卫星反演湿度场及其在暴雨预报中的初步应用分析

THE STUDY OF PRELIMINARY APPLICATION OF SATELLITE-DERIVED RELATIVE HUMI-DITY IN RAINSTORM FORECAST

文中分别用两套相对湿度初始场方案 ,对 5个暴雨个例进行了预报对比试验 ,探讨了利用GMS- 5卫星图象反演的高分辨率湿度场资料改进初始场 ,以提高暴雨中尺度数值预报水平的可能性。结果表明 :( 1 )引入卫星反演湿度场能较好地反映出实际湿度场的水平中尺度结构 ,尤其是有效地增强了对降水有重要影响的高湿区的分析 ,且与实际探空资料更为接近 ;( 2 )引入卫星反演湿度场资料改进初始场一般会使 2 4 h的降水预报评分有所提高 ,特别是具有使预报前期 ( 0～ 6h)的降水场向观测场逼近的作用 ,但对预报后期 ( 6～ 2 4 h)降水场的影响具有不确定性。

The feasibility of improving rainstorm precipitation forecast by making an attempt to augment high resolution satellite derived relative humidity to initial conditions is studied. In forecast contrast experiments, total five rainstorm cases were tested with two relative humidity initial condition sources respectively.(1)CNMC T63L16 model generated global large scale assimilated forecast data with resolution 1.875 latitude×1.875 longitude;(2)GMS 5 four spectrum derived data with high resolution 0.052 latitude × 0.0633 longitude. The result indicates: (1) satellite derived relative humidity could better depict real relative humidity horizontal mesoscale structures, significantly it could enhance the analysis of high humidity regions where precipitation is sensible to and it agrees better with radiosonde observations;(2)the introduction of satellite derived relative humidity to initial conditions usually could improve 24h precipitation forecast scores, specifically, in earlier forecast stage(0 6h), it could nudge model generated precipitation field to observational field, nevertheless, in later forecast stage(6 24h) its effect on precipitation is not clear.