Abstract:The generalized extreme value (GEV) distribution is a popular model for analyzing and forecasting extreme weather data. To increase prediction accuracy, spatial information is often pooled via a latent Gaussian process (GP) on the GEV parameters. Inference for GEV‐GP models is typically carried out using Markov Chain Monte Carlo (MCMC) methods, or using approximate inference methods such as the integrated nested Laplace approximation (INLA). However, MCMC becomes prohibitively slow as the number of spatial locations increases, whereas INLA is applicable in practice only to a limited subset of GEV‐GP models. In this article, we revisit the original Laplace approximation for fitting spatial GEV models. In combination with a popular sparsity‐inducing spatial covariance approximation technique, we show through simulations that our approach accurately estimates the Bayesian predictive distribution of extreme weather events, is scalable to several thousand spatial locations, and is several orders of magnitude faster than MCMC. A case study in forecasting extreme snowfall across Canada is presented. Résumé La distribution de valeurs extrêmes généralisée (GEV) est couramment utilisée pour l'analyse et la prévision des phénomènes météorologiques extrêmes. Pour améliorer la précision prédictive, les informations spatiales sont souvent regroupées via un processus gaussien latent (GP) appliqué aux paramètres GEV. L'inférence des modèles GEV‐GP utilise généralement les méthodes de Monte‐Carlo par chaînes de Markov (MCMC) ou des approximations comme l'approximation de Laplace imbriquée intégrée (INLA). Toutefois, les méthodes MCMC deviennent très lentes quand le nombre de sites spaciaux augmente, et l'INLA ne s'applique qu'à un nombre restreint de modèles GEV‐GP. Dans cet article, les auteurs réexaminent l'approximation de Laplace classique pour ajuster les modèles GEV spatiaux. En la combinant à une technique d'approximation de la covariance spatiale induisant la parcimonie, ils montrent par simulation que leur approche estime avec précision la distribution prédictive bayésienne des événements météorologiques extrêmes, qu'elle s'adapte à plusieurs milliers de sites spatiaux, et qu'elle est nettement plus rapide que les méthodes MCMC. L'article présente également une étude de cas portant sur la prévision des chutes de neige extrêmes au Canada.

A penalized maximum likelihood approach for m-year precipitation return values estimation with lattice spatial data

Assessing the Performance of the Independence Method in Modeling Spatial Extreme Rainfall

Penalized Maximum Likelihood Estimators for the Nonstationary Pearson Type 3 Distribution

L-Moment-Based Regional Frequency Analysis of Annual Extreme Precipitation and its Uncertainty Analysis

GAMLSS-based nonstationary modeling of extreme precipitation in Beijing–Tianjin–Hebei region of China

Improved return level estimation via a weighted likelihood, latent spatial extremes model

Penalized quasi-maximum likelihood estimation for extreme value models with application to flood frequency analysis

Performance evaluation of a multi-site weather generator coupling maximum entropy resampling for estimating the probability distribution of annual maximum daily rainfall in the Loess Plateau

Assessing the Gumbel-Logistic Model’s performance in modeling long-term rainfall series in a high-latitude plain region

Fast and scalable inference for spatial extreme value models

Predictive Inference for Spatio-temporal Precipitation Data and Its Extremes

A Note on Spatial-Temporal Lattice Modeling and Maximum Likelihood Estimation

Poisson Autoregressive Process Modeling Via the Penalized Conditional Maximum Likelihood Procedure

Enhancing Extreme Precipitation Forecasts through Machine Learning Quality Control of Precipitable Water Data from Satellite FengYun-2E: A Comparative Study of Minimum Covariance Determinant and Isolation Forest Methods

Comparison of Mapping Approaches for Estimating Extreme Precipitation of Any Return Period at Ungauged Locations

A Deep Learning Synthetic Likelihood Approximation of a Non-stationary Spatial Model for Extreme Streamflow Forecasting

Modeling of Multisite Precipitation Occurrences Using Latent Gaussian-based Multivariate Binary Response Time Series

Spatial Bayesian hierarchical modeling of precipitation extremes over a large domain

Modelling Extremes of Spatial Aggregates of Precipitation using Conditional Methods

Hierarchical low rank approximation of likelihoods for large spatial datasets

Multivariate Regression Analysis and Statistical Modeling for Summer Extreme Precipitation over the Yangtze River Basin, China