A Bayesian Spatio-Functional Clustering Model Based on Wavelet Smoothing, with Application to Climate Change Study

• Activity Number: 681
• Type: Topic Contributed
• Date/Time: Thursday, August 8, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Section on Nonparametric Statistics
• Author(s): Zhen Zhang and Chae Young Lim*+ and Tapabrata Maiti
• Companies: Michigan State University and Michigan State and Michigan State University
• Keywords: Spatial clustering ; Functional clustering ; Bayesian wavelet smoothing ; Shrinkage priors ; Dimension reduction ; Climate change study

Abstract

In climate change study, the infrared spectral signatures of climate change have recently been conceptually adopted and widely applied to identifying and attributing atmospheric composition change. We propose a Bayesian hierarchical model for spatial and functional clustering of the climate model data as surrogates for measured spectra to assess climate change. Our model allows spatio-functional dependence and functional covariates with cluster-specific fixed effect functions that are regularized using wavelet basis. Non-informative priors are extensively elicited for both the clustering priors and the covariance structure of random effect functions, and multiple shrinkage priors for fixed effects are adopted and investigated via simulation studies. Dimension reduction is achieved by assuming conditional independence between clusters for random effect functions. The model is applied to the spectral signatures of climate change that were observed globally, and produces spatial clustering map that is compared with traditional clustering techniques and variations of the proposed model. The model fitting utilizes high-performance parallel computing and sparse matrix algorithms.