Crop monitoring by advanced SAR remote sensing techniques : modelling and experimental analysis
(2005)
Files
Blaes.pdf
Restricted Access - Adobe PDF
- 12.93 MB
Details
- Authors
- Supervisors
- Defourny, Pierre
- Abstract
- Satellite remote sensing is a key source of frequent information for agricultural production forecasting. Only Synthetic Aperture Radar (SAR) sensor can provide systematic observation thanks to the microwave insensitivity to the cloud cover. This research investigates actual information requirements for agricultural monitoring, current information sources and advanced SAR potentialities. Performances of advanced SAR techniques to crop type identification and crop growth estimation in operational conditions were studied through experimental and modelling research. Operational complementarities between SAR and optical high resolution images for crop type discrimination are first demonstrated leading to actual use by the Belgian Ministry of Agriculture. The efficiency concept is introduced as an alternative indicator of classification performance in the context of crop area control. Simulated ENVISAT Wide Swath time series was also found promising for crop identification at regional scale. Crop growth is demonstrated to be best monitored by 1-day interferometric coherence. Indeed canopy cover and plant height were found very highly correlated to signal coherence for 4 different crops and weakly influenced by soil water content. In order to further understand backscattered signal from crops a unique and very dedicated field data set was collected for maize and wheat. A new method of soil roughness measurement based on photogrammetric and geostatistical techniques also allowed describing the agricultural soil anisotropy by a bi-dimensional roughness characterisation. Two fully polarimetric radiative transfer models are validated using SAR observations acquired in various configurations (i.e. polarisations and incidence angles) simultaneously with the field measurements. Based on these validated simulations polarimetric indices correlated to maize leaf area index and biomass are developed to dissociate the vegetation backscattering from the soil water content contribution.
- Affiliations
Citations
Blaes, X. (2005). Crop monitoring by advanced SAR remote sensing techniques : modelling and experimental analysis. https://hdl.handle.net/2078.5/97524