利用包含大尺度稳定凝结潜热和对流凝结潜热信息的改进的湿Q矢量, 同时考虑地形抬升和地表摩擦作用, 完成对湿Q矢量释用技术的改进研究。将改进后的湿Q矢量释用（QMVIP）技术应用于华东区域中尺度模式（基于WRF V3.1, 简称WRF模式）, 并针对2010年6－9月汛期, 统计检验了QMVIP技术定量降水预报（QPF）效果, 且与WRF模式QPF效果进行了对比分析。结果表明, 0~72 h内, 对于0.1 mm·(24h)-1以上、 10.0 mm·(24h)-1以上及25.0 mm·(24h)-1以上降水预报来讲, TS评分分别提高了8.61%, 17.65%和29.57%; 正确率分别提高了8.44%, 6.49%和1.88%; 空报率分别下降了16.79%, 15.05%和10.74%; 漏报率分别下降了9.75%, 3.78%和10.48%； 对于50.0 mm·(24h)-1以上降水预报来讲, TS评分提高了65.04%, 二者平均正确率在0~72 h内几乎相同, 漏报率下降了12.94%, 空报率上升了12.67%。因此, 在总体上QMVIP技术较WRF模式改善了QPF效果。
A kind of moist Q vector interpretation technique applied to quantitative precipitation forecast (QPF), specifically, based on wind, temperature and dew-point temperature output by numerical prediction model, the vertical velocity can be obtained by solving the Omega equation whose forcing term is moist Q vector divergence, then the rain amount can be calculated by combining vapor, whereby dynamic interpretation method of QPF is produced. The moist Q vector interpretation technique will be improved by using the revised moist Q vector consisting of convective vapor condensational potential heating besides synoptic scale stable vapor condensational heating and by considering the roles of orographic lifting and surface friction. The modified moist Q vector interpretation technique (termed as QMVIP) is applied to Eastern China regional mesoscale numerical prediction model (which is based on WRF V3.1 and hereafter termed as WRF), and the QPF effect of QMVIP is verified with comparison to the counterpart of WRF from June to September 2010. The results of forecast statistical verification in 72h show that the test scores (TSs) and forecast accuracy of QMVIP forecast are higher than the counterparts of the WRF in the context of the rain with intensity over 0.1 mm·(24h)-1, 10.0 mm·(24h)-1 and 25.0 mm·(24h)-1, on average by 8\^61%, 17.65% and 29.57% respectively for TSs and by 8.44%, 6.49% and 1.88% respectively for accuracy, meanwhile, the false-alarm and miss rates of QMVIP forecast are lower than those of the WRF, on average by 16.79%, 15.05% and 10.74% respectively for false-alarm and by 9.75%, 3\^78% and 10.48% respectively for miss-alarm. Furthermore, for the rain with intensity over 50.0 mm·(24h)-1, TS is increased by 65.04%, accuracy is almost the same and miss rates are decreased by 12.94%, while false-alarm is increased by 12.67%. Therefore, QMVIP technique improves QPF effect in comparison to WRF on the whole. Finally, the further perfection of QMVIP technique and the improvement of operational QPF effect are discussed constructively.
[1]Sokol Z. MOS-based precipitation forecasts for river basins[J]. Wea Forecasting, 2003, 18: 769-781.
[2]Olson D A, Junker N W, Korty B. Evaluation of 33 years of quantitative precipitation forecasting at the NMC[J]. Wea Forecasting, 1995, 10: 498-511.
[3]Kumar A, Maini P, Singh S. An operational model for forecasting probability of precipitation and Yes/No forecast[J]. Wea Forecasting, 1999, 14: 38-48.
[4]Ebert E E, Damrath U, Wergen W, et al. The WGNE assessment of short-term quantitative precipitation forecasts[J]. Bull Amer Meteor Sci, 2003, 84: 481-492.
[5]Charba J P, Samplatsky F G. High-resolution GFS-based MOS quantitative precipitation forecasts on a 4 km grid[J]. Mon Wea Rev, 2011, 139: 39-68.
[6]刘还珠, 赵声蓉, 陆志善, 等. 国家气象中心气象要素的客观预报－MOS系统[J]. 应用气象学报, 2004, 15(2): 181-191.
[7]薛纪善. 和预报员谈数值预报[J]. 气象, 2007, 33(8): 3-11.
[8]Maini P, Kumar A, Rathore L S, et al. Forecasting maximum and minimum temperatures by statistical interpretation of numerical weather prediction model output[J]. Wea Forecasting, 2003, 18: 938-952.
[9]王迎春, 刘凤辉, 张小玲, 等. 北京地区中尺度非静力数值预报产品释用技术研究[J]. 应用气象学报, 2002, 13(3): 312-321.
[10]梁钟清, 金龙, 巩远发. 低纬度地区局地暴雨的神经网络预报方法研究[J]. 热带气象学报, 2009, 25(4): 458-464.
[11]胡邦辉, 刘丹军, 王学忠, 等. 最小二乘支持向量机在云量预报中的应用[J]. 气象科学, 2011, 31(2): 187-193.
[12]曾晓青, 邵明轩, 王式功, 等. 基于交叉验证技术的KNN方法在降水预报中的试验[J]. 应用气象学报, 2008, 19(4): 471-478.
[13]Gel Y R. Comparative analysis of the local observation-based (LOB) method and the nonparametric regression-based method for gridded bias correction in mesoscale weather forecasting[J]. Wea Forecasting, 2007, 22: 1243-1256.
[14]漆梁波, 曹晓岗, 夏立, 等. 上海区域要素客观预报方法效果检验[J]. 气象, 2007, 33(9): 9-18.
[15]Donswell C A. Flash flood forecasting: An ingredients based methodology[J]. Wea Forecasting, 1996, 11: 560-581.
[16]矫梅燕. 关于提高天气预报准确率的几个问题[J]. 气象, 2007, 33(11): 3-8.
[17]夏建国, 张芬馥. 利用有限区域数值预报产品改进暴雨客观预报的技术研究[C]. 暴雨业务预报方法和技术研究. 北京: 气象出版社, 1996: 34-36.
[18]刘还珠. 台风暴雨天气预报的现状和展望[J]. 气象, 1998, 24(7): 5-9.
[19]岳彩军, 寿亦萱, 寿绍文, 等. 湿Q矢量释用技术及其在定量降水预报中的应用[J]. 应用气象学报, 2007, 18(5): 666-675.
[20]岳彩军, 寿亦萱, 寿绍文, 等. Q矢量的改进与完善[J]. 热带气象学报, 2003, 29(3): 308-316.
[21]Sinclair M R. A diagnostic study of the extratropical precipitation resulting from tropical cyclone Bola[J]. Mon Wea Rev, 1993, 121: 2690-2707.
[22]姜勇强, 王元. 地形对1998年7月鄂东特大暴雨鞍型场的影响[J]. 高原气象, 2010, 29(2): 297-308.
[23]徐辉, 金荣花. 地形对2008年初湖南雨雪冰冻天气的影响分析[J]. 高原气象, 2010, 29(4): 957-967.
[24]董美莹, 陈联寿. 地形影响热带气旋"泰利"降水增幅的数值研究[J]. 高原气象, 2011, 30(3): 700-710.
[25]于群, 周发璓, 汤子东, 等. 冬季山东半岛局地性降水气候的形成[J]. 高原气象, 2011, 30(3): 719-726.
[26]马学谦, 孙安平. 祁连山区降水的大气特征分析[J]. 高原气象, 2011, 30(5): 1392-1398.
[27]朱乾根, 林锦瑞, 寿绍文, 等. 天气学原理和方法[M]. 北京: 气象出版社, 1992, 322-464.
[28]Miller B I, Chase P P, Jarvinen B R. Numerical prediction of tropical weather systems[J]. Mon Wea Rev, 1972, 100: 825-835.
[29]岳彩军. \!海棠\"台风降水非对称分布特征成因的定量分析[J]. 大气科学, 2009, 33(1), 51-70.
[30]曹钰, 苗春生, 岳彩军, 等. 引入对流凝结潜热作用对非均匀饱和大气中非地转湿Q矢量的改进研究[J]. 高原气象, 2012, 31(1): 76-86.
[31]Kuo H L. On formation and intensification of tropic cyclones through latent heat release by cumulus convection[J]. J Atmos Sci, 1965, 22: 40-63.
[32]Kuo H L. Further studies of the parameterization of the influence of cumulus convection on large-scale flow[J]. J Atmos Sci, 1974, 31: 1232-1240.
