540 – Sampling Using Indirect and Non-Standard Frame Information
Bayesian Multivariate Estimate of Global Temperature Trends
Dongchu Sun
University of Missouri
Shawn Ni
University of Missouri - Columbia
Paul L. Speckman
University of Missouri - Columbia
In this paper, we consider a multivariate smoothing problem with correlated errors and correlated derivatives of the curves. Full Bayesian inference is introduced for the smoothing spline, the unknown covariance matrix of the errors and a symmetric smoothing parameter matrix. A prior is proposed on the symmetric smoothing parameter matrix, and Markov Chain Monte Carlo (MCMC) algorithms are developed for Bayesian computation. The proposed method is then applied to estimate the trends of abnormal surface temperature in ten geographical zones.