Speaker: Eryn Blagg, PhD Candidate, Department of Statistics, Iowa State University

Title: Conditional Simulation for Spatial Hierarchical Generalized Linear Mixed Models Containing a Covariate Effort

Abstract: In ecological data, the effort of sampling is an important covariate in understanding the number of mammals seen in a given location. The number of mammals seen is conditional on the sampling effort within the domain. One way to predict the number of mammals in a given area is using conditional simulation to simulate a spatial field. This idea can be applied to simulation of an effort field as well. In this talk, we model conditional simulation of an effort surface along with a data surface using a hierarchical generalized linear mixed model. Not only can we predict the count value at unobserved locations, but we can also simulate an effort surface along the entire domain. We apply this model to predict the total number of Steller sea lions and Northern fur seals conditional on effort, in the Bering Sea.