Ella Colby ’26, Computer Science

Ella Colby ’26

Major

Computer Science

Project Title

The Interface of Hydrology and Machine Learning: Generating Better Information for Decision Makers and Educating the Decision Makers of the Future

Presentation Link

View Ella's Presentation

Minor(s): Dance, Finance, Statistics and Machine Learning

Physics-based modeling is an accurate, yet computationally expensive method to predict groundwater conditions. Machine Learning (ML) models have increased in popularity as a hydrological modeling method that can potentially complement or provide an alternative to expensive physics-based models. I worked on measuring the constraints of ML models’ ability to predict groundwater conditions, in particular, water transit time. To do this, I developed programs that would train different ML models based on data from one region and then test the models’ efficacy on a different region. I found that while ML models do not accurately predict water transit time when trained and tested in this manner, combining datasets improved the models’ accuracy. Interestingly, my research also highlighted how permutations and curvatures disproportionately affected ML models’ predictions. This experience allowed me to learn a lot about how research is conducted and how to implement various ML models in Python. I am excited to continue exploring how ML can change the landscape of hydrology and continue working with my incredible mentors in the upcoming semester.