Deforestation Triggers State‐Dependent Cloud Regimes in the Amazon

Abstract

The Amazon's arc of deforestation is a hotspot of rapid land‐cover change. Using 20 years of MODIS imagery combined with cloud property retrievals, CERES radiative fluxes, and ERA5 data, we show that deforestation consistently enhances low‐level cloudiness over non‐forest relative to nearby forest during the Amazon dry season. Non‐forest‐preference cloud scenes overwhelm those of forest‐ or no‐preference, with MODIS imagery often revealing clouds strikingly echoing the underlying land cover pattern. While the sign of this cloud response is robust, its magnitude is state‐dependent, varies with the background atmospheric state. Within the non‐forest‐preference cloud scenes, two distinct cloud regimes emerge: a warmer and drier regime, where higher sensible heat and lower latent heat produce sparse, shallow clouds with low cloud fraction (CF), liquid water path (LWP), and optical thickness (COT); and a cooler and moister “deepening” regime, where enhanced latent heat supports clustered, deeper clouds with higher CF, LWP, COT, greater precipitation, and stronger net cooling at the top‐of‐atmosphere (TOA). Our findings reveal that while deforestation triggers an enhancement in cloudiness, the nature of this response is tightly coupled to the background climate conditions, with implications for both the radiative energy budget and the hydrological cycle. As the Amazon trends toward a warmer and drier climate, a shift in cloud regime prevalence could diminish the TOA cooling and precipitation currently provided by the deepening cloud regime.