ClimateBench: A benchmark for data-driven climate projections
Abstract
While some of the most complex Earth System Models have simulated a small selection of Shared Socioeconomic Pathways it is impractical to use these expensive models to fully explore the space of possible future scenarios. Such policy explorations therefore mostly rely on one-dimensional box models, or simple pattern scaling approaches to approximate the physical climate response to a given scenario.
Here we present ClimateBench - a benchmark dataset based on a suite of CMIP, AerChemMIP and DAMIP simulations performed by NorESM2, and a set of baseline machine learning models that emulate its response to a variety of forcers. These surrogate models can accurately predict annual mean global distributions of temperature, diurnal temperature range and precipitation (including extreme precipitation) given a wide range of emissions and concentrations of carbon dioxide, methane and aerosols. We discuss the accuracy and interpretability of these emulators and consider their robustness to physical constraints such as total energy conservation. We will also introduce a novel, physically constrained, Gaussian process regression approach which exactly reproduces the global mean temperature change predicted by a box-model. Explicitly linking (physical) box models and spatial emulators in this way opens a wide range of opportunities to improve prediction, consistency and mathematical tractability. Further, by combining with existing forcing emulators, this approach can seamlessly combine top-down and bottom-up (process based) views on the underlying full complexity models (NorESM2 in this case) and ultimately bridge the gap between simulation and understanding.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFMGC22I0688W