Spatiotemporal NDVI-SAR Analysis Using Sentinel-1 and Full-Polarimetric ALOS-2 Imagery in The Küre Mountains Forests, Türkiye

Abstract

This study investigates the temporal relationship between SAR backscatter (ALOS-2 PALSAR-2 and Sentinel-1) and NDVI across oriental beech, Uluda? fir, and mixed forests in the Küre Mountains of Türkiye. Results show that correlation patterns are strongly shaped by forest type and seasonal phenology. Fir forests exhibit stable and moderately strong correlations - particularly with ALOS VH (r ? 0.59) - due to their evergreen canopy and low seasonal variability. In contrast, beech forests display pronounced seasonal fluctuations, with correlations peaking during the growing season (r ? 0.53) and declining during senescence (r ? 0.04). Mixed forests show intermediate yet more variable behavior, reflecting their heterogeneous composition. ALOS L-band SAR consistently outperforms Sentinel-1 C-band, especially in coniferous monitoring, underscoring the significance of wavelength and polarization. These findings highlight the importance of tailoring SAR-optical integration to forest type and season and suggest that multi-sensor methods can improve monitoring in complex ecosystems. © 2025 IEEE.