From Global Mean Temperature to Regional Climate Impacts

There is strong motivation in Integrated Assessment Modelling to generate methods for quickly approximating regional climate impacts. Here, we use climate impacts from the Inter-Sectoral Impacts Model Intercomparison Project phase 2b (ISIMIP2b) to emulate impacts based on global mean surface temperature anomaly, beginning with crop yields and extending the analysis to other sectors. Unlike many existing methods, we emulate impacts directly, rather than first emulating regional climate. We find that a second order polynomial function of global mean temperature is a very good predictor of changes in regional crop yield, in line with existing research on damage function literature. Socio-economic and other drivers must be treated with care in the process. By using multiple driving climate models and impact models that are available for many impacts in ISIMIP, we are also able to sample uncertainty in the severity in changes in impacts with increasing warming.

We built our climate impacts emulator for the 10 regions considered by the IPCC for the Sixth Assessment Working Group 3 report, but the selection of regions is flexible, and could be applied to any existing IAM. Our impacts emulator can also be used to construct updated damage functions for use in economic models and cost-benefit IAMs.

This approach will be used to generate climate impact functions within the newly created FRIDA integrated assessment model, which seeks to account for feedbacks between all components of the human-Earth system – such as the climate, food systems, and energy production – as these are key drivers of the response to anthropogenic forcing.