Study Of Planetary Boundary Layer Schemes In The WRF Model In Lower Reaches Of The Yangtze River

WRF is the latest numerical program model to be adopted by NOAA’s National Weather Service as well as the U.S. military and private meteorological services. It is also being adopted by meteorological services worldwide. Accurately simulating the meteorological processes within the planetary boundary layer(PBL) is important for rainfall simulations, planetary boundary processes have attracted much attention for its importance in rainfall simulations. WRFV3.1.1have seven different PBL schemes, each has its advantages and disadvantages.Thus, it is necessary to get a suitable scheme for rainfall simulations in the Lower Yangtze River regions of China. Furthermore, this paper identified five parameters in the relatively better PBL scheme(hereafter QNSE scheme) and examined resonably chosen set of parameters.By using Mellor-Yamada-Janjic, Quasi-Normal Scale Elimination,Yonsei University, Asymmetric Convective Model version2, Mellor-Yamada-Nakanishi and Niino Level2.5,Mellor-Yamada-Nakanishi and Niino Level planetary boundary layer schemes in the Weather Research and Forecasting numerical Model, seven sets of model simulations over three rainstorm cases in the lower Yangtze River were performed. This paper emphatically analyzed and compared the simulation ability of seven planetary boundary layer parameterizations to the distribution of24-hour total rainfall, boundary layer structure of the rainstorm area, fundamental fields of meteorological elements, Statistical testing results of total rainfall and fundamental fields of meteorological elements between simulations and observations indicated that the Quasi-Normal Scale Elimination(QNSE) planetary boundary layer(PBL) scheme is superior to the other six schemes.The empirical constants (A1A2, B1, B2, C1) that affect the vertical mixing within PBL in the model are the proportional coefficients of turbulent length scales and master turbulent length scale, Those constants are controlled by turbulent Prandtl numbers, they can be used with any other reasonable constants, as long as they are more suitable to model simulations in the Lower Yangtze River Regions. To know whether there exists more suitable parameters for the rainfall simulations in the Lower Yangtze River Regions, we first investigate the influence of the Quasi-Normal Scale Elimination PBL parameters to the rainfall simulations in Lower Yangtze River Regions, and then we conducted a sensitivity test over three characteristic cases for the parameters in QNSE PBL scheme in the version3.1.1of the Weather Research Forecasting Model. For each case, six different treatments of model parameters were used. Threat score and hit rate measures over three cases for each QNSE PBL scheme with specified constants were used to verify the rainfall fields. The results indicate that QNSE1.1with revised constants has higher scores over other QNSE experiments. Furthermore, we compared PBL height, potential temperature and water vapor mixing ratio (hereafter qvapor) profiles of six QNSE experiments with different values of parameters. A comparison of mean absolute bias of wind speed within PBL for different experiments were performed by using the wind profiler data. Results indicate that the QNSE PBL scheme with revised constants performs comparably better than other QNSE PBL schemes, suggesting that parameters used in QNSE with revised constants are more suitable for rainfall simulations in the WRFV3.1.1model in the Lower Yangtze River Regions of China.