This page will look better in a browser that supports web standards, but its content is accessible to any browser or Internet device.

4.5: Computer Modeling

Robert Twilley is also taking the lead in the Brown Marsh Project's computer modeling efforts. Because investigative teams weren't in place to observe the dieback as it was beginning to unfold, the design and implementation of a computer model will allow researchers to create a variety of "hindcast" scenarios, scenarios not intended to predict the future but reconstruct the past.

Given the enormous range of data types the modeling team has to work with and the variety of scales that have to be depicted in their effort, the modeling work underway is, without a doubt, the most highly integrated single task in the entire Brown Marsh Project.

To create a viable dynamic model, Twilley drew on the resources of several modeling teams that had never worked together. The first of these was LSU's's landscape modeling group, a group that worked at the 10,000-square-kilometer scale with Swenson and Gryme's climate factors such as wind conditions, salinities, water levels, and river flow.

With this model in place, Twilley enlisted the help of UL's Ehab Meselhe, a modeler who works at the one-square-kilometer scale. Meselhe took the LSU team's conditions and ran his smaller "segment" model to produce far more localized pictures of water level and salinity based upon landscape elevation contours derived from Handley's mapping team. However, because these models were largely geared toward flooding simulations, they simply went to zero on the hydrological scale and stopped. In other words, they had no parameters built into them to help recreate the impact of water deficits.

Hence, Twilley used a variant of a preexisting hydrological model he had previously developed that could account for negative hydrological coefficients, one that he had tweaked to simulate mangrove response during drawdown periods. Yet despite this extensive integration of model types and scales, there was still one crucial aspect missing: a vegetative model that included smooth cordgrass. For this final aspect of the effort, Twilley called on Jim Morris, a plant habitat modeling specialist from the University of South Carolina whose work focuses on Spartina grasses. With his help, they overlaid his model on top of the others so that it would "grow" based upon the values plugged in to the fully integrated suite.