A Path to Improving Simulated Properties of Low Clouds Over the Beaufort Sea Using Airborne in Situ Observations of Subgrid‐Scale Variability

Arctic low clouds influence Arctic system evolution through their effects on the atmosphere and surface. Unfortunately, atmospheric models and retrospective analysis (reanalysis) struggle to simulate Arctic low cloud occurrence and properties. A potential source of this problem is the representations of subgrid‐scale (SGS) variability within large‐scale condensation schemes used to determine grid‐scale (GS) cloud properties. We utilize cloud and thermodynamic data of primarily liquid low‐level clouds from two aircraft campaigns conducted in the Arctic marine, sea ice environment to better characterize SGS variability and its relationship to GS cloud water. We examine the influence of two assumed SGS parameters on the estimation of GS cloud water: the total water SGS PDF shape, and the SGS total water PDF width (ΔQT). The results show that the conventional assumptions of shape and ΔQT lead to substantial errors in GS cloud water and cloud fraction estimates, particularly failing to handle partly cloudy conditions. The observed total water PDF is triangular with Gaussian‐like tails, which we replicate with a Gaussian function. ΔQT is sensitive to GS relative humidity, and so we allow it to vary as an empirically derived linear function of GS relative humidity. Both alterations result in an improved representation of GS cloud water and cloud fraction, though the details vary significantly between the aircraft campaigns. Both alterations show good promise for improving the model representation of Arctic boundary‐layer clouds.