Compensating Errors in Cloud Radiative and Physical Properties over the Southern Ocean in the CMIP6 Climate Models

Abstract The Southern Ocean is covered by a large amount of clouds with high cloud albedo. However, as reported previous climate model intercomparison projects, underestimated cloudiness and overestimated absorption solar radiation (ASR) over the lead to substantial biases in sensitivity. present study revisits this long-standing issue explores uncertainty sources latest CMIP6 models. We employ 10-year satellite observations evaluate radiative effect (CRE) physical properties five models that provide comprehensive output cloud, radiation, aerosol. simulated longwave, shortwave, net CRE at top atmosphere are comparable CERES observations. Total fraction (CF) also reasonably CMIP6, but comparison liquid (LCF) reveals marked spatial pattern seasonal variations. discrepancies between MODIS become even larger other macro- micro-physical properties, including water path (LWP), optical depth (COD), effective radius, well aerosol (AOD). underestimation both LWP radius (regional means ∼20% 11%, respectively) results relatively smaller bias COD, impacts COD LCF cancel out each other, leaving ASR predicted CMIP6. An error estimation framework employed, different signs sensitivity errors from CF corroborate notions there compensating modeled shortwave CRE. Further correlation analyses geospatial patterns reveal most relevant factor determining observations, while too sensitive COD. relationships LWP, analyzed explore possible Our calls for more rigorous calibration detailed future development projection.