基于风云三号气象卫星微波亮温资料反演东北地区土壤湿度及其对比分析

Soil Moisture Retrieval over Northeast China Based on Microwave Brightness Temperature of FY3B Satellite and Its Comparison with Other Datasets

土壤湿度在陆面过程中发挥着重要作用,是联系陆地水循环和能量循环的纽带和关键环节。遥感技术可以实时地对土壤湿度进行长期、大区域动态监测,已成为监测土壤湿度的有效手段。本文基于风云三号气象卫星(FY3B)微波亮温资料,利用基于微波能量辐射传输方程的陆面参数反演模型(LPRM)反演了中国东北地区的土壤湿度(FY3BLPRM)。把反演获得的FY3BLRPM土壤湿度与中国气象局农业气象站土壤湿度观测资料、NCEP和ERA-Interim土壤湿度再分析资料、欧洲空间局多卫星融合ECV(Essential Climate Variable)土壤湿度产品和国家卫星气象中心FY3B官方土壤湿度产品(FY3Boffical)进行了对比分析。结果表明:(1)本文反演的FY3B土壤湿度与农业气象站观测资料有较高的一致性;卫星土壤湿度整体上呈现自西向东逐渐增加的空间变化,与农业气象站观测资料较为一致;在季节变化上,两者高度相关,在大部分地区的相关系数都达到0.7以上。(2)在大兴安岭和东北三省东部等地表相对湿润的地区,FY3BLPRM土壤湿度与NCEP、ERA-Interim和FY3Boffical土壤湿度呈现较强的负相关;再分析资料和FY3Boffical土壤湿度均无法反映该地区土壤干湿状况的季节性变化,甚至与观测资料呈反位相变化特征。FY3BLPRM土壤湿度与欧洲空间局研发的多卫星融合ECV土壤湿度产品在绝大多数地区有较好的一致性。本文基于风云三号极轨气象卫星微波亮温反演的土壤湿度资料,可用于干旱监测、数值同化与天气预报、水文水资源等领域。

Soil moisture is of critical importance in land surface processes, which plays a key role in linking the terrestrial water and energy cycles. Remote sensing has become a promising technique for obtaining large-scale soil moisture information. In this study, the authors use the Land Parameter Retrieving Model （LPRM）, which is based on radiative transfer equations, to retrieve surface soil moisture from FY3B microwave brightness temperature images over Northeast China （FY3BLPRM）. The FY3BLPRM soil moisture is then compared with other datasets including in-situ observations at agrometeorological stations in China, the NCEP and ERA-Interim reanalysis of soil moisture, the multi-satellite soil moisture product from European Space Agency （ECV data）, and the FY3B soil moisture product officially produced in the National Satellite Meteorological Center of China （FY3Bomcal）. It is found that the spatial and temporal variations of the derived FY3BLPRM soil moisture agree well with that of in-situ observations. FY3BLPRM and in-situ observations both show higher soil moisture content in eastern Northeast China than in western Northeast China. The two datasets are highly correlated in most regions with correlation coefficients more than 0.7 mainly because they agree well with each other in seasonal variation. However, the FY3BLeRM is highly negatively correlated with the FY3Borncial and the NCEP and ERA-Interim datasets in some relatively humid regions such as the surrounding areas of Da Hinggan Mountains and eastern part of the study area. Further analysis suggests that the FY3BLPRM can realistically represent the seasonal variability of in-situ observations, whereas the other three datasets cannot. In most of the study area, the derived FY3BLPRM agrees very well with the multi-satellite ECV data. The derived FY3BLPRM soil moisture can be applied for drought monitoring and hydrology and water resource studies. It also provides realistic soil moisture information for data assimilation in numerical weather forecast studies.