Probabilistic flood hazard assessment framework development : extreme rainfall analysis

Abstract

This report introduces a framework for probabilistic flood hazard assessment (PFHA) whose basis leverages recent advances in the science of spatial extremes. The framework basis includes a latent variable model (LVM) or a max-stable process application wherein for either case model inference is likelihood based. The framework is flexible in that it can leverage robust approaches to quantify model uncertainty while also supporting the capacity to readily combine additional relevant data types; for example, historical and/or paleoflood data for flood frequency analyses. This report profiles applications of Bayesian inference for flood hazard curve development for at-site and spatial LVM analyses. Pointwise spatial model development using an LVM or a max-stable process requires the parameters of the model characterizing the pointwise extremes to vary spatially as a function of gridded covariate data relevant to the hydrometeorological extreme under consideration. Recent advances in mathematical regularization facilitate spatial pointwise model reduction. The PFHA framework accommodates the multiple model parameterizations encapsulated within a given LVM or max-stable process deployment by generalizing model choice using information criteria. DISCLAIMER: The contents of this report are not to be used for advertising, publication, or promotional purposes. Citation of trade names does not constitute an official endorsement or approval of the use of such commercial products. All product names and trademarks cited are the property of their respective owners. The findings of this report are not to be construed as an official Department of the Army position unless so designated by other authorized documents.