基于平均滤波算法的北疆春季日最低气温格点预报产品订正分析

本文选用中国气象局下发的0.05°×0.05°的国家级格点预报指导产品和中国气象局陆面数据同化系统（CLDAS）逐时实况数据资料，使用三种平均滤波方法分别订正北疆地区08时、20时起报的240h内逐24h 最低气温的格点预报指导产品，并分别对比检验订正前后共8种产品的预报效果。检验结果表明: 订正后的预报产品相比原始格点预报指导产品的均方根误差均明显降低，气温预报准确率及稳定性均显著提高。三种订正算法均随着海拔高度越高订正效果越好，且随着预报时效延长订正效果减弱。三种算法中集成订正效果略优。

Correction analysis of grid forecast products of spring minimum temperature in Northern Xinjiang Based on Average filter algorithm

This paper selects the 0.05°×0.05°national grid forecast guidance products issued by the China Meteorological Administration and the hourly real-time data of the CMA Land Data Assimilation System (CLDAS),which uses three average filtering methods to respectively correct the grid-point forecasting guidance products reported 08AM and 20PM for the minimum temperature of 24hours within 240hours of the northern Xinjiang region, and respectively compares the forecast effects of 8 products before and after the correction. The test results show that the corrected forecast product has a significantly reduced mean squared error compared to the original grid-point forecast guidance product. The accuracy and stability of the temperature forecast are significantly improved. The three correction algorithms all have a better correction effect as the altitude increases, and the correction effect weakens as the forecast time period increases. The integrated correction effect of the three algorithms is slightly better.