Modelling spatial heteroskedasticity by volatility modulated moving averages

TL;DR

This paper develops a framework using volatility modulated moving averages to model spatial heteroskedasticity, providing theory, simulation, inference methods, and an application to climate data.

Contribution

It introduces a novel approach for modeling spatial heteroskedasticity with volatility modulated moving averages, including theoretical foundations and practical inference techniques.

Findings

Effective modeling of spatial heteroskedasticity demonstrated

Simulation studies validate the proposed methods

Application to sea surface temperature anomalies shows practical utility

Abstract

Spatial heteroskedasticity refers to stochastically changing variances and covariances in space. Such features have been observed in, for example, air pollution and vegetation data. We study how volatility modulated moving averages can model this by developing theory, simulation and statistical inference methods. For illustration, we also apply our procedure to sea surface temperature anomaly data from the International Research Institute for Climate and Society.