Rice mapping and crop growth monitoring using ERS SAR data in Guangdong Province, China

The capability of the ERS SAR to acquire images independently of cloud coverage or daylight conditions is of major significance in the context of agricultural applications, which are time critical. The accurate identification of rice crops using satellite radar has been shown to be dependent on the availability of images acquired during specific time windows through the growing season. For rice growing areas in the tropics, experience shows that high resolution optical satellite sensors cannot provide the desired information due to constraints related to cloud cover and revisit schedules. During 1997, ground data were collected on several ERS overpass dates in the Guangzhou Region of Guangdong Province where rice production is both irrigated plain and terraced. Parameters measured included rice crop growth and performance indicators. ERS-2 PRI images were calibrated, gee-coded and filtered, so allowing backscatter temporal signatures to be developed. Filtering and per pixel based classification performance have been investigated using supervised and neural network classifiers. Initial results have shown that it is possible to classify irrigated rice using images acquired on two dates and the difference image. However terrace rice has proved more difficult to classify due to interactions between the SAR beam and local topography. Standard beam mode Radarsat imagery has been acquired to investigate the effect of polarisation and incidence angle on rice classification performance.