• VIS albedo shows strong negative correlations with SPEI over grasslands and savannas. • EBF exhibits weak VIS and shortwave albedo responses, whereas NIR albedo is negatively linked to SPEI. • PV, NPV and soil background dominate NIR albedo variability in grassland and savannas. • Vegetation indices decline under extremely wet conditions in the northern Amazon. Vegetation responses to drought in the Amazon have been extensively studied; however, the impacts of diverse hydroclimatic conditions (i.e., drought and wet) on albedo across its spectral domains (visible, near-infrared, and shortwave) remain poorly documented. To address this gap, we investigated the effects of hydroclimatic variability on spectral albedo, vegetation greenness, and albedo-driven radiative forcing across three major land surface types − evergreen broadleaf forest (EBF), grassland (GRA), and savannas (SAV) − in the Amazon during 2005 − 2016. The analysis integrated three satellite-based albedo products: MCD43C3 (MODIS), C3S derived from SPOT/VEGETATION and PROBA-V, and GAC43 (AVHRR). Additional, three vegetation indices products (GIMMS3g NDVI, MOD13C2 NDVI, and MOD13C2 EVI) were analyzed. Overall, the visible (VIS) albedo exhibited strong and significant negative correlations with standardized precipitation evapotranspiration index (SPEI) over GRA and SAV sites, while shortwave albedo demonstrated similar but slightly weaker responses. In contrast, VIS and shortwave albedo responses over EBF were much less pronounced across all three products, whereas near-infrared (NIR) albedo consistently showed negative correlations with SPEI. At GRA and SAV surfaces, NIR responses were more complex, exhibiting product-dependent and spatially heterogeneous patterns that likely reflect interactions among photosynthetic vegetation (PV), non-photosynthetic vegetation (NPV), and soil background contributions. Vegetation greenness, as captured by the three vegetation index products, displayed significantly positive correlations with SPEI for GRA and SAV surfaces, consistent with strong sensitivity to hydroclimatic variability. Over EBF, however, correlations were generally weaker. Moreover, MODIS vegetation indices (MOD13C2 NDVI and MOD13C2 EVI) revealed negative responses under extremely wet conditions (SPEI > 2.0) in the northern Amazon, suggesting that radiation may become a limiting factor for vegetation growth under such conditions. Our findings on hydroclimatic-induced changes in albedo and vegetation greenness under both drought and wet conditions support understanding of land-atmosphere interactions and provide valuable insights for improving land surface models.
Li et al. (Mon,) studied this question.