- A carefully calibrated primitive-equation model from 41°N to 48°N is used to study the poleward undercurrent off the US west coast. Chapter 2 describes poleward flow over the slope from Eulerian and Lagrangian perspectives. The model is robust, in the sense of several characteristics being qualitatively consistent with observational and modeling studies. For example, poleward flow reaches the surface at the right times, and is an undercurrent at the right times and in the right locations, the cross-shelf scale is 25km. The robust representation also points to qualities that do not conform to the typical way we think of undercurrents. None the less, it is very likely that the reader would find in observational studies (provided they measure deep enough) that poleward flow over the slope persists year round, albeit deeper than about 400m in spring, that poleward flow over the slope typically reaches 1–1.5 km deep, and that poleward flow tends to reach the surface during upwelling months, except when and where there is a coastal jet; numerical models tend to be qualitatively consistent with this representation. A variety of references in the discussion of chapter 2 support this description. This work benefitted from analyzing a relatively small domain and short (one year) simulation, allowing attention to be paid to the details that resulted in a somewhat unusual description, despite supporting evidence spanning 6 decades of scientific literature. Chapter 2 also gives a definition of the undercurrent using Eulerian and Lagrangian information, and is used in chapter 3 to study the dynamics. Additional descriptions of the model coastal flow, for example cross-shelf and vertical connectivity, may be of interest to researchers studying hypoxia, fisheries or upwelling-related productivity.
Chapter 3 investigates model-undercurrent dynamics by analyzing momentum and vorticity balances. First order momentum balances are as expected: decisively geostrophic in the zonal equation, and geostrophic with ageostrophic contributions in the meridional equation. In the second order two-dimensional balance, the zonal equation is mainly a balance between the ageostrophic pressure gradient (pressure gradient minus Coriolis acceleration), except during upwelling when zonal wind is negligible and advection takes its place. Through the year, the meridional equation is mainly a balance between the ageostrophic pressure gradient and meridional wind. Bottom stress increases in importance during upwelling months. The three-dimensional vorticity balance does not explain the undercurrent because there is a term that is proportional to the undercurrent (identical in vertical and cross-shelf structure) yet it is an order of magnitude smaller than the leading balance; the signal of interest is buried in noise. Vorticity from the depth-integrated balance supports this result: topographic waves and an advective signal are considerably more energetic than the undercurrent signal of interest. Vorticity from both the depth-averaged and depth-integrated momentum equations reveal that the main balance is an arrested topographic wave with an additional term, advection of planetary vorticity; equivalently, the main balance is topographic Sverdup with an additional contribution from Ekman pumping due to bottom stresses. Evidence is presented to advance the hypothesis that the main undercurrent balance is the arrested topographic wave, as suggested by Pedlosky and Csanady in the 1970’s and 1980’s, with a small barotropic contribution (<10%) due to Sverdrup balance. This dynamical explanation is consistent with the description of poleward flow reaching the surface during upwelling months, except when and where there is an upwelling coastal jet.
Each of chapters 2 and 3, is a standalone manuscript, from abstract through discussion.