C.5.2.8 Modeling (PIs: Signell, McGillicuddy, Lynch)

Rationale: Interconnection between the physical, biological and chemical aspects of the ecology of Alexandrium stems from: 1) redistribution of water column constituents by currents through advective processes, and 2) mediation of biological and chemical reactions by the environment. This duality poses a challenge to understanding observed cell distributions, since differentiation between the sources of variability requires a detailed knowledge of property distributions in space and time and of the processes by which the environment controls rates of biological and chemical transformations. A purely observational approach to such problems is logistically impractical, but the coupling of models with observations offers a useful alternative.

Approach: ECOHAB-GOM will yield a set of multiscale quasi-synoptic maps of physical, biological and chemical constituents of the water column, together with time series of physical and bio-optical properties at mooring sites. The overall objective of the modeling component is to provide a basis for interpretation of the measurements in a suite of multiscale models. Simple representations of Alexandrium population dynamics will be incorporated into the circulation models so that the interactions of these biological processes with transport phenomena can be studied to elucidate the mechanisms controlling PSP outbreaks. The modeling program consists of five elements (Fig. 5).

I. Casco Bay. A model of the Casco Bay region will be developed based on the Blumberg-Mellor (1987) primitive equation model. The objective is to study the dynamics of the source region in a domain that encloses the region of high-resolution field studies (Fig. 4D). Short-term tidally resolving simulations will aid interpretation of the WMCC survey data with the aim of understanding the fine scale structure of the river plume, Alexandrium abundance, and their inter-relationship. Vertical structure will be included, allowing Alexandrium vertical migration to be included.

II. WMCC. The plume advection hypothesis states that downstream patterns in Alexandrium abundance result from variations in the outflow of cells from the source region and fluctuations in the WMCC. Earlier efforts to explain the observed downstream patterns have been limited by imprecise knowledge of the temporal variability in the source region. Thus, better understanding of the dynamics of the source region resulting from (I) above may lead to an improved ability to represent source region fluctuations in previous models of the western GOM. This would set the stage for a fruitful retrospective analysis of the 1993-94 WMCC data using our existing model (see web page) with either an improved parameterization of the source region or perhaps a nested Casco Bay submodel.

III. WMCC + EMCC. The WMCC and EMCC surveys will provide unprecedented coverage of a coastal region which encompasses two distinct zones of toxicity separated by a toxin free zone. The model used in (I) and (II) above will be configured into a domain encompassing the two survey regions to provide a context for dynamically interpolating the survey data and exploring the linkages between the EMCC and WMCC. The model domain (III; Fig. 5) extends well offshore of EMCC and WMCC, abuts Nova Scotia on the east, and includes Mass Bay to the southwest. Nova Scotia is critical since this is the primary inflow region in our domain. The Massachusetts Bay region is included because it allows us to include the influence of the Merrimack River. Simulations of various types will be conducted, ranging from initial value problems initialized with one survey and run forward to the next, to more sophisticated approaches using data assimilation.

IV. Large Scale GOM. Studies of short term (intra-annual) variations in Alexandrium abundance such as those described above can be conducted in the context of the near-coastal circulation of the WMCC and EMCC. However, lower frequency phenomena must be addressed in terms of the larger scale circulation in which the coastal currents are embedded. One important set of such questions is: How are Alexandrium populations maintained over long time scales? Does the large scale/low frequency circulation provide a mechanism for retention within the GOM and eventual re-seeding of the source region? The required modeling tools exist, such as in the climatological high resolution finite element model of Lynch et al. (1996). Velocity and diffusivity fields extracted from the climatological solutions can drive a separate advection - diffusion - reaction (ADR) module suitable for use with a passive tracer such as Alexandrium.

V. Empirical Toxicity Predictive Models. While the complex three dimensional models described above are useful for scientific study, their utility for management purposes is somewhat limited by logistical considerations. However, insight gained from such modeling will allow us to articulate the mechanisms underlying PSP outbreaks into simpler dynamic models driven by a small number of observable parameters. Essential processes will probably include wind forcing, river discharge, and recent observations of toxicity. One parameter to be given special emphasis is a calculated upwelling index which is useful as an indicator of the dynamics of PSP events in northwest Spain (Fraga et al. 1993) and California (Anderson et al., unpub.). In the WMCC, wind-driven upwelling and downwelling events play a key role in the cross-shelf transport and alongshore advection of Alexandrium blooms and thus directly influence nearshore toxicity. Calculation of an upwelling index would be straightforward using data that are readily available see Appendix 1), and could be used to update a simple predictive model on a regular basis. As we gain scientific knowledge of the causes of PSP outbreaks, we will develop simple predictive tools and evaluate their skill using historical data.