- Thursday, October 29, 2015
- 3:40 PM–4:30 PM
- North Hall 276
Underwater noise concerns environmentalists and policymakers because it may adversely affect whales, causing them to change their normal behavior or even injuring their auditory systems. Noise sources at sea are many, including shipping, military activities, and geophysical exploration. Animal-borne tags have been developed to track whale movements and behavior underwater, but finding coherent ways to analyze the complex tag data remains a challenge. Often, scientists simply want to know what a whale was doing in general, and whether that changed when the whale was exposed to certain sounds. State-switching models are one novel way to organize all these data and obtain an estimate of the “behavior state” of the whale at any given time. Using as an example a semi-Markov model for sperm whale behavior and responses to noise, I will introduce state-switching models (including the key idea of the state transition probability matrix), talk about how to fit these models to data, and detail a few interesting features of the sperm whale model.
Photo by Ari Friedlaender, taken under NMFS permit 14534
Refreshments precede the talk at 3:30pm in NH-282.