Assimilating GNSS PWV and radiosonde meteorological profiles to improve the PWV and rainfall forecasting performance from the Weather Research and Forecasting (WRF) model over the South China

Abstract

Precipitable Water Vapor (PWV) and rainfall play important roles in the meteorological processes. In this study, we have investigated the impact of assimilation of Global Navigation Satellite System (GNSS) PWV and radiosonde profiles on the performance of Weather Research and Forecasting (WRF) model in forecasting PWV and rainfall over the South China region for April 01, 2020 to May 31, 2020. PWV observations derived from 213 GNSS stations and meteorological profiles recorded by 23 radiosonde stations are assimilated into WRF model. Five different WRF schemes are adopted: WRF scheme 0: no data assimilation (DA); WRF scheme 1: assimilation of PWV from 76 GNSS stations; WRF scheme 2: assimilation of PWV from 213 GNSS stations; WRF scheme 3: assimilation of meteorological profiles from 23 radiosonde stations; WRF scheme 4: assimilation of both PWV from 213 GNSS stations and meteorological profiles from 23 radiosonde stations. PWV observations derived from 170 independent GNSS (have not been used in assimilation) and rainfall data recorded by 648 surface meteorological stations are used to evaluate WRF forecasting performance in PWV and rainfall, respectively. The results indicate that all DA schemes improve the WRF forecasting performance for both PWV and rainfall. For the first 6 h after data assimilation, WRF schemes 1 to 4 improve the PWV forecasting accuracy by 11.6%, 14.5%, 2.9%, and 14.8%, respectively. For the accumulated rainfall within the first 6 h after data assimilation, WRF scheme 2 and WRF scheme 4 have a similar performance and outperform other DA schemes while the WRF scheme 4 is the best one among all five schemes. WRF scheme 4 improves the rainfall forecast probability of detection and equitable threat score by 0.085 and 0.057, respectively.