Influence Of Different Microphysical Parameterization Schemes On Forecast Of Heavy Rainfall

At present, the mesoscale observation and monitor are not completely in China, the main methods of studying mesoscale disaster weather system are also the high- resolution, nonhydrostatic mesoscale numerical models. From the prediction of numerical weather in China in operations at current situation, the horizontal resolution of the models was still not very precisely. But when the precise weather forecast to be come true, the needs of the operational forecast will increasingly rise for mesoscale numerical forecast productions. So they need us to rise the resolving capability and change the physical process in those models, making sure that the physical process and the resolving capability are matching. A lot of researches indicate that it is very important for the cloudy microphysical process to describe and show the precipitation process. This paper was used the next- generation nonhydrostatic mesoscale numerical models—Weather Research and Forecasting(WRF), rising the resolving capability to 4-12 kilometer, and contrastive analyzed different microphysical schemes in July 4th～5th 2003, which could influence the forecast outcome of a heavy rainfall weather process at jianghuai region in China, and take advantage of satellite cloud picture, TRMM satellite data, radar echo picture, the data of the earth automatic precipitation station and so on, utilized the outcome to examine the result of the model.The result make clear that the resolving capability was to reach 12 kilometer, the different imperceptibility microphysical processes was a little lower while forecasting the amount of precipitation, and some discrepancy while forecasting the center position and precipitation's strength, and some influence for the cumulus convective precipitation. The result of Lin scheme was inosculate with the center position, but it was reasonable that forecasted the amount of precipitation was still a little lower; The result of WSM6 scheme was the same as Lin scheme, but the center position was some depart west and north; the result of Thompson scheme was at the east center position from the Lin; the result of Ferrier scheme was simulated the lowest amount of precipitation for 1/5 to the actually happening. When the resolving capability reach to 4 kilometer, The result of Lin scheme was the same as the actually happening, the WSM3 changed a little; the WSM6 needed to change the strength of the precipitation; the Thompson was the same as the actually center position; the Ferrier was raised the strength of the precipitation. In the wake of raise the horizontal resolution of the models, the strength of the precipitation had some increased at a certain extent, the Lin, WSM6,Ferrier were sensitive to the resolving capability, the WSM3 and Thompson had a little increase depart from others. The Thompson scheme had improved the center position of forecast obviously, the WSM3 was not very sensitive to the resolving capability, this illustrated that the rising of the resolving capability could increase the sensitive of the models. When we see from the characteristic distribution of space-time, except from the Thompson's precipitation was appeared a little earlier, the other four plans could simulate the characteristic distribution of space-time of rainfall, but the later period of the rainfall had some difference from the fact, it maybe have something with the integral timeliness of the models. From the forecasting of cloud parameter, the WSM6 was the same as the Lin, but the content of the ice and cloud,the snow,the sleet were more than the Lin. It may be the main reason that the WSM6 weaker than the Lin on precipitation. The sleet forecasting of Thompson scheme was more than other two, moreover, the former period was stronger than the later period, so maybe it was the cause of the strength of precipitation which the former period was stronger than the later.