| 摘要: |
| 该文采用EC细网格2 m温度预报场及国家站实况资料，通过滑动平均法、双权重滑动平均法、多项式拟合法以及最佳系数法对EC细网格2 m温度在黔南州的预报进行订正分析，结果显示：4种方法订正后最高气温7 d平均准确率分别提升了12.70%~17.84%，最低气温7 d平均准确率分别提升了1.14%~2.86%。对于高温预报，最佳系数法订正效果最优，其次是多项式拟合法，对于最低气温，前3 d采用滑动平均法订正效果最明显，第4 d开始则采用最佳系数法订正，订正效果更明显。4种订正方法均在7—9月效果最明显。黔南州西部及中部地区多项式拟合法订正效果更好，州东南部地区最佳系数法与滑动平均法订正效果更好，但在120 h时效后滑动平均法的订正效果明显下降，州北部地区最佳系数法订正效果更为明显。 |
| 关键词: 黔南；滑动平均；双权重滑动平均；多项式拟合；最佳系数 |
| DOI： |
| 投稿时间：2021-03-03 |
| 基金项目:贵州省气象局科研业务项目（黔气科登[2020]07-10号）：EC细网格2 m温度的误差分析及订正研究 |
|
| Correction of Temperature Prediction by EC Fine Grid in Qiannan Prefecture |
| ZHAO Jie,CHEN Bo,MO Yidong,PAN Qixue,GU Xin |
| (Qiannan Meteorological Bureau of Guizhou Province,Duyun 558000 ,China) |
| Abstract: |
| In this paper, the EC fine grid 2-meter temperature forecast field and the national meteorological stations real-time observation data are used to revise and analyze the EC fine grid 2-meter temperature forecast in Qiannan Prefecture the moving average method, bi-weight moving mean method, polynomial fitting method and optimal coefficient method. The results show that after the four methods were revised, the 7-days average accuracy rate of the maximum temperature increased by 12.70% to 17.84% and the 7-days average accuracy rate of the minimum temperature increased by 1.14% to 2.86%. For high temperature forecast, the optimal coefficient method has the best correction effect, followed by the polynomial fitting method. For the lowest temperature,the moving average method has the most obvious correction effect in the first 3 days, and the optimal coefficient method has has the most obvious correction effect from the fourth day. The effect of the four correction methods was most obvious from July to September. The correction effect of the polynomial fitting method is better in the western and central areas of Qiannan Prefecture. The correction effect of the optimal coefficient method and the moving average method is better in the southeastern areas of Qiannan Prefectur. However, the correction effect of the moving average method is significantly reduced after 120-hours.The correction effect of the optimal coefficient method is more obvious in the northern areas of the Qiannan. |
| Key words: Qiannan Prefecture；moving average；bi-weight moving mean；polynomial fitting；optimal coefficient |
