Modelling the tails of distributions is important in many fields, such as environmental science, hydrology, insurance, engineering and finance, where the risk of unusually large or small events are of interest. This thesis applies extreme value models in neonatal and finance studies and develops novel extreme value modelling for financial applications, to overcome issues associated with the dependence induced by volatility clustering and threshold choice. The instability of preterm infants stimulates the interests in estimating the underlying variability of the physiology measurements typically taken on neonatal intensive care patients. The stochastic volatility model (SVM), fitted using Bayesian inference and a particle filter to capture the on-line latent volatility of oxygen concentration, is used in estimating the variability of medical measurements of preterm infants to highlight instabilities resulting from their under-developed biological systems. Alternative volatility estimators are considered to evaluate the performance of the SVM estimates, the results of which suggest that the stochastic volatility model provides a good estimator of the variability of the oxygen concentration data and therefore may be used to estimate the instantaneous latent volatility for the physiological measurements of preterm infants. The classical extreme value distribution, generalized pareto distribution (GPD), with the peaks-over-threshold (POT) method to ameliorate the impact of dependence in the extremes to infer the extreme quantile of the SVM based variability estimates. Financial returns typically show clusters of observations in the tails, often termed “volatility clustering” which creates challenges when applying extreme value models, since classical extreme value theory assume independence of underlying process. Explicit modelling on GARCH-type dependence behaviour of extremes is developed by implementing GARCH conditional variance structure via the extreme value model parameters. With the combination of GEV and GARCH models, both simulation and empirical results show that the combined model is better suited to explain the extreme quantiles. Another important benefit of the proposed model is that, as a one stage model, it is advantageous in making inferences and accounting for all uncertainties much easier than the traditional two stage approach for capturing this dependence. To tackle the challenge threshold choice in extreme value modelling and the generally asymmetric distribution of financial data, a two tail GPD mixture model is proposed with Bayesian inference to capture both upper and lower tail behaviours simultaneously. The proposed two tail GPD mixture modelling approach can estimate both thresholds, along with other model parameters, and can therefore account for the uncertainty associated with the threshold choice in latter inferences. The two tail GPD mixture model provides a very flexible model for capturing all forms of tail behaviour, potentially allowing for asymmetry in the distribution of two tails, and is demonstrated to be more applicable in financial applications than the one tail GPD mixture models previously proposed in the literature. A new Value-at-Risk (VaR) estimation method is then constructed by adopting the proposed mixture model and two-stage method: where volatility estimation using a latent volatility model (or realized volatility) followed by the two tail GPD mixture model applied to independent innovations to overcome the key issues of dependence, and to account for the uncertainty associated with threshold choice. The proposed method is applied in forecasting VaR for empirical return data during the current financial crisis period.

Extreme value theory (EVT) has wide applicability in several areas like hydrology, engineering, science and finance. Across the world, we can see the disruptive effects of flooding, due to heavy rains or storms. Many countries in the world are suffering from natural disasters like heavy rains, storms, floods, and also higher temperatures leading to desertification. One of the best known extraordinary natural disasters is the 1931 Huang He flood, which led to around 4 millions deaths in China; these were a series of floods between Jul and Nov in 1931 in the Huang He river.Several publications are focused on how to find the best model for these events, and to predict the behaviour of these events. Normal, log-normal, Gumbel, Weibull, Pearson type, 4-parameter Kappa, Wakeby and GEV distributions are presented as statistical models for extreme events. However, GEV and GP distributions seem to be the most widely used models for extreme events. In spite of that, these models have been misused as models for extreme values in many areas.The aim of this dissertation is to create new modifications of univariate extreme value models.The modifications developed in this dissertation are divided into two parts: in the first part, we make generalisations of GEV and GP, referred to as the Kumaraswamy GEV and Kumaraswamy GP distributions. The major benefit of these models is their ability to fit the skewed data better than other models. The other idea in this study comes from Chen, which is presented in Proceedings of the International Conference on Computational Intelligence and Software Engineering, pp. 1-4. However, the cumulative and probability density functions for this distribution do not appear to be valid functions. The correction of this model is presented in chapter 6.The major problem in extreme event models is the ability of the model to fit tails of data. In chapter 7, the idea of the Chen model with the correction is combined with the GEV distribution to introduce a new model for extreme values referred to as new extreme value (NEV) distribution. It seems to be more flexible than the GEV distribution.

L'étude et la gestion des risques littoraux sont plébiscitées par notre société au vu des enjeux économiques et écologiques qui y sont impliqués. Ces risques sont généralement réponse à des conditions environnementales extrêmes. L'étude de ces phénomènes physiques repose sur la compréhension de ces conditions rarement (voire nullement) observées.Dans un milieu littoral, la principale source d'énergie physique est véhiculée par les vagues. Cette énergie est responsable des risques littoraux comme l'érosion et la submersion qui évoluent à des échelles de temps différentes (événementielle ou long-terme). Le travail réalisé, situé à l'interface de l'analyse statistique, de la géophysique et de l'informatique, vise à apporter des méthodologies et outils aux décideurs en charge de la gestion de tels risques.En pratique, nous nous intéressons à mettre en place des méthodes qui prennent en compte non seulement un site ponctuel mais traitent les problématiques de façon spatiale. Ce besoin provient de la nature même des phénomènes environnementaux qui sont spatiaux, tels les champs de vagues.L'étude des réalisations extrêmes de ces processus repose sur la disponibilité d'un jeu de données représentatif à la fois dans l'espace et dans le temps, permettant de projeter l'information au-delà de ce qui a déjà été observé. Dans le cas particulier des champs de vagues, nous avons recours à la simulation numérique sur calculateur haute performance (HPC) pour réaliser un tel jeu de données. Le résultat de ce premier travail offre de nombreuses possibilités d'applications.En particulier, nous proposons à partir de ce jeu de données deux méthodologies statistiques qui ont pour but respectif de répondre aux problématiques de risques littoraux long-termes (érosion) et à celles relatives aux risques événementiels (submersion). La première s'appuie sur l'application de modèles stochastiques dit max-stables, particulièrement adapté à l'étude des événements extrêmes. En plus de l'information marginale, ces modèles permettent de prendre en compte la structure de dépendance spatiale des valeurs extrêmes. Nos résultats montrent l'intérêt de cette méthode au devant de la négligence de la dépendance spatiale de ces phénomènes pour le calcul d'indices de risque.La seconde approche est une méthode semi-paramétrique dont le but est de simuler des champs spatio-temporels d'états-de-mer extrêmes. Ces champs, interprétés comme des tempêtes, sont des amplifications contrôlées et bi-variés d'épisodes extrêmes déjà observés. Ils forment donc des tempêtes encore plus extrêmes. Les tempêtes simulées à une intensité contrôlée alimentent des modèles physiques événementiels à la côte, permettant d'aider les décideurs à l'anticipation de ces risques encore non observés.Enfin et depuis la construction de ces scenarii extrêmes, nous abordons la notion de pré-calcul dans le but d'apporter en quasi-temps réel au décideur et en tant de crise une prévision sur le risque littoral.L’ensemble de ce travail s'inscrit dans le cadre d'un besoin industriel d’aide à la modélisation physique : chainage de modèles numériques et statistiques. La dimension industrielle de cette thèse est largement consacrée à la conception et au développement d’un prototype de plateforme de modélisation permettant l’utilisation systématique d’un calculateur HPC pour les simulations et le chainage de modèles de façon générique.Autour de problématiques liées à la gestion du risque littoral, cette thèse démontre l'apport d'un travail de recherche à l'interface de plusieurs disciplines. Elle y répond en conciliant et proposant des méthodes de pointe prenant racine dans chacune de ces disciplines. / Studies and management of coastal hazards are of high concerns in our society, since they engage highly valuable economical and ecological stakes. Coastal hazards are generally responding to extreme environmental conditions. The study of these physical phenomena relies on the understanding of such environmental conditions, which are rarely (or even never) observed.In coastal areas, waves are the main source of energy. This energy is responsible of coastal hazards developed at different time-scales, like the submersion or the erosion.The presented work, taking place at the interface between Statistical Analysis, Geophysics and Computer Sciences, aiming at bringing forward tools and methods serving decision makers in charge of the management of such risks.In practice, the proposed solutions answer to the questionings with a consideration of the space dimension rather than only punctual aspects. This approach is more natural considering that environmental phenomena are generally spatial, as the sea-waves fields.The study of extreme realisations of such processes is based on the availability of a representative data set, both in time and space dimensions, allowing to extrapolating information beyond the actual observations. In particular for sea-waves fields, we use numerical simulation on high performance computational clusters (HPC) to product such a data set. The outcome of this work offers many application possibilities.Most notably, we propose from this data set two statistical methodologies, having respective goals of dealing with littoral hazards long-terms questionings (e.g., erosion) and event-scale questionings (e.g., submersion).The first one is based on the application of stochastic models so-called max-stable models, particularly adapted to the study of extreme values in a spatial context. Indeed, additionally to the marginal information, max-stable models allow to take into account the spatial dependence structures of the observed extreme processes. Our results show the interest of this method against the ones neglecting the spatial dependence of these phenomena for risk indices computation.The second approach is a semi-parametric method aiming at simulating extreme waves space-time processes. Those processes, interpreted as storms, are controlled and bi-variate uplifting of already observed extreme episodes. In other words, we create most severe storms than the one already observed. These processes simulated at a controlled intensity may feed littoral physical models in order to describe a very extreme event in both space and time dimensions. They allow helping decision-makers in the anticipation of hazards not yet observed.Finally and from the construction of these extreme scenarios, we introduce a pre-computing paradigm in the goal of providing the decision-makers with a real-time and accurate information in case of a sudden coastal crisis, without performing any physical simulation.This work fits into a growing industrial demand of modelling help. Most notably a need related to the chaining of numerical and statistical models. Consequently, the industrial dimension of this PhD.~is mostly dedicated to the design and development of a prototype modelling platform. This platform aims at systematically using HPC resources to run simulations and easing the chaining of models.Embracing solutions towards questionings related to the management of coastal hazard, this thesis demonstrates the benefits of a research work placed at the interface between several domains. This thesis answers such questionings by providing end-users with cutting-edge methods stemming from each of those domains.

Extreme Value Theory nds application in problems concerning low probability but high consequence events. In hydrology the study of heavy rainfall is important in regional ood risk assessment. In particular, the N-year return level is a key output of an extreme value analysis, hence care needs to be taken to ensure that the model is accurate and that the level of imprecision in the parameter estimates is made explicit. Rainfall is a process that evolves over time and space. Therefore, it is anticipated that at extreme levels the process would continue to show temporal and spatial correlation. In this study interest is in whether any trends in heavy rainfall can be detected for the Western Cape. The focus is on obtaining the 50-year daily winter rainfall return level and investigating whether this quantity is homogenous over the study area. The study is carried out in two stages. In the rst stage, the point process approach to extreme value theory is applied to arrive at the return level estimates at each of the fteen sites. Stationarity is assumed for the series at each station, thus an issue to deal with is that of short-range temporal correlation of threshold exceedances. The proportion of exceedances is found to be smaller (approximately 0.01) for stations towards the east such as Jonkersberg, Plettenbergbay and Tygerhoek. This can be attributed to rainfall values being mostly low, with few instances where large amounts of rainfall were observed. Looking at the parameters of the point process extreme value model, the location parameter estimate appears stable over the region in contrast to the scale parameter estimate which shows an increase towards in a south easterly direction. While the model is shown to t exceedances at each station adequately, the degree of uncertainty is large for stations such as Tygerhoek, where the maximum observed rainfall value is approximately twice as large as the high rainfall values. This situation was also observed at other stations and in such cases removal of these high rainfall values was avoided to minimize the risk of obtaining inaccurate return level estimates. The key result is an N-year rainfall return level estimate at each site. Interest is in mapping an estimate of the 50-year daily winter rainfall return level, however to evaluate the adequacy of the model at each site the 25-year return level is considered since a 25 year return period is well within the range of the observed data. The 25-year daily winter rainfall return level estimate for Ladismith is the smallest at 22:42 mm. This can be attributed to the station's generally low observed winter rainfall values. In contrast, the return level estimate for Tygerhoek is high, almost six times larger than that of Ladismith at 119:16 mm. Visually design values show di erences between sites, therefore it is of interest to investigate whether these di erences can be modelled. The second stage is the geostatistical analysis of the 50-year 24-hour rainfall return level The aim here is to quantify the degree of spatial variation in the 50-year 24-hour rainfall return level estimates and to use that association to predict values at unobserved sites within the study region. A tool for quantifying spatial variation is the variogram model. Estimation of the parameters of this model require a su ciently large sample, which is a challenge in this study since there is only fteen stations and therefore only fteen observations for the geostatistical analysis. To address this challenge, observations are expanded in space and time and then standardized and to create a larger pool of data from which the variogram is estimated. The obtained estimates are used in ordinary and universal kriging to derive the 50-year 24-hour winter rainfall return level maps. It is shown that 50-year daily winter design rainfall over most of the Western Cape lies between 40 mm and 80 mm, but rises sharply as one moves towards the east coast of the region. This is largely due to the in uence of large design values obtained for Tygerhoek. In ordinary kriging prediction uncertainty is lowest around observed values and is large if the distance from these points increases. Overall, prediction uncertainty maps show that ordinary kriging performs better than universal kriging where a linear regional trend in design values is included.