Sean Haight, Andrew Staal, Aycin Iplikci-Arodirik, Sayantan Roy
Salmon populations on the Pacific Coast of North America have been in decline for the last 40 years. Yet salmon have important economic, cultural, and environmental impacts. Indigenous peoples have fished salmon since time immemorial, while commercial and recreational salmon fisheries are worth ~$641 million annually, being a substantial employer in British Columbia. Negative environmental impacts of declining salmon stocks include the disruption of energy and nutrient transfers between the Pacific Ocean and freshwater rivers, and the disruption of the food chain for salmon predators.
Our goal is to study salmon viability, i.e. salmon survival rates, using data science. Here we identify environmental, predatory, and human factors that have the largest impact on salmon populations in the Strait of Georgia. We reduce to a minimal set of features that have the most effect on salmon viability, and build a linear regression model using estimated features.