Detailseite
Projekt Druckansicht

Representation of the convective atmospheric boundary layer during cold-air outbreaks in regional models: a joined study based on observations, Large Eddy Simulation and mesoscale modelling

Fachliche Zuordnung Physik, Chemie und Biologie des Meeres
Förderung Förderung von 2010 bis 2014
Projektkennung Deutsche Forschungsgemeinschaft (DFG) - Projektnummer 171803021
 
Erstellungsjahr 2015

Zusammenfassung der Projektergebnisse

Cold-air outbreaks (CAO) play an important role for the air-ice-ocean interaction over huge regions of the polar oceans. This has been shown already by previous work. Thus, the focus of the present project was on the uncertainties in the representation of the multi scale physical CAO processes in models. Open questions to be answered concerned the impact of roll vortices on the dynamic characteristics of the boundary layer, the representation of CAO evolution in regional scale models dependent on the spatial resolution, and the impact of external forcing on the CAO and roll evolution characteristics. Furthermore, we studied the smallscale feedback mechanisms between the ocean mixed layer and convection rolls. We investigated these questions using previous and new observations by AWI, mesoscale models (NH3D and METRAS), and a large eddy simulation model (PALM). The latter model was run by IMUK on a massively parallel computer which allowed a very high resolution using 50 m grid spacing in a domain of 16 × 190 km2 . Furthermore, boundary conditions were implemented that were more realistic than the cyclic boundary conditions in many earlier investigations. The two mesoscale models were run at AWI using horizontal grid spacings between 30 km as in regional climate models and 1 km as in high resolution mesoscale models. One of the most important results was the identification and explanation of a jet-like structure (IBJ) of the wind field that occurs often over the open ocean during CAOs. Its strength was found to depend on the flow orientation relative to the orientation of the ice edge. The convective development and the IBJ could be simulated with the mesoscale models in close agreement with the LES and observations when adequate turbulence closures were used. But it was also found that with a horizontal grid spacing of more than 30 km the IBJ is underestimated. Independent on the representation of the IBJ a further result was that an improvement of convective boundary layer details in the mesoscale models were possible by implementing explicitly vertical entrainment through the capping inversion and by using the so-called eddy diffusivity/mass flux closure. The comparison between results of the mesoscale models and LES revealed the unexpected result that the closures which were originally developed for cases with no convection rolls can be used also for cold-air outbreaks with rolls without further modification related to the roll structure. This could be explained by LES runs showing that although rolls can contribute significantly to the total vertical turbulent fluxes, these fluxes do not differ between roll and non roll-cases when the large scale meteorological forcing is the same in both cases. In other words, (forced) roll convection does not increase vertical transports, but takes over a part of the unorganized turbulent transport. A future further consideration of the parametrization topic will be possible, e.g., within the Grey Zone project (of WMO) for which the data sets of 21 LES runs of our project were delivered which differ by the used resolution and by the meteorological forcing. The LES data were used also to investigate a possible smallscale feedback mechanism between convection rolls and the oceanic mixed layer. It was shown that such a feedback exists in the ocean, however, its amplitude was so small that a relevant impact vice versa to the atmosphere does not exist.

Projektbezogene Publikationen (Auswahl)

 
 

Zusatzinformationen

Textvergrößerung und Kontrastanpassung