In south-western United States, soil moisture data is important for drought studies in the region which is experiencing a drought for many years, whereas in South Florida, water stage data is required by hydrologists to monitor the hydrological flow in wetlands. Soil moisture data and water stage data are not sufficiently available due to sparse monitoring stations. Installation of dense measuring stations over an extended area is costly and labor intensive. Therefore, there is a need to develop an alternative method of measuring soil moisture and water stage.

Microwave remote sensing has proven to be a useful tool in the measurement of various surface variables from space. This research explores the capability of microwave remote sensing to measure soil moisture and water stage on the earth from space. Tropical Rainfall Measuring Mission Precipitation Radar (TRMMPR) provides the Ku-band backscatter measurements that are used to measure soil moisture and water stage. Models that relate soil moisture and water stage to TRMMPR backscatter (σ°) are developed. The dependence of σ° on the dielectrical and physical characteristics of the land surface is used as the basis of this research. The soil moisture content affects σ° by changing the dielectric constant of the surface whereas the vegetation density affects σ° by changing the physical characteristics of the surface. Vegetation density in the model is represented by Normalized Difference Vegetation Index (NDVI). Dependence of σ° on partial submergence of vegetation in inundated areas is used to measure water stage in wetlands of South Florida. The effects of the exposed vegetation above the water surface on the model are assessed by comparing two cases of model run (a) that includes NDVI in the model, and (b) that does not include NDVI in the model.

Eleven years of data is used in this research where 75% of the data is used for calibration of the model and 25% of the data is used for validation. The estimated values of soil moisture and water stage are compared to the observed values and the performance of the models is assessed by calculating correlation coefficients, calculating root mean square errors, and plotting non-exceedance probability plots for the absolute error between observed and modeled values.

The soil moisture and water stage models work reasonably well and are able to estimate soil moisture and water stage with low errors. The soil moisture model works better in low vegetated areas because low vegetation allows the incident radiation to penetrate through the canopy cover and provide measurements from underlying surfaces. The water stage model works better in shrublands where there are no tree trunks and the model has an immediate impact from the vegetation canopy. This research provides an alternate way of measurement of soil moisture and water stage using remote sensing.