Spatial classification of precipitation from operational radar data

More Info
expand_more

Abstract

This research is intended to improve on existing rainfall estimation algorithms derived from operational radar data as well as to facilitate the study of spatial and temporal variability of rainfall. Using the C-band radar data of the KNMI at De Bilt (the Netherlands) several operational products with a grid of 256 by 256 pixels are produced with a resolution of 2.5km × 2.5km from which precipitation characteristics can be derived. Commonly a single Z-R relation is applied on CAPPI fields to create an estimate of rainfall. The proposed algorithm tries to make a distinction between convective and stratiform precipitation types and therefore will offer a chance to improve upon the estimation of precipitation. First an algorithm based on reflectivity intensity is to provide a first guess of convective and stratiform areas. Secondly separate reflectivity areas (RAs) are identified and tracked in consecutive images to follow their development. Within the first step CAPPI, lightning, echo-top height and Vertically Integrated Liquid (VIL) data can be used to identify more accurately different precipitating areas. Future plans also involve incorporating volume scans directly into the algorithm.