Improvement of Vegetation Water Content Estimation Over the Tibetan Plateau Using Field Measurements

Abstract

Vegetation water content (VWC) plays a significant role in the retrieval of soil moisture using microwave remote sensing, which further supports applications such as weather forecasting, flood prediction, and landslide early-warning. In the current SMAP algorithm, the commonly used vegetation index (i.e., the NDVI) was utilized to determine the VWC. This study evaluates the accuracy of SMAP VWC product, together with other Jackson's algorithm using NDVI and Paloscia's method using LAI. Comparing to ground observation, SMAP overestimates VWC, while Jackson's method, in which stem factor is not included, performs better than other two. By using Jackson's method, we found the accuracy of soil moisture retrieved from SMAP could be improved.