Abstract
This study evaluates the relationships between Sentinel-1 radar vegetation indices (modified radar vegetation index [mRVI], modified Radar Forest Degradation Index [mRFDI]) and Sentinel-2 optical indices (normalized differential vegetation index [NDVI], ratio vegetation index [RVI]) across diverse land covers in Gilan province, Iran—a region characterized by humid subtropical climates, frequent cloud cover, and threatened ecosystems. Using Google Earth Engine, we analyzed summer 2023 data, comparing indices in non-vegetated, sparse, dense, and mixed landscapes. Results revealed near-perfect inverse correlations between radar indices (mRVI–mRFDI: r ≈ −0.99) across all environments, confirming their structural consistency. Radar–optical correlations, however, were landcover dependent: negligible in non-vegetated (|r| < 0.15) and dense-vegetated (|r| < 0.05), but moderate in mixed zones (e.g., mRVI–NDVI: r = 0.42) where structural (radar) and biochemical (optical) signals partially aligned. Sparse vegetation showed transitional ties (mRVI–RVI: r = 0.39), highlighting early sensor synergy. Radar excelled in detecting degradation and soil moisture under cloud cover, while optical indices tracked chlorophyll dynamics but faltered in dense canopies. These findings underscore the complementary roles of Synthetic Aperture Radar and optical sensors: integrated use is critical in homogeneous areas (e.g., forests), while mixed landscapes benefit from inherent synergies, reducing reliance on resource-intensive fusion. This study advances tailored multi-sensor strategies for cloud-prone regions, enhancing vegetation monitoring accuracy to support sustainable management in ecologically sensitive zones like Gilan.