Evolution of Atmospheric Water Vapor and Cloud Liquid Water During Non- and Pre-Precipitation Conditions over the Middle Yangtze River Basin in the Warm Season

Quantifying the distribution and spatiotemporal variation of water vapor and liquid water is of great significance for understanding the atmospheric thermodynamic processes during extreme meteorological events. The water vapor and liquid water data obtained from ground-based measurements by three MP-3000A microwave radiometers (MWRs) over the middle reaches of the Yangtze River Basin were analyzed. Firstly, a comparison between MWRs and radiosonde was conducted, and the co-located observation results indicated that MWRs used in this study feature high detection accuracy and favorable consistency. The integrated water vapor (IWV) measured by one of MWRs (Serial No. 3115) was with the best performance for IWV observation, and the bias and RMSE were 0.22 cm and 0.18 cm. In addition, the detection biases of integrated liquid water (ILW) between three MWRs in pre-precipitation were smaller than those in non-precipitation. All three instruments captured the diurnal variation characteristics of vapor density (VD) and liquid water content (LWC) profiles. The variation in ILW and IWV in different stations showed that ILW maintained low values before precipitation and increased sharply during the pre-precipitation stage, indicating strong indicative significance for rainfall occurrence. The ILW increment was more remarkable in Wuhan station, where mostly covered with urban and water body underlying surfaces. However, the magnitude of IWV variation before precipitation was smaller than that of ILW, especially in Jingzhou station. Under non-precipitation condition, VD and LWC vertical profiles at the three stations were relatively stable. Before precipitation, they exhibited substantial increases with obvious spatial discrepancies: sharp growth in Wuhan, moderate enhancement in Xianning, and slight increment in Jingzhou. Overall, atmospheric water vapor and liquid water increase significantly before precipitation, and their distribution spatiotemporal differences are closely related to local underlying surfaces and precipitation characteristics, which can provide meaningful references for short-term precipitation forecasting.