Emergence of a climate oscillation in the Arctic Ocean due to global warming

nature climate change Article https://doi.org/10.1038/s41558-024-02171-3 Emergence of a climate oscillation in the Arctic Ocean due to global warming Received: 16 April 2023 Soong-Ki Kim 1 & Soon-Il An 1,2,3 Accepted: 24 September 2024 Published online: xx xx xxxx Check for updates Global warming is expected to be able to trigger abrupt transitions in various components of the climate system. Most studies focus on abrupt changes in the mean state of the system, while transitions in climate variability are less well understood. Here, we use multimodel simulations to show that sea-ice loss in the Arctic can trigger a critical transition in internal variability that leads to the emergence of a new climate oscillation in the Arctic Ocean. The intensified air–sea interaction due to sea-ice melt causes an oscillatory behaviour of surface temperatures on a multidecadal timescale. Our results suggest that a new mode of internal variability will emerge in the Arctic Ocean when sea ice declines below a critical threshold. Future projections from the latest generation of global climate models show that the Arctic will experience dramatic changes due to global warming1,2. Arctic sea-ice cover is projected to rapidly decrease and become practically ice-free in summer within the twenty-first century under all levels of anthropogenic emissions scenarios3,4. This massive sea-ice loss would expand the area of open ocean and intensify the interaction between the atmosphere and the ocean over the Arctic. As demonstrated by the cases of El Niño/Southern Oscillation5, Indian Ocean Dipole6 and Atlantic Multidecadal Oscillation7, air–sea interaction processes and associated coupled feedbacks are some of the key ingredients for climate oscillation. Large-scale air–sea interaction processes give rise to a mode of climate oscillation, a recurring cycle of climate variables that deviate from the background climate noise. Notably, ref. 5 demonstrated that El Niño/Southern Oscillation theoretically ceases to occur if the air–sea interaction process is sufficiently weakened. This suggests that climate oscillation mode can be switched on or off following changes in background climatological state. Therefore, the Arctic sea-ice melt due to global warming may activate a new climate oscillation in the Arctic Ocean that does not currently manifest in the present sea-ice-covered state. Although atmospheric scale variabilities in the Arctic have been previously well recognized8–17, the hypothetical air–sea coupled mode, which is expected to emerge in a warm climate, has not yet been systematically explored in current literature. In this Article, we show multiple lines of evidence that a climate oscillation characterized by a multidecadal variation in annual surface temperature can emerge in the Arctic due to global warming. We use a total of 134 simulations run from three climate model intercomparison project archives—the Climate Model Intercomparison Project Phase 6 (CMIP6)18,19, Long Run Model Intercomparison Project (LongRunMIP)20 and Pliocene Model Intercomparison Project Phase 2 (PlioMIP2)21–23. Our analysis will focus mainly on the annual mean surface temperature anomaly over the central Arctic Ocean (80° N–90° N) (the domain does not include the marginal Arctic seas such as the Barents Sea and Greenland Sea). Evidence from the CMIP6 archive We begin the analysis of annual surface temperature variability in the projected warm climate using the historical and Shared Socioeconomic Pathway 5–8.5 (SSP5–8.5) experiment output from the CMIP6 archive. The SSP5–8.5 is the high greenhouse gas emissions scenario for forcing fossil-fuelled development with a radiative forcing level of 8.5 W m–2 in 2100. We use merged historical and SSP5–8.5 runs spanning 1850 to 2099 for 41 models (Supplementary Table 1 and Methods). The multimodel ensemble mean shows a 5 °C increase in global mean surface temperature (GMST) and 95% decrease in summer Arctic sea-ice area from 1850 to 2099 (Fig. 1a,c). Thus, the CMIP6 ensemble here shows the response of Arctic surface temperature variability in a warm climate where sea ice is mostly diminished (Fig. 1d). We perform a spectral analysis on the annual mean surface temperature anomaly of each ensemble member (Methods). The CMIP6 analysis provides ensemble-wise evidence for the emergence of a multidecadal oscillation in the projected warm climate in the twenty-first century. The spectral analysis shows that the Irreversible Climate Change Research Center, Yonsei University, Seoul, Republic of Korea. 2Department of Atmospheric Sciences, Yonsei University, Seoul, Republic of Korea. 3Division of Environmental Science and Engineering, Pohang University of Science and Technology (POSTECH), Pohang, Republic of Korea. e-mail: 1 Nature Climate Change Article https://doi.org/10.1038/s41558-024-02171-3 Ensemble mean 18 16 14 1900 1950 2000 2050 2100 0 c Ensemble mean –10 –20 –30 1850 1900 1950 Year Temperature anomaly (°C) 1 0.30 0 –1 0.25 –2 –3 1900 2100 10 5 Threshold of practically ice-free 0 1850 1900 1950 1950 2000 2050 0.20 2100 Year 2000 2050 2100 Year e 1.0 0.035 Ensemble mean 100 yr moving variance 0.030 0.025 0.5 0.020 0 0.015 –0.5 0.010 –1.0 0.005 1850 1900 1950 2000 Year 2050 0 2100 Temperature anomaly variance (°C2) 0.35 Temperature anomaly variance (°C2) Ensemble mean 100 yr moving variance 2 1850 2050 Ensemble mean Year d 3 2000 Temperature anomaly (°C) 1850 10 Sea-ice area (106 km2) b 20 Temperature (°C) Temperature (°C) a ACCESS-CM2 ACCESS-ESM1-5 AWI-CM-1-1-MR BCC-CSM2-MR CAMS-CSM1-0 CESM2 CESM2-WACCM CIESM CMCC-CM2-SR5 CMCC-ESM2 CNRM-CM6-1 CNRM-CM6-1-HR CNRM-ESM2-1 CanESM5 CanESM5-CanOE E3SM-1-1 EC-Earth3 EC-Earth3-CC EC-Earth3-Veg EC-Earth3-Veg-LR FGOALS-f3-L FGOALS-g3 FIO-ESM-2-0 GFDL-ESM4 GISS-E2-1-G HadGEM3-GC31-LL IITM-ESM INM-CM4-8 INM-CM5-0 IPSL-CM6A-LR MCM-UA-1-0 MIROC-ES2L MIROC6 MPI-ESM1-2-HR MPI-ESM1-2-LR MRI-ESM2-0 NESM3 NorESM2-LM NorESM2-MM TaiESM1 UKESM1-0-LL Fig. 1 | Changes in the Arctic temperature for 1850–2099 in the CMIP6 historical and SSP5–8.5 experiment. a–c, Changes in global and Arctic climate. a, GMST. b, Arctic surface temperature. c, September Arctic sea-ice area. The dashed grey line is the threshold of a practically ice-free state (sea-ice area less than 106 km2). All three variables are 40-year moving means, which show their long-term trend. The black thick line is a multimodel ensemble mean. d, Changes in the Arctic surface temperature anomaly. The anomaly is defined as a deviation from the quadratic trend of Arctic surface temperature for 1850–2099 (Methods). Left axis: time series of Arctic surface temperature anomaly (coloured lines). A 10-year low-pass filter is applied to display multidecadal variability. Right axis: multimodel ensemble mean of the 100-year moving variance of Arctic surface temperature anomaly (black line). e, Same as d, but for (...truncated)