Narragansett Bay features a complex marine system of organisms, nutrients, chemicals and currents, all interdependent and potentially affected by even the slightest change of environment. For decades, ocean scientists and engineers have been making observations detailing the inner workings of these systems, but like an incomplete DNA sequence, gaps of knowledge remain.
The challenge for ocean modelers like Brown University’s Dr. Baylor Fox-Kemper, associate professor of earth, environmental and planetary sciences at Brown University, centers on developing sophisticated computer programs that provide scientific insights of coastal ecosystems changing due to climate variability.
The Consortium’s modeling team is building the Ocean State Ocean Model (OSOM), an ’end-to-end‘ simulation of the Narragansett Bay ecosystem that will help researchers better define the ecological and physical dynamics at play in this complex marine system.
“We will simulate everything from the circulation of the Bay up to societal issues,” details Fox-Kemper. “Data on salinity, temperature, phytoplankton, zooplankton, fisheries, and how humans interact with the bay will all be incorporated. It won’t be a single piece of software that does all that, but a series of modules that illustrate the bay environments.”
OSOM’s development is in its nascent stages as Consortium researchers from Brown and the University of Rhode Island figure out how to mesh data from many disciplines, the variables and time scales of which are not all uniform.
“There’s different ‘currency’, we might say,” Fox-Kemper explains. “In physics, for example, our currency is momentum and energy as we think about things like mass and kinetic movement throughout the water column. For marine biologists, however, the currency is usually nutrient levels and plankton count. For fisheries researchers, they are dealing with calories among species. We are learning how to put these pieces together.
“That’s why a large, collaborative network is important as no one group can assemble all the different pieces together. We need all the complementing disciplines so that when we make a model, we know their currencies and how to convert them.”
The Consortium's Dr. David Ullman, associate research scientist at URI’s Graduate School of Oceanography, is developing models which characterize the many interactions in Narragansett Bay’s ecology, from chemicals such as nitrates and phosphates to classes and sizes of phytoplankton.
Image courtesy David Ullman
The ability to hindcast, or simulate models of Narragansett Bay’s past environments, highlights an important feature of the OSOM that will allow researchers to test the accuracy of observational data from previous years.
“Scientists usually write papers about observations they have made in the past,” says Fox-Kemper. “If last November, for example, was a productive month for phytoplankton growth in Narragansett Bay, then you may want to have a cotemporaneous model about things like water circulation to inform your paper. That is a hindcast setting we are hoping to offer.”
Hindcasting furthermore will provide OSOM users the chance to conduct ‘reanalysis’ studies through which observational data are examined alongside a computer model, thus offering a clearer picture of the past dynamics in a given ecological system.
Hindcast v. To test a mathematical model with data from a past event or environment
“In atmospheric reanalysis, we are creating more sophisticated computer models for storms even when observations stay stationary, which allows us to learn more about past events although we are not adding new data,” Fox-Kemper notes. “The models help us to reduce guesswork as we trust in the physics of the ocean and fill out our understanding even with limited information.”
As scientists improve the world’s knowledge and approach to climate variability, developing forecast models of Narragansett Bay environments poses a crucial but challenging prospect for the Consortium’s modeling team.
“When you predict the weather, people have a sense of how accurate that is,” says Fox-Kemper. “Similarly, when you predict something about the bay, people should be able to assess whether it was accurate.”
Fox-Kemper notes that Rhode Island is at an advantage because of the many environmental datasets collected on Narragansett Bay. Through OSOM’s hindcasting capabilities, he says, forecast models of future environmental change can be tested. The timescale for such forecasts will start small, however.
Not all ecological change is human-induced, but if we can predict correlation versus causation, then maybe we can have a better sense of what’s coming before it gets here. -Dr. Baylor Fox-Kemper, Brown University
“Projecting decades into the future based on a system built to project weeks into the future seems crazy, but let’s run next week before we get there and see if we’re doing a good job,” says the Brown researcher. “If we can develop an accurate model of last week’s marine dynamics, we can gain greater confidence about predicting into the future, if only in small time increments at the start.”
Once the OSOM is fully operational, the Consortium’s modeling team hopes that scientists can rely on its open-source modeling capabilities to make well-informed inferences about Narragansett Bay’s changing ecosystems, research from which will augment future decision-making from Rhode Island communities and their leaders.
Fox-Kemper says: “This is a really sophisticated modeling system we’re building, so the question becomes: can we create a model in which it’s possible to switch human decisions on and off, and see the potential consequences? Not all ecological change is human-induced, but if we can predict correlation versus causation, then maybe we can have a better sense of what’s coming before it gets here.”