Performance and evaluation of a sectional cloud microphysical model (CARMA Cloud) in the Community Earth System Model (CESM2)

Clouds are critical to Earth’s climate system, but simulations in climate models remain highly uncertain. The cloud model, MG2, in the Community Earth System Model (CESM2/CAM6) separates cloud-sized particles from raindrops and snow using two-moment microphysics. Here, we evaluate a new sectional cloud model (Zhu et al., 2026) uniting liquid cloud-sized particles with rain and ice/graupel cloud-sized particles with snow within single size distributions (CESM2/CARMA Cloud). CARMA Cloud better represents microphysical processes than the two-moment model, but at additional computational cost. We compare MG and CARMA Cloud simulations with observations. CARMA Cloud simulates liquid cloud effective radius, with a peak of ~ 14 µm over ocean and ~ 12 µm over land, as well as raindrop diameter with a peak near 1 mm, similar to observations. CARMA Cloud, like aircraft observations, has ~ 31% of cloud mass outside the range of the Moderate Resolution Imaging Spectroradiometer (MODIS) observations, questioning its utility for liquid water path (LWP) observations. Drizzle is ~ 12% of liquid water and rain ~ 3%, a large fraction of ice mass comes from riming. Both CARMA Cloud and MG simulations show LWP increases with cloud number density in agreement with aircraft data. CARMA Cloud ice water size distributions show significant mass larger than sizes from satellite data. Both cloud models simulate comparable cloud radiative forcing, but deep convective clouds may not be high enough in CESM2. Global albedo in CARMA Cloud strongly depends on aerosol-cloud interactions, suggesting possible deficiencies in aerosol activation rates and coalescence kernels.