A Study on the Dynamics of Antarctic Blizzards: Roles of Cyclones in the Southern Ocean and the Continental Topography

Blizzards, i.e. severe snow storms, are classified by Japanese Antarctic Research Expeditions into three classes A-C according to visibilities, wind speeds and durations at Syowa Station (69.0°S, 39.6°E). The number of blizzards happening in an average year is about 25. It is known that blizzards often occur when cyclones approach Syowa Station (Sato and Hirasawa, 2007). The PANSY radar, the first Mesosphere-Stratosphere-Troposphere (MST) radar in the Antarctica, started continuous observations of the troposphere and lower stratosphere on 30 April 2012. The PANSY radar observes vertical profiles of three-dimensional winds at Syowa Station continuously with fine height and time resolutions of Δz@150 m and Δt@1 min. That radar has observed unusually strong and deep vertical wind disturbances (VWDs) extending from the lower troposphere to the tropopause during several blizzard events (Sato et al., 2013). In this study, we use a global non-hydrostatic model NICAM (Satoh et al., 2008) to examine two class-B blizzard events in 15-18 June 2012 associated with the two cyclones from the middle latitudes. Our focuses are on the mechanisms of the cyclone developments and their relationship on the blizzard events and the strong VWDs.

We used a stretch-grid option of NICAM to enable a horizontal mesh size of about 5 km near Syowa Station under a limited computational environment. The center of the stretched grid was at 65°S, 15°E to set the resolution in the upstream region of Syowa Station higher. Data from NCEP FNL was interpolated to make the initial condition. The simulation was made over 120 h from 0000 UT 14 June 2012. Simulated outgoing long radiation was similar to the observed infrared satellite images. Comma-shaped bright clouds were found when the cyclones were in the middle latitudes, while only low clouds remained at the vortex centers when they reached the coastal region of the Antarctic continent. Compared to the reanalysis data (MERRA), the time evolution during the development and decay phases of the cyclones was quite realistic, although the eastward movements were slightly faster. Interestingly, strong VWDs similar to the PANSY radar observations appeared near Syowa Station, when the cyclones reached the coastal region of the Antarctica. These results indicate that the NICAM has an ability to simulate the atmosphere in the Antarctic.

We analyzed the simulation data and found a similarity of the dynamical structure and its time evolution between the two cyclones. Longitude-altitude cross sections of the pressure and temperature indicated that the cyclones were baroclinic in the middle latitudes. The cyclone structure became gradually barotropic as they approached the Antarctica. The vorticities at the centers of the two cyclones at a 1 km height increased when the cyclones were located to the north and to the south of 60°S, whereas those at 5 km increased only when the cyclones were to the north of 60°S. Next, the vorticity budget was examined. It is shown that the convergence and vertical advection of the environmental vorticities largely contributed to the cyclone developments, while the vorticity tilting and frictions worked to decay the cyclones. The high vorticities at 1 km were initially elongated in the east-west direction in the baroclinic stage. The wind associated with the cyclone modified the elongated high vorticity region to a spiral shape, causing strong cyclonic winds in the peripheral region of the cyclone system. The self-advection by the strong cyclonic winds to the Antarctic coast brought strong easterly winds sufficient to be categorized as class B blizzards.

Next, the dynamics of the atmospheric disturbances in the Antarctic coastal region was examined. In the simulation, strong VWDs similar to the radar observation appeared near Syowa Station, when the cyclones reached the coast of the Antarctica. The disturbances had the horizontal scale about 200 km × 30 km and were located to the lee side of the coastal mountains. Such strong vertical winds were scarcely observed over the ocean except for spotty cumulus regions in the simulation. Moreover, strong vertical winds were rarely present, when the cyclones were located far from the Antarctic continent, or when the easterly winds along the coast were weak. These results suggest that both the coastal steep topography and the strong easterly winds associated with the cyclones made the strong vertical winds. Froude number of the flow was close to unity, suggesting that the downslope winds and hydraulic jumps are possible mechanisms to cause VWDs observed by the PANSY radar. The radar observation also shows that significant negative vertical flux of zonal momentum was associated with the VWDs. This feature was not simulated by the model experiment.

The results in this study suggest the following mechanism of the blizzards in the Antarctic coastal regions. A cyclone having experienced the baroclinic development in the extratropical region is barotropically developed again while it approaches the coast of the Antarctica. The associated strong westward winds at the coast sometimes have Froude number reaching to unity and enhance downslope winds to cause a blizzard. Strong VWDs are observed at the station when a front edge of the downslope winds reaches there. Since such steep topography are commonly observed in the whole Antarctic coastal regions, these strong vertical winds may be generated everywhere and, therefore, possibly make a significant impact on the momentum and heat budgets of the Antarctic atmosphere.