Uncertainty in Climate Impacts Projections due to Model Internal Variability
Abstract
Uncertainty in climate projections is classically understood to be driven by three components; scenario uncertainty, inter-model uncertainty, and intra-model uncertainty (natural variability). Socioeconomic studies of climate impacts have increasingly accounted for the first two sources of uncertainty in recent years, but little attention has been paid to the role of internal variability. Instead, standard practice generally consists of using single realizations of climate models to construct impacts projections, despite the fact that a given single, scenario-driven run poorly captures the full breadth of interannual to decadal variability inherent in the climate system. Especially in near-term projections, where internal variability dominates, consequences of plausible climate extremes may be left unaccounted for. Furthermore, the relationship between climate variables and socioeconomic outcomes is often non-linear; as a result, the effect of internal variability on impact uncertainty may be suppressed or enhanced beyond the ensemble spread of modeled climate variables. In this study, we use simulations from multiple large ensembles to determine the magnitude of uncertainty due to internal variability on results from classical econometric studies on future heat mortality, corn yields, and per-capita GDP. We find that additional uncertainty from natural variability can be substantial. Relative increases in uncertainty are strongest in projections of the first half of the 21st century, and when the relationship between the climate and socioeconomic variables is particularly non-linear. We recommend future impacts studies acknowledge the role of internal model variability when assessing the uncertainty in their results.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFM.A51E..03L