Theses Doctoral

# Quantifying and Understanding the Linkages between Clouds and the General Circulation of the Atmosphere

Lipat, Bernard

Due to the wide range of physical scales involved, clouds cannot be fully resolved in models of the global climate, and so are parameterized. The resultant model deficiencies in simulating important cloud processes within the current climate are strongly implicated in the large uncertainty in model predictions of future climate changes. Previous work has highlighted the uncertainties in predictions of future climate related to thermodynamic cloud changes, understanding of which requires detailed observations of small-scale cloud microphysics. In this thesis, we argue that understanding the linkages between mid-latitude clouds and the general circulation of the atmosphere can advance efforts to constrain their response to climate forcing. We make this argument with three main methods of analysis: 1) observations, 2) state-of-the-art general circulation models, and 3) experiments with an idealized model of the global climate.
First, we perform a comprehensive investigation of the observed inter-annual relationships between clouds, their radiative effects, and key indices of the large-scale atmospheric circulation. Using reanalysis data and satellite retrievals, we find a relationship between the edge of the Hadley circulation (HC) and the high cloud field that is largely robust against season and ocean basin. In contrast, shifts of the mid-latitude eddy-driven jet latitude, which had been the focus of previous work on the coupling between mid-latitude clouds and circulation, only correlate with the high cloud field in the wintertime North Atlantic. In that season and basin, poleward shifts of the circulation are associated with anomalous shortwave cloud radiative warming. During all seasons in the Southern Hemisphere, however, poleward shifts of the circulation are associated with anomalous shortwave cloud radiative cooling.
Second, we examine Coupled Model Intercomparison Project phase 5 (CMIP5) model output to evaluate the models' simulation of the inter-annual co-variability between the Southern Hemisphere HC extent and the shortwave cloud radiative effect. In the control climate runs, during years when the HC edge is anomalously poleward, most models reduce their cloud cover in the lower mid-latitudes (approximately 30$^\circ$S - 45$^\circ$S) and allow more sunlight to warm the region, although we find no such shortwave radiative warming in observations. We correlate these biases in the co-variability between the HC extent and shortwave cloud radiative anomalies with model biases in the climatological HC extent. Models whose climatological HCs are unrealistically equatorward compared to the observations exhibit weaker climatological subsidence in the lower mid-latitudes and exhibit larger increases in subsidence there with poleward HC extent shifts than models with more realistic climatological HCs. This behavior, based on control climate variability, has important implications for the model response to forcing. In 4$\times$CO$_2$-forced runs, models with unrealistically equatorward HCs in the control climatology exhibit a stronger shortwave cloud radiative warming response in the lower mid-latitudes and tend to have larger values of equilibrium climate sensitivity than models with more realistic HCs in the control climatology.
The above correlative analyses suggest that uncertainty in the linkages between mid-latitude clouds and the general circulation of the atmosphere contributes to uncertainty in the model response to forcing. Finally, we use simulations of the global climate in an idealized aquaplanet model to show that the biases in the climatological Southern Hemisphere circulation do indeed contribute to much of the model spread in the cloud-circulation coupling. We find that for the same 1$^\circ$ latitude poleward shift, simulations with narrower climatological HCs exhibit stronger mid-latitude shortwave cloud radiative warming anomalies than simulations with wider climatological HCs. The shortwave cloud radiative warming anomalies result predominantly from a subsidence warming of the planetary boundary layer, which decreases low-level cloud fraction and is stronger for narrower HCs because of a tighter mean meridional circulation. A comparison of the spread across aquaplanet simulations with that across CMIP5 models suggests that about half of the model uncertainty in the mid-latitude cloud-circulation coupling stems from this impact of the circulation on the large-scale temperature structure of the boundary layer, and thus can be removed by improving the representation of the climatological circulation in models. Therefore, a more realistic representation of the Hadley circulation in models can improve their representation of the linkage between mid-latitude clouds and the atmospheric circulation in the current climate and increase overall confidence in predictions of future climate.

## Files

• Lipat_columbia_0054D_14809.pdf application/pdf 13.5 MB Download File