Statistical Analysis and Storm Sampling Approach for Lakes Michigan and St. Clair, Great Lakes Coastal Flood Study, 2012 Federal Inter-Agency Initiative

Abstract

A new methodology is proposed for Great Lakes flood hazard mapping. The methodology includes a process for sampling and screening storm events and computing water level probabilities based on high-fidelity modeling of significant storm events. A technical analysis framework is provided to construct accurate extremal distributions of total water levels and to accurately estimate base flood elevations. High-resolution, high-fidelity modeling of all historical storm events is simply not feasible due to time, computational and funding constraints. Therefore, the recommended approach is to screen and sample historical events to select the minimum number of events required to accurately model the total water level extremal distributions. This study focused on evaluating the validity of the recommended statistical analysis and storm sampling approach, and determining the adequate storm sample size. For this purpose, several tasks were performed, including: computation of storm-surrogate waves and water levels; definition of full sample and composite storm sets; and evaluation of the statistical analysis approach for a record length of 50 years. It was determined that the ideal number of events that should be sampled to accurately define water level extremal distributions is roughly 150 storms. Also, the prioritization of waves and water levels in storm sampling was extensively evaluated. It was found that approximately 25 percent to 30 percent of all selected storms were both highly ranked surge events and highly ranked wave events, thus minimizing the effects of different event prioritization ratios. Extreme water levels, corresponding to one percent and 0.2 percent annual chance of exceedance, had negligible variation regardless of the event prioritization ratio.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 2012
Accession Number
ADA569629

Entities

People

  • Bruce A. Ebersole
  • Jeffrey A. Melby
  • Norberto C. Nadal-Caraballo

Organizations

  • Engineer Research and Development Center

Tags

Communities of Interest

  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Accuracy
  • Barometric Pressure
  • Climate Change
  • Computations
  • Department Of Homeland Security
  • Flood Hazards
  • Floods
  • Great Lakes
  • High Resolution
  • Meteorology
  • Probability
  • Reliability
  • Sea Level Rise
  • Statistical Analysis
  • Statistics
  • Storm Surges
  • Test And Evaluation

Fields of Study

  • Environmental science

Readers

  • Atmospheric Science/Meteorology
  • Riverine Ecology
  • Statistical inference.