Multitemporal analysis of SAR images over the tropics for agricultural applications

The potential of multitemporal SAR data obtained from ERS-1 (C-band) and JERS-1 (L-band) satellites for monitoring of rice and sugarcane is investigated at a study area in Kanchanaburi, Thailand. The possibility of discrimination of agricultural crops from other cover types such as water, urban areas, bush and shrubs is explored. The mean values of radar backscatter are computed and the variations in the temporal backscatter are discussed. Both qualitative and quantitative measures are employed to investigate the separability of different cover types. The results show that the use of at least two appropriately timed imaging dates during the growing season is sufficient for rice field inventory. Difficulty is, however, encountered in an easy separation of sugarcane from shrub