Resolving cloud microphysical heterogeneity with Sentinel-2: implications for aerosol-cloud interaction studies

Cloud radiative effects depend critically on microphysical properties, that in turn are influenced by aerosol-cloud interactions, which remain the dominant source of uncertainty in anthropogenic radiative forcing estimates. Cloud droplet number concentration, Nd, is a key parameter for constraining these phenomena using satellite observations. However, current satellite-based Nd retrievals suffer from substantial biases in comparison with in-situ measurements. A fundamental limitation is that even specialized satellites, such as MODIS, typically observe at scales of hundreds of meters to kilometers. At these resolutions, retrievals necessarily average over fine-scale cloud variability associated with differing cloud life-cycle stages, spatially varying dynamical forcing, and heterogeneity in aerosol conditions at cloud base. Averaging this sub-pixel heterogeneity limits our ability to understand aerosol-cloud interactions or to evaluate high-resolution cloud models.

The Clouds Decoded project, funded by the Advanced Research + Invention Agency (ARIA), retrieves cloud properties from Sentinel-2 (S2) imagery at ~60 m resolution. This resolution captures spatial structures at scales where key microphysical processes operate: cloud edges, gradients in optical depth and effective radius, and spatial heterogeneity patterns that inform sub-grid parameterizations in climate models. These high-resolution observations can also complement ground-based and aircraft measurements when S2 overpasses are available.

In this contribution, we present case studies demonstrating how S2 observations can characterize cloud heterogeneity at scales previously invisible to satellite sensors. Can observations at this spatial resolution help resolve discrepancies in satellite-derived aerosol-cloud relationships and reduce uncertainties in aerosol-cloud interaction estimates?