- Investigating Sea Level Rise and Variability at Tide-Gauge Stations using Supervised Machine Learning
- Project Year:
2019
- REU Student (s):
Amin Fadel | Stockton University NJ
- Student 1 Institution:
Stockton University
- Project Mentor:
Robert Kopp
- Project Mentor Area:
Department of Earth and Planetary Sciences
- Project Abstract:
Sea level change is an important issue that will impact millions of people living on United States coastlines by the end of this century (Sweet et al, 2017). This project seeks to develop a model for sea level variability that can make more appropriate predictions with respect to the different components. Using data collected from tide gauge stations located along the United States coastlines, we created a model that predicts monthly local sea level variability. This was accomplished using Gaussian Process Modeling, a machine learning technique that allows us to capture the different components of sea level change. These components include the trend, seasonal cycle, and interannual variability. Using this method of modeling also allows us to create more accurate predictions, which leads to a more compelling visualization of sea level change than simply using linear regression. Analysis with our model allows for comparisons between different aspects of sea level variability among the different stations.
