Crop NDVI time series construction by fusing Sentinel-1, Sentinel-2, and environmental data with an ensemble-based framework

Dairong Chen, Haoxuan Hu, Chunhua Liao, Junyan Ye, Wenhao Bao, Jinglin Mo, Yue Wu, Taifeng Dong, Hong Fan, Jie Pei

Abstract: Highlights•A robust NDVI estimation RF model was established by inputting SAR and auxiliary environmental data.•An improved feature importance measurement method based on RF was proposed to suggest model input parameters.•An ensemble-based uncertainty quantification method was proposed to optimize NDVI time series construction results.•Temperature-related data can effectively improve the accuracy and generalization ability of the SAR-NDVI relationship model.•The weighted least squares (WLS) approach outperformed the baseline method in NDVI time series construction.