NCEP/NCAR再分析温度·位势高度和风速场资料在上对流层/下平流层的适用性评估

Evaluation of NCEP/NCAR Reanalysis Temperature,Geopotential Height and Wind Field Data in the Upper Troposphere and Lower Stratosphere by Radiosonde Data

利用全球169个探空站资料检验NCEP/NCAR再分析资料,分析全球(主要是陆地地区)上对流层互下平流层位势高度、温度以及风速场标准误差的分布特征。结果表明,再分析风速场资料标准误差分布主要受急流区的影响,在急流区与实际风场存在明显的差异;并且风速场标准误差的分布具有明显的冬夏季节差异,在海岸线附近平均偏差较大;再分析位势高度、温度场的标准误差高值区主要集中在东半球低纬地区(印度洋沿岸),标准误差分布与其平均偏差分布基本一致,所以标准误差可以由平均偏差很好的解释;再分析温度、位势高度资料内陆地区标准误差偏低,高值区主要沿海岸线分布,风速场平均误差在海岸线附近比较大,这与探空站数据的质量地区差异及陆地观测站密集、海洋观测站较少密切相关。

Based on sounding measurements in 160 radiosonde stations over the world,the uncertainty of geopotential heights,temperature and wind fields in the upper troposphere and lower stratosphere in the NCEP/NCAR reanalysis data were analyzed. The analysis showed that the distribution of standard error of reanalysis wind field in comparison to sounding observations is related to the westerly jet,where the reanalysis wind field has significant difference with soundings. The distribution of standard error in the wind field has significant seasonal difference in winter and summer,and the average deviation was larger along the coastline. Larger standard error of reanalysis geopotential height and temperature is mainly distributed in low-latitude regions in the Eastern Hemisphere,such as the Indian Ocean coast,and smaller standard error is found in inland regions. The distribution of standard error in temperature field has a similar distribution with the average bias,which suggests that the average bias plays an important role in the formation of standard error. The standard error of reanalysis temperature and geopotential height inland was low,high value area distribute along the coastline. The average bias of wind field is larger along the coastline,which is possibly related to the regional differences of sounding data quality and the fact that there are more observing stations in inland areas and less at sea.