Soil moisture estimation using empirical models upgraded with SOMs on TerraSAR-X data

Inhalt:
For public safety reasons is essential early prediction of natural disasters like floods and landslides, which can be based on soil moisture estimation. Our goal is to evaluate quality of the Dubois and Shi empirical models used in soil moisture parameters retrieved from TerraSAR-X satellite data. In order to compare and validate estimated results, field measurements were compared to estimated values. Results are further improved using Self-organizing map in order to filter out artefacts and fields with inaccurate soil moisture estimation.