Seasonal variations in the spatial‐temporal dependence of total column ozone

Abstract

SUMMARYThe Total Ozone Mapping Spectrometer (TOMS) is a satellite-based instrument that measures total column ozone on a daily basis over a fairly dense spatial grid with near global coverage. Statistical models for the spatialtemporal variations in total column ozone provide insights into ozone dynamics, are valuable for obtaining inferences on long-term trends in ozone levels, and can be used to predict ozone levels at unobserved points in space-time. However, developing such a model is complicated by the seasonally varying nature of the space-time dependence and the partial confounding of spatial and temporal variation caused by the sun-synchronous orbit of the satellite. This work considers methods for describing, modeling and estimating the seasonal patterns in the dependence structure for measurements at a single latitude. Applying one of these models to the synoptic prediction of total column ozone at a latitude shows that there is at least the potential for substantially improved predictions by exploiting the high spatial density of observations TOMS provides.

Analytic and Asymptotic Properties of Multivariate Generalized Linnik’s Probability Densities

“…This stochastic model has two parameters α and ν which can give separate characterizations of the fractal dimension and Hurst effect. It has found applications in many modeling problems [19,46,47,48,58,59,67,70]. In this context, the function under investigation p α,ν,n (x) appears as the spectral density function of the stochastic model [49].…”

Spatial variation of total column ozone on a global scale

“…It is in fact more common to study the gridded Level 3 version of the TOMS data, in which the Level 2 values are interpolated daily onto a grid of 1 • latitude × 1.25 • longitude for latitudes between 50 • S and 50 • N (with fewer grid points per latitude near the poles) [16]. For example, [5,8,12] and [26] consider statistical models for the spatial-temporal variation of various subsets of the Level 3 data. To a rough approximation, Level 3 observations can be treated as if they were taken at local noon [5].…”

Gaussian fields and Gaussian sheets with generalized Cauchy covariance structure

“…The main aim of this paper is to study GFGCC and its anisotropic counterpart, which we call Gaussian sheet with generalized Cauchy covariance (GSGCC). In view of the fact that GFGCC is widely used in geostatistics and other applications [11,8,48,21,50,40,41,49], it will be useful to consider its properties in more detail. In these existing applications, usually only the covariance structure of GFGCC are used, and the sample path properties of GFGCC are rarely mentioned.…”