Maximum likelihood estimation of the dynamic coregionalization model with heterotopic data

46Citations
Citations of this article
12Readers
Mendeley users who have this article in their library.
Get full text

Abstract

The information content of multivariable spatio-temporal data depends on the underlying spatial sampling scheme. The most informative case is represented by the isotopic configuration where all variables are measured at all sites. The opposite case is the completely heterotopic case where different variables are observed only at different locations. A well known approach to multivariate spatio-temporal modelling is based on the linear coregionalization model (LCM). In this paper, the maximum likelihood estimation of the heterotopic spatio-temporal model with spatial LCM components and temporal dynamics is developed. In particular, the computation of the estimates is based on the EM algorithm and two solutions are proposed: one is based on the more cumbersome exact maximization of the a posteriori expected log likelihood and the other is an approximate closed-form solution. Their properties are assessed in terms of bias and efficiency through an example of air quality dinamic mapping using satellite data and a Monte Carlo simulation campaign based on a large data set. © 2011 John Wiley & Sons, Ltd.

Cite

CITATION STYLE

APA

Fassò, A., & Finazzi, F. (2011). Maximum likelihood estimation of the dynamic coregionalization model with heterotopic data. Environmetrics, 22(6), 735–748. https://doi.org/10.1002/env.1123

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free