Rice Mapping and Sowing Date Estimation Using CSK SAR Data

The use of multi-polarization features (intensities at the HH, VV channels and their ratio) from the X-band COSMO-Sky Med SAR images collected over an agricultural test site in Mekong Delta, Vietnam, is investigated in this study in order to develop methodology for rice monitoring. The proposed approach aims to detect rice areas and to estimate the sowing dates. The rice mapping is based on a classification of rice fields or non-rice fields and the sowing date estimation is performed with a novel procedure based on a rice training signature obtained from trends of ratio between HH and VV intensities related to rice fields.