STEM with Technical Writing is one of the most unique but work-intensive courses at MAMS. As such, students are forced to learn proper time management, as well as how to run a long-term project. In STEM I, students work on the Independent Research Project, which can be a scientific experiment, engineering design proposal, or mathematical conjecture. It is a great opportunity for students to study topics they themselves are interested in. The scientific and engineering processes are followed. First, potential problems and/or research questions to focus on via brainstorming and research are identified. Then, projects are constructed and carried out, with some students even using WPI laboratory facilities and other cutting-edge resources. Then, their results are analyzed, and conclusions are presented at the STEM Fair in February.
Computational Modeling of Phytoplankton Dynamics with Climatic and Ecological Ramifications
This study develops a series of computational models delineating the causes and effects of changing characteristics in phytoplankton populations.
Key Takeaways:
1. Accuracy of time series models for environmental parameters tested (e.g., oxygen, pH, etc.) varies greatly.
2. There exist significant directional relationships between parameters and phytoplankton primary production and biomass, though they are not effective predictors of these metrics.
3. At a global scale, pH, followed by salinity and pressure, are the most influential parameters for these aspects of phytoplankton populations.
Abstract
Phytoplankton lie at the base of marine food webs and are major regulators of climate and biogeochemical cycling, accounting for over half of primary production and the absorption of 30% of carbon emissions. With global warming modifying ocean conditions, understanding the drivers and impacts of changing phytoplankton dynamics is crucial. However, one-factor experiments have limited applicability due to the heterogeneity in oceanic conditions and biological responses and preferences among different phytoplankton groups. Conversely, multi-factor experiments produce confounding results. Therefore, a computational approach was taken wherein a series of models was developed. All data were derived from the NOAA’s comprehensive World Ocean Database (WOD). Total oceanic chlorophyll concentration was used as an indicator for primary production. To assess accuracy in forecasting capabilities and potential impacts on primary production, for multiple environmental parameters, a time series was developed using sinusoidal regression, as was a linear regression model of the directional relationship held with chlorophyll. Model fitness was variable, as R-Squared values for the first and second set of models ranged from 0.077 to 0.847, and 0.07 to 0.54, respectively. Subsequently, driving parameters behind chlorophyll levels were identified using principal component analysis. Results indicated pH, followed by salinity and pressure, as the most influential parameters. Overall results indicate that the proposed computational apparatus is viable for analyzing phytoplankton dynamics, but that iteration in the form of model modification and greater data implementation is necessary. This apparatus could serve as a significant tool for policymaking related to aquatic ecosystem management.
Graphical Abstract
Research Question
How can the causes and effects of changing phytoplankton dynamics be computationally modeled?
Hypothesis
The causes and effects of changing phytoplankton populations can be observed by developing a series of computational models that incorporates and forecasts data on environmental factors, identifies driving parameters, and assesses individual parametric impact on primary production, in turn providing environmental and climatic implications.
Background Graphic
Background
Methodology Graphic
Methodology
Results: Tables and Figures
Analysis and Discussion
Conclusion
References
Poster