INLA goes extreme: Bayesian tail regression for the estimation of high spatio-temporal quantiles

02/04/2018
by   Thomas Opitz, et al.
0

This work has been motivated by the challenge of the 2017 conference on Extreme-Value Analysis (EVA2017), with the goal of predicting daily precipitation quantiles at the 99.8% level for each month at observed and unobserved locations. We here develop a Bayesian generalized additive modeling framework tailored to estimate complex trends in marginal extremes observed over space and time. Our approach is based on a set of regression equations linked to the exceedance probability above a high threshold and to the size of the excess, the latter being modeled using the generalized Pareto (GP) distribution suggested by Extreme-Value Theory. Latent random effects are modeled additively and semi-parametrically using Gaussian process priors, which provides high flexibility and interpretability. Fast and accurate estimation of posterior distributions may be performed thanks to the Integrated Nested Laplace approximation (INLA), efficiently implemented in the R-INLA software, which we also use for determining a nonstationary threshold based on a model for the body of the distribution. We show that the GP distribution meets the theoretical requirements of INLA, and we then develop a penalized complexity prior specification for the tail index, which is a crucial parameter for extrapolating tail event probabilities. This prior concentrates mass close to a light exponential tail while allowing heavier tails by penalizing the distance to the exponential distribution. We illustrate this methodology through the modeling of spatial and seasonal trends in daily precipitation data provided by the EVA2017 challenge. Capitalizing on R-INLA's fast computation capacities and large distributed computing resources, we conduct an extensive cross-validation study to select model parameters governing the smoothness of trends. Our results outperform simple benchmarks and are comparable to the best-scoring approach.

READ FULL TEXT
research
04/01/2020

Bayesian space-time gap filling for inference on extreme hot-spots: an application to Red Sea surface temperatures

We develop a method for probabilistic prediction of extreme value hot-sp...
research
03/21/2023

Analytical Conjugate Priors for Subclasses of Generalized Pareto Distributions

This article is written for pedagogical purposes aiming at practitioners...
research
12/10/2019

Spatial hierarchical modeling of threshold exceedances using rate mixtures

We develop new flexible univariate models for light-tailed and heavy-tai...
research
12/03/2017

Bayesian prior elicitation and selection for extreme values

A major issue of extreme value analysis is the determination of the shap...
research
03/09/2020

evgam: An R package for Generalized Additive Extreme Value Models

This article introduces the R package evgam. The package provides functi...
research
06/05/2023

On Tail Decay Rate Estimation of Loss Function Distributions

The study of loss function distributions is critical to characterize a m...
research
09/06/2023

A Semiparametric Generalized Exponential Regression Model with a Principled Distance-based Prior for Analyzing Trends in Rainfall

The Western Ghats mountain range holds critical importance in regulating...

Please sign up or login with your details

Forgot password? Click here to reset