ECHO Post-Doctoral Fellowship Solicitation Request (CHL Internal Submission)

Project Title:
Improving Statistical Elements of Probabilistic Coastal Compound Hazard Analysis

Research Advisor: Meredith L. Carr, PhD, PE, 
Dr. Meredith L. Carr is a Research Civil Engineer and member of the Coastal Hazards Group in the Harbors, Entrances, & Structures Branch at CHL. Dr. Carr’s research focuses on probabilistic compound coastal hazards analysis for precipitation-induced riverine and coastal storm (storm surge and wave) hazards in support of local and district coastal flood risk management.

Proposed Length of Time:
2-year position, with option to extend 1 year.

Location: Vicksburg, MS

Project Background:
USACE needs a technically robust method for probabilistic assessment of coastal storm hazards that includes compound flood processes, (i.e. precipitation-induced flooding as well as storm surge) to properly characterize risk at coastal/inland transition zones. Regional coastal projects, R&D studies, flood-risk mapping, and emergency response activities require quantifying coastal compound hazard to plan, design, and ascertain risk. Several studies have shown that neglecting compound, multi-hazard effects can underestimate the annual exceedance probability or expected likelihood of a given water level (e.g., Bevacqua et al., 2017; Moftakhari et al., 2019). The effort to properly describe the full probability space of coastal hazards is challenged by the relative low frequency of Tropical Cyclone (TC) occurrence, combined with limited record lengths and low geospatial data resolution. Full description of coastal compound hazards is further aggravated by more severe limitations of these kinds of limitations, further complicating assessment of the compound hazard.

The goal of this work is to develop a robust framework for extending Joint Probability Methods (JPM) to include rainfall processes for analysis of compound coastal storm and precipitation hazards. The Coastal Hazards System (CHS) is a national-scale effort for quantification of coastal storm hazards along U.S. coastlines. The foundation of the CHS is its Probabilistic Coastal Hazard Analysis (PCHA) framework (Nadal-Caraballo et al., 2020). PCHA is a comprehensive statistical and probabilistic framework for characterization of regional storm climatology, joint probability analysis of atmospheric forcing and storm responses, high-resolution numerical modeling, machine learning, and quantification of associated aleatory and epistemic uncertainties. The CHS database and web tool store and distribute high-fidelity modeling and PCHA results for regional coastal hazard storm studies conducted as part of the CHS. The CHS-PCHA provides a platform and starting point to extend current capabilities to address multi-hazard compound flooding problems, using similar characterization of storm climatology, high-resolution numerical modeling, and advanced JPM approaches. This framework also provides synthetic TCs to drive TC rainfall models that predict spatially-distributed rainfall rates from atmospheric parameters (e.g. TCRM, Lu et al. 2018). TC rainfall is used to drive the inland flooding models which, coupled with the coastal flooding models, simulates the full physical response of the compound system. Using copula theory for a more complete representation of the joint probability and Gaussian Process Metamodeling (GPM) machine learning (Zhang et al., 2018) to expand the storm set, these advancements expand compound PCHA to better define the physical and probabilistic hazard space. This framework is being piloted and parameters and methods being optimized to provide a comprehensive representation of the compound hazard at the coast and in the transition zone.

Project Goals:
This Project seeks proposals for improving the probabilistic and statistical elements of the Probabilistic Coastal Compound Hazard Analysis (PCCHA) Framework in development at CHL. Topics could include improvement of correlation and joint probability representation between atmospheric JPM parameters, investigation and implementation of other parameters to improve compound hazard evaluation, implementing stochastic methods to improve TC rainfall limitations and/or application to hydrologic models, probabilistic exploration of the methods within the framework, and approaches to uncertainty. Proposals should include a description of the topic the candidate intends to focus on and any techniques to be applied.

Anticipated Skillsets:
Seeking applicants with a PhD in statistics and data science, civil engineering, flood risk, atmospheric science, meteorology or related disciplines.  The successful candidate should have strong skills in applying foundation concepts of probability and statistics to assess frequency, severity, and uncertainty of natural hazards. A solid background in applying and improving statistical techniques, joint probability methods and machine-learning approaches in Matlab and/or Python is preferred. Experience crossing interdisciplinary specialties and in the areas of inland flooding, coastal flooding, atmospheric sciences and precipitation would be beneficial.


Bevacqua, E., Maraun, D., Hobæk Haff, I., Widmann, M., and Vrac, M. 2017. Multivariate statistical modelling of compound events via pair-copula constructions: analysis of floods in Ravenna (Italy), Hydrol. Earth Syst. Sci., 21, 2701–2723,

Lu, P., N. Lin, K. Emanuel, D. Chavas, and J. Smith. 2018. Assessing Hurricane Rainfall Mechanisms Using a Physics-Based Model: Hurricanes Isabel (2003) and Irene (2011). Journal of Atmospheric Science 75: 2337–2358.

Moftakhari, H., J. E. Schubert, A. AghaKouchak, R. A. Matthew, and B. F. Sanders. 2019. Linking Statistical and Hydrodynamic Modeling for Compound Flood Hazard Assessment in Tidal Channels and Estuaries. Advances in Water Resources 128): 28–38.  

Nadal-Caraballo, N.C. M. O. Campbell, V. M. Gonzalez, M. J. Torres, J. A. Melby, A. A. Taflanidis. 2020. Coastal Hazards System: A Probabilistic Coastal Hazard Analysis Framework. Journal of Coastal Research; 95 (SI): 1211–1216.

Zhang, J., Taflanidis, A.A., Nadal-Caraballo, N.C., Melby, J.A., Diop, F. 2018. Advances in surrogate modeling for storm surge prediction: storm selection and addressing characteristics related to climate change. Nat Hazards 94, 1225–1253

Click Here for Additional Information