摘要
格点气象要素预报是中国气象局目前的主推业务和未来天气预报的发展方向。本文基于欧洲中期数值预报中心ECMWF高分辨率模式2 m温度预报资料,在传统中央气象台站点温度指导预报SCMOC和回归方法建立的站点温度预报的基础上,提出"站点订正值向格点传递"的方法来订正格点温度预报。结果表明:(1)SCMOC站点最高、最低温度24~168 h预报误差<2℃准确率分别平均高于ECMWF的10.0%和23.1%,ECMWF存在较大的系统性偏差,最低温度预报偏高,最高温度预报偏低。(2)"站点订正值向格点传递"方法能够订正模式格点温度预报的系统误差,且整体上不改变原ECMWF温度预报场的空间形态和原模式对地形的刻画特征。(3)利用研究区域内98个县级站SCMOC温度预报,订正ECMWF格点场,返回到区域内1289个乡镇站进行检验,结果24 h最低、最高温度<1℃的准确率较ECMWF分别提高22.8%和11.9%,<2℃的准确率则分别提高29.7%和17.4%。最低(高)温度绝对误差平均减小0.99℃(0.69℃),平均误差(ME)下降到0.7℃(-0.9℃)以内。(4)通过一元线性回归,得到98个县级站的温度预报,返回差值场来订正格点场,也能较好地订正ECMWF的系统性误差。对比两种方法,SCMOC差值传递在最低温度订正方面有较大的优势,而回归方法的最高温度订正效果较好。此外,回归方法能够较好地改善逐时温度预报效果。该方法已成功运用于陕西省精细化格点预报业务系统中。
The grid element forecasting is the main business of China Meteorological Administration, and also the future development trend of weather forecasting operation. This article proposes the method of ＂station corrected value transfer to grid point＂ which uses traditional station temperature forecast issued by SCMOC of Central Meteorological Observatory and station temperature established by the regression method to correct the grid forecast data of ECMWF high resolution model 2 m temperature. The results show that （1） the accuracy rate of maximum and minimum temperature deviation less than 2℃ in 24-168 h of SCMOC was higher than that of ECMWF by 10.0℃ and 23. 1% respectively. There was large systematic deviation in ECMWF temperature forecast, whose minimum temperature forecast was higher and maximum temperature forecast was lower. （2） The ＂station corrected value transfer to grid point＂ method could correct the systematic deviation of ECMWF temperature forecast, and at the same time, keep the spatial pattern of forecasted field and topography characterization described by the original model unchanged. （3） Using SCMOC temperature forecasts from 98 county stations in the study area to correct ECMWF grid forecast data and returned the results to 1289 village stations for testing, we found that the accuracy rate of 24 h minimum and maximum temperature deviation -1℃ increased by 22.8% and 11.9% compared to ECMWF, and the accuracy of deviation %2℃ increased by 29.7% and 17.4%. The absolute error of the minimum （maximum） temperature decreased 0.99 ℃ （0.69 ℃） and the mean error de- creased 0.7℃ （-0.9℃）. （4） Using the temperature forecast of the 98 county stations by the regression method to correct the grid field could correct the systematic deviation of ECMWF as well. Comparing the two methods, SCMOC difference transfer has a great advantage in minimum temperature correction, and the regression method is better in maximum temperature correction. In addition, the regression method could improve the hourly temperature forecast effect. This method has been successfully applied to Shaanxi fining grid forecasting system.
出处
《气象》
CSCD
北大核心
2017年第12期1584-1593,共10页
Meteorological Monthly
基金
陕西省自然科学基金(2016JM4020和2016JM4011)
中国气象局预报业务关键技术发展专项(YBGJXM2017:03-13)
陕西省气象局重点和面上科研项目(2016Z-1
2016M-1)共同资助
关键词
格点温度
站点订正值传递
SCMOC
一元回归
grid temperature, station corrected value transfer, SCMOC, simple regression
作者简介
潘留杰，主要从事天气预报与研究工作．Email：pljmtgh57245@sina．com;通信作者：薛春芳，主要从事天气气候研究．Email：xuecf2002@163．com
