Abstract
This work considers a generalization of the INAR(1) model to the panel data first order Seemingly Unrelated INteger AutoRegressive Poisson model, SUINAR(1). It presents Bayesian and classical methodologies to estimate the parameters of Poisson SUINAR(1) model and to forecast future observations of the process. In particular, prediction intervals for forecasts - classical approach - and HPD prediction intervals - Bayesian approach - are derived. A simulation study is provided to give additional insight into the finite sample behaviour of the parameter estimates and forecasts.

Abstract
In this work we consider the problem of forecasting integer-valued time series, modelled by the INAR(1) process introduced by McKenzie (1985) and Al-Osh and Alzaid (1987). The theoretical properties and practical applications of INAR and related processes have been discussed extensively in the literature but there is still some discussion on the problem of producing coherent, i.e. integer-valued, predictions. Here Bayesian methodology is used to obtain point predictions as well as confidence intervals for future values of the process. The predictions thus obtained are compared with their classic counterparts. The proposed approaches are illustrated with a simulation study and a real example.

Abstract
The basic statistical features of radon time series from continuous radon monitoring at the Elat granite, Israel are analysed. A similar analysis is carried out for ancillary and possibly related geophysical parameters for the Elat area. The results show that air temperature, precipitable water and longwave radiation time series exhibit constant variance over the analyzed period, while radon time series, atmospheric pressure, short-wave radiation and total electron content exhibit heteroscedasticity. Furthermore, for radon and shortwave radiation the variability is associated with the overall mean level, while for atmospheric pressure such an association is not present. The analyzed radon time series not only are non-stationary but also nonlinear, reflecting the complex dynamics of radon emanation and transport in natural subsurface systems.

Abstract
The detection of a trend in a time series and the evaluation of its magnitude and statistical significance is an important task in geophysical research. This importance is amplified in climate change contexts, since trends are often used to characterize long-term climate variability and to quantify the magnitude and the statistical significance of changes in climate time series, both at global and local scales. Recent studies have demonstrated that the stochastic behavior of a time series can change the statistical significance of a trend, especially if the time series exhibits long-range dependence. The present study examines the trends in time series of daily average temperature recorded in 26 stations in the Tuscany region (Italy). In this study a new framework for trend detection is proposed. First two parametric statistical tests, the Phillips-Perron test and the Kwiatkowski-Phillips-Schmidt-Shin test, are applied in order to test for trend stationary and difference stationary behavior in the temperature time series. Then long-range dependence is assessed using different approaches, including wavelet analysis, heuristic methods and by fitting fractionally integrated autoregressive moving average models. The trend detection results are further compared with the results obtained using nonparametric trend detection methods: Mann-Kendall, Cox-Stuart and Spearman's rho tests. This study confirms an increase in uncertainty when pronounced stochastic behaviors are present in the data. Nevertheless, for approximately one third of the analyzed records, the stochastic behavior itself cannot explain the long-term features of the time series, and a deterministic positive trend is the most likely explanation.

Abstract
The characterisation and quantification of long-term sea-level variability is of considerable interest in a climate change context. Long time series from coastal tide gauges are particularly appropriate for this purpose. Long-term variability in tide gauge records is usually expressed through the linear slope resulting from the fit of a linear model to the time series, thus assuming that the generating process is deterministic with a short memory component. However, this assumption needs to be tested, since trend features can also be due to non-deterministic processes such as random walk or long range dependent processes, or even be driven by a combination of deterministic and stochastic processes. Specific methodology is therefore required to distinguish between a deterministic trend and stochastically-driven trend-like features in a time series. In this chapter, long-term sea-level variability is characterised through the application of (i) parametric statistical tests for stationarity, (ii) wavelet analysis for assessing scaling features, and (iii) generalised least squares for estimating deterministic trends. The results presented here for long tide gauge records in the North Atlantic show, despite some local coherency, profound differences in terms of the low frequency structure of these sea-level time series. These differences suggest that the long-term variations are reflecting mainly local/regional phenomena. © 2008 Springer-Verlag Berlin Heidelberg.

Abstract
This work addresses the autoregressive modelling of sea level time series from TOPEX/Poseidon satellite altimetry mission. Datasets from remote sensing applications are typically very large and correlated both in time and space. Multivariate analysis methods are useful tools to summarise and extract information from such large space-time datasets. Multivariate autoregressive analysis is a generalisation of Principal Oscillation Pattern (POP) analysis, widely used in the geosciences for the extraction of dynamical modes by eigen-decomposition of a first order autoregressive model fitted to the multivariate dataset of observations. The extension of the POP methodology to autoregressions of higher order, although increasing the difficulties in estimation, allows one to model a larger class of complex systems. Here, sea level variability in the North Atlantic is modelled by a third order multivariate autoreerressive model estimated by stepwise least squares. Eigen-decomposition of the fitted model yields physically-interpretable seasonal modes. The leading autoregressive mode is an annual oscillation and exhibits a very homogeneous spatial structure in terms of amplitude reflecting the large scale coherent behaviour of the annual pattern in the Northern hemisphere. The phase structure reflects the seesaw pattern between the western and eastern regions in the tropical North Atlantic associated with the trade winds regime. The second mode is close to a semi-annual oscillation. Multivariate autoregressive models provide a useful framework for the description of time-varying fields while enclosing a predictive potential.