On the Minimum Scale of Surface Fronts

Abstract

A series of numerical experiments of a surface front forced by stretching deformation using Clark's nonhydrostatic model at very high resolution is presented. These simulations are compared to those reported by Williams who used hydrostatic models at lower resolution. The main purpose was to determine whether this front would collapse (in the absence of friction) to a scale similar to that reported by Shapiro et al. (most of the temperature gradient contained in 200 m). The question is whether there is a natural physical process in the frontal dynamics which limits the frontal collapse in the absence of diffusion processes. For this front we could not find a natural limiting process, although the mechanism discussed by Orlanski et al. appears to be operating. The minimum scale is determined by the vertical resolution. At the vertical and horizontal resolutions we tried, the vertical resolution determined the scale because the slope of the front is so shallow. Some of the structure found by Cullen and Purser by extending the semigeostrophic models beyond the initial development of a discontinuity is apparent in our solutions.