Estimating soil wetness using satellite data

Abstract

Improved estimates of soil wetness were obtained using observations from both the NIMBUS-7 Scanning Multichannel Microwave Radiometer (SMMR) and the NOAA-7 Advanced Very High Resolution Radiometer (AVHRR). SMMR 6.6 GHz frequency, horizontal polarization, brightness temperature (T_BH) was first correlated with soil wetness, as computed using an Antecedent Precipitation Index (API) model, for a number of SMMR ground resolution areas involving a fairly wide range of vegetation densities. The API generally accounted for more than 70 per cent of the observed temporal variability in T_BH, with linear correlations being significant at the 1 per cent level. The regression slope of T_BH versus API correlated well, at the 1 per cent level, with a vegetation index derived from AVHRR visible and near-infrared observations. The regression intercept was found to correlate less satisfactorily, but was significant at the 5 per cent level. These linear regression results were used to develop a diagnostic model for soil wetness using SMMR and AVHRR data only. The model was found to be useful in describing four levels of soil wetness as compared to three levels when vegetation was not considered.