Linkage Between Deep Convection, Large-scale Circulation and Low Cloud Feedback

It is well established that shortwave cloud feedback associated with marine boundary layer cloud (MBLC) fraction change is the primary contributor to the uncertainty in equilibrium climate sensitivity (ECS). Su et al. (2014) showed a metric that represents the model performance in capturing the spatial structures of zonal-mean cloud fraction and relative humidity associated with the Hadley Circulation can be used to constrain ECS. Qu et al. (2018) confirmed that the Su metric is correlated with the low cloud feedback. The Su metric is already in use by some groups testing Coupled Model Intercomparison Project Phase 6 (CMIP6) models, but process-oriented diagnostics that identify steps in the underlying physical pathways require development.

Work Summary: The proposed Type 1 project aims to establish the physical processes that link
deep convection, the Hadley Circulation and low cloud feedback, develop process-oriented model diagnostics to characterize CMIP6 model representation of these pathways, and apply multiple observations to constrain these pathways so as to constrain ECS. Based on existing knowledge and a set of perturbed physical experiments (PPEs) (e.g., Schiro et al. 2019) in which deep convective parameters were altered, we hypothesize three potential pathways for deep convection and large-scale circulation to modify low cloud fraction (LCF) change in a warmer climate: (1) the temperature-stability pathway that hinges on the tropospheric temperature anomalies propagated by wave dynamics, (2) the moisture-mixing pathway that may depend on shallow ascent and subgrid-scale mixing of moisture between the free troposphere and the MBL, and (3) the radiation stability pathway that involves longwave radiation mediated subsidence control on LCF. We will analyze CMIP6 simulations to determine the relative contribution of each pathway to the model spread in ECS and use satellite observations and reanalysis datasets to assess CMIP6 model representations of these processes and develop diagnostic tools to trace the sources of model errors.
A statistical hierarchical framework will be employed to estimate the probability of ECS given
present-day climate simulations of the physical pathways and their observations following Bayes’ theorem (Bowman et al. 2018). In addition, we will employ the Pareto-optimal technique to place multi-objective constraints on ECS (Langenbrunner and Neelin 2017a; 2017b) and compare the results from the two methods. This proposed work builds upon the strong expertise of the proposal team on model-observation diagnostics and evaluations of climate models using satellite observations especially related to deep convection and its parameterizations. We will leverage the process-oriented diagnostics approach and framework in the MAPP Model Diagnostics Task Force (MDTF) and contribute new diagnostics to augment the existing software package, helping to accelerate model improvements and reduce the uncertainties in climate projections. Relevance to competition: The proposed study targets the intimate linkage between deep convection and low cloud feedback through large-scale circulation and addresses one of the grand challenges put forward by the World Climate Research Programme on “Clouds, Circulation and Climate Sensitivity”. It is aligned with the MAPP solicitation as one of the “investigations to constrain climate model sensitivity focusing on clouds, convection and aerosol processes and their role within the coupled Earth system”. It will “develop key process-level metrics and diagnostics using relevant observations to accelerate the improvement of models” and enhance evaluation capabilities of the MDTF software package. This proposed study addresses the MAPP’s goals to “use observations to develop direct and indirect constraints on models’ climate sensitivity and apply them to reduce model projections’ uncertainty focusing on temperature projections.”