More accurate quantification of model-to-model agreement in externally forced climatic responses over the coming century

Abstract

Separating how model-to-model differences in the forced response (UMD) and internal variability (UIV) contribute to the uncertainty in climate projections is important, but challenging. Reducing UMD increases confidence in projections, while UIV characterises the range of possible futures that might occur purely by chance. Separating these uncertainties is limited in traditional multi-model ensembles because most models have only a small number of realisations; furthermore, some models are not independent. Here, we use six largely independent single model initial-condition large ensembles to separate the contributions of UMD and UIV in projecting 21st-century changes of temperature, precipitation, and their temporal variability under strong forcing (RCP8.5). We provide a method that produces similar results using traditional multi-model archives. While UMD is larger than UIV for both temperature and precipitation changes, UIV is larger than UMD for the changes in temporal variability of both temperature and precipitation, between 20° and 80° latitude in both hemispheres. Over large regions and for all variables considered here except temporal temperature variability, models agree on the sign of the forced response whereas they disagree widely on the magnitude. Our separation method can readily be extended to other climate variables.

Ubiquity of human-induced changes in climate variability

“…Other studies have employed individual model simulations, small (≤10 members) ensembles, or CMIP multi-model ensembles (12)(13)(14)(15)(16)(17) to address whether surface temperature and precipitation variability may change under global warming. To date Large Ensemble studies of changes in variance have mainly focused on specific quantities, timescales, or regions (7,(18)(19)(20)(21). However, to our knowledge, the full power of the Large Ensemble framework has not been applied to gauge broad-scale forced changes in climate statistics, including changes in variance, spectrum, pattern, phase and covariance, for a wide range of quantities, regions, and timescales.…”

Ubiquity of human-induced changes in climate variability

“…4b), the standard deviations calculated separately over all years spanning 1960-1989 and 2070-2099 were first calculated, and then averaged over the two respective periods. The intention with the calculation of both standard deviations and correlations is to harness the full power of the Large Ensemble, and is analogous to the empirical orthogonal function (EOF) EOF-E snapshot method previously applied with empirical orthogonal functions (EOFs) (Maher et al, 2018). 270…”

“…Other studies have employed individual model simulations, small (≤10 members) ensembles, or CMIP multi-model ensembles (Rind et al, 1989;Raisanen, 2002;Huntingford et al, 2013;Screen, 2014;Stouffer and Wetherald, 2007;Wetherald, 2009) to address whether surface temperature and precipitation variability may change under global warming. To date Large Ensemble studies of changes in variance have mainly focused on specific 50 quantities, timescales, or regions (Deser et al, 2020;Pendergrass et al, 2017;Maher et al, 2019;Haszpra et al, 2020;Maher et al, 2021). However, to our knowledge, the full power of the Large Ensemble framework has not been applied to gauge broad-scale forced changes in climate statistics, including changes in variance, spectrum, patterns, phase, and variance, for a wide range of quantities, regions, or timescales.…”

“…CC BY 4.0 License. in forced changes exhibits pronounced differences between models (Maher et al, 2021) (their Supplementary Figures 7 and 360 8). Thus it is our hope that our work will motivate further investigations of forced change in Earth system variance across a broad range of timescales under existing archives of large ensemble simulations (Deser et al, 2020;Schlunegger et al, 2020).…”

Surface temperature variability in climate models with large and small internal climate variability

“…However, despite its importance, it is very difficult to quantify the characteristics of surface temperature variability associated with ICV in climate models, including in phases 3, 5, and 6 of the Coupled Model Intercomparison Project (CMIP3, CMIP5, and CMIP6, Meehl et al ., 2007; Taylor et al ., 2012; Deser et al ., 2012b, Maher et al ., 2021). Such difficulties are mainly due to the notion that there are various sources related to the variability in the surface temperature simulated in climate models, including ICV, model biases, and external factors that include anthropogenic forcing, stratospheric ozone concentration, and land use change (Lehner et al ., 2020).…”