Chesapeake hypoxia modeling (click image to access project description)

Dr. Michael Kemp's graduate advisor, H.T. Odum, encouraged his students to have a specialty that could be applied to tackling complex regional and global problems. "I thought long and hard about that," says Michael. "One skill I had was how to build models, systems of equations that could be solved to mimic the dynamic of a natural environment."

In small lakes, ecosystem modeling needs only a modest connection to physics. In oceanography, models have detailed descriptions of the movement of water, the mixing of water, and the transport of water, salt and heat. Those descriptions then become the template upon which oceanographers put the ecosystem. "The early oceanographic models were described with a few simple ecological parts – nutrients, phytoplankton," says Michael. "We took a little different tack in coastal systems and modeled more details about the ecosystem – more organisms, more chemical compounds, more details of the biogeochemistry but less on the physics."

"The purpose is being able to glue it together into a dynamic processing tool that helps us learn what's going on."

Today there's an attempt to build models that combine detailed descriptions of both physics and ecosystems. Michael still constructs the simpler models – "toy" models, as his physical oceanographer friends refer to them. "We built the details of the chemistry, biology and the ecology in with generalized physics," he says. "Now we have the unit models which can be placed into the physical model describing transport and mixing between over space – describing cubes of water. And inside each one of those cubes we can put a little ecosystem model."

In addition to physics, coastal scientists have other challenges that deep sea oceanographers don't always have to deal with – such as the bottom. "In the shallow systems that we operate in, the bottom can dominate," says Michael. "We have seagrasses growing, and they change the physics. They attenuate waves and tidal currents. We also and reefs where the biology dominates the physics."

Michael collaborates with colleagues whom he calls the "real" modelers. "I came out of a graduate school where we were learning a lot of details about a lot of things," he says. "Our goal is to pull that knowledge together, working with others to build a dynamic processing tool that helps us learn what's going on."