江苏省土壤湿度观测资料与再分析资料对比分析研究

利用江苏省2010—2015年的60个站点土壤湿度观测资料,对欧洲中心ERA-Interim再分析资料(ERA)和美国宇航局再分析资料(MERRA)的两套土壤湿度数据在江苏地区的可靠性进行了评估。结果表明:相比于ERA再分析资料,MERRA较好地再现出江苏省次表层年平均土壤湿度的空间分布特征,但是两种资料的次表层和深层土壤湿度的数值均小于观测。ERA和MERRA基本都能揭示出江苏省次表层土壤湿度的季节变化特征,但是深层土壤湿度与观测仍有较大差距。在时间演变方面,ERA次表层土壤湿度与站点观测在研究时段内较为接近,EOF分析揭示出1979—2016年江苏省次表层土壤湿度存在区域一致型与南北偶极型两个主要的年代际变率模态。但是对于深层土壤湿度时间演变而言,两种再分析资料都与观测有较大的差距。总体而言,再分析资料的次表层土壤湿度与站点观测较为接近,但是由于再分析资料陆面模式中地下水等影响深层土壤湿度的关键过程刻画较为简单,使得深层土壤湿度与观测有较大的差距。

Comparative analysis of soil moisture observations and reanalysis in Jiangsu Province

Based on soil moisture observations of 60 stations data in Jiangsu province from 2010 to 2015, this paper evaluated the reliability of two sets of soil moisture data in the European Centre for Medium Range Weather Forecasts (ECMWF) ERA-Interim reanalysis data and the Modern-Era Retrospective Analysis for Research and Applications (MERRA) reanalysis data. The results indicated that MERRA performed better than ERA in reproducing the spatial distribution feature of yearly mean soil moisture of sub-surface layer in Jiangsu Province. The seasonal spatial characteristics of sub-surface layer of soil moisture were captured by both ERA and MERRA, but not for the deep layer. The analysis of time variation showed the ERA-Interim can well reproduce temporal characteristics of the sub-surface soil moisture, and the empirical orthogonal function analysis (EOF) show that the two decadal modes of the sub-surface soil moisture are characterized by a single positive center and an south-north dipole structure for the period from 1979 to 2016. But for the soil moisture of deep layer, both ERA and MERRA data show the poor ability to describe the temporal characteristics. In general, the sub-surface soil moisture of reanalysis data is close to the observations, but some key processes that could have an impact on the deep soil moisture simulations, such as the groundwater process, are relatively simple in most of the land surface models. This will lead to a large bias of the deep layer of soil moisture in reanalysis data.