Hostname: page-component-7b7567c8c-fvjsw Total loading time: 0 Render date: 2026-11-16T12:36:42.374Z Has data issue: false hasContentIssue false

A perturbation approach to understanding the effects of turbulence on frontogenesis

Published online by Cambridge University Press:  25 November 2019

Abigail S. Bodner*
Affiliation:
Department of Earth, Environmental and Planetary Sciences, Brown University, Providence,RI02912, USA
Baylor Fox-Kemper
Affiliation:
Department of Earth, Environmental and Planetary Sciences, Brown University, Providence,RI02912, USA
Luke P. Van Roekel
Affiliation:
Theoretical Division, Fluid Dynamics and Solid Mechanics, Los Alamos National Laboratory, Los Alamos, NM87545, USA
James C. McWilliams
Affiliation:
Department of Atmospheric and Oceanic Sciences, University of California, Los Angeles,CA90095-1565, USA
Peter P. Sullivan
Affiliation:
National Center for Atmospheric Research, Boulder, CO80307-3000, USA
*
†Email address for correspondence: abigail_bodner@brown.edu

Abstract

Ocean fronts are an important submesoscale feature, yet frontogenesis theory often neglects turbulence – even parameterized turbulence – leaving theory lacking in comparison with observations and models. A perturbation analysis is used to include the effects of eddy viscosity and diffusivity as a first-order correction to existing strain-induced inviscid, adiabatic frontogenesis theory. A modified solution is obtained by using potential vorticity and surface conditions to quantify turbulent fluxes. It is found that horizontal viscosity and diffusivity tend to be readily frontolytic – reducing frontal tendency to negative values under weakly non-conservative perturbations and opposing or reversing front sharpening, whereas vertical viscosity and diffusivity tend to only weaken frontogenesis by slowing the rate of sharpening of the front even under strong perturbations. During late frontogenesis, vertical diffusivity enhances the rate of frontogenesis, although perturbation theory may be inaccurate at this stage. Surface quasi-geostrophic theory – neglecting all injected interior potential vorticity – is able to describe the first-order correction to the along-front velocity and ageostrophic overturning circulation in most cases. Furthermore, local conditions near the front maximum are sufficient to reconstruct the modified solution of both these fields.

Information

Type
JFM Papers
Copyright
© 2019 Cambridge University Press
Figure 0

Table 1.Dimensionless expressions for quantities of interest following ST13 framework in the semi-geostrophic limit, which implies that the along-front velocity is purely geostrophic.

Figure 1

Figure 1.Frontal formation of the dimensionless zeroth-order buoyancy field (a), together with the cross-frontal planes showing the along-front velocity $v$ (b), and overturning streamfunction $\unicode[STIX]{x1D713}$ (c). The cross-frontal, vertical and time axes correspond to the dimensionless $x,z$ and $t$ axes respectively.

Figure 2

Figure 2.Cross-frontal profiles of the zeroth-order along-front velocity $v^{0}$ (shading) at two times defined as early frontogenesis ($t=5$, a) and late frontogenesis ($t=18$, b). Note the different colour bar axes. Superimposed in black contours is the buoyancy field, with intervals of 0.1 in non-dimensional units. Note the singularity developing during late frontogenesis. Here, $x$ and $z$ are the dimensionless cross-frontal and vertical axes, respectively, in real space.

Figure 3

Figure 3.Contours show buoyancy as in figure 2, shading shows the zeroth-order frontal tendency $T_{b}^{0}$.

Figure 4

Figure 4.Contours show buoyancy as in figure 2, shading shows the zeroth-order streamfunction $\unicode[STIX]{x1D713}^{0}$.

Figure 5

Figure 5.Total frontal tendency $T_{b}=T_{b}^{0}+\unicode[STIX]{x1D700}T_{b}^{1}$ for each of the forcing cases, over a variety of $\unicode[STIX]{x1D700}$ (shades of blue), evaluated at the surface, where the zeroth-order frontal tendency maximum (black) becomes infinitely strong. For each forcing case, the maroon line represents the $\unicode[STIX]{x1D700}$ for which $T_{b}=0$. The green vertical line represents the time where the limit of the perturbation approach is reached for each forcing case: horizontal viscosity ($t=11.1$), vertical viscosity ($t=8$), horizontal diffusivity ($t=8.4$), vertical diffusivity ($t=9.1$).

Figure 6

Figure 6.Cross-frontal plane of the total frontal tendency $T_{b}=T_{b}^{0}+\unicode[STIX]{x1D700}T_{b}^{1}$ during early frontogenesis ($t=5$) with the appropriate arrest $\unicode[STIX]{x1D700}$ for each forcing case: horizontal viscosity ($\unicode[STIX]{x1D700}=0.25$), vertical viscosity ($\unicode[STIX]{x1D700}=0.53$), horizontal diffusivity ($\unicode[STIX]{x1D700}=0.04$), vertical diffusivity ($\unicode[STIX]{x1D700}=0.51$). Black ticks indicate the location of the maximum zeroth-order frontal tendency corresponding to that measured in figure 5 and for the sense of where the front is strongest.

Figure 7

Figure 7.Same as in figure 6 for the total along-front velocity $v=v^{0}+\unicode[STIX]{x1D700}v^{1}$.

Figure 8

Figure 8.Same as in figure 6 for the total streamfunction $\unicode[STIX]{x1D713}=\unicode[STIX]{x1D713}^{0}+\unicode[STIX]{x1D700}\unicode[STIX]{x1D713}^{1}$.

Figure 9

Figure 9.Cross-frontal planes of the first-order along-front velocity $v^{1}$ at time $t=5$ for all forcing cases, calculated from the full system (a)–(d), the IQG system (e)–(h), the SQG system (i)–(l), the SQG system with a point source approximation (m)–(p) and the sum of the SQG and IQG systems (q)–(t). Note that for plotting purposes the IQG system is multiplied by a constant.

Figure 10

Figure 10.Same as in figure 9 for the first-order streamfunction $\unicode[STIX]{x1D713}^{1}$. Note that for plotting purposes also here the IQG system is multiplied by a constant.