PNNL

Abstract

Simple climate models (SCMs) are computationally efficient and capable of emulating global mean output of more complex Earth system models (ESMs). In doing so, SCMs can play a critical role in climate research as stand-ins for the computationally more ex- pensive models, especially in studies involving low resolution, computationally efficient climate data. Here we use Hector v2.0 to emulate the multi-forcing historical and RCP scenario output for 31 concentration and seven emission driven ESMs. When calibrat- ing Hector, sufficient comparison data must be used to constrain the model, otherwise climate and/or carbon parameters affecting physically related processes may be able to tradeoff with one another, making it impossible successfully calibrate Hector to individ- ual ESMs. Ultimately we were able to design two successful calibration protocols, one for concentration-driven Hector and another for emissions-driven Hector. These findings are relevant to other SCM calibration exercises including Bayesian approaches, where insufficient comparison data could have subtle impacts on the joint marginal probabil- ity density. Plain Language summary We calibrated a Simple Climate Model (SCM), Hector, to efficiently reproduce or emulate the output of more complex climate models. Using SCMs as emulators is a com- mon practice in the climate science and climate-economic research communities and are often central in international reports, such as the IPCC assessment reports. In order to use Hector as a satisfactory emulator we found that Hector must be constrained with multiple output variables to produce calibration fits that rely on physically reasonable parameter values.