EVALUATION OF THE PARAMETERIZATION SCHEMES AND NUDGING TECHNIQUES IN GRAPES FOR WARM SECTOR TORRENTIAL RAINS USING SURFACE OBSERVATIONS

Abstract:

Sensitivities of parameterization schemes were conducted based on the Global/Regional Assimilation and Prediction System (GRAPES) model. Surface observations were used to evaluate the simulations and to improve the model’s ability to simulate the extreme precipitation over southern China on 20 July 2016. The results showed that GRAPES captured the large-scale precipitation over southern China but failed to predict the extreme precipitation over Xinyi. The model showed a systematic cold biases by adopting different parameterization schemes. In particular, the ECMWF analyses data showed a strong cold bias over Guangdong province and Guangxi Region. Observational nudging results showed that the surface temperature could largely help to alleviate the cold bias. The alleviation in the warm sector accounted for main improvement by the nudging scheme, and the RMSE was reduced by 1.56 degree from 3.25 degree to 1.69 degree by 1-h simulation and with 1.3 degree alleviation by 2-h simulation. Sensitivities using different parameterizations and the nudging scheme showed that the model’s underestimation of the precipitation was still present despite improvements in the predictions of surface temperature.