Spatiotemporal Data Fusion for Precipitation Nowcasting
This work addresses the challenge of providing accurate precipitation forecasts for regions lacking ground-based radar, which is crucial for improving global weather prediction services.
The paper tackles the problem of global precipitation nowcasting by proposing a data fusion pipeline that combines radar and satellite observations, using a novel inpainting algorithm with soft masking to address limitations in regions without ground-based radar coverage.
Precipitation nowcasting using neural networks and ground-based radars 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 observations. We propose the data fusion pipeline based on computer vision techniques, including novel inpainting algorithm with soft masking.