What spatial resolution do we need to resolve shallow cumulus cloud microphysics?

Cloud microphysical processes on sub-centimeter scales strongly affect precipitation formation, cloud radiative properties, and ultimately large-scale climate. However, both in situ observations and numerical models typically rely on spatial averaging at scales far larger than those at which these processes occur. We present airborne, high-resolution holographic measurements of marine shallow cumulus clouds over the North Atlantic collected during EUREC4A, that provide new insight into the spatial variability of cloud droplet populations.
The Max Planck CloudKite holography system on a tethered balloon samples large localized samples (~10 cm³) of cloud only separated by 10 cm along horizontal cloud transects.

Previous analyses of this dataset have revealed that droplet clustering is a highly localized phenomenon occurring in hotspots on meter and sub-meter scales, with important implications for collision–coalescence rates. Here, we extend the perspective to droplet size distributions. We quantify the scales at which and how often averaged size distributions are representative of local distributions. We find that representativeness is rare: spatial averaging often obscures substantial local variability in droplet size distributions.

This variability indicates distinct locally different growth histories and implies correspondingly different local microphysical process rates. Our results demonstrate that commonly resolved scales in both in situ measurements and model representations are insufficient to capture key aspects of warm-cloud microphysics, highlighting the need to account for small-scale variability when interpreting observations and developing parameterizations.