Numerical Modeling of Monterey Bay Circulation and Ecosystem Dynamics

The physical characteristics of the marine environment (e.g., ocean currents, temperatures, and salinities) play a major role in the distribution and evolution of all material within the Sanctuary. This includes both beneficial material, such as the phytoplankton that form the base of the food web, and hazardous material, such as spilled oil. Observing the physical state of the area that makes up the Sanctuary is a daunting task, particularly since the relevant ocean currents span a wide range of scales from breaking waves and small eddies near coastal rocks and promontories out to kilometers-wide deep flows along the continental slope.

The complexity and vast extent of motions in the ocean dictate that some type of model must be coupled to the necessarily limited direct observations. The model itself may take on many different forms. One example is the physical notion that, due to the earth’s rotation, wind blowing along the coast from the northwest will produce a net offshore current near the surface. For decades, this conceptual model has enabled predictions of coastal upwelling—which is needed to replace the surface waters—based simply on estimates of alongshore winds. It is now understood that much of the productivity within the Sanctuary depends on the nutrients that are upwelled into the lighted surface waters by this type of circulation. It is also understood, however, that this upwelling model is an oversimplification. Upwelling is observed to occur in isolated locations along the coast and nutrient-rich waters are spread horizontally by upwelling-related currents.

For several years, research institutions around Monterey Bay have deployed observing systems that go beyond the simple coastal wind measurements. Monterey Bay Aquarium Research Institute (MBARI), for example, has maintained two or three deep-ocean moorings that report subsurface temperature, salinity, and current information along with surface meteorological data in real time via radio links to shore. MBARI and the Naval Postgraduate School (NPS) have conducted regular ship-based transects across the Sanctuary. NPS and others have deployed high-frequency (HF) radars along the shoreline that produce maps of surface currents over much of the Sanctuary.

A new federal program called the National Ocean Partnership Program (NOPP) has made it possible to coordinate and expand many of these observing systems. In addition, numerical modeling experts from as far away as the University of Southern Mississippi and UCLA have begun to develop sophisticated computer models for the region. Once perfected, these models have the potential to act as “dynamic interpolators” that can fill in among the sparse observations. The physical circulation models also have the capability of hosting embedded ecosystem models, which take the environmental information as input for equations predicting trophic-level interactions within the food web.

The first NOPP project, called the Innovative Coastal-Ocean Observing Network (ICON; see www.oc.nps.navy.mil/~icon), was begun in 1998 and is focused on circulation modeling and the coupling to physical data. The second NOPP project, called Simulations of Coastal Ocean Physics and Ecosystems (SCOPE; see www.mbari.org/bog/NOPP), was initiated in 2000. It is building on the modeling results from ICON and moving beyond them to incorporate ecosystem models. In this case, Monterey Bay is almost unique in the wealth of historical bio-chemical observations that have been taken alongside the physical observations. These data will be needed to validate the developing ecosystem algorithms. The many ICON and SCOPE partner institutions are listed in the table below.

Finally, it is important in this short note to outline the extreme challenge that modeling Monterey Bay circulation represents. Because computers continue to become more and more capable, and because the fundamental physical equations describing ocean currents are known, it may appear straightforward to build and use such a numerical model. This appearance is far from accurate. Briefly, modeling coastal ocean currents is made difficult by a number of factors, some of which are internal to the numerical code while others are external. Internal limitations include the need to have sufficient grid resolution to allow for the interaction among scales and the breakdown of scales through turbulence. Despite today’s computers, it is not possible to run models that cover hundreds of square kilometers with grid resolutions of just a few meters, which would be needed to resolve all of the interacting scales. Beyond that, grid resolutions of less than one centimeter would be needed to resolve the turbulent scales. Hence, it is always necessary to include some overly simplified parameterization of the small-scale currents and turbulence as part of any ocean model.

ICON model surface velocities during an upwelling event (courtesy I. Shulman). The coastal model receives boundary information from a navy regional model, which itself is embedded within a global navy model.

External limitations include the need to have accurate wind forcing, which can itself require high-resolution atmospheric models to produce the variability observed near the coastline. It is also necessary to provide inflow information along all of the open boundaries in a coastal ocean model. This requirement may be the most limiting effect of all. For example, a single snapshot of surface velocities from the ICON model is shown in the accompanying figure. The currents reflect the influence of strong upwelling centers located just north of Santa Cruz and offshore of Point Sur. This complex circulation is realistic and is largely driven by the wind forcing applied to the ICON model. However, the information supplied to the model along the offshore open boundaries will influence and overwhelm this circulation pattern within hours to days, which means that the source of boundary information becomes as important as the internal workings of the ICON model and the model’s wind forcing.

In the case of the ICON model and the next-generation model under development within SCOPE, open boundary information is derived from a second, regional model that covers the entire West Coast with less resolution. The regional model itself is embedded within a lower-resolution global model (see figure). Information passed through this hierarchical scheme includes such effects as thermocline deepening due to coastally trapped waves initiated by El Niño events in the tropical Pacific. In this way, accurate circulation modeling within the Sanctuary really represents the need to model the entire north Pacific Ocean.

Despite the difficulties, the combined modeling and observing components of ICON and SCOPE represent an exciting beginning toward the goal of tracking and predicting ocean movements and productivity throughout the Sanctuary.

Jeffrey D. Paduan(1) and Francisco P. Chavez(2)
(1)Naval Postgraduate School
(2)Monterey Bay Aquarium Research Institute