Time Series Analysis and Applications to Geophysical Systems -

. .

Résumé :
Time series methods are essential tools in the analysis of many geophysical systems. This volume, which consists of papers presented by a select, international group of statistical and geophysical experts at a Workshop on Time Series Analysis and Applications to Geophysical Systems at the Institute for Mathematics and its Applications (IMA) at the University of Minnesota from November 12-15, 2001 as part of the IMA's Thematic Year on Mathematics in the Geosciences, explores the application of recent advances in time series methodology to a host of important problems ranging from climate change to seismology. The works in the volume deal with theoretical and methodological issues as well as real geophysical applications, and are written with both statistical and geophysical audiences in mind. Important contributions to time series modeling, estimation, prediction, and deconvolution are presented. The results are applied to a wide range of geophysical applications including the investigation and prediction of climatic variations, the interpretation of seismic signals, the estimation of flooding risk, the description of permeability in Chinese oil fields, and the modeling of NOx decomposition from thermal power plants.

Sommaire:
Interpretation of Seismic Signals.-Nonparametric deconvolution of seismic depth phases.- State space approach to signal extraction problems in seismology.- Improved signal transmission through randomization.- Online analysis of seismic signals.- Temperature Data.- Nonstationary time series analysis of monthly global temperature anomalies.- A test for detecting changes in mean.- Spatio-temporal modelling of temperature time series.- Modeling North Pacific climate time series.- Assortment of Important Time Series Problems and Applications.- Skew-elliptical time series with application to flood risk.- Hidden periodicities anaysis and its application in geophysics.- The innovation approach to the identification of nonlinear causal models in time series analysis.- Non-Gaussian time series driven by fractional Gaussian noise.- List of workshop participants.