1:45 PM - 3:15 PM
[ACG30-P12] Delay effect of teleconnection exited by Tropical Atlantic convection on Arctic sea ice
Keywords:Tropical Atlantic, convection, teleconnection, Arctic sea ice
1. Introduction
Arctic sea ice fluctuates from year to year, and it has the potential to bring cold waves to Japan in years with little sea ice. This topic has been actively discussed among researchers in recent years, and it is considered important to understand the causes of Arctic sea ice variability. Several studies have examined the reasons for the interannual variability of Arctic Sea ice, suggesting that ENSO and Atlantic SST excite teleconnections that affect Arctic Sea ice at the same time. However, no consideration has been given to how tropical Atlantic convective activity affects sea ice several months later. Therefore, this study aims to clarify the process by which convective activity in the tropical Atlantic activates teleconnection and affects Arctic Sea ice with a delay.
2. Data and Methods
Outgoing longwave radiation (OLR) from the National Oceanic and Atmospheric Administration (NOAA) was used as an index of convective activity, JRA-55 from the Japan Meteorological Agency 55-year long-term reanalysis data was used as atmospheric data, and HadISST was used as ocean data. The analysis period was October to December from 1979 to 2021 (43 years), and monthly average data were used for each. The December sea ice index, which was created by averaging the sea ice concentration in the area framed in green in Fig. 1a, was subjected to lag regression analysis against the October atmospheric field data and OLR. Next, we performed a lag regression analysis of the October OLR index against the atmospheric and oceanic field data in October, November, and December.
Finally, we conducted an experiment using a linear baroclinic model (LBM: Watanabe and Kimoto 2000) to confirm the steady-state response of the atmosphere to heat sources associated with convective activity in the tropical Atlantic. A heat source of up to 0.8 K/day was given to the middle troposphere in the tropical Atlantic according to the distribution of diabatic heating calculated by regression analysis.
3. Results
Regressing the sea ice index in December to geopotential height at 300 hPa in October and the OLR, a wave train pattern was observed and convective activity was active in the tropical Atlantic Ocean. We, therefore, regressed the tropical Atlantic OLR index for October to a geopotential height of 300 hPa for the same month. A similar pattern was observed when the December sea ice index was regressed to the 300hPa geopotential height.
Next, October’s tropical Atlantic OLR index was regressed on the SST and SIC for October, November, and December. High-temperature anomalies in the Arctic SST persisted from October to December, and sea ice decreased as the season progressed.
An LBM experiment was performed to confirm the steady-state response of the 300 hPa surface geopotential height to the tropical Atlantic heat source. As a result, a wave train pattern similar to the result of regression analysis appeared.
4. Summary
The regression analysis and LBM results suggest that teleconnection excited by convective activity in the tropical Atlantic may have a delayed effect on sea ice in the Arctic region. Although this study hypothesizes that the persistence of SSTs affects Arctic sea ice, other reasons are possible and should be discussed further in the future.
Arctic sea ice fluctuates from year to year, and it has the potential to bring cold waves to Japan in years with little sea ice. This topic has been actively discussed among researchers in recent years, and it is considered important to understand the causes of Arctic sea ice variability. Several studies have examined the reasons for the interannual variability of Arctic Sea ice, suggesting that ENSO and Atlantic SST excite teleconnections that affect Arctic Sea ice at the same time. However, no consideration has been given to how tropical Atlantic convective activity affects sea ice several months later. Therefore, this study aims to clarify the process by which convective activity in the tropical Atlantic activates teleconnection and affects Arctic Sea ice with a delay.
2. Data and Methods
Outgoing longwave radiation (OLR) from the National Oceanic and Atmospheric Administration (NOAA) was used as an index of convective activity, JRA-55 from the Japan Meteorological Agency 55-year long-term reanalysis data was used as atmospheric data, and HadISST was used as ocean data. The analysis period was October to December from 1979 to 2021 (43 years), and monthly average data were used for each. The December sea ice index, which was created by averaging the sea ice concentration in the area framed in green in Fig. 1a, was subjected to lag regression analysis against the October atmospheric field data and OLR. Next, we performed a lag regression analysis of the October OLR index against the atmospheric and oceanic field data in October, November, and December.
Finally, we conducted an experiment using a linear baroclinic model (LBM: Watanabe and Kimoto 2000) to confirm the steady-state response of the atmosphere to heat sources associated with convective activity in the tropical Atlantic. A heat source of up to 0.8 K/day was given to the middle troposphere in the tropical Atlantic according to the distribution of diabatic heating calculated by regression analysis.
3. Results
Regressing the sea ice index in December to geopotential height at 300 hPa in October and the OLR, a wave train pattern was observed and convective activity was active in the tropical Atlantic Ocean. We, therefore, regressed the tropical Atlantic OLR index for October to a geopotential height of 300 hPa for the same month. A similar pattern was observed when the December sea ice index was regressed to the 300hPa geopotential height.
Next, October’s tropical Atlantic OLR index was regressed on the SST and SIC for October, November, and December. High-temperature anomalies in the Arctic SST persisted from October to December, and sea ice decreased as the season progressed.
An LBM experiment was performed to confirm the steady-state response of the 300 hPa surface geopotential height to the tropical Atlantic heat source. As a result, a wave train pattern similar to the result of regression analysis appeared.
4. Summary
The regression analysis and LBM results suggest that teleconnection excited by convective activity in the tropical Atlantic may have a delayed effect on sea ice in the Arctic region. Although this study hypothesizes that the persistence of SSTs affects Arctic sea ice, other reasons are possible and should be discussed further in the future.