-
Here, we perform sensitivity experiments with the Community Atmospheric Model, version 5 (CAM5) to validate our hypothesis. A detailed introduction of CAM5 can be found in Neale et al (2010). The resolution used here is 0.937° × 1.25° in the horizontal directions, with 30 hybrid sigma-pressure levels in the vertical direction.
We design four experiments, including one control run (CTRL) and three sensitivity experiments (Table 1). In CTRL, the boundary conditions of SST and SIC are prescribed as the climatological monthly mean during 1980–2016. In the sensitivity experiment of BSIC (Exp. BSIC), we first calculate the anomalous composite SIC over the Bering Sea (55°–70°N, 165°–195°E) based on the high and low years of BSIC, and then these composite BSIC anomalies in MAMJ are added to the CTRL sea ice boundary condition. In Exp. BOK, we consider the composite BSIC anomalies in MAMJ in the Bering Sea, as well as in the Okhotsk Sea (45°–60°N, 140°–160°E). We design Exp. BOK in order to improve the simulated responses in the North Pacific region. Finally, we conduct a SST sensitivity experiment (Exp. ASST). In Exp. ASST, we first calculate the monthly SST differences in JJA over the North Atlantic (15°–55°N) between high and low BSIC years, and then add these differences in the CTRL SST boundary condition. Exp. ASST is designed to investigate the impact of the dipole SST pattern in the central North Atlantic on summer atmospheric circulation over Eurasia. In all experiments, other boundary conditions are identical and fixed in the year 2000. CTRL runs for 40 years. All the sensitivity experiments are initialized from the 10th year in CTRL and run for 30 years. We analyze the last 30-year model outputs.
SST SIC CTRL Global monthly climatology during 1980–2016 Monthly climatology during 1980–2016 in Northern Hemisphere Exp BSIC Same as CTRL Composite monthly SIC anomalies over the Bering Sea [55°–70°N, 165°E–155°W] during MAMJ added to CTRL Exp BOK Same as CTRL Composite monthly SIC anomalies over the Bering Sea [55°–70°N, 165°E–155°W] and the Sea of Okhotsk [45°–60°N,140°–150°E] during MAMJ added to CTRL Exp ASST Composite monthly SST anomalies over the North Atlantic [15°–55°N, 80°W–0°] during JJA added to CTRL Same as CTRL Table 1. Experimental designs and boundary conditions.
From these experimental designs, the differences between the Exp. BSIC and the CTRL reveal the effect on the atmosphere of high MAMJ BSIC; the differences between the Exp. BOK and the CTRL show the atmospheric changes in response to the MAMJ sea ice anomalies over North Pacific (Bering sea and the Okhotsk sea) between the high and low BSIC years; the differences between the Exp. ASST and the CTRL indicate how the atmosphere over the Eurasian continent responds to the North Atlantic dipole SST pattern in summer.
-
Figure 6 shows the simulated SLP anomalies from two sea ice anomaly forcing experiments, Exp. BSIC and Exp. BOK. In Exp. BSIC, the simulation for the North Atlantic part is quite similar to the results obtained from ERAI (Fig. 3a). In MAMJ, there is a high-pressure anomaly over the east coast of 60°N in the North Atlantic with a low-pressure anomaly centre over the west coast of the North Atlantic (black box in Fig. 6a). The simulated result is quite well matched with the result in ERAI over the North Atlantic region, but the Exp. BSIC shows a limited similarity with ERAI over the North Pacific. The features of the intensified Aleutian low and Okhotsk high, as seen in Fig. 3a, are not able to be reproduced in the Exp. BSIC. Therefore, another experiment (Exp. BOK) is performed to try to improve simulation responses for the North Pacific region. The MAMJ SLP anomaly in the Exp. BOK produces the intensified Aleutian low and Okhotsk high (Fig. 6b), resembling the pattern in ERAI (Fig. 3a). It is worth noting, that the SLP anomaly over the North Atlantic in Exp. BOK is qualitatively consistent with Exp. BSIC. We see that the positive SLP anomaly centre over 60°N in the North Atlantic is also produced in Exp. BOK (Fig. 6b). This indicates that the MAMJ SLP anomaly can also persist into June in Exp. BOK. This indicates that the MAMJ SLP anomaly over the North Atlantic is mainly driven by the MAMJ Bering SIC changes, not the Okhotsk SIC.
Figure 6. Anomalies of SLP (shadings, hPa) and horizontal wind at 925 hPa (vectors, m s−1). (a) Exp. BSIC in MAMJ, relative to CTRL, and (b) Exp. BOK in MAMJ, relative to CTRL, both where dots and purple vectors denote changes with significance higher than 95% confidence level based on Student’s t-test.
In Exp. BSIC, the WAF at 200 hPa shows a wave train that originates from the mid-high latitude North Pacific. This wave train propagates to the central North Atlantic (about 40°N) and then towards the east coast of the North Atlantic (60°N). The wave train contains the stationary wave energy and leads to a positive geopotential height anomaly over the North Atlantic (60°N, Fig. 7a). This wave train is also shown in Exp. BOK (Fig. 7b), which is more similar to the results found in ERAI (Figs. 5a and b). Compared to the result obtained from ERAI (Figs. 5a and b), the negative geopotential anomaly centre over the North Pacific is weaker and moves further southward in the Exp. BSIC (Fig. 7a). Therefore, unlike the result in ERAI (Fig. 5c), the deep, wide low geopotential anomaly system over 180°–135°W is not shown in the meridional profile averaged along 40°–60°N in Exp. BSIC, and neither is the upward WAF from 180° (Fig. 7d). In Exp. BOK, the simulated position of the negative geopotential anomaly over the North Pacific is much more similar to the ERAI than in Exp. BSIC. We also see the WAF in Exp. BOK through a similar pathway (Fig. 7e) to the ERAI (Figs. 5c and d).
Figure 7. Simulated anomalies in geopotential height (shading, gpm) and WAF (vectors, m2 s–2). (a) Anomalies in geopotential height at 200 hPa in MAMJ for Exp. BSIC, relative to CTRL. (b) same as (a), but for Exp. BOK, (c) anomalies in geopotential height at 200 hPa in JJA for Exp. ASST, relative to CTRL. (d) Anomalies in vertical-horizontal cross section averaged along 40°–60°N in MAMJ for Exp. BSIC, relative to CTRL. (e) same as (d), but for Exp. BOK, (f) anomalies in vertical-horizontal cross section averaged along 45°–55°N in JJA for Exp. ASST, relative to CTRL.
From the above analysis, we find that the atmospheric circulation anomalies pattern over the North Atlantic in ERAI is mainly caused by changes in MAMJ BSIC rather than sea ice changes in the Sea of Okhotsk. The high MAMJ BSIC anomaly causes the positive SLP anomaly centre over 60° N in the North Atlantic during MAMJ. The high SLP anomaly in the North Atlantic corresponding with the anti-cyclonic circulation favors the formation of the SST dipole-pattern in the central North Atlantic in JJA. Considering the MAMJ sea ice changes in both the Sea of Okhotsk and the Bering Sea can improve the simulation skill of the stationary wave pathway.
-
The WAF at 200 hPa shows a zonally orientated wave train originating from the Atlantic Ocean and propagating to the west (Fig. 7c), which looks similar to the wave train shown in ERAI (Figs. 5e and f). On the meridional profile averaged along 45°–55°N, the WAF moves upwards from the lower layer over the eastern coast of the North Atlantic and is transmitted down to the Baikal−NEC, where a deep low-pressure anomaly system is formed (Fig. 7f). When the dipole SST pattern is forced in the Exp. ASST, the experiment basically reproduces the large-scale atmospheric circulation anomalies at mid-latitudes over the Eurasian continent in summer (Fig. 8). The simulated T2m field (Fig. 8a) shows a negative anomaly at Lake Baikal and a significant positive anomaly over the Sea of Japan. At 500 hPa, a robust negative geopotential height anomaly occurs over Baikal−NEC and a positive anomaly centre appears near the Sea of Japan (Fig. 8b), which is consistent with the result in ERAI (Figs. 4c and d). The Exp. ASST also closely reproduced the JJA water vapor transport originating from the tropical Pacific and the Sea of Japan and converging in Baikal−NEC (Fig. 8c compared to Fig. 2d).
Figure 8. Simulated atmospheric circulation changes in JJA for Exp. ASST, relative to CTRL. (a) Changes in JJA T2m (°C). (b) Changes in JJA geopotential height (gpm) at 500 hPa. (c) Changes in JJA VQ (kg m−1 s−1). Purple vectors denote changes with significance higher than 95% confidence level based on Student’s t-test.
SST | SIC | |
CTRL | Global monthly climatology during 1980–2016 | Monthly climatology during 1980–2016 in Northern Hemisphere |
Exp BSIC | Same as CTRL | Composite monthly SIC anomalies over the Bering Sea [55°–70°N, 165°E–155°W] during MAMJ added to CTRL |
Exp BOK | Same as CTRL | Composite monthly SIC anomalies over the Bering Sea [55°–70°N, 165°E–155°W] and the Sea of Okhotsk [45°–60°N,140°–150°E] during MAMJ added to CTRL |
Exp ASST | Composite monthly SST anomalies over the North Atlantic [15°–55°N, 80°W–0°] during JJA added to CTRL | Same as CTRL |