Blending of radar and gauge rainfall measurements: a preliminary analysis of the impact of radar errors

Abstract

Several methodologies have been proposed to combine radar and raingauge measurements with the aim of generating improved quantitative precipitation estimates (QPEs). These methods are based on interpolating point raingauge measurements (implicitly assumed to be “the truth”) and benefiting from the structure of the rainfall field as depicted by the radar. The use of a non-parametric approach based on radar measurements has been recently demonstrated, showing the benefits in the interpolation of raingauge measurements under the hypotheses of the Kriging approach. Several experiments have been carried out over a large number of cases and a variety of regions, Kriging with an external drift (i.e. the radar description of the rainfall field) being the approach showing more robust and (overall) better performance. Here, the impact of the discrepancies between two almost-collocated radars on the blended QPE fields was investigated.