The Influence of Temperature-Dependent Precipitation Efficiency on Cloud Feedbacks and Climate Sensitivity

Equilibrium climate sensitivity estimates from current climate models have long been associated with a large spread, ranging from 2.0 to 4.5 K. Recent studies found that differences in the treatment of precipitation efficiency – specifically the autoconversion rate for clouds to form rain – contribute to the large inter-model spread. (e.g. Zhao et al 2015). It is common for convective parameterization schemes in climate models to carry a constant precipitation efficiency, though such values can vary significantly from one model to the other. Mauritsen and Stevens (2015) explore a similar mechanism, noting that a strong temperature-dependent precipitation efficiency increases hydrological sensitivity and produces negative cloud and water vapor feedbacks in the ECHAM6 model. However, changes to cloud feedbacks depend heavily on a combination of cloud microphysics and changes in cloud extent. To pursue this question, we employ temperature-varying precipitation efficiencies of various strengths in CESM (version 1.2.2 coupled with a slab ocean), a model on the higher end in the range of climate sensitivities. We find that each simulation, with increasing temperature-dependent precipitation efficiency, reaches a higher climate sensitivity to a doubling of CO₂, contrary to the findings by Mauritsen and Stevens (2015). Using the radiative kernel technique (e.g., Soden et al., 2008), we present an analysis of the climate feedbacks that contribute to this unexpected result. Finally, we present a sensitivity study in which we investigate the extent to which these feedbacks are sensitive to cloud microphysical properties like the thermodynamic phase and size of the detrained cloud hydrometeors from deep convection.