Relating soil moisture to TRMMPR backscatter in Southern United States

Document Type



Soil Moisture is an important variable in hydrological cycle. It plays a vital role in agronomy, meteorology, and hydrology. In spite of being an important variable, soil moisture measuring stations are sparse. This is due to high cost involved in the installation of dense network of measuring stations required to map a comprehensive spatio-temporal behavior of soil moisture. Hence, there is a need to develop an alternate method to measure soil moisture. This research relates soil moisture (SM) to backscatter (σ°) obtained from Tropical Rainfall Measuring Mission Precipitation Radar (TRMMPR) and Normalized Difference Vegetation Index (NDVI) obtained from Advanced Very High Resolution Radiometer. SM data is obtained from Soil Climate Analysis Network (SCAN). σ° measurements are normalized at an incidence angle of 10° at which it has the highest sensitivity to SM. An empirical model that relates SM to normalized σ° and NDVI is developed. NDVI takes into account the different vegetation densities. The relationship between model variables is approximated to be linear. The model is applied to data from 1998 to 2008 where 75% of the data is used for calibration and the remaining 25% for validation. Figure 1 shows the comparison of observed and modeled soil moisture for a site with low vegetation. Even though the model underestimates the soil moisture content, it captures the signal well and produces peaks similar to the observed soil moisture. The model performs well with a correlation of 0.71 and root mean square error of 4.0%. The accuracy of the model depends on vegetation density. Table 1 summarizes the model performance for different vegetation densities. The model performance decreases with the increase in vegetation as the leaves in the vegetation canopy attenuate the incident microwaves which reduces the penetration depth and subsequently the sensitivity to soil moisture. This research provides a new insight into the microwave remote sensing of soil moisture.


Fresh Water Studies | Geographic Information Sciences