On the impact of forced roll convection on vertical turbulent transport in cold air outbreaks

Abstract

We investigated the impact of roll convection on the convective boundary layer and vertical transports in different cold air outbreak (CAO) scenarios using large eddy simulations (LES). The organization of convection into rolls was triggered by upstream heterogeneities in the surface temperature, representing ice and water. By changing the sea ice distribution in our LES, we were able to simulate a roll and a nonroll case for each scenario. Furthermore, the roll wavelength was varied by changing the scale of the heterogeneity. The characteristics of the simulated rolls and cloud streets, such as aspect ratios, orientation of the roll axes, and downstream extensions of single rolls agreed closely with observations in CAO situations. The vertical turbulent fluxes, calculated for each simulation, were decomposed into contributions from rolls and from unorganized turbulence. Even though our results confirmed that rolls triggered by upstream heterogeneities can substantially contribute to vertical turbulent fluxes, the total fluxes were not affected by the rolls.

Meteorology and oceanography of the Atlantic sector of the Southern Ocean—a review of German achievements from the last decade

“…Figures from Gryschka et al (2014) which demonstrate that the observed development of the convective boundary layer during CAOs, including turbulent fluxes, are well reproduced with grid sizes of 4-15 km when adequate non-local turbulence closures are used. These closures do not need to account explicitly for roll convection, which can be explained by the findings of Gryschka et al (2014) described above. Furthermore, Chechin et al (2013) found that the formation of a wind maximum in the convective layer over open water with strong impact on the turbulent fluxes requires grid sizes smaller than 30 km for an accurate reproduction.…”

“…Kristovich et al 1999). However, as discussed in Gryschka et al (2014), this needs to be proven. One reason is that it is not possible to find an observed reference case without rolls for the same largescale forcing.…”

“…This can be achieved by prescribing the surface temperatures with or without upstream heterogeneities as was done in the LES parameter study of Gryschka et al (2014). The authors carried out 27 LES for 12 different CAO scenarios, covering a wide parameter range typical for CAOs.…”

“…The solid black lines (appearing as almost one line, a, c) represent the subgrid-scale contributions to the total turbulent fluxes. (Spikes in the profiles (left column) below 100 m appear only due to the analysis of fluxes in post processing and do not occur within the model physics (see Gryschka et al (2014)). Figures from Gryschka et al (2014) which demonstrate that the observed development of the convective boundary layer during CAOs, including turbulent fluxes, are well reproduced with grid sizes of 4-15 km when adequate non-local turbulence closures are used.…”

“…(Spikes in the profiles (left column) below 100 m appear only due to the analysis of fluxes in post processing and do not occur within the model physics (see Gryschka et al (2014)). Figures from Gryschka et al (2014) which demonstrate that the observed development of the convective boundary layer during CAOs, including turbulent fluxes, are well reproduced with grid sizes of 4-15 km when adequate non-local turbulence closures are used. These closures do not need to account explicitly for roll convection, which can be explained by the findings of Gryschka et al (2014) described above.…”