Models. The stratosphere-resolving (that is, “high-top”) AGCM used in this study is the SC-WACCM4, a component of the National Center for Atmospheric Research (NCAR) Community Earth System Model version 1.2 (CESM1), with specified chemistry rather than the full interactive chemistry version WACCM4 (29). The removal of interactive chemistry does not change the climatology and variability of the atmospheric circulation below 55 km but much reduces the computational cost, which makes SC-WACCM4 suitable for studies of stratosphere-troposphere dynamical coupling (29). Like WACCM4, SC-WACCM4 has 66 vertical levels and a horizontal resolution of 1.9° latitude by 2.5° longitude, and the model lid is at 5.1 × 10−6 hPa (approximately 140 km). For comparison to a “low-top” model, the CAM4, which shares the same physical parameterizations as WACCM4 (SC-WACCM4), was used. CAM4 has 26 vertical levels, and the model lid is at about 3.5 hPa (30, 31).

Experiments. The modeling experiments in this study are summarized in table S1. Sea surface temperature (SST) and SIC were prescribed to define the model’s surface boundary condition. In the control run (hereafter referred to as CTRL), the repeating climatological seasonal cycle of SST and SIC, from the CESM1-WACCM4 historical outputs (average of seven ensembles) from the CMIP5 and averaged during 1980–1999, were prescribed. In the first perturbation run, referred to as BKS_FL (full response), the settings are identical to those of the CTRL, but the SIC in BKS is replaced by that in the CMIP5 Representative Concentration Pathway (RCP) 8.5 outputs (average of four ensembles) averaged during 2080–2099 (see Fig. 2). The SST in BKS was also replaced with RCP8.5 SST in the open water areas that used to be covered by sea ice in CTRL run, similar to previous studies (13). The second perturbation experiment, BKS_TP, has the same SST and SIC forcings as those in BKS_FL run, but a nudging method was used to largely remove the contribution of the stratospheric pathway and to isolate the tropospheric adjustment by itself (tropospheric pathway). In this nudged experiment, the zonal means of temperature, zonal wind, meridional wind, and specific humidity above 90 hPa were nudged toward those in the CTRL with a time scale of 6 hours. The fields were fully nudged above 54 hPa (the nudging coefficient is 1), and the nudging strength was linearly decreased to zero between 54 and 90 hPa, with no nudging applied below 90 hPa. The tapering region in the vertical, that is, the lower stratosphere (54 to 90 hPa), was chosen to relax the strength of nudging over some vertical scale to prevent issues that would arise at the sharp boundary between no nudging and nudging (32). The nudging was performed at every time step of the model integration, but the target states from the CTRL run were read in every 6 hours, and the model fields were nudged toward the linear interpolation between consecutive target states, which, in this case, were the full time-evolving zonal means of the CTRL simulation. On the basis of Hitchcock and Haynes (32), both the vertical levels of the nudging and the tapering that we used here and the nudging time scale of 6 hours are reasonable choices that achieve what we aim for. In the BKS_TP run, we, therefore, numerically suppressed stratosphere-troposphere dynamical coupling that occurred via a change in the zonal mean stratospheric circulation, allowing the tropospheric pathway to be viewed in isolation. To explicitly quantify the contribution of the stratospheric pathway, another nudging experiment, referred to as BKS_SP (stratospheric pathway), was conducted. The BKS_SP run was the same as the CTRL run, except that the zonal mean state in the stratosphere was nudged toward that in the BKS_FL run with the same nudging coefficients as in the BKS_TP run. The BKS_SP run provides an opportunity to examine the impacts of the perturbed stratospheric circulation by the BKS sea ice anomaly on tropospheric circulation and surface climate when direct surface sea ice forcing is absent.

In addition, to examine the impact of the nudging method, we conducted an additional experiment in which all the settings were the same as the CTRL, except that nudging was applied in the stratosphere and above to nudge the zonal mean state to that of the CTRL itself, referred to as CTRL_NUDG. The results of CTRL_NUDG (zonal wind in December for example) showed that nudging neither changed the zonal mean states nor markedly altered the stationary planetary-scale wave structure (fig. S6). A theoretical discussion of the nudging impact can be found in the study of Hitchcock and Haynes (32).

We also conducted the ARC run, in which the pan-Arctic SIC and SST changes were prescribed. For comparison, two additional experiments with the low-top model version CAM4, that is, CTRL_CAM4 and BKS_CAM4, were performed. These two experiments were the same as CTRL and BKS_FL, except that CAM4 was used.

The anomaly (or response) in the perturbation run was defined as the difference between each perturbation run and the CTRL run. Because the most significant midlatitude response to early winter sea ice changes occurred during the following winter (25, 33), we focused on the responses in DJF. All of the experiments were integrated for 60 model years, with the first 10 years discarded as spin-up. There are thus 49 whole winters in the last 50 model years analyzed in each run in this study. We checked the robustness of the results with two different statistical methods (see the “Statistical analysis” section). The results showed that the conclusions in this study were robust. In addition, we tested whether the same response was obtained when dividing our 49 winters data set in two halves, and it was.

Cold air outbreak. A CAO event is defined as two or more consecutive days during which area-averaged Siberian daily SAT (60°E to 140°E and 45°N to 65°N; highlighted by the red box in Fig. 3) is at least 1 standard deviation (σ) below the DJF mean SAT in the CTRL run, where σ is the average of the daily standard deviation of Siberian daily SAT in 49 winters in the CTRL run (34). The Siberian High index is defined as the weighted area-averaged sea-level pressure (SLP) over 65°E to 100°E and 60°N to 75°N, the region where the maximum in the leading empirical orthogonal function mode of winter SLP in the CTRL run is located (fig. S4C).

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.