Meteorological Institute, Center for Marine and Climate Research, University of Hamburg, Bundesstr. 55, 20146 Hamburg, Germany
The development of polar mesocyclones is strongly affected by sea ice cover: the contrast of ice covered areas and open water causes regions of enhanced baroclinicity and convection. In return, mesocyclones influence the ice drift and thus, sea ice cover, by the divergence and convergence of their wind fields and high wind velocities. During a cyclone passage an increased ratio between sea ice drift and wind velocity was measured. This possibly indicates that sea ice is broken up during cyclone passage. The aim of the present study is to investigate the effect of ocean and sea ice properties on a baroclinic mesocyclone development when atmosphere - sea ice interactions are considered. Furthermore, the impact of the cyclone passage on sea ice cover and its feedback on cyclone development is studied.
The investigations are based on idealised model studies performed with a coupled atmosphere - sea ice model. The coupled model consists of the atmospheric model METRAS, which is a mesoscale nonhydrostatic model applied in the mesoscale- and - range. The sea ice model is based on the sea ice model by Hibler and modified for the mesoscale range. The basic case is an idealised simulation of a mesocyclone passing through Fram Strait. The mesocyclone initially moves along the ice edge and slightly deepens. Afterwards, it crosses an ice covered area and starts to decay. Based on this scenario sensitivity studies are performed. The impact of the large-scale cyclone track, the ice edge, the ice thickness, the ocean flow velocity and the ocean surface temperature are investigated. The simulations are repeated with sea ice cover kept constant in time.
Cyclones affect the sea ice cover (Figure 1). It is reduced along the cyclone track by divergent sea ice drift in the vicinity of the cyclone center. Additionally, in some cases, the floe length is shortened, because floes are broken up due to internal forces. In the basic simulation, the sea ice concentration is reduced by about 9 % (Figure 1a). In the simulations with 75 % sea ice concentration (Figure 1b) and in the simulation with increased water surface temperature resulting in higher wind velocities, the maximum reduction is about 18 %. The reduction of sea ice concentration along the cyclone track gets stronger the lower the initial sea ice concentration and the higher the wind velocity is. The change of sea ice distribution might be important concerning the organisation of ship tracks.
The areas of reduced sea ice concentration increase the average heat flux. According to the stronger break up, the increase in heat flux gets stronger with a lower initial sea ice concentration, resulting in an approximately doubled average heat flux. In contrast, with a high initial sea ice concentration, changes are in the order of 10 %. The increase of the heat flux averaged over the whole evaluation domain and 24 hours might be also important on the global scale, if it takes place in regions with a lower sea ice cover and a high cyclone frequency. Besides, the changed spatial distribution of heat fluxes affects the boundary layer development and, consequently, the local weather.
One important reason for the deepening of the mesocyclone is the region of enhanced baroclinicity near the ice edge (Figure 2a), that develops due to the temperature contrast between sea ice and areas of open water. In this case, the mesocyclone development is mainly influenced by the position of the large-scale cyclone track relative to the ice edge. If the distance gets smaller than about 50 km to 100 km, depending on the strength of the baroclinic region, the local cyclone track is influenced by the ice edge and the cyclone deepens within the baroclinic region (Figure 3). The baroclinic region is weaker if the sea ice concentration in the ice covered area is reduced. On the other hand, an increase of water surface temperature does not increase the baroclinicity. The wind velocity is stronger and, considering atmosphere-sea ice interactions, the stronger break up of sea ice cover near the ice edge reduces the temperature gradient and results in a less intense mesocyclone development at the ice edge (Figure 3). The deepening at the ice edge is mainly determined by the large-scale cyclone track relative to the ice edge and furthermore by sea ice concentration as well as temperature difference between water and sea ice. In the simulations with lower sea ice concentration and increased water surface temperature, when the cyclone passage remarkably reduces the sea ice concentration along the cyclone track, baroclinicity at the ice edge is also affected.
Figure 2:Horizontal Cross-Section of Eady Growth Rate after 10 Hours Simulation Time in the Basic simulation (left) and after 17 Hours Simulation Time in the Sensitivity Study with Homogenous Initial Ice Coverage (right).
The mesocyclone not only deepens at the ice edge, but also when passing ice covered areas. This deepening takes place just in case of considering atmosphere - sea ice interactions. It is caused by the reduction of sea ice concentration along the cyclone track. These areas of reduced sea ice concentration and, thus, increased heat fluxes produce regions of enhanced baroclinicity (Figure 2b), that do not occur if atmosphere - sea ice interaction is neglected (Figure 4). The mesocyclone is deepening if the break up of sea ice is strong, namely in case of lower sea ice concentration or stronger winds. In case of higher sea ice concentration the break up of sea ice is weak and, therefore, the effect on mesocyclone development negligible. Due to the limited evaluation period, it is not possible to quantify the full effect on mesocyclone development. But the investigations show that in case of lower sea ice concentration and with high wind velocities, the effect of the cyclone - sea ice interaction on mesocyclone development has to be considered.
Figure 4: Horizontal Cross-Section of Eady Growth Rate after 17 Hours Simulation Time in the Simulation with Homogenous Initial Ice Cover and with Fixed Ice Cover (left) and after 17 Hours Simulation Time in the Simulation with Homogenous Initial Ice Cover including Sea Ice Dynamics (right).