin, Vadim Lebedev

Abstract: Precipitation nowcasting using neural networks and radar data has become one of the key components of modern weather prediction services, but it is limited to the regions covered by ground-based radars. Truly global precipitation nowcasting requires fusion of radar and satellite data. We propose the data fusion pipeline based on computer vision techniques, including novel inpainting algorithm with soft masking.

Keywords: precipitation nowcasting, data fusion, inpainting

TL;DR: We propose a data fusion pipeline to include satellite data into precipitation nowcasting.