ERA-Interim气温数据在中国区域的适用性评估

运用中国756个观测站点的逐月平均气温数据,对比分析了ERA-Interim再分析资料的误差。结果发现:ERA-Interim再分析资料能够很好地反映观测值的年际变化,相关性达到0.955~0.995。ERA-Interim在580个站点的冷偏差或暖偏差小于1℃,占站点总数的76.7%,可信度较高。64个站点的冷偏差或暖偏差大于5℃,可信度较低。ERA-Interim在东部地区的暖偏差多于西部地区,冷偏差的高值主要集中在西部地区的高海拔站点。海拔低于200 m的站点偏差最小,适用性好,多数海拔3 000 m以上的站点呈现较大冷偏差,适用性较差。通过回归分析发现,观测站点与ERA-Interim格点的高度差是导致误差的主要原因,因此通过高程校正能够有效降低误差,提高ERA-Interim适用性。

Verification of ERA-Interim Reanalysis Data over China

In this paper, the monthly average temperatures derived from 756 meteorological stations over China were compared with ERA-Interim reanalysis data for the verification. The bias analysis indicated that the ERA-Interim reanalysis data could capture the inter-annual variability of observations very well with high temporal correlation values （0. 955 -0. 995 ）. ERA-Interim shows the good reliability at 580 （76. 7% ） stations with cooler or wanner bias smaller than 1℃. The weak reliability was found at 64 stations with cooler or wanner bias higher than 5 ℃. ERA-Interim reanalysis data haswarmer bias in eastern China than in western China, and greater cooler biases distribute at the high el- evated stations in the west of China. ERA-Interim is more suitable for lower elevated stations （ 〈 200m） than the majority of high stations （ 〉3 000 m）. The main reason for the bias is height differences among elevations of observations and ERA-Interim grid heights using linear regression analysis. Therefore. the bias could be significantly reduced via the elevation correction method.