Understanding subtropical cloud feedbacks in anthropogenic climate change simulations of CMIP5 models

Abstract:

Subtropical marine boundary layer clouds over the eastern subtropics are poorly simulated by climate models and contribute substantially to inter-model differences in climate sensitivity. The aim of the present study is to better understand inter-model differences in projected cloud changes and to constrain the cloud feedback to warming. To do this, we compute independent relationships of cloud properties (cloud fraction, cloud-top height, and cloud radiative effect) to interannual variations in sea surface temperature, estimated inversion strength, horizontal surface temperature advection, free-tropospheric humidity, and subsidence using observations and as simulated by models participating in the Coupled Model Intercomparison Project phase 5. Each relationship is considered to be independent because it represents the association between some cloud property and a meteorological parameter when the other parameters are held constant. We approximate modelled cloud trends in climate change simulations as the sum of the simulated cloud/meteorology relationships multiplied by the respective meteorological trends. We compare these estimated cloud trends to the sum of the observed cloud/meteorology relationships multiplied by the simulated meteorological trends. This method allows us to better understand the sources of inter-model differences in projected cloud changes, including whether cloud/meteorology relationships or meteorological trends dominate the spread of cloud changes. We approximate the true cloud trend due to climate change as the sum of the observed cloud/meteorology relationships multiplied by the multi-model mean meteorological trends. The results may provide an observational and model constraint on climate sensitivity.