Improving Remote Sensing Ecological Assessment in Arid Re-gions: Revealing Divergent Environmental Drivers with a Du-al-Index Framework in the Tarim Basin

In this study, we developed a zonally adaptive ecological assessment framework tailored to the heterogeneous landscape of the Tarim River Basin. Utilizing MODIS remote sensing products, we constructed a Remote Sensing Ecological Index for desert zones by incorporating surface albedo to enhance sensitivity to arid ecosystem dy-namics (ARSEI). Meanwhile, the traditional Remote Sensing Ecological Index (RSEI) was applied in non-desert regions. These two indices were harmonized through standardized transformation to generate a Composite Remote Sensing Ecological Index (CoRSEI), enabling an integrative ecological quality assessment across the entire basin. Building upon this, multivariate statistical methods and machine learning models were employed to explore spatial and temporal patterns of ecological quality change and to identify region-specific environmental drivers. This integrated approach provides new insights into the interactions between ecological processes and environmental condi-tions under arid climate constraints, offering theoretical support for regional ecological monitoring and adaptive management.