Online Program Home
My Program

Abstract Details

Activity Number: 488 - Nonstationary and Anisotropic Spatial Processes
Type: Contributed
Date/Time: Wednesday, August 1, 2018 : 8:30 AM to 10:20 AM
Sponsor: Section on Statistics and the Environment
Abstract #328376 Presentation
Title: Approximate Bayesian Inference for Big Spatial Data Using Non-Stationary Spectral Simulation
Author(s): Hou-Cheng Yang*
Keywords: Bayesian Statistics; Spatial Statistics

It is increasingly understood that the assumption of stationarity is unrealistic for many spatial processes. For example, in oceanography, environmental, and atmospheric sciences, non-stationarity may occur because of interactions with covariates (e.g., rainfall at one location may occur more around bodies of water even though nearby observations are over land). In this article, we combine dimension expansion with a spectral method to model big non-stationary spatial fields. Specifically, we use Mej ??a and Rodr ??guez-Iturbe (1974)'s spectral approach to approximately simulate from a covariogram at locations that have an expanded dimension. We bring the Bayesian hierarchical framework to dimension expansion, which originally has only been modeled using a method of moments approach. Our method is both full rank and non-stationary, and can be applied to big spatial data because it does not involve storing and inverting large covariance matrices. Additionally, we have fewer parameters than many other non-stationary spatial models. We demonstrate the wide applicability of our approach using a simulation study, and an application using ozone data obtained from the NASA.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program