NDVI and EVI Estimation of Root Zone Soil Moisture in East Texas

The soil surface layer is a critical boundary between land and atmosphere, and soil moisture is a critical condition affecting interaction of land surface and atmosphere. The root zone can be defined as the top 100 cm of the soil layer. Remotely sensed data can indirectly measure soil moisture, but the signal only penetrates the top few centimeters, so soil moisture at deeper layers must be estimated. Remotely sensed vegetation indices (VI) are a valid method to estimate soil moisture at deeper layers. The objective of this study is to determine the potential of Terra-MODIS derived NDVI and EVI to estimate root zone soil moisture at one site in east Texas, representing a humid climate condition. The relationship between NDVI and EVI, NDVI and root zone soil moisture, and EVI and root zone soil moisture was investigated for a 30 day time period in a humid climate in East Texas. Soil Climate Analysis Network soil moisture and Terra-MODIS daily surface reflectance data were used in this study. Several problems with the surface reflectance data required data adjustment using linear regression. Pearson product-moment correlation was used to estimate the strength of the relationships with same-day VIs and soil moisture, and with one- and two-day VI time lags. NDVI was not correlated with same-day soil moisture at any depth, but was correlated with soil moisture with a one-day time lag. EVI was correlated with same-day soil moisture at 10 and 20cm depths, and at 50cm with a one- or two-day time lag. The results of this study indicate that EVI may be a better predictor of soil moisture than NDVI in humid climates, and that NDVI and EVI must be derived using established algorithms in order to be useful in predicting root zone soil moisture. A long-term study using quality controlled MODIS products is warranted.