A5020692507- Atmospheric Ozone and Climate
- Atmospheric chemistry and aerosols
- Climate variability and models
- Financial Risk and Volatility Modeling
- Hydrology and Drought Analysis
- Statistical Methods and Inference
- Atmospheric and Environmental Gas Dynamics
- Soil Geostatistics and Mapping
- Stochastic processes and financial applications
- Probabilistic and Robust Engineering Design
- Spatial and Panel Data Analysis
- Risk and Portfolio Optimization
- Statistical Methods and Bayesian Inference
- Air Quality Monitoring and Forecasting
- Insurance, Mortality, Demography, Risk Management
- Advanced Multi-Objective Optimization Algorithms
- Optimal Experimental Design Methods
- Hydrology and Watershed Management Studies
- Fault Detection and Control Systems
- Financial Markets and Investment Strategies
- Advanced Control Systems Optimization
- Bayesian Methods and Mixture Models
- Agricultural Economics and Policy
- Flood Risk Assessment and Management
- Control Systems and Identification
Laboratoire de Mathématiques Jean Leray
2005-2024
École Centrale de Nantes
2021-2024
Université de Montpellier
2011-2020
Centre National de la Recherche Scientifique
2012-2020
Laboratoire de Mathématiques d'Orsay
2010-2015
Institut Montpelliérain Alexander Grothendieck
2011-2014
École Polytechnique Fédérale de Lausanne
2008-2013
University of Padua
2012
Institut National de la Recherche Scientifique
2006-2007
Laboratoire d’HYdrologie et de GEochimie
2006
The areal modeling of the extremes a natural process such as rainfall or temperature is important in environmental statistics; for example, understanding extreme crucial flood protection. This article reviews recent progress statistical spatial extremes, starting with sketches necessary elements value statistics and geostatistics. main types models thus far proposed, based on latent variables, copulas max-stable processes, are described then compared by application to data set Switzerland....
The last decade has seen max-stable processes emerge as a common tool for the statistical modeling of spatial extremes. However, their application is complicated due to unavailability multivariate density function, and so likelihood-based methods remain far from providing complete flexible framework inference. In this article we develop inferentially practical, fitting derived composite-likelihood approach. procedure sufficiently reliable versatile permit simultaneous marginal dependence...
Abstract Motivation Approximate Bayesian computation (ABC) has grown into a standard methodology that manages inference for models associated with intractable likelihood functions. Most ABC implementations require the preliminary selection of vector informative statistics summarizing raw data. Furthermore, in almost all existing implementations, tolerance level separates acceptance from rejection simulated parameter values needs to be calibrated. Results We propose conduct likelihood-free...
Composite likelihoods are increasingly used in applications where the full likelihood is analytically unknown or computationally prohibitive. Although some frequentist properties of maximum composite estimator akin to those estimator, Bayesian inference based on its early stages. This paper discusses when one uses Bayes' formula. We establish that using a results proper posterior density, though it can differ considerably from stemming likelihood. Building previous work ratio tests, we use...
History: The latest developments of extreme value theory focus on the functional framework and much effort has been put in max-stable processes regular variations.Paralleling univariate theory, this work focuses exceedances a stochastic process above high threshold their connections with generalized Pareto processes.More precisely we define an exceedance through homogeneous cost show that limiting (rescaled) distribution is -Pareto whose spectral measure can be characterized.Three equivalent...
We show how to perform full likelihood inference for max‐stable multivariate distributions or processes based on a stochastic expectation–maximization algorithm, which combines statistical and computational efficiency in high dimensions. The good performance of this methodology is demonstrated by simulation the popular logistic Brown–Resnick models, it shown provide time improvements with respect direct computation likelihood. Strategies further reduce burden are also discussed.
Since many environmental processes such as heat waves or precipitation are spatial in extent, it is likely that a single extreme event affects several locations and the areal modelling of extremes therefore essential if dependence has to be appropriately taken into account. This paper proposes framework for conditional simulations max-stable give closed forms Brown-Resnick Schlather processes. We test method on simulated data an application rainfall around Zurich temperature Switzerland....
Abstract. We present the first spatial analysis of "fingerprints" El Niño/Southern Oscillation (ENSO) and atmospheric aerosol load after major volcanic eruptions (El Chichón Mt. Pinatubo) in extreme low high (termed ELOs EHOs, respectively) mean values total ozone for northern southern mid-latitudes (defined as region between 30° 60° north south, respectively). Significant influence on extremes was found warm ENSO phase both hemispheres during spring, especially towards latitudes, indicating...
Abstract. We use statistical models for mean and extreme values of total column ozone to analyze "fingerprints" atmospheric dynamics chemistry on long-term changes at northern southern mid-latitudes grid cell basis. At each cell, the r-largest order statistics method is used analysis events in low high (termed ELOs EHOs, respectively), an autoregressive moving average (ARMA) model corresponding value analysis. In describe dynamical chemical state atmosphere, include important covariates:...
Abstract. In this study the frequency of days with extreme low (termed ELOs) and high EHOs) total ozone values their influence on mean trends are analyzed for world's longest record (Arosa, Switzerland). The results show (i) an increase in ELOs (ii) a decrease EHOs during last decades (iii) that overall trend 1970s 1980s is strongly dominated by changes these events. After removing extremes, time series shows reduced (reduction factor 2.5 annual mean). Excursions events reveal "fingerprints"...
Abstract. In this study ideas from extreme value theory are for the first time applied in field of stratospheric ozone research, because statistical analysis showed that previously used concepts assuming a Gaussian distribution (e.g. fixed deviations mean values) total data do not adequately address structure extremes. We show methods appropriate to identify extremes and describe tails Arosa (Switzerland) series. order accommodate seasonal cycle ozone, daily moving threshold was determined...
The statistical modeling of spatial extremes has been an active area recent research with a growing domain applications. Much the existing methodology, however, focuses on magnitudes extreme events rather than their timing. To address this gap, article investigates notion extremal concurrence. Suppose that daily temperatures are measured at several synoptic stations. We say concurrent if record maximum occur simultaneously, is, same day for all It is important to be able understand,...
We apply methods from extreme value theory to identify events in high (termed EHOs) and low ELOs) total ozone describe the distribution tails (i.e. very values) of five long-term European ground-based time series. The influence these on observed mean values, trends changes is analysed. results show a decrease EHOs an increase ELOs during last decades, establish that downward trend column 1970–1990s strongly dominated by frequency events. Furthermore, it shown clear 'fingerprints' atmospheric...
Regional flood frequency analysis is a convenient way to reduce estimation uncertainty when few data are available at the gauging site. In this work, model that allows non‐null probability regional fixed shape parameter presented. This methodology integrated within Bayesian framework and uses reversible jump techniques. The performance on stochastic of new estimator compared two other models: conventional index approach. Results show proposed absolutely suited only target Moreover, unlike...
While univariate nonparametric estimation methods have been developed for estimating returns in mean-downside risk portfolio optimization, the problem of handling possible cross-correlations a vector asset has not addressed selection. We present novel multivariate optimization procedure using kernel-based estimators conditional mean and median. The method accounts covariance structure information from full set returns. also provide two computational algorithms to implement estimators. Via...