<
doi:  10.12013/qxyjzyj2019-028
DERF 2.0模式对江西月尺度气候预测能力的检验评估

Evaluation on monthly climate prediction in Jiangxi by the Second Generation Monthly Dynamic Extended Range Forecast Model
摘要点击 161  全文点击 69  投稿时间:2019-06-30  修订日期:2019-08-10
查看HTML全文  查看全文  查看/发表评论  下载PDF阅读器
基金:  中国气象局预报员专项(编号:CMAYBY2020 069); 江西省气象局重点项目“延伸期(16—30 d)逐日预报技术研究和业务建立”; 江西省气象局重点项目“江西省雾霾短期气候预测技术及大气污染输送—转化机理研究”; 江西省气象局面上项目“江西省月内强降温过程预测技术研究”.
中文关键词:  气候模式,预测能力,检验,偏差,原因
英文关键词:  DERF 2.0 model  climate prediction  evaluation  prediction deviation  cause
        
作者中文名作者英文名单位
马锋敏Ma Fengmin江西省气候中心
谢佳杏Xie Jiaxing江西省气候中心
唐传师Tang Chuanshi南昌市气象局
引用:马锋敏,谢佳杏,唐传师.2019,DERF 2.0模式对江西月尺度气候预测能力的检验评估[J].气象与减灾研究,42(3):170-177
中文摘要:
      基于国家气候中心第二代月动力延伸预测模式(DERF 20)输出的环流场、地面气象要素场和NCEP/NCAR全球再分析场以及江西省83站气象观测数据,利用空间距平相关系数(ACC)、距平符号一致率(PC)、趋势异常综合评分(PS)三种方法,检验了DERF 20模式对1983—2018年江西月降水和气温的预测能力,并进一步分析了模式对江西主汛期(6月)异常降水事件预测能力及造成偏差的可能原因。结果表明: 1) 月气温ACC总体上高于降水,冬季气温和春季降水ACC预报技巧相对较高;模式预测结果总体上能够反映出气温和降水的主要趋势,月气温、降水距平符号一致率均高于55%;模式对江西月降水和气温异常有一定的预测能力,降水PS评分比气温稍好,基本保持在70分左右。2) 江西主汛期降水异常偏少时,模式的预测能力较好,而降水异常偏多时,则相对较差,这可能与模式对东亚阻塞高压系统的低估,以及对西太平洋副热带高压模拟的西伸脊点偏西、脊线偏北、强度偏强有关。
Abstract:
      Based on the output data of the Second Generation Monthly Dynamic Extended Range Forecast Model (DERF 20) of National Climate Center, NCEP/NCAR reanalysis data and observed data from gauge stations from 1983 to 2016, the prediction of monthly precipitation and temperature by DERF 20 model are evaluated and analyzed using Anomaly Correlation Coefficient (ACC), Anomaly Symbol Consistency rate (PC), and Trend Anomaly Inspection Evaluation (PS) methods. Moreover, the prediction ability and deviation of anomaly precipitation in June are also discussed. The results indicate that the monthly temperature predicted by ACC is better than that of the precipitation. The performance of ACC on winter temperature and spring precipitation forecast is significantly better than others. DERF 20 model can reflect the primary trend of monthly temperature and precipitation in general, and their PC is greater than 55%. The anomaly precipitation and temperature can be simulated partly by the model. The PS of monthly precipitation is nearly 70, which is slightly better than that of monthly temperature. It has good performance when the precipitation in main flood seasons is poor, otherwise the model performs poorly. Comparison with NCEP/NCAR reanalysis data reveals that block systems in mid high latitudes are on average underestimated by DERF 20 model. The prediction is stronger for intensity, by north for ridge line, and by west for west ridge point than the observation. These may result in the bias of precipitation.
主办单位:江西省气象学会 单位地址:南昌市高新开发区艾溪湖二路323号
联系电话: (0791)82713175传真:0791-82713175 邮编:330096 Email:qxyjzyj@163.com
本系统由北京勤云科技发展有限公司设计