Keywords
surface soil moisture, remote sensing, artificial neural networks
Start Date
1-7-2008 12:00 AM
Abstract
Spatial distributed information of isochronal surface soil moisture is very important to compensate the inaccuracy of initial conditions and the uncertainty of parameters in hydrological models at the landscape scale. In this paper the conceptual procedure to derive spatial distributed surface soil moisture values from synthetic aperture radar data by using ancillary optical remote sensing data is presented. Different biophysical vegetation parameters like vegetation water content and leaf area index overlay the soil moisture information on the microwave signal and hamper the application of many existing models. Therefore the objective of the proposed study is to test the performance of artificial neural networks to extract soil moisture information from radar data. Multi- and hyperspectral remote sensing data provide spatial distributed information about above ground vegetation parameters and can thereby used as ancillary network input to support the soil moisture extraction. The results of the study are expected to provide an improved database (initial conditions, plant parameters, etc.) for hydrological models.
Radar-based surface soil moisture retrieval over agricultural used sites – A multi-sensor approach
Spatial distributed information of isochronal surface soil moisture is very important to compensate the inaccuracy of initial conditions and the uncertainty of parameters in hydrological models at the landscape scale. In this paper the conceptual procedure to derive spatial distributed surface soil moisture values from synthetic aperture radar data by using ancillary optical remote sensing data is presented. Different biophysical vegetation parameters like vegetation water content and leaf area index overlay the soil moisture information on the microwave signal and hamper the application of many existing models. Therefore the objective of the proposed study is to test the performance of artificial neural networks to extract soil moisture information from radar data. Multi- and hyperspectral remote sensing data provide spatial distributed information about above ground vegetation parameters and can thereby used as ancillary network input to support the soil moisture extraction. The results of the study are expected to provide an improved database (initial conditions, plant parameters, etc.) for hydrological models.
