Inhaltsbereich
Forschungsprojekte
Sentinel-Synergy-Study S3
| Fachgebiet: Datenassimilation, Fernerkundung, Geländekampagne, Geländemessungen, Geo-Informationssysteme Gefördert durch: European Space Agency Förderkennzeichen: 4000119682/17/NL Projektleitung: Projektwissenschaftler: , Thomas Kaminski (The InversionLab), Tristan Quaife (Uni Reading), Joris Timmermans (UCL) Laufzeit: 05/2017 - 04/2019 |
The overall concept of the project activity aims at a most consistent retrieval of the spatio-temporal dynamics of the land surface conditions using complementary optical and SAR
observations in a coherent retrieval scheme. Time series of multispectral surface reflectances in the optical EM spectrum are combined with multi-polarized time series of the backscattering coefficients from SAR data in a joint retrieval approach.
The retrieval scheme is based on the concept of the ESA EO-LDAS system (Lewis et al., 2012)
which estimates the land surface state on the minimization of a cost function within a Bayesian
Optimum estimation framework. Radiative transfer models in the optical and SAR domain are used as forward operators to map the surface state to the observational space. Information on land use and vegetation phenology is obtained in a post-retrieval process from the derived land surface state. This approach is rather different from existing ones, as the land cover/use information is not obtained from the spectral signatures themselves, but rather from the surface state variables.