A Simple Method for Soil Moisture Determination From LST–VI Feature Space Using Nonlinear Interpolation Based on Thermal Infrared Remotely Sensed Data

Abstract

Soil moisture is an important parameter that is widely used for drought monitoring and characterizing environment. Based on the universal triangular space constructed by the normalized land surface temperature (LST) and normalized difference vegetation index (NDVI), this study proposes a simple method to estimate surface soil moisture (SSM) using a nonlinear interpolation function with theoretical limiting edges derived from the energy balance principle. As an interpolation function, the quadratic polynomial equation is used to reveal the relationships among soil moisture, LST, and fractional vegetation cover (FVC), which are determined and validated by the simulation data from the Noah land surface model (Noah LSM). The developed function is much simpler than Carlson's function because the number of coefficients is reduced by half. The proposed method is applied to the North China Plain. The estimation results are also validated by the in situ measurements of soil water content at 10 and 20 cm depths at the Yucheng meteorological station; the corresponding biases are as high as 0.02 and 0.12 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> /m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> , respectively, and the root-mean-square errors (RMSEs) are approximately 0.04 and 0.10 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> /m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> , respectively. The results indicate that this method has high-estimation accuracy and overcomes the weakness induced by the effects of atmospheric conditions on different days. This study provides insight into the comparable soil moisture estimation from remotely sensed data on a regional scale.