Remote Sensing Retrieval of Soil Moisture in the Wetland Area of a Flood Plain-a case study of Honghe National Nature Reserve

摘要

Soil moisture is an important indicator for wetland degradation. It is significant to do research on spatial and time variability of soil moisture and on how to monitor soil moisture by remote sensing in wetland, which is still a lack of research at present. In this paper, the Honghe National Nature Reserve is taken as an example. After analyzing all kinds of information about soil moisture reversion and comparing different quantitative remote sensing models, we introduce Ts/NDVI model and rectify it by many monitoring data to adapt to the soil moisture retrieving in the wetland study area. Supported by Landsat TM data and other high-resolution data, we inverse the factors (e.g. the vegetation index, the surface temperature, the albedo and so on) needed in the model by remote sensing techniques. And further on, the process of soil moisture is modulated and described quantitavely. Then we use the new model to monitor soil moisture in Honghe National Nature Reserve in June、July、 August and September of 2008. Based on the monitoring data of some typical fields, the retrieved information is validated and the result show that the model is very well to obtain the depth soil moisture in wetland. The results of this study show that the improved T_s/EVI space can extract more accurate spatial/temporal distribution of soil moisture information in wetland. It is useful for the quantitative investigation of the four typical growing season's soil moisture variations as well as analysis of habitat characteristics of the humidity gradient.