“…Many studies have used single-temporal SAR backscatter characteristics to monitor rice (Gao et al, 2019), whereas single-temporal SAR data usually results in low identification accuracy due to missing key phenological stage information (Lopez-Sanchez et al, 2012;Li et al, 2014). Compared with single-temporal SAR data, multi-temporal or time-series SAR data can capture phenological information about rice in the whole growth cycle, thereby contributing to improving rice identification accuracy (Yang et al, 2017(Yang et al, , 2018Csorba et al, 2019;Chandra Paul et al, 2020;Pang et al, 2021;Zhan et al, 2021). At present, most studies on rice identification using SAR have focused on the flat terrain areas where rice fields are concentrated and large-sized, such as the Mekong Delta (Bouvet and Le Toan, 2011;Clauss et al, 2018), Bangladesh (Panigrahy et al, 2012), Vijayawada in India (Mandal et al, 2020), and Northeast China and the Middle and Lower Yangtze Valley Plain (Zhan et al, 2021).…”