Mapping and monitoring rice crop with RADARSAT data

The aim of this paper is to assess the use of RADARSAT data for rice field mapping and monitoring. Ground data have been collected over one of the most important irrigated flood plain of Indonesia and analyzed along with RADARSAT data. The radar backscatter coefficient /spl sigma//spl deg/ of rice fields appears to have a significant temporal variation. However, the dynamic is found lower than that of the ERS data due to a higher backscatter at HH than at VV polarization at early stages. At higher incidence angle (43/spl deg/ vs. 23/spl deg/), the /spl sigma//spl deg/ of the fine resolution F3 mode reaches a saturation level earlier compared to the standard resolution mode, which thus appears to be more appropriate for rice field mapping. The mapping result, based on the /spl sigma//spl deg/ temporal change, has been found to have an accuracy of 87% compared to the existing map. In order to follow the growing cycles and improve the management of irrigated perimeters, a methodology for Golongan mapping has also been developed. Despite a lower dynamic of RADARSAT data vs rice growth compare to ERS, the results appear promising and rice mapping operations using RADARSAT data can be foreseen.