RESEARCH ARTICLE Impingement of buoyancy-driven flows at a stratified interface Aline J. Cotel Æ Yuriko Kudo Received: 7 December 2005 / Revised: 27 December 2007 / Accepted: 2 January 2008 / Published online: 19 February 2008 Ó Springer-Verlag 2008 Abstract Laser-induced fluorescence (LIF) and parti- cle-image velocimetry (PIV) are used to study both thermals and plumes impinging on a stratified interface. Data are obtained for a central slice of the flow near the stratified interface. Both the thermal and plume are generated by releasing fresh water at the bottom of a tank filled with two layers of salt water of different densities. Thermals and plumes are studied at Reynolds numbers ranging from 3,000 to 8,000, above the value for the mixing transition, a Schmidt number of about 600, and Richardson numbers from 1 to 22. The Richardson and Reynolds numbers are based on the thermal or plume characteristics (size and vertical velocity) before impingement and the initial density difference across the interface. Laser-induced fluorescence (LIF) is used to determine the maximum penetration height, rebound distance and lateral spreading velocity. The vorticity results obtained from the PIV data reveal the vortical structure near impingement. When the thermal impinges upon the stratified interface, a baroclinic eddy generated at the interface appears to merge with eddies comprising the thermal itself to form a vortex ring. This ring remains near the interface, moving mainly along the lateral or horizontal direction away from the region of impinge- ment. These results suggest that lateral transport is significant for thermals impinging on stratified interfaces, and that ignoring such transport may greatly underesti- mate overall transport and mixing in such flows. 1 Introduction Transport in stratified environments is highly dependent on turbulent processes such as entrainment and mixing. An accurate representation of these processes is essential, for example, in improving the accuracy of weather forecasting, pollutant-transport models and global-climate change pre- dictions. Present-day models still lack an adequate representation of the turbulence dynamics and mixing in stratified flows. One example of this is weather forecasting. Obtaining accurate weather forecasts in some cases is difficult, in particular where smaller-scale effects such as those due to variations in local topography or the presence of large bodies of water are significant. The objective of this work is to report on the production of vorticity at impingement and how interface dynamics affects lateral transport in stratified flows. This paper specifically addresses the following questions: • How far does a thermal or plume penetrate past the initial location of the stratified interface? • How does the baroclinic vorticity generated at the interface interact with the initial vorticity of either the plume or thermal? • How is lateral transport related to the production of baroclinic vorticity generated at the time of impingement? • How does the Richardson number affect vorticity dynamics and the related questions? Some of these issues have already been resolved for a vortex ring (Linden 1973; Dahm et al. 1989) and for a jet (Cotel et al. 1997), but this work is, to our knowledge, the first to address these questions for buoyant flows such as thermal and plume. Dahm et al. (1989) described the generation of vorticity and the perturbation of the interface A. J. Cotel (&)  Y. Kudo Department of Civil and Environmental Engineering, University of Michigan, Ann Arbor, MI 48109-2125, USA e-mail: acotel@umich.edu 123 Exp Fluids (2008) 45:131–139 DOI 10.1007/s00348-008-0469-5