Statistical Modeling of Disease in Ecological Communities Using Partial Differential Equations

• Activity Number: 62 - Modeling and Inference Using Stochastic Differential Equations
• Type: Topic Contributed
• Date/Time: Sunday, July 29, 2018 : 4:00 PM to 5:50 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Trevor Hefley* and Haoyu Zhang and Robin Russell and Anne Ballmann
• Companies: Kansas State University and Kansas State University and United States Geological Survey and United States Geological Survey
• Keywords: oint species distribution model; inhomogeneous point process; spatio-temporal statistics

Abstract

Dynamic spatio-temporal statistical models based on partial differential equations (PDE) are used in fields such as ecology and epidemiology. Although the approach is well-established, PDE based statistical models can be challenging to apply to multivariate data. For example, the presence or absence of a disease may be reported for many species of wildlife, but the susceptibility to a common pathogen may depend on the characteristics of a specific species. We present a joint modeling approach using a latent PDE based dynamical model. This approach borrows strength across processes that are common to all species such as the location and time a pathogen was introduced. Our work is motivated by the need to understand the ecology and forecast the spread of white-nose syndrome. White-nose syndrome is a devastating disease of cave-hibernating bats that was first detected in North America in 2006. Since the first detection, the disease has spread throughout the eastern and midwestern United States endangering many species of bats.