Author Information

Samuel Seth Demel
Juan Du

Abstract

A valid covariance structure is needed to model spatio-temporal data in various disciplines, such as environmental science, climatology and agriculture. In this work we propose a collection of spatio-temporal functions whose discrete temporal margins are some autoregressive and moving average (ARMA) models, obtain a necessary and sufficient condition for them to be covariance functions. An asymmetric version of this model is also provided to account for space-time irreversibility property in practice. Finally, a spatio-temporal model with AR(2) discrete margin is fitted to wind data from Ireland for estimation and prediction, which are compared with some general existing parametric models in terms of likelihood and mean squared prediction error.

Keywords

Autoregressive and moving average process; Fourier transform; Spatio-temporal covariance function; Stationary

Creative Commons License


This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Share

COinS
 
May 1st, 9:30 AM

SPATIO-TEMPORAL COVARIANCE MODELING WITH SOME ARMA TEMPORAL MARGINS

A valid covariance structure is needed to model spatio-temporal data in various disciplines, such as environmental science, climatology and agriculture. In this work we propose a collection of spatio-temporal functions whose discrete temporal margins are some autoregressive and moving average (ARMA) models, obtain a necessary and sufficient condition for them to be covariance functions. An asymmetric version of this model is also provided to account for space-time irreversibility property in practice. Finally, a spatio-temporal model with AR(2) discrete margin is fitted to wind data from Ireland for estimation and prediction, which are compared with some general existing parametric models in terms of likelihood and mean squared prediction error.