A Decision-Tree Approach to Identifying Paddy Rice Lodging with Multiple Pieces of Polarization Information Derived from Sentinel-1

Lodging is one of the typical abiotic adversities during paddy rice growth. In addition to affecting photosynthesis, it can seriously damage crop growth and development, such as reducing quality hindering automated harvesting. It is, therefore, imperative accurately in good time acquire crop-lodging areas for yield prediction, agricultural insurance claims, disaster-management decisions. However, accuracy requirements monitoring remain challenging due complicated impact factors. Aiming at identifying lodging on Shazai Island, Guangdong, China, caused by heavy rainfall strong wind, a decision-tree model was constructed using multiple-parameter information from Sentinel-1 SAR images situ samples. The innovatively combined five backscattering coefficients with polarization decomposition parameters quantified importance each parameter feature. found that method coupled be used obtain an accurate distribution rice-lodging areas. results showed that: (1) Radar capture changes lodged rice. radar best distinguish are VV, VV+VH, VH/VV, Span. (2) Span strongest feature importance, which shows necessity adding classification model. (3) dual-polarized database effectively extract area overall 84.38%, total precision 93.18%. These observations guide future use SAR-based assessment post-disaster management.