The ultimate objective of the United Nations Framework Convention on Climate Change (UNFCCC) to ensure “…stabilization of greenhouse gas emissions at a level that would prevent dangerous anthropogenic interferences with the climate system…” not only implies an obligation to protect humanity’s means of livelihood, but also a clear mandate for research that provides a sound scientific basis for the normative decision of what “dangerous anthropogenic interferences with the climate system” means and where it starts. The relevance is reflected, for example, in the review process of the 1.5°C / 2°C target initiated by the UNFCCC Cancun agreements. Meanwhile, Greenhouse gas emission budgets in line with a given limit of global mean warming have been quantified and Integrated Assessment Models (IAM) provide estimates of mitigation costs for various temperature limits. However, there is only fragmentary and often merely qualitative knowledge of the impacts of different levels of global warming that can be avoided by a more ambitious mitigation policy.
In this situation a temperature-stratified probabilistic assessment of multi-sectoral impacts (e.g. in the water, agriculture, biomes, and health sectors) is urgently needed to facilitate sensible political decisions regarding mitigation and adaptation strategies. In particular, the change in frequency and intensity of extreme climatic events, potentially followed by extreme impacts, is of great societal interest.
The development of a functional relationship between global mean temperature (GMT) and multisectoral impacts is particularly attractive within the context of economic modeling. While Integrated Assessment Models (IAMs) are usually employed to find cost-optimized mitigation strategies given a GMT target, a representation of damages is only very roughly included, or sometimes not at all. In particular, there is no computationally efficient tool to describe spatially resolved impacts associated with the emission pathways generated within these models. Current Atmosphere Ocean General Circulation Models (AOGCMs) providing spatially explicit climate projections are too computationally expensive. In contrast, IAMs generally include only simplified climate models, providing GMT projections without any spatial resolution. However, based on this general set-up, a fast impact emulator providing spatially resolved impacts in terms of GMT change will not only close an important gap with respect to current impact assessments, but also represent a unique way to integrate impacts into economic models.
In comparison to a “localized” impact assessment, the economic perspective demands a truly global view of both the climate and the economic systems, requiring the impact emulator to account for the temporal and spatial correlation of impacts. As often driven by large-scale circulation patterns such as the Northern Annular Mode (NAM) or the El Niño Southern Oscillation (ENSO), extreme events are not expected to occur independently. From the economic perspective this is critical, since simultaneous impacts could have non-linear consequences on prices and welfare that potentially propagate through the world economic system.
Based primarily on the next generation of climate model projections provided by CMIP5 (Coupled Model Intercomparison Project Phase 5) in the context of the IPCC’s Fifth Assessment Report (AR5), on associated multi-model impact projections provided by ISI-MIP (Inter-Sectoral-Impact-Model-Intercomparison-Project), and on in-depth analysis based on the ecohydrological land surface model LPJmL, we will develop a simplified but powerful impact emulator to:
1. Fill the gap in quantitative impact assessment for different levels of mean global warming, and
2. Pioneer the integration of quantified, potentially non-linear consequences of critical simultaneous impacts into an economic model of the agricultural world market.
With an initial focus on impacts on crop yields, the emulator will be designed to be highly adaptable to cover other extreme events in other sectors (hurricanes and flooding events). The emulator will allow for spatially resolved impact projections in terms of GMT. It will account for the spatial correlation of extreme events that is due to large-scale circulation patterns. In addition the probabilistic projections will cover the uncertainties 1) in GMT changes as projected by the simple climate model MAGICC6.0, 2) associated with regional climate changes and changes in extreme events per degree of global warming as provided by a multi-climate model assessment based on the newly available CMIP5 data set, and 3) associated with multi-crop-model simulations provided by ISI-MIP.
The emulator is intended to be computationally fast, i.e. hundreds of impact simulations possible within days, no restrictions on the number of emissions scenarios, allowing for the possibility of integration into economic models.
A precursor of EXPACT allowing for projections of 1) regional temperature and precipitation changes, 2) coral bleaching, and 3) regional sea level rise for user defined emission pathways is available here as part of the PRIMAP online tool live.PRIMAP.