Sasquatch Statistics: Spatial Point Regression with R
Regular talk, 2:00 - 3:00 PM
Can we use spatial statistics and R to explain why Bigfoot sightings cluster where they do? Are these clusters driven by landscape, people, or pure legend? Bigfoot sightings have been reported across the PNW for decades, but are they randomly scattered, or spatially predictable? This presentation will use a dataset of Bigfoot sightings to demonstrate a high-level spatial point regression workflow in R.
Starting with raw latitude–longitude coordinates, we’ll walk through how to build a reproducible pipeline to model sightings as an inhomogeneous Poisson point process. Environmental covariates such as forest cover, elevation, and population density will be used to estimate a spatially varying intensity surface. Rather than diving deep into theory, the focus is on practical implementation: fitting models, generating predictions, and visualizing results.
If time allows, we’ll end with a demonstration of a Shiny app that makes the modelling process interactive, allowing users to toggle covariates and explore how the predicted “Bigfoot intensity” surface changes in real time.
Although the subject of Bigfoot is cryptozoological, this workflow is broadly applicable to ecology, epidemiology, criminology, and other spatial point pattern problems, demonstrating how R enables end-to-end spatial modelling in a single ecosystem.
 | Pronouns: she/herVancouver, BC, Canada |