Abstract
Satellite-based hyperspectral Earth observation data combined with physically based radiative transfer models have the strong potential to support sustainable agriculture by providing accurate spatial and temporal information of important vegetation biophysical and biochemical variables such as leaf chlorophyll content. To meet this goal, possible error sources in the modelling should be minimized. Thus, the capability of a model to reproduce the measured spectral signals has to be tested before applying any retrieval algorithm. For an exemplary demonstration, the PROSAIL model was employed to emulate the setup of the future EnMAP hyperspectral sensor in the visible and near-infrared (VNIR) spectral region with a 6.5nm spectral sampling distance. Model uncertainties were determined to subsequently exclude those wavelengths with highest mean absolute error (MAE) between model simulation and spectral measurement. For this purpose data from two campaigns were exploited (1) from Nebraska-Lincoln (maize and soybean) and (2) from Munich-North-Isar (maize and winter wheat). A significant increase of accuracy for leaf chlorophyll content (LCC, mu gcm(-2)) estimations could be obtained, with relative RMSE decreasing from 26% (full VNIR range) to 15% (optimized VNIR) for maize and from 77% to 29% for soybean, respectively. We therefore recommend applying a specific model-error threshold (MAE0.01) to stabilize the retrieval of crop biochemical variables.
| Dokumententyp: | Zeitschriftenartikel |
|---|---|
| Fakultät: | Geowissenschaften > Department für Geographie |
| Themengebiete: | 500 Naturwissenschaften und Mathematik > 550 Geowissenschaften, Geologie |
| ISSN: | 2512-2789 |
| Sprache: | Deutsch |
| Dokumenten ID: | 67817 |
| Datum der Veröffentlichung auf Open Access LMU: | 19. Jul. 2019, 12:23 |
| Letzte Änderungen: | 04. Nov. 2020, 13:50 |
- Dokument bearbeiten
