Correlation between atmospheric stability, dynamics, and cloud properties in observations, re-analyses, and models

Clouds, especially low clouds, are key to our ability to understand and predict climate. They are an important component of the physical climate system and contribute significantly to the difference in climate projections by General Circulation Models (GCMs). Cloud predictors, such as atmospheric stability and large-scale circulation, are often used in model parametrizations. This thesis evaluates the performance of the latest Canadian atmospheric GCM (CanAM-4.1), in particular with respect to its cloud simulation. Its output is compared to observations, re-analyses, and its predecessor (AGCM-3). The analysis focuses on low clouds in the tropical band (30 degree South to 30 degree North) over the ocean. Results show that CanAM-4.1 systematically performs better than AGCM-3 (when compared to observations). Variability between observational datasets is also shown to be much smaller than variability between observations and models (or re-analyses). A model-to-satellite approach is used, i.e. the CFMIP Observation Simulator Package (COSP), and reduces observations-CanAM-4.1 differences in low cloud fractions. Results are not as unambiguous for high clouds. Three cloud regimes (stratiform, convectiform, and storm track) are well reproduced by all datasets, i.e. CanAM-4.1, AGCM-3, the ECMWF Interim Re-Analysis (ERA-Interim), and the ECMWF 40 years Re-Analysis (ERA-40). Conditional sampling of low cloud fractions as a function of the Lower Tropospheric Stability (LTS), Estimated Inversion Strength (EIS), and vertical velocity at the 500 hPa level (ω₅₀₀) show good agreement with observations. Overall, conclusions are not sensitive to using EIS rather than LTS, except for the storm track regime. In comparison to observations from the International Satellite Cloud Climatology Project (ISCCP), CanAM-4.1, AGCM-3, and ERA-40 underestimate the low cloud fraction in stratiform regimes. ERA-Interim is shown to reproduce particularly well low cloud regimes and the relationship between large-scale circulation and stability.Science, Faculty ofEarth, Ocean and Atmospheric Sciences, Department ofGraduat