Relating TRMM precipitation radar land surface backscatter response to soil moisture in the Southern United States

SummarySoil moisture is an important variable in the hydrological cycle and plays a vital role in agronomy, meteorology, and hydrology. It regulates the exchange of water and heat between land surface and atmosphere and thus plays an important role in the development of weather patterns. It is difficult to obtain a comprehensive spatio-temporal map of soil moisture because of expensive installation of soil moisture measuring instruments.In this paper, a model to estimate soil moisture (ms) using Tropical Rainfall Measuring Mission Precipitation Radar (TRMMPR) backscatter (σ°) and Normalized Difference Vegetation Index (NDVI) is developed for the Southern United States. Soil moisture data from Soil and Climate Analysis Network (SCAN) stations is used to calibrate and validate the model.The estimated values of ms compare well with the ground measurements of soil moisture. The model works well for various landcovers but works best for low density vegetated areas (closed shrubland). All the soil moisture estimates in this landcover have an absolute error of less than 8%. The model performance deteriorates with increase in vegetation density (crops and forest). Overall, the model performance is satisfactory for all landcover types with RMSE less than 6.3% and absolute error of 10% or less for 90% of the estimates. Estimation of soil moisture over a large area with low error provides another use of TRMMPR data.

► We present a linear model to estimate soil moisture using TRMMPR backscatter and NDVI.
► Modeled values are compared with observed soil moisture over six landcover types.
► The model works well in arid and semi-arid regions and the model performance deteriorates with increase in vegetation density.
► This research presents a method to measure soil moisture using TRMM backscatter data.