Improving Constraints on Climate System Properties with Additional Data and New Statistical and Sampling Methods (Invited Presentation)

We use the updated MIT Earth System Model (MESM) to derive the joint probability distribution function for Equilibrium Climate sensitivity (ECS), an effective heat diffusivity (K_v), and the net aerosol forcing (F_aer). Climate sensitivity (ECS) in the model is set through an adjustment to the cloud feedback parameter. The vertical diffusion coefficient, K_v, represents the mixing of heat anomalies into the deep ocean by all mixing processes. The net anthropogenic aerosol forcing parameter, F_aer, estimates the contribution of aerosol cooling to the global energy budget. Using a new 1800-member ensemble of MESM runs, we derive PDFs by comparing model outputs against historical observations of surface temperature and global mean ocean heat content. We focus on how changes in (i) the MESM model, (ii) recent surface temperature and ocean heat content observations, and (iii) estimates of internal climate variability will all contribute to uncertainties. We show that estimates of ECS increase and estimates of F_aer are less negative. These shifts result partly from new model forcing inputs but also from including recent temperature records that lead to higher values of ECS and K_v. We show that the parameter distributions are sensitive to the internal variability in the climate system and that using a single estimate of variability can lead to PDFs that are too narrow. When considering these factors, we derive our best estimate for the joint probability distribution for the climate system properties. We estimate the 90-percent confidence intervals for climate sensitivity as 2.7-5.4 ^oC with a mode of 3.5 ^oC, for K_v as 1.9-23.0 cm²s^-1 with a mode of 4.41 cm²s^-1, and for F_aer as -0.4 – -0.04 Wm^-2 with a mode of -0.25 Wm^-2. Lastly, we estimate TCR to be between 1.4 and 2.1 ^oC with a mode of 1.8 ^oC.