资源环境科技发展态势分析平台

The Weather Research and Forecasting (WRF) model's data assimilation (WRFDA) hybrid ensemble-variational data assimilation (EnVar) system is used to examine the performance of the EnVar method for assimilating cloud liquid/ice water path products. To add flow-dependent features to background error covariance (BEC) of hydrometeors, hydrometeors mixing ratios (Q(c), Q(i), Q(r), Qs) are extended into analysis state vector for the "alpha" control variable. Then the BEC in the updated WRFDA-EnVar system combines the static hydrometeors BEC and flow-dependent hydrometeors BEC derived from ensemble forecasts. The updated system is evaluated by performing a series of single observation tests and two-weeks cycling assimilation and forecasting experiments by assimilating Cloud Liquid/Ice Water Path from NASA. The single observation tests show that the flow-dependent and multivariate SEC is introduced into the updated WRFDA-EnVar system by including extended hydrometeors analysis variables. The cycling assimilation and forecasting experiments demonstrate that by using the updated system included hydrometeors analysis variables, the root mean square errors (RMSEs) of analysis and forecasts are reduced and the Fractions Skill Scores (FSSs) of the precipitation forecasts are increased when compared with 3DVar method and the EnVar method without hydrometeors analysis variables. The diagnostics for a local severe rainfall case in the two-weeks cycling assimilation and forecasting experiments further show that through the application of the EnVar method included hydrometeors analysis variables, the convective available potential energy (CAPE) and humidity are increased effectively, and then better forecasts in terms of spatial distribution and intensity in accumulated precipitation are obtained, as well as cloud component.