Emulating Earth System Model temperatures: from global meantemperature trajectories to grid-point level realizations on land

Abstract. Earth System Models (ESMs) are invaluable tools to study the climate system's response to specific greenhouse gas emission pathways. Large single-model initial-condition and multi-model ensembles are used to investigate the range of possible responses and serve as input to climate impact and integrated assessment models. Thereby, climate signal uncertainty is propagated along the uncertainty chain and its effect on interactions between humans and the Earth system can be quantified. However, generating both single-model initial-condition and multi-model ensembles is computationally expensive. In this study, we assess the feasibility of geographically-explicit climate model emulation, i.e., of statistically producing large ensembles of global spatially and temporally correlated land temperature field time series at a negligible computational cost which closely resemble ESM runs spanning from 1870 to 2099. For this purpose, we develop a modular framework that consists of (i) a global mean temperature emulator, (ii) a local mean temperature emulator, and (iii) a local residual temperature variability emulator. We first show that to successfully mimic single-model initial-condition ensembles of yearly temperature, it is sufficient to train on a single ESM run, but separate emulators need to be calibrated for individual ESMs given fundamental inter-model differences. We then emulate 40 climate models of the Coupled Model Intercomparison Project Phase 5 (CMIP5) to create a super-ensemble , i.e., a large ensemble that closely resembles a multi-model initial-condition ensemble. Furthermore, the thereby emerging ESM-specific emulator calibration parameters provide essential insights on inter-model divergences across a broad range of scales which can be viewed as a model ID of core properties of each ESM. Our results highlight that, for temperature at the spatio-temporal scales considered here, it is likely more advantageous to invest computational resources into generating multi-model ensembles rather than large single-model initial-condition ensembles. Such multi-model ensembles can then be extended to super-ensembles with geographically-explicit temperature emulators like the one presented here.