Dataset Information

Striking stationarity of large-scale climate model bias patterns under strong climate change.

ABSTRACT: Because all climate models exhibit biases, their use for assessing future climate change requires implicitly assuming or explicitly postulating that the biases are stationary or vary predictably. This hypothesis, however, has not been, and cannot be, tested directly. This work shows that under very large climate change the bias patterns of key climate variables exhibit a striking degree of stationarity. Using only correlation with a model's preindustrial bias pattern, a model's 4xCO₂ bias pattern is objectively and correctly identified among a large model ensemble in almost all cases. This outcome would be exceedingly improbable if bias patterns were independent of climate state. A similar result is also found for bias patterns in two historical periods. This provides compelling and heretofore missing justification for using such models to quantify climate perturbation patterns and for selecting well-performing models for regional downscaling. Furthermore, it opens the way to extending bias corrections to perturbed states, substantially broadening the range of justified applications of climate models.

SUBMITTER: Krinner G

PROVIDER: S-EPMC6156650 | biostudies-literature | 2018 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Striking stationarity of large-scale climate model bias patterns under strong climate change.

Krinner Gerhard G Flanner Mark G MG

Proceedings of the National Academy of Sciences of the United States of America 20180904 38

Because all climate models exhibit biases, their use for assessing future climate change requires implicitly assuming or explicitly postulating that the biases are stationary or vary predictably. This hypothesis, however, has not been, and cannot be, tested directly. This work shows that under very large climate change the bias patterns of key climate variables exhibit a striking degree of stationarity. Using only correlation with a model's preindustrial bias pattern, a model's 4xCO<sub>2</sub> ...[more]

PMID: 30181268

Similar Datasets

Project description:We modelled the distribution of two vulnerable steppe birds, Otis tarda and Tetrax tetrax, in the Western Palearctic and projected their suitability up to the year 2080. We performed two types of models for each species: one that included environmental and geographic variables (space-included model) and a second one that only included environmental variables (space-excluded model). Our assumption was that ignoring geographic variables in the modelling procedure may result in inaccurate forecasting of species distributions. On the other hand, the inclusion of geographic variables may generate an artificial constraint on future projections. Our results show that space-included models performed better than space-excluded models. While distribution of suitable areas for T. tetrax in the future was approximately the same as at present in the space-included model, the space-excluded model predicted a pronounced geographic change of suitable areas for this species. In the case of O. tarda, the space-included model showed that many areas of current presence shifted to low or medium suitability in the future, whereas a northward expansion of intermediate suitable areas was predicted by the space-excluded one. According to the best models, current distribution of these species can restrict future distribution, probably due to dispersal constraints and site fidelity. Species ranges would be expected to shift gradually over the studied time period and, therefore, we consider it unlikely that most of the current distribution of these species in southern Europe will disappear in less than one hundred years. Therefore, populations currently occupying suitable areas should be a priority for conservation policies. Our results also show that climate-only models may have low explanatory power, and could benefit from adjustments using information on other environmental variables and biological traits; if the latter are not available, including the geographic predictor may improve the reliability of predicted results.

Dataset Information

Striking stationarity of large-scale climate model bias patterns under strong climate change.

Publications

Striking stationarity of large-scale climate model bias patterns under strong climate change.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets