Home > Press Releases > details

Press Releases

July 15, 2022
JAMSTEC
Atmosphere and Ocean Research Institute, The University of Tokyo
Research Institute for Applied Mechanics, Kyushu University

Newly identified mechanism may explain Arctic sea-ice decline
— Remote effects of Gulf Stream warming —

1. Key Points

The rate of winter sea-ice decline in the Barents–Kara Sea, on the Atlantic side of the Arctic Ocean, tends to be underestimated even in the most recent simulations.
We successfully simulated winter sea-ice decline in the Barents–Kara Sea over the past several decades by modifying the sea surface temperature variations in the Gulf Stream region using observational data.
The results of this study can reduce the uncertainty in Arctic climate projections with respect to anthropogenic greenhouse gas and aerosol emissions.

2. Overview

Yoko Yamagami, of the Research Center for Environmental Modeling and Application, Japan Agency for Marine–Earth Science and Technology (JAMSTEC), Masahiro Watanabe of the Atmosphere and Ocean Research Institute, the University of Tokyo, Masato Mori of the Research Institute for Applied Mechanics, Kyushu University, and Jun Ono of the Institute of Arctic Climate and Environment Research, JAMSTEC conducted and analyzed climate simulations that reproduced past conditions in the Arctic. They found that variability in the sea surface temperatures (SSTs) of the Gulf Stream region is responsible for the observed reduction in winter sea-ice in the Barents–Kara Sea, on the Atlantic side of the Arctic Ocean (Fig. 1). Winter sea-ice in the Barents–Kara Sea has declined significantly in recent decades and such changes are believed to affect extreme weather events in the mid-latitudes of the Northern Hemisphere, including Japan. Meanwhile, the rate of winter sea-ice decline in the Barents–Kara Sea has generally been underestimated, even in the latest simulations (e.g., CMIP6, ※1), but the factors contributing to this rate remains unclear.

In the study reported here, the impact of surface water temperatures in the Gulf Stream region on the rate of winter sea-ice decline in the Barents–Kara Sea was investigated by comparing the results of two simulations. The first simulation (hereafter, “HIST experiment”) was used to reproduce past climates, while the second (hereafter, “NAGA experiment”) was used with corrected SST variations in the Gulf Stream region of the North Atlantic Ocean (Fig. 1) to the observed data. The NAGA experiment was found to reproduce the rate of sea-ice decline within a realistic range (Fig. 2).

In the Barents–Kara Sea, oceanic currents transport more heat from the Atlantic domain, driving significant increases in SSTs (Fig. 3). The analysis of CMIP6 climate simulations also revealed a statistically significant correlation, wherein the faster SSTs rose in the Gulf Stream region, the faster the sea-ice in the Barents–Kara Sea declined. The results of this study suggest that variations in SST in the Gulf Stream region may control the long-term variability of Arctic sea-ice. The research team plans to further improve their simulations and investigate the relationship between the Gulf Stream and the Arctic Ocean in climate change projections.

This work was supported by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan through the “Integrated Research Program for Advancing Climate Models” (Grant no. JPMXD0717935457), the MEXT “advanced studies of climate change projection” (Grant no. JPMXD0722680395), the “Arctic Challenge for Sustainability Project II” (ArCS II) (Grant no. JPMXD1420318865), and the Japan Society for the Promotion of Science Grants-in-Aid for Scientific Research (KAKENHI, Grant nos. JP20H05729 “Deep Numerical Analysis Climate project,” and JP19H05703, JP22H01299, JP22H04487, and JP22K14098).

This study was published in Nature Communications on July 15, 2022 (JST).

Title:
Barents-Kara sea-ice decline attributed to surface warming in the Gulf Stream
DOI:
10.1038/s41467-022-31117-6
Authors:
Yoko Yamagami1, Masahiro Watanabe2, Masato Mori3, Jun Ono1
Affiliation:
1. JAMSTEC
2. Atmosphere and Ocean Research Institute, the University of Tokyo
3. Research Institute for Applied Mechanics, Kyushu University

【Supplemental information】

※1
CMIP6 (Coupled Model Intercomparison Project Phase 6): A comparison projcet in simulations using a climate model that simulates Earth’s climate. The model is based on physical laws and includes the various atmospheric, oceanic, and terrestrial phenomena that make up the global climate system, and is being developed by research institutes worldwide. In the study reported here, large ensemble data from 39 CMIP6 model simulations of historical (1970–2014) climate conditions were conducted, with 372 members in total.
1

Fig. 1. Trends in SSTs observed from 1970–2017 and the climatological mean percentage of sea-ice covering the sea surface in winter during that period. The areas outlined in white were defined as the Barents–Kara Sea and the Gulf Stream regions in this study. The former is part of the Arctic Ocean and faces the North Atlantic Ocean via the Norwegian Sea in the west and the Laptev Sea in the east. The Gulf Stream flows from the equator to the north along the eastern coast of the United States.

2

Fig. 2. Box-and-whisker plots of (a) trends in SST in the Gulf Stream region and (b) trends in sea-ice concentrations (i.e., percentage of sea-ice covering the sea surface) in the Barents–Kara Sea based on observations, the HIST and NAGA experiments, and CMIP6. The period was 1970–2017 for all except for CMIP6, for which it was 1970–2014. The ensemble means, indicated by the horizontal lines in the boxes, show that the trends in SSTs and sea-ice concentrations for both the HIST experiment and CMIP6 were small compared to those observed. In the NAGA experiment, trends in SSTs were concentrated near the observed values due to the correction for SST variability (a) and the average trend in decreasing sea-ice concentrations was similar to the observed value (b).

3

Fig. 3. (a) Differences in sea-ice concentration (colors) and sea-ice drift (vectors) trends between the NAGA and HIST experiments (NAGA – HIST). The hatching and black vectors indicate statistically significant trends in the NAGA experiment and the black solid box indicates the Barents–Kara Sea region as defined in this study. (b) Similar to (a) but showing SSTs (colors) and surface current velocities (vectors). The black dashed lines indicate cross sections (70°N, 20°E) where horizontal heat transport was examined in (c) and (d). (c) Trend in ocean heat transport over a cross section at 20°E, where a significant velocity difference was found. (d) Similar to (c) but at 70°N. In the NAGA experiment, SSTs increased in the area where sea-ice was more reduced (a and b) and SSTs increased greatly from the Norwegian Sea to the Barents Sea, with a difference in current velocity toward the interior Barents Sea (b). The trends in heat transport by the ocean at the western boundary of the Barents–Kara Sea (20°E) and in the interior Norwegian Sea (70°N) indicate that in the NAGA experiment, more heat was being supplied to the Barents-Kara Sea than in the HIST experiment, suggesting that the oceanic dynamics leads to reduction of Barents–Kara sea-ice.

Contacts:

(For this study)
Yoko Yamagami, Postdoctoral Researcher, Climate Model Development and Application Group (CliM-DAG), Research Center for Environmental Modeling and Application (CEMA), Research Institute for Global Change(RIGC), JAMSTEC
Prof. Masahiro Watanabe, Department of Climate Variability Research, Division of Climate System Research, Atmosphere and Ocean Research Institute, the University of Tokyo
(For press release)
Press Office, Marine Science and Technology Strategy Department, JAMSTEC
Inquiry Form