doi:  10.3878/j.issn.1006-9585.2018.18082
沙尘预报客观检验中最优实况资料和圆插值半径选择研究

The reaserch of selecting the best obersvation and the best radius of circular interpolation during the objective verification of the sand forecast
摘要点击 480  全文点击 143  投稿时间:2018-06-26  修订日期:2018-09-06
查看HTML全文   下载PDF   查看/发表评论  下载PDF阅读器
基金:  
中文关键词:  沙尘  检验  圆插值  最优插值半径  适合检验实况资料
英文关键词:  Sand  Verification  Circular interpolation  The best radius of interpolation  The best observations for the obejective verification
  
作者中文名作者英文名单位
孟庆涛MENG Qing-Tao国家气象中心
引用:孟庆涛..沙尘预报客观检验中最优实况资料和圆插值半径选择研究[J].气候与环境研究
Citation:MENG Qing-Tao..The reaserch of selecting the best obersvation and the best radius of circular interpolation during the objective verification of the sand forecast[J].Climatic and Environmental Research(in Chinese)
中文摘要:
      本文使用2012~2014年每年3~5月预报CUACE模式地面沙尘浓度格点产品,预报员预报等级产品和实况地面观测沙尘等级和特征站观测PM10资料,对2012~2014年共25次沙尘过程值,采取三年总样本计算TS评分(针对CUACE模式格点产品,选取不同半径对格点产品进行圆插值),对格点产品适合的圆插值半径和预报产品适合的检验实况资料进行研究,结论如下:(1)如实况采用沙尘等级,模式插值半径与TS评分在浮尘或扬沙等级成近似线性下降关系,在沙尘暴以上等级TS评分与插值半径关系几乎不大,最优插值半径可选为最小插值半径0.5度;如采用PM10资料,TS评分在浮尘或扬沙等级根据半径不同变化较为剧烈,沙尘暴以上等级TS评分与插值半径关系不大,可以采用浮尘或扬沙平均线3.5度为平均最优插值半径。如果沙尘过程较弱,插值半径可适当减小,以1度最合适;(2)对CUACE模式预报产品,沙尘等级观测TS评分可用性要优于PM10资料,但两者在沙尘暴以上等级评分差别不大,总体上CUACE模式最优检验实况资料可选为观测沙尘等级资料。对预报员沙尘预报产品,观测沙尘等级TS评分要远高于PM10,预报员沙尘预报产品最优检验实况资料为观测沙尘等级资料。
Abstract:
      By utilizing the ground sand concentration netgrid products forecasted by the CUACE numerical model, the sand rank products forecasted by the forecaster, the observed ground sand rank and the observed PM10, this article calculats the TS score from 2012 to 2014(for the netgrid CUACE numerical model products, different radius was choosed by using the circular interpolation to generate the interpolated station forecast sand), which contains totally 25 days with sand weather, to select the best radius of the circular interpolation and the best observations used for the objective verification of the sand forecast. Several conculsions are given below: (1) for the sand rank observations, the radius are approximately linear declining with the TS score in the float dust rank, the TS score rarely changes when the radius changes in the sand rank above the sandstorm, the best radius is selected to the minmum radius, namely 0.5 degree; for the PM10, the TS score changes greatly when the radius changes in the float dust, it also rarely changes when the radius changes in the sand rank above the sandstorm, the best radius can be selected as the averaged 3.5 degree. If the sand is very weak, the radius can be properly decrease to 1 degree; (2) for the CUACE numerical model prodct, the usability of the TS score of the observed sand rank is better than the observed PM10, while the two is nearly the same in the rank above the sandstorm, so generally the best observation used for the objective verification of the CUACE numerical model product is the observed ground sand rank. For the sand rank forcasted by the forecaster, the TS score of the observed ground sand rank is much higher than the PM10, the best observation for the verification is the observed ground sand rank.
主办单位:中国科学院大气物理研究所/中国气象学会 单位地址:北京市9804信箱
联系电话: 010-82995048,010-82995413传真:010-82995048 邮编:100029 Email:qhhj@mail.iap.ac.cn
本系统由北京勤云科技发展有限公司设计
京ICP备09060247号