Surrogate models, also referenced as metamodels, have emerged as attractive data-driven, predictive models for storm surge estimation. They are calibrated based on an existing database of synthetic storm simulations and can provide fast-to-compute approximations of the expected storm surge, replacing the numerical model that was used to establish this database. This paper discusses specifically the development of a kriging metamodel for the prediction of peak storm surges. For nearshore nodes that have remained dry in some of the synthetic storm simulations, a necessary first step, before the metamodel calibration, is the imputation of the database to address the missing data corresponding to such dry instances to estimate the so-called pseudo-surge. This imputation is typically performed using a geospatial interpolation technique, with the k nearest-neighbor (kNN) interpolation being the one chosen for this purpose in this paper. The pseudo-surge estimates obtained from such an imputation may lead to an erroneous classification for some instances, with nodes classified as inundated (pseudo-surge greater than the node elevation), even though they were actually dry. The integration of a secondary node classification surrogate model was recently proposed to address the challenges associated with such erroneous information. This contribution further examines the above integration and offers several advances. The benefits of implementing the secondary surrogate model are carefully examined across nodes with different characteristics, revealing important trends for the necessity of integrating the classifier in the surge predictions. Additionally, the combination of the two surrogate models using a probabilistic characterization of the node classification, instead of a deterministic one, is considered. The synthetic storm database used to illustrate the surrogate model advances corresponds to 645 synthetic tropical cyclones (TCs) developed for a flood study in the Louisiana region. The fact that various flood protective measures are present in the region creates interesting scenarios with respect to the groups of nodes that remain dry for some storms behind these protected zones. Advances in the kNN interpolation methodology, used for the geospatial imputation, are also presented to address these unique features, considering the connectivity of nodes within the hydrodynamic simulation model.
Coastal regions within the U.S. Army Corps of Engineers (USACE) areas of responsibility are highly vulnerable to coastal storms, with more than $1 trillion in damages in the last two decades. Tropical cyclones (TCs), extratropical cyclones (XC), and other extreme storm events impose a variety of coastal hazards including storm surge, waves, rainfall, and wind, which often lead to devastating impacts to coastal communities. Compounding these events with increased storminess (e.g., frequency, intensity) and sea level rise is expected to exacerbate damages, therefore driving the need for more resilient flood and coastal storm risk management efforts to protect coastal communities, critical infrastructure, and ecosystems. The Coastal Hazards System (CHS) (https://chs.erdc.dren.mil) was developed by the USACE in direct response to the need for high quality hazard information to support coastal resiliency. CHS is a multi-agency, national scale effort for the accurate, efficient, and consistent quantification of coastal storm hazards for all U.S. coastlines and other locations of significance for national security. The primary objective of the CHS is the development and provision of high-fidelity, high-resolution probabilistic coastal storm hazard and process-based numerical modeling data covering the full probability space of coastal storm responses (e.g., storm surge, waves) for direct application in coastal engineering and risk assessment activities.
Surrogate models are widely recognized as effective, data-driven predictive tools for storm surge risk assessment. For such applications, surrogate models (referenced also as emulators or metamodels) are typically developed using existing databases of synthetic storm simulations, and once calibrated can provide fast-to-compute approximations of the storm surge for a variety of downstream analyses. The storm surge predictions need to be established for different geographic locations of interest, typically corresponding to the computational nodes of the original numerical model. A number of inland nodes will remain dry for some of the database storm scenarios, requiring an imputation for them to estimate the so-called pseudo-surge in support of the surrogate model development. Past work has examined the adoption of kNN (k-nearest neighbor) spatial interpolation for this imputation. The enhancement of kNN with hydraulic connectivity information, using the grid or mesh of the original numerical model, was also previously considered. In this enhancement, neighboring nodes are considered connected only if they are connected within the grid. This work revisits the imputation of peak storm surge within a surrogate modeling context and examines three distinct advancements. First, a response-based correlation concept is considered for the hydraulic connectivity, replacing the previous notion of connectivity using the numerical model grid. Second, a Gaussian Process interpolation (GPI) is examined as alternative spatial imputation strategy, integrating a recently established adaptive covariance tapering scheme for accommodating an efficient implementation for large datasets (large number of nodes). Third, a data completion approach is examined for imputation, treating dry instances as missing data and establishing imputation using probabilistic principal component analysis (PPCA). The combination of spatial imputation with PPCA is also examined. In this instance, spatial imputation is first deployed, followed by PPCA for the nodes that were misclassified in the first stage. Misclassification corresponds to the instances for which imputation provides surge estimates higher than ground elevation, creating the illusion that the node is inundated even though the original predictions correspond to the node being dry. In the illustrative case study, different imputation variants established based on the aforementioned advancements are compared, with comparison metrics corresponding to the predictive accuracy of the surrogate models developed using the imputed databases. Results show that incorporating hydraulic connectivity based on response similarity into kNN enhances the predictive performance, that GPI provides a competitive (to kNN) spatial interpolation approach, and that the combination of data completion and spatial interpolation emerges as the recommended approach.
The quantification of coastal hazards caused by tropical cyclones (TCs) faces challenges due to insufficient historical observations, particularly for intense hurricanes, and other data limitations. In the wake of Hurricane Katrina in 2005, a hybrid probabilistic/process-based modeling approach known as the Joint Probability Method (JPM) went through in-depth scruntinity and significant improvements, spurred by the Interagency Performance Evaluation Taskforce (IPET 2009). As a result of this forensic study, IPET developed the JPM with Optimal Sampling (JPM-OS) technique for the analysis of extreme coastal flooding resulting from hurricanes. This collaborative effort involved several U.S. federal agencies, the private sector, and academic researchers. The IPET’s JPM-OS became a standard for U.S. federal agencies, including the U.S. Army Corps of Engineers (USACE) and the Federal Emergency Management Agency (FEMA), for assessing coastal storm risks in areas susceptible to hurricanes. However, this led to the development of two distinct JPM-OS variants, each with different advantantages and limitations. Since then, JPM has informed numerous storm surge and hurricane hazard studies led by the federal government, academia, and private sector. During the last decade, considerable effort was made to address limitations of existing JPM-OS approaches. Advancements in multi-variate statistics and machine learning (ML) led to the development of a more robust version of the JPM called JPM Aided by Metamodel Prediction (JPM-AMP).
Flooding in many coastal regions is exacerbated due to complex interactions of multiple flood drivers which can occur concurrently or sequentially during coastal storm events, resulting in compound flood hazards. The Texas Gulf Coast (USA) is especially vulnerable to compound flooding due to frequent tropical cyclones (TCs) that bring strong winds, driving storm surge and heavy rainfall leading to inland flooding. Devastation caused by recent TCs, such as Hurricane Harvey, have underscored the need to accurately quantify combined coastal and inland flood hazards to support coastal resilience. Traditionally, the quantification of probabilistic coastal storm hazards in the Gulf of Mexico has employed the Joint Probability Method (JPM), considering TCs as the primary drivers of surge and wave hazards (Nadal-Caraballo et al. 2019). However, these approaches typically overlook or simplify the flood hazard from TC- induced rainfall. Recent studies of rainfall-surge pairing analyses using bivariate copula methods on observational datasets have evaluated the joint flood hazards resulting from rainfall and storm surge (e.g., Kim et al., 2022). Such statistical analysis of observations can be a critical tool for evaluating compound coastal flood hazards. Nevertheless, existing studies generally neglect uncertainty or are limited to bootstrapping/resampling approaches. Substantial uncertainty exists due to limited observations of extreme events, particularly when characterizing the low-frequency tails of flood hazard curves. In this study, we performed a bivariate copula analysis at outlets of large rivers along the Texas Gulf coast. Our objective was to investigate the uncertainties associated with the bivariate copula approach and establish confidence intervals of the hazard curves.