Effect of GNSS RO on prediction of the 2021 Henan Rainstorm

Abstract

Abstract Precise prediction of the extremely heavy rainstorm is still challenging due to less or low spatial-temporal measurements. Nowadays, space-borne Global Navigation Satellite System (GNSS) radio occultation (RO) provides high spatial-resolution atmospheric parameters, which may improve the prediction precision of heavy rainfalls. In this paper, the impact of GNSS radio occultation on forecasting the heavy precipitation event is assessed for the extremely massive rainfall in Henan, China, on July 20, 2021. The GNSS radio occultation data from Constellation Observing System for Meteorology, Ionosphere, and Climate-2 (COSMIC-2), MetOp-A/B/C, Fengyun (FY)-3C GNOS are applied for assimilation in Weather Research and Forecasting Model Data Assimilation (WRFDA) of three-dimensional framework (3DVAR) system using the local refractivity operator. Control Experiment (CNTL) and RO are designed to examine the roles of GNSS radio occultation, and RO+GNOS is conducted to further evaluate the impact of GNSS RO data onboard FY-3C on this extreme rainfall. The fractions skill score (FSS) is used to quantify the accuracy of predicted precipitation at given thresholds. The 24-h forecast result shows that the experiments with assimilating GNSS radio occultation data produce better precipitation forecasts with regard to the distribution and the amount due to more precise initial conditions of the moisture field. In general, RO and RO+GNOS have similar increments for a more accurate humidity field near Henan and more explicit water vapor channels, and thus their predictions outperform CNTL. Compared with RO and CNTL, RO+GNOS exhibits the higher marked FSSs for heavy rainfall forecast at 50 mm and 100 mm thresholds, with average advancements of 7.76% and 32.55% for the 50 mm threshold, and 10.50% and 47.39% for 100 mm threshold, respectively. For the 48-h and 72-h forecasts, three experiments exhibit similar results that RO+GNOS gives the best performance in rainfall predictions, followed by RO and then CNTL. Overall results imply that GNSS radio occultation data has a noticeable enhancement for the prediction of this record-breaking rainfall, and data from GNOS onboard FY-3C plays an indispensable role.