干旱半干旱区土壤含水量反演与验证

Validation of soil moisture retrieval in arid and semi-arid areas

基于MODIS遥感影像和表观热惯量法,以新疆为研究区,建立了适用于干旱半干旱区1 m土体的土壤含水量反演模型。模型根据高表观热惯量,高土壤含水量,低表观热惯量,低土壤含水量这一理论,通过日地表温差和宽波段反照率确定土壤含水量的时空变化。假设通过1 m土体的土壤水通量正比于上下底层土壤含水量的差值,利用水平衡方程建立土壤表面和底层土壤含水量关系方程,并利用中国土壤类型特点确定优化模型。通过验证结果表明,壤土和壤质粘土这两类土壤含水量接近真实值,砂土在区域验证中,模拟与实测差值为2.16%,整个模型模拟精度较好,能够准确地从时空上反演干旱半干旱地区1 m土体的土壤水分情况。

The objective of this study is to present an approach based on the thermal inertia approach to estimate the profile of soil moisture content（SMC） in the top 1m soil column using Moderate Resolution Imaging Spectroradiometer（MODIS） imagery.In the approach,both diurnal Land Surface Temperature（LST） difference and broadband albedo（α0） are used to determine the spatial-temporal variation of SMC.The SMC retrieval rationale is that the high Apparent Thermal Inertia（ATI） values represent the high soil moisture contents,and the opposites are true for the low ATI values.The approach assumes that the water flux in the top 1m soil column is proportional to the difference in SMC at the surface and bottom of the soil column.The water balance equation can then be applied over the soil column,which results in the establishment of a relationship between the areal mean of SMC in the soil column and the remotely sensed surface SMC.The established SMC relationship is calibrated using the experimental data and a new soil texture classification technique based on the existing Chinese soil type classification.The SMC relationship is validated over a typical arid and semi-arid area in Xinjiang.The observed soil texture data is used in the application of the SMC relationship.Results show that the SMC relationship has the best performance for the loam and loamy clay types of soils,as the values of estimated SMC are very much close to the Time Domain Reflectometry（TDR） observations.In comparison with the TDR observation,a marginal relative error of 2.16% is found in the estimated SMC in a subarea with the sandy type of soil.Our result demonstrates the potential of using the established SMC relationship to map the spatial-temporal variation of SMC in the top 1m soil column using the MODIS imagery over arid and semi-arid areas.