Project goals are to:

1. develop regional circulation model of CGOA (based on SCRUM code);
2. develop basin-scale circulation model of NE Pacific (SEOM);
3. couple (1) and (2); and
4. incorporate regional NPZ and IBM models in CGOA.

Status:

The CGOA regional model is operational. See the figures below. Each contains shaded and contoured bathymetry, with velocity vectors from the topmost layer of the model (in m/s). Every other current vector is plotted; the mean grid spacing of this non-telescoped area is approximately 10 kilometers. What the model has done here is simply to spin up for a day and a half in response to an initial Levitus temperature and salinity field (interpolated to the model grid using OA routines). The bathymetry is a nice blend of ETOPO5 and more accurate, finer scale data. Much of the Alaska Coastal Current has emerged (not much in the way of an Alaskan Stream, though). Many transients are still evident. We are gradually developing longer runs, as we adjust both the time step parameters and the number of vertical levels (here only four).

The Pacific basin model has been redesigned for implicit time-stepping of the free sea surface, to allow affordable extended integrations of the multi-layer SEOM code. (Speed-ups of a factor of seven have been obtained.) A target grid -- with variable resolution down to about 25 km in the CGOA -- has been developed (see figure below). The layered code is now being tested. Following this, wind-driven and tidally forced basin-scale simulations will be run.

Open boundary conditions have now been improved to allow cleaner simultaneous forcing by tides and low-frequency circulation. These improvements should allow straightforward one-way nesting of the regional and basin-scale models, with outputs of the basin-scale model providing boundary conditions for the regional model. A method is now in place for the proper time phasing of the many tidal components, corresponding to specific years in the regional simulations.

The NPZ model has been developed and coded in 1-D form. Stability issues are currently being addressed.

A method has been worked out for transferring surface information from the physical and NPZ models to the individual based model (IBM) of salmon, which has been running for some time now with lower-resolution (OSCURS wind-driven) physical output. Rand will be on the west coast again this summer, for collaboration with Hermann and Hinckley on the physical-biological coupling. He will also be conferring with Jack Helle, for biological field data relevant to intialization/verification of the models.