ECMWF和GRAPES(Global/Regional Assimilation and Prediction System)预报产品是国内目前主要的应用服务产品。为了了解ECMWF和GRAPES预报产品的性能,使用户在实际应用中,根据需求可选择性地应用上述预报产品,本文利用中国气象局2421...ECMWF和GRAPES(Global/Regional Assimilation and Prediction System)预报产品是国内目前主要的应用服务产品。为了了解ECMWF和GRAPES预报产品的性能,使用户在实际应用中,根据需求可选择性地应用上述预报产品,本文利用中国气象局2421个国家级自动站和8155个地面天气站(骨干站)逐时观测资料对2017年7月和11月、2018年1月和4月的ECMWF确定性预报模式(C1D)和我国研发的区域数值预报模式GRAPES_MESO、全球数值预报模式GRAPES_GFS的气温、地表温度、湿度、风速预报资料在中国区域的适用性进行了评估。结果表明:与各观测要素实况相比,3个模式均存在系统误差。地表温度预报易低估、风速预报易高估;3个模式预报能力普遍存在明显的区域差异、季节差异和昼夜变化。青藏地区3个模式预报能力明显低于其他地区。3个模式气温、风速的预报能力春季最差,湿度预报能力夏季最优,地表温度白天的预报能力秋冬季低于春夏季。GRAPES_MESO模式气温、风速的预报能力没有明显的昼夜变化;在分析的所有气象要素中,3个模式均为湿度的预报准确率最低,GRAPES_MESO模式的地表温度预报准确率最高,GRAPES_GFS模式和C1D模式风速预报准确率最高。展开更多
Based on the monthly precipitation data from 43 stations in North China from 1979 to 2008,the variation characteristics of precipitation in North China in recent 30 years were analyzed by means of empirical orthogonal...Based on the monthly precipitation data from 43 stations in North China from 1979 to 2008,the variation characteristics of precipitation in North China in recent 30 years were analyzed by means of empirical orthogonal function(EOF)decomposition,Morlet wavelet transform and Mann-Kendall test.The results showed that the spatial distribution of annual and seasonal precipitation was basically identical in North China,while the annual and summer precipitation from the middle of 1980s to the middle of 1990s were obviously more than these in other periods,and there was great annual variation in spring precipitation in 1990s,while autumn precipitation was higher from 1980s to 1990s and then went down after the beginning of 21st century,which was opposite to winter precipitation,namely there was more winter precipitation from 1980s to 1990s and fewer winter precipitation after the beginning of 21st century.In addition,the annual and summer precipitation changed abruptly in 1997,and there was no obvious change in spring precipitation and autumn precipitation,while winter precipitation had an abrupt change in 2000.Meanwhile,wavelet analysis revealed that the variation period of annual and seasonal precipitation was 3-4 years.展开更多
文摘ECMWF和GRAPES(Global/Regional Assimilation and Prediction System)预报产品是国内目前主要的应用服务产品。为了了解ECMWF和GRAPES预报产品的性能,使用户在实际应用中,根据需求可选择性地应用上述预报产品,本文利用中国气象局2421个国家级自动站和8155个地面天气站(骨干站)逐时观测资料对2017年7月和11月、2018年1月和4月的ECMWF确定性预报模式(C1D)和我国研发的区域数值预报模式GRAPES_MESO、全球数值预报模式GRAPES_GFS的气温、地表温度、湿度、风速预报资料在中国区域的适用性进行了评估。结果表明:与各观测要素实况相比,3个模式均存在系统误差。地表温度预报易低估、风速预报易高估;3个模式预报能力普遍存在明显的区域差异、季节差异和昼夜变化。青藏地区3个模式预报能力明显低于其他地区。3个模式气温、风速的预报能力春季最差,湿度预报能力夏季最优,地表温度白天的预报能力秋冬季低于春夏季。GRAPES_MESO模式气温、风速的预报能力没有明显的昼夜变化;在分析的所有气象要素中,3个模式均为湿度的预报准确率最低,GRAPES_MESO模式的地表温度预报准确率最高,GRAPES_GFS模式和C1D模式风速预报准确率最高。
基金Supported by National Key Technology R&D Program(2007BAC-29B02)
文摘Based on the monthly precipitation data from 43 stations in North China from 1979 to 2008,the variation characteristics of precipitation in North China in recent 30 years were analyzed by means of empirical orthogonal function(EOF)decomposition,Morlet wavelet transform and Mann-Kendall test.The results showed that the spatial distribution of annual and seasonal precipitation was basically identical in North China,while the annual and summer precipitation from the middle of 1980s to the middle of 1990s were obviously more than these in other periods,and there was great annual variation in spring precipitation in 1990s,while autumn precipitation was higher from 1980s to 1990s and then went down after the beginning of 21st century,which was opposite to winter precipitation,namely there was more winter precipitation from 1980s to 1990s and fewer winter precipitation after the beginning of 21st century.In addition,the annual and summer precipitation changed abruptly in 1997,and there was no obvious change in spring precipitation and autumn precipitation,while winter precipitation had an abrupt change in 2000.Meanwhile,wavelet analysis revealed that the variation period of annual and seasonal precipitation was 3-4 years.
