Publications

Xia, L; Song, XN; Leng, P; Wang, Y; Hao, YB; Wang, YF (2019). A comparison of two methods for estimating surface soil moisture based on the triangle model using optical/thermal infrared remote sensing over the source area of the Yellow River. INTERNATIONAL JOURNAL OF REMOTE SENSING, 40(6-May), 2120-2137.

Abstract
As an important surface parameter, surface soil moisture (SSM) plays a significant role in water resources management, crop growth, land degradation, and vegetation coverage as well as global climate change studies. In particular, the triangle model based on the spatial relationship between the land surface temperature (LST) and vegetation index from optical/thermal infrared remote sensing data has been widely used for the estimation of SSM. In the present study, Moderate Resolution Imaging Spectroradiometer (MODIS) remote sensing data, including the MOD11A2 LST and MOD13A2 normalised differential vegetation index (NDVI) from July to October, 2008 to 2010, over the source area of the Yellow River (SAYR) in the alpine vegetation region of the Tibetan Plateau, are selected to construct the LST/NDVI triangle space. Two methods based on the triangle space, namely, the temperature vegetation drought index (TVDI) method and second order polynomial method, is used to estimate the SSM at the regional scale. Finally, an analysis between the estimated SSM and ground-measured SSM is carried out to explore not only the quality of these two methods in triangle space but also the applicability of these two methods in the alpine vegetation region of the Tibetan Plateau. In addition, the spatial distribution of SSM in the study area is investigated. The results of the validation of the estimated SSM results using ground-measured data show that the accuracy of the second order polynomial method is significantly higher than that of the TVDI method. The TVDI method root mean square error (RMSE) is 0.072m(3)m(-3), coefficient of determination R-2 is 0.461, and bias is -0.022m(3)m(-3), while the second order polynomial method RMSE is 0.075m(3)m(-3), R-2 is 0.519, and bias is 0.007m(3)m(-3). Moreover, the SSM decreases from west to the east; this distribution can be obtained with the estimated results by using either of the two methods. In addition, the SSM estimations by using the two methods are in good correlation with the Climate Change Initiative (CCI) soil moisture product. The present study indicates that the second order polynomial method can simulate the change in SSM in the triangle space more realistically and effectively and that the second order polynomial method is more suitable for the estimation of SSM over the alpine vegetation region of the Tibetan Plateau.

DOI:
10.1080/01431161.2018.1519287

ISSN:
0143-1161