Validation of GPM Dual-Frequency Precipitation Radar (DPR) Rainfall Products over Italy

Abstract The Ka–Ku Dual-Frequency Precipitation Radar (DPR) and the Microwave Imager on board the Global Precipitation Measurement (GPM) mission core satellite have been collecting data for more than 3 years, providing precipitation products over the globe, including oceans and remote areas where ground-based precipitation measurements are not available. The main objective of this work is to validate the GPM-DPR products over a key climatic region with complex orography such as the Italian territory. The performances of the DPR precipitation rate products are evaluated over an 18-month period (July 2015–December 2016) using both radar and rain gauge data. The ground reference network is composed of 22 weather radars and more than 3000 rain gauges. DPR dual-frequency products generally show better performance with respect to the single-frequency (i.e., Ka- or Ku-band only) products, especially when ground radar data are taken as reference. A sensitivity analysis with respect to season and rainfall intensity is also carried out. It was found that the normal scan (NS) product outperforms the high-sensitivity scan (HS) and matched scan (MS) during the summer season. A deeper analysis is carried out to investigate the larger discrepancies between the DPR-NS product and ground reference data. The most relevant improvement of the DPR products’ performance was found by limiting the comparison to the upscaled radar data with a higher quality index. The resulting scores in comparison with ground radars are mean error (ME) = −0.44 mm h−1, RMSE = 3.57 mm h−1, and fractional standard error (FSE) = 142%, with the POD = 65% and FAR = 1% for rainfall above 0.5 mm h−1.