Vol. 59 No. 3 (2020)

‘Preferred Trajectories’ defined by mass and potential vorticity conservation

Julio Sheinbaum
Departamento de Oceanografía Física/ CICESE, Ensenada, Baja California, México
Jorge Castro
Instituto d Universidad del Mar/ Puerto Angel, Oaxaca, México

Published 2020-07-01


  • circulación geostrófica de gran escala,
  • compresibilidad,
  • restricciones de dirección,
  • superficies neutrales
  • large-scale geostrophic,
  • compressibility,
  • direction constraints,
  • neutral surfaces

How to Cite

Ochoa, J., Badan, A., Sheinbaum, J., & Castro, J. (2020). ‘Preferred Trajectories’ defined by mass and potential vorticity conservation. Geofísica Internacional, 59(3), 195-207. https://doi.org/10.22201/igeof.00167169p.2020.59.3.2094


Most schemes to estimate ‘absolute’ geostrophic velocities, in the absence of actual velocity measurements, use directional constraints of the flow at different vertical levels. These constraints allow the determination, often as a least square problem, of the integration constants in the thermal-wind equation. Examples of such directions are those defined by the intersection of constant potential temperature and isohaline surfaces, which under appropriate approximation are material surfaces. Here we show that under adiabatic, non-diffusive, geostrophic, hydrostatic motions, but allowing for compressibility, hence the focus being on the large-scale circulation, a pair of orthogonality constraints, much closely related with dynamical balances than the conservation of potential temperature and salinity, yield a flow direction. These constraints are the conservation of ‘local’ potential density and potential vorticity, in their reduction consistent with such approximations. ‘Neutral’, ‘Orthobaric’, and ‘Topobaric’ surfaces are approximately material surfaces defined as a result of the conservation of ‘local’ potential density, but in order to build such global surfaces additional assumptions are required. The conservation constraint is an inexact differential equation that cannot define, uniquely, a global surface. Here we explicitly show that to define the flow direction, there is no need to build global surfaces out of inexact differentials, as would also be the case with the iso-potential vorticity surfaces, thus avoiding additional assumptions. The ‘Preferred Trajectories’ are then well-defined paths as integrals along this flow direction without being the intersection of global surfaces. Some examples are included for illustrative purposes. Further analysis including error propagation are beyond the scope of this work and left for future studies.