Complementarity of 4 Polarizations in C-band SAR Imagery to Estimate Biophysical Variables for Crop Monitoring

The overall objective of this paper is to estimate crop parameters thanks to 4 polarizations in SAR signal. More especially, the Leaf area Index (LAI) is a key parameter for the coupling of earth observation with crop growth modelling. Based on a SAR data set and intensive field campaigns the performance of LAI estimation from SAR time series is assessed for two different crops, winter wheat and maize. The study is based on very large remote sensing data set including ERS time series and 1 RADARSAT image acquired in Belgium (13 ERS/SAR images from April to July 2007, and from mid-June to mid-July 2008 and 1 Quad-Pol standard RADARSAT image on the 30 June 2008) complemented by comprehensive field campaigns. Simultaneously with the SAR data acquisition, detailed field measurements of crop/soil parameters (plant density, height, LAI, canopy cover and top soil moisture) for a set of large fields during the 2007 (about 25 fields per crop) and 2008 (10 maize fields and 6 winter wheat fields) crop growth seasons have been completed. The research was completed on a site of 60 x 60 km distributed in Belgium. The field measurements have been carried out according to the same protocol every 2 to 3 weeks during the crop growing seasons. Previous studies completed in Belgium already demonstrated the performance of SAR data at C-band to retrieve LAI by inverting the locally adjusted Water Cloud model. In the current research, first the robustness of the Water Cloud model calibrated for a given crop in a given site is assessed through a cross-validation between years. Furthermore, 4 polarizations of High Resolution SAR data are compared and the potential complementarities of these are investigated in details for the different growth stages of the crops. Finally the current results are compared with the performance of optical data for LAI estimate in the perspective of their assimilation in crop growth models to forecast the forthcoming yield. This research is supported in the framework of the international GLOBAM project, a globally distributed agricultural experiment to enhance the crop monitoring by remote sensing.