The PISCO-UCSC US west coast modeling effort consists of an implementation of the Regional Ocean Modeling System (ROMS) at various resolutions. The outermost domain extends from the Baja peninsula to Vancouver Island, is driven by oceanic properties derived from the global state estimation product based at MIT and JPL, called ECCO (Estimating the Circulation and Climate of the Ocean). In the most studied configuration, this outer grid exists at a nominal 10 km resolution and then drives a middle grid at approximately 3 km resolution. In turn, an innermost grid covers the coastal domain from the Big Sur coast to Pt. Arena at about 1 km grid spacing. All three model domains have 42 levels in the vertical and are forced at the surface by atmospheric properties generated by the COAMPS model and provided to us by Jim Doyle at NRL.
Our approach is to use the realistic physical ocean model, now well evaluated against multiple data sources including PISCO and MBARI moorings, to drive a very large number of numerical drifters representing larvae. Thousands of numerical floats are seeded at multiple levels within the central California coastal region and tracked for time periods consistent with pelagic larval duration phase of commercially important rockfish and barnacle larvae. This seeding recurs each day or two days depending on the model configuration. Analysis of the statistics of the trajectories provide useful metrics for quantifying larval dispersal. Thus far, only the simplest approaches to larval behavior has been considered: (a) buoyancy regulation keeping larvae at fixed depths in the water column, transported by the horizontal currents only, and (b) no control, in which individuals are advected by fully 3-dimensional velocities.
In collaboration with NOAA’s National Marine Fisheries Service SWFSC in Santa Cruz, CA, PISCO has documented the seasonal variability in alongshore and crossshore larval transports. Based on the four year period from Jan. 2000 through Dec. 2003 and using isobaric floats on the middle grid, we find a strong semiannual pattern in mean trajectory direction, net displacement, fractional cross-shore displacement, and duration of retention within specified geographical bounds. We have found evidence for enhanced retention in the coastal zone south of Pt. Reyes during both upwelling and non-upwelling periods. Details of the results can be found in Petersen et al. (submitted).
PISCO is analyzing neutrally buoyant larvae statistics from the 1 km innermost grid. We are currently working with the annual and coastwide averages which enable improved estimates of standard measures of dispersion. We find velocity decorrelation times of approximately 1 and 3 days in the cross- and along-shore directions, respoectively. We estimate eddy diffusivities from our model in the range of 4600 m2/s in the alongshore direction and 800 m2/s in the cross-shore direction. We have applied a simple recruitment model based on a competency period for barnacle larvae to investigate connectivity. This recruitment model also enables an annual mean estimate for alongshore dispersion of potentially succesful recruits of approximately 100 km.
PISCO has also extended the simple theory for turbulent dispersion to account for the presence of a coastal boundary. In short, a side boundary creates anisotropy in dispersion (evident for example in our results above). We have found that a mixing length theory can be applied in a coastal environment where the appropriate length scale for turbulent dispersion is the internal deformation radius which varies with shelf depth. Details of both this theory and some results from (2) are nearly in final manuscript form.
In support of the 2007 biophysical field study in Monterey Bay, we have implemented a new version of the west coast model, combining the spatial extent of the full outer domain with the higher resolution of the previously middle grid (nominally 3 km.) The goal is to provide context of the larger California Current System along with improved local dynamical fields over what can be achieved with smaller central coast grids. Initial results appear excellent, and we are in the process of evaluating output with respect to detailed PISCO field observations. Two examples of sea surface temperature from June 2007 are shown in Figure 1.
Figure 1. The surface temperature from the ROMS model on June 21, 2007 (left) and three days later June 24, 2007 (right). Within 3 days, the Northern Monterey Bay region transitioned from relaxation with uniformly warm waters in the broader coastal environment to a more characteristic upwelling circulation with cool waters from Año Nuevo dividing Bay interior waters from those of the broader shelf/slope.