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

Abstract

By analyzing large ensemble simulations using the Community Earth System Model (CESM_LE), the Max Planck Institute Earth System Model Grand Ensemble (MPI_GE), and Coupled Model Intercomparison Project phase 5 (CMIP5) climate models, we quantified internal climate variability (ICV) of surface temperature in each model based on the spread of simulated global mean surface temperature from the ensemble mean. Then, we examined the characteristics of simulated surface temperature variability in climate models with large and small ICV in the present climate and in a future climate. Both the CESM_LE and MPI_GE members with large ICVs tended to simulate larger surface temperature variability at low latitudes, including El Niño and Southern Oscillation (ENSO) variability, and larger cooling and warming trends of the global mean surface temperatures than those with small ICVs in the present climate. Similar characteristics were observed in CMIP5 climate models with large and small ICVs in the present climate. This implies that surface temperature variability including extreme climate events should be cautiously examined in climate models with large and small ICVs. On the other hand, the characteristics of surface temperature variability simulated in the CMIP5 climate models with large or small ICVs were similar from the present climate to future climate with magnitude of ICVs. This was in contrast to that simulated in the CESM_LE and MPI_GE, in which the magnitude of ICV changes between the present climate and the future climate. We inferred that these differences between CMIP5 climate models and large ensemble simulations could primarily be attributed to intermodel differences in the CMIP5 climate models, including model physics and parameterizations.

Internal Climate Variability in the Present Climate and the Change in ENSO Amplitude in Future Climate Simulations

“…This might be due to the fact that ENSO variability largely contributes to the variability of global surface temperature in the present climate (Thompson et al, 2009). Yeh et al (2021) also argued that surface temperature variability at low latitudes (30 • S−30 • N) could be considered as a key source for determining the intensity of ICV in climate models. It is found that the ENSO amplitude simulated in the Strong_ICV group either decreases (90%) or stay (10%) the same in the future out of 10-ensemble members.…”

“…Note that both CESM-LENS and GFDL-LENS simulate the maximum ENSO amplitude in boreal winter (figure not shown). Following Yeh et al (2021), the ICV of global surface temperature in each member is defined as the square root of global (60 • N−60 • S, 0-360 • E) surface temperature including both land and ocean following the equation:…”

“…M, N denote the number of total grid point corresponding at 0-360 • E and 60 • S−60 • N, respectively. The reason of why we focused on 60 • S−60 • N instead of 90 • S−90 • N is that the global ICV is largely determined in the surface temperature variability in low latitude regions (Yeh et al, 2021) and to exclude large model biases of unknown physics and parameters on polar regions (Hyun et al, 2017). Note that the main results are similar if the region expanded to 90 • S−90 • N. Hereafter, this metric will be identified by the acronym ICV, which stands for ICV of global surface temperature.…”

“…Thus, it is not immediately clear whether such relationship holds for ICV and ENSO amplitude change in general. Recently, Yeh et al (2021) quantified ICV of surface temperature in large ensemble simulations based on the spread of simulated global mean surface temperature from their ensemble mean. Following this methodology, we defined a metric for the intensity of ICV based on global surface temperature in the present climate using the 35-member Community Earth System Model Large Ensemble (CESM-LENS) and the 30-members Geophysical Fluid Dynamics Laboratory Large Ensemble (GFDL-LENS).…”

Unraveling the Role of the Interdecadal Pacific Oscillation in Recent Tropical Expansion via Large‐Ensemble Simulations

“…While previous studies based on the analysis of CMIP models have enhanced our understanding of internal variability (Allen et al., 2012; N. A. Davis & Birner, 2017; Grise & Davis, 2020; Grise et al., 2019; Tao et al., 2015), it is challenging to entirely disentangle the influence of external forcings from the internal variability with a limited number of realizations (Deser et al., 2020). Compared with CMIP models, a single model initial‐condition large ensemble with sufficient independent realizations is an effective tool for estimating the relative contribution of internal variability (Deser et al., 2020; Yeh et al., 2021). This occurs because, in initial‐condition large‐ensemble simulations, all the members share the same external forcing but are initiated with differing initial states (Maher et al., 2019).…”