Approximate Bayesian inference for analysis of spatiotemporal flood frequency data

DOI10.1214/21-AOAS1525zbMATH Open1498.62285OpenAlexW3142820938WikidataQ114135206 ScholiaQ114135206MaRDI QIDQ2154186FDOQ2154186

Authors Árni V. Jóhannesson, Stefan Siegert, Raphaël Huser, Haakon Bakka, Birgir Hrafnkelsson

Publication date 14 July 2022

Published in The Annals of Applied Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1907.04763

zbMATH Keywords

approximate Bayesian inferenceflood frequency analysislatent Gaussian modelmultivariate link functionmax-and-smoothspatiotemporal extremes

Mathematics Subject Classification ID

Bayesian inference (62F15) Directional data; spatial statistics (62H11) Inference from spatial processes (62M30) Statistics of extreme values; tail inference (62G32) Applications of statistics to environmental and related topics (62P12) Geostatistics (86A32)

Abstract: Extreme floods cause casualties, and widespread damage to property and vital civil infrastructure. We here propose a Bayesian approach for predicting extreme floods using the generalized extreme-value (GEV) distribution within gauged and ungauged catchments. A major methodological challenge is to find a suitable parametrization for the GEV distribution when covariates or latent spatial effects are involved. Other challenges involve balancing model complexity and parsimony using an appropriate model selection procedure, and making inference using a reliable and computationally efficient approach. Our approach relies on a latent Gaussian modeling framework with a novel multivariate link function designed to separate the interpretation of the parameters at the latent level and to avoid unreasonable estimates of the shape and time trend parameters. Structured additive regression models are proposed for the four parameters at the latent level. For computational efficiency with large datasets and richly parametrized models, we exploit an accurate and fast approximate Bayesian inference approach. We applied our proposed methodology to annual peak river flow data from 554 catchments across the United Kingdom (UK). Our model performed well in terms of flood predictions for both gauged and ungauged catchments. The results show that the spatial model components for the transformed location and scale parameters, and the time trend, are all important. Posterior estimates of the time trend parameters correspond to an average increase of about $1.5\mathrm{\%}$ per decade and reveal a spatial structure across the UK. To estimate return levels for spatial aggregates, we further develop a novel copula-based post-processing approach of posterior predictive samples, in order to mitigate the effect of the conditional independence assumption at the data level, and we show that our approach provides accurate results.

Recommendations

• A regional Bayesian POT model for flood frequency analysis
• Bayesian Spatial Modeling of Extreme Precipitation Return Levels
• INLA goes extreme: Bayesian tail regression for the estimation of high spatio-temporal quantiles
• Generalized extreme value additive model analysis via mean field variational Bayes
• Bayesian inference for multivariate extreme value distributions

Cites work

• A Test of Goodness of Fit
• A hierarchical max-infinitely divisible spatial model for extreme precipitation
• Advanced spatial modeling with stochastic partial differential equations using R and INLA
• An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach
• An introduction to statistical modeling of extreme values
• Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations (with discussion)
• Bayesian hierarchical modeling of extreme hourly precipitation in Norway
• Bayesian modeling of air pollution extremes using nested multivariate max-stable processes
• Computationally efficient spatial modeling of annual maximum 24‐h precipitation on a fine grid
• Constructing Priors that Penalize the Complexity of Gaussian Random Fields
• Continuous spatial process models for spatial extreme values
• Extremes on river networks
• Generalized Additive Models for Exceedances of High Thresholds With an Application to Return Level Estimation for U.S. Wind Gusts
• Generalized Additive Models for Location, Scale and Shape
• INLA goes extreme: Bayesian tail regression for the estimation of high spatio-temporal quantiles
• LGM split sampler: an efficient MCMC sampling scheme for latent Gaussian models
• Likelihood-based inference for max-stable processes
• Max-and-smooth: a two-step approach for approximate Bayesian inference in latent Gaussian models
• Max-infinitely divisible models and inference for spatial extremes
• Penalising model component complexity: a principled, practical approach to constructing priors
• Space–Time Modelling of Extreme Events
• Spatial hierarchical modeling of precipitation extremes from a regional climate model
• Spatial regression models for extremes
• Statistical modeling of spatial extremes
• Statistics for spatial data
• Uncertainty quantification in complex simulation models using ensemble copula coupling

Cited in

(5)

• Bayesian Spatial Clustering of Extremal Behavior for Hydrological Variables
• Smooth copula-based generalized extreme value model and spatial interpolation for extreme rainfall in Central Eastern Canada
• A regional Bayesian POT model for flood frequency analysis
• A combined statistical and machine learning approach for spatial prediction of extreme wildfire frequencies and sizes
• The flood algorithm -- a multivariate, self-organizing-map-based, robust location and covariance estimator

This page was built for publication: Approximate Bayesian inference for analysis of spatiotemporal flood frequency data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2154186)