Improving Forecasts of the “21⋅7” Henan Extreme Rainfall Event Using a Radar Assimilation Scheme that Considers Hydrometeor Background Error Covariance

Abstract

Abstract On 20–21 July 2021, a record-breaking rainfall event occurred in Henan Province, China, and a maximum hourly accumulated precipitation of 201.9 mm was recorded at Zhengzhou Meteorological Station. To improve the prediction of such extreme rainfall and to better understand the impacts of the radar reflectivity assimilation on forecasting, we assimilated radar reflectivity data using the hydrometeor background error covariance (HBEC) that includes vertical and multivariate correlations and then diagnosed the dynamic, thermal, and microphysical forecasts of this event. The results show that the radar reflectivity assimilation based on the HBEC properly transferred the observed radar reflectivity to the analysis of hydrometeors and other model states, and clearly improved the heavy rainfall forecast. The diagnosis of the dynamic and thermal forecasts indicated that the reflectivity assimilation based on the HBEC improved the convective environments of the precipitation systems, with stronger cold pools near the surface and deeper and wetter updrafts near Zhengzhou station, when compared with the experiment that did not assimilate radar reflectivity and the experiment that assimilated radar reflectivity without using the HBEC. The diagnosis of the microphysical forecasts further shows that assimilating reflectivity data using HBEC contributed to higher conversion rates of water vapor and cloud water to graupel and higher conversion rates of graupel and cloud water to rainwater, when compared with the other experiments. These improvements of both convective environments and microphysical processes within the convections ultimately enhanced the forecasts of this extreme rainfall event.