Biological Forecast Model

Our goal is to predict biological dynamics in Lunenburg Bay — that is, the distributions and growth of phytoplankton — using mathematical models guided directly by measurements from our coastal observatory. The models describing ecological dynamics are embedded within the coastal circulation model of the Bay. Oceanographic and optical measurements from the observing system are used to keep the models close to reality while constantly assessing their accuracy.

Prediction of biological dynamics for Lunenburg Bay relies on ecosystem models that mathematically describe the interactions of the ecosystem components: plankton, nutrients and non-living detritus organic matter. To allow for the movement and mixing of these freely floating ecosystem components, the models are coupled to the physical circulation model for Lunenburg Bay. Biological models are forced by observed light and temperature. They must also be linked to the far-field conditions outside the bay – this is done through coupling with other models, as well as through the use of additional observation sources (e.g. from satellites). The accuracy of model forecasts is assessed through comparison with the biological observations obtained by optical sensors deployed in the observing system, and from water sampling data. The aim is to identify the best ecosystem model in terms of its predictive ability, while taking into account its ultimate application in forecast systems that will rely on automated measurements. In addition, new statistical data assimilation approaches are being developed to more systematically synthesize the model results and observations with the goal of improved prediction.

BioForecastModel

Here is an animation of the simulated concentration of phytoplankton in Lunenburg Bay over 30 days. Warm colours correspond to a modest bloom, and light blue is about average for the bay during the summer. This simulation has been done with a coupled biological-physical model of Lunenburg Bay. The physical model, CANDIE, is the barotropic version (T=13°C and S=30 psu) and it has been forced by M2 tide only. This experimental run simulates sewage effluent in Lunenburg Harbour, inducing a higher concentration of phytoplankton inshore. Such an inshore source of nutrients allows us to retrieve a significant part of observed patterns in observed time-series of phytoplankton over the Bay. More coupled biological-physical experiments are being tested to see how well they explain observed patterns of phytoplankton and nutrients in the bay. This should improve our understanding of this marine ecosystem so as to reach the ultimate goal of forecasting its biological dynamics.