摘要
基于1983—2021年河南“三夏”期间(5月21日—6月20日)观测降水资料、NCEP/NCAR再分析资料及国家气候中心第二代月动力延伸预报模式(BCC_DERF2.0)5月20日起报的“三夏”期间降水和环流场,把河南省18个地市作为预报地区,检验模式对1983—2017年“三夏”期间河南省降水和环流的预测能力,用系统误差订正基于降水回归预测模型的解释应用和两者集合的方法订正了模式输出降水。主要结论如下:(1)模式输出降水量的气候态较观测降水量的气候态存在较大的正偏差,豫北的正偏差大于豫南的;时间相关系数(TCC)检验的高技巧区主要位于豫西北和豫中。(2)可以用模式输出的500 hPa青藏高原、西太平洋和贝加尔湖附近的高度场作为其中12个预报地区的预测因子建立回归模型预测降水。(3)检验订正期(1983—2017年)三种方法中各预报地区平均的同号率,系统误差订正的最高,TCC是集合订正后的明显高于其他的。各年平均的PS评分三种订正方法订正的均明显高于模式输出的。独立样本检验期(2018—2021年)系统误差订正后丰水年的PS评分减少,而枯水年的评分增加。各年集合订正的PS评分绝大多数介于系统误差订正和降水回归预测模型的评分之间,是最稳妥的订正方法。
Based on the observed rainfall,NCEP/NCAR reanalysis data,BCC_DERF2.0 model products(starting from May 20)including atmospheric circulation and precipitation in busy summer season(May 21 to June 20)during 1983-2021 in Henan Province,18 prefectural-level cities of Henan are taken as forecast areas,the prediction skills of BCC_DERF2.0 model for precipitation and atmospheric circulation during correction period(1983-2017)are tested.Bias correction,prediction regression models and integrated correction are carried out for the correction of the model-simulated precipitation.The results shows that:(1)There are large positive biases between the climate state of model-simulated precipitation and that of observed precipitation,and the positive biases in northern Henan are larger than those in southern Henan.The high-skill areas of time correlation coefficient(TCC)test are mainly concentrated in the northwestern and central parts of Henan.(2)Geopotential height over the Tibetan Plateau,western Pacific and Baikal at 500 hPa output by the model can be used as predictors for 12 forecast areas to establish regression models to predict precipitation.(3)Among the three correction methods in the correction period,the average ratio of same symbols of each forecast area is the highest in the systematic error correction,while TCC of integrated correction are significantly higher than others.The annual average PS score of the three correction methods is significantly higher than that of the model output.After the systematic error correction in the independent sample test period(2018-2021),the PS score of bias correction in wet years decreases while that in dry year increases.The PS scores of integrated correction are between that of bias correction and prediction regression models,so the method of integrated correction is the most reliable correction method.
作者
竹磊磊
杨婷
贺哲
许蓬蓬
Zhu Leilei;Yang Ting;He Zhe;Xu Pengpeng(CMA·Henan Key Laboratory of Agrometeorological Support and Applied Technique,Zhengzhou 450003,China;Henan Climate Center,Zhengzhou 450003,China;Henan Meteorological Observatory,Zhengzhou 450003,China)
出处
《气象与环境科学》
2023年第3期18-25,共8页
Meteorological and Environmental Sciences
基金
河南省气象局气象科学技术研究重点项目(KZ201705)
中国气象局·河南省农业气象保障与应用技术重点开放实验室开放研究基金项目(AMF201509)
中国气象局气候变化专项(CCSF201911)。
关键词
三夏
气候模式
系统误差订正
降水回归预测模型
集合订正
DERF2.0
busy summer season
climate model
systematic bias correction
application of linear regression
integrated correction
DERF2.0
