Forests and clouds as a coupled system: moving beyond boundary conditions

Mounting evidence indicates concurrent changes in forest carbon uptake and cloudiness across tropical, temperate, and boreal biomes, pointing to emerging shifts in forest–atmosphere coupling with the potential to amplify climate feedbacks. At the same time, a growing body of research highlights the critical role of within-canopy microclimate, where forests generate strong vertical gradients in radiation, temperature, humidity, and turbulence. These gradients buffer climatic extremes, regulate ecosystem functioning, and control phenological responses that differ markedly between overstory and understory environments. Together, these findings demonstrate that forests actively regulate their internal microclimate while interacting dynamically with the atmosphere above.

Despite these advances, forests and clouds are still largely studied and modelled as separate components, with land treated primarily as a lower boundary condition rather than as an active, three-dimensional driver of atmospheric dynamics. This conceptual separation limits our ability to understand and predict coupled carbon–cloud–climate feedbacks, particularly under ongoing climate change where both carbon uptake and cloud regimes are shifting.

Here, we propose a conceptual and methodological shift towards treating forests and clouds as an integrated, dynamically coupled system. We outline a first-principles framework that bridges biological, chemical, and physical processes across spatiotemporal scales, explicitly linking radiative perturbations, stomatal responses, turbulence, atmospheric chemistry, and cloud formation. In the talk, I will present observations from comprehensive field campaigns, including LIAISE and CloudRoots-Amazon22, integrated with large-eddy simulations to resolve canopy–boundary layer–cloud interactions. This combined observational–modelling approach enables a process-based understanding of how forest structure and function feed back on atmospheric dynamics and cloud development.

Embedding such observationally constrained, canopy-resolving representations into Earth System Models offers a pathway to reduce uncertainties in projections of carbon uptake, cloud dynamics, and their combined influence on climate, ultimately improving our capacity to predict biosphere–atmosphere feedbacks in a changing world.