基于数值模式误差分析的气温预报方法
投稿时间:2019-03-05  修订日期:2019-06-14  点此下载全文
引用本文:
摘要点击次数: 24
全文下载次数: 0
作者单位E-mail
蔡凝昊 江苏省气象台,南京大气科学联合研究中心 nu911@163.com 
俞剑蔚 江苏省气象台,南京大气科学联合研究中心 radargroup@foxmail.com 
智协飞 南京大气科学联合研究中心南京信息工程大学  
基金项目:北极阁开放研究基金—南京大气科学联合研究中心基金资助项目( NJCAR2016ZD04) ; 华东区域气象科技协同创新基金合作项目(QYHZ201602)
中文摘要:本文采用欧洲中期天气预报中心(ECMWF)全球确定性预报模式地面气温和全国国家地面站点观测资料,对模式初值场误差、历史误差以及卡尔曼滤波预测误差与实况误差之间的相关性进行分析,设计了4种回归方案订正日最高、最低气温预报偏差,并与欧洲中心、中央气象台和全国城镇的预报产品进行了检验对比。结果表明:采用了模式近1-3天最高(最低)气温和模式最高(最低)气温历史平均误差、初值场误差以及卡尔曼滤波反演误差作为预报因子的改进方案效果最优,经对其2017年日最高和最低气温的预报检验,预报准确率较欧洲中心原始模式预报有较明显提高,也明显优于中央气象台指导预报。同时,与当前业务中质量最好的全国城镇预报相比,最高气温预报平均绝对偏差(Mean Absolute Error, MAE)较全国城镇预报低8.24%~13.97%,预报准确率提高1.24%~3.57%,日最低气温平均绝对偏差较城镇预报低9.43%~17.69%,预报准确率提高1.77%~2.72%。在3天的预报中,对24小时时效内预报相对于48小时和72小时的改进幅度更大,订正效果更加明显。
中文关键词:偏差订正,线性回归,初值场误差,卡尔曼滤波
 
Temperature forecasting method based on numerical model bias analysis
Abstract:Based on the European Centre for Medium-Range Weather Forecasts (ECMWF) of 2 m surface air temperature and automatic weather observatory data of China. The correlation between the initial field bias, the historical bias, the Kalman filter predicted bias and the real-time bias is analyzed. Four kinds of daily maximum and minimum temperature forecast regression schemes are designed. The comparison with ECMWF, CMA, and provincial forecasting has been investigated. The results show that the improved scheme predicted temperature, historical bias, initial field bias and Kalman filter inversion bias as the predictor is optimal. Its forecast quality for the maximum and minimum temperatures in 2017 is significantly better than that of ECMWF and CMA. Compared with the best-performing provincial forecasting, the maximum temperature MAE is 8.24%~13.97% lower than the provincial forecast, the forecast accuracy is increased by 1.24%~3.57%, and the daily minimum temperature MAE is 9.43%~17.69% lower than the provincial ones. The forecast accuracy rate increased by 1.77%~2.72%. And it has the greatest improvement within 1 day, that is, the correction effect decreases with the forecast lead time considerably.
keywords:bias correction  linear regression  initial field bias, Kalman filter
  查看/发表评论  下载PDF阅读器