Storm Surge Model Evaluation

Airborne LIDAR data have already resulted in considerable improvements in hurricane storm surge mapping. These data are an order of magnitude more accurate vertically (6 inches compared to 5 feet) and are much higher resolution horizontally (by two to three orders of magnitude) than commonly available topographic information. High resolution topographic data combined with an updated numerical model with similar resolution result in much more realistic characterization of the storm surge hazard.

A number of 2-D and 3-D hydrodynamic models have been developed for storm surges. Model theory, performance, and computation schemes for different models were evaluated, and performance compared for selected hurricanes including Andrew, Hugo and Betsy. Tide and wave influences on storm surge prediction and grid resolution were also investigated. IHRC researchers selected three numerical models for evaluation, including the latest developments in the field:

  • SLOSH
  • ADCIRC
  • HRSM
STORM SURGE MODEL COMPARISON
SLOSH

  • Model has been tested extensively and used widely
  • Model domain is limited
  • Resolution is limited (0.5 -7km)
  • Difficult to simulate convoluted shorelines
  • Tide and wave set up not included
  • Difficult to incorporate features that block or accelerate storm surge flooding
  • Capacity to model overland flooding

ADCIRC (Version 27)

  • Model domain is extremely flexible
  • Resolution can reach finer than 50×50 m
  • Very good to simulate convoluted shorelines
  • Model includes tide, but no wave setup yet
  • Flexible to incorporate features like highways and canals
  • Capacity to model overland flooding

HRSM

  • Model algorithm is based on POM, ECOM, and EFDC which are well studied
  • Model domain is very flexible
  • Resolution can reach 50×50 m
  • Good to simulate convoluted shorelines
  • Model includes tide, but no wave setup yet
  • Good to incorporate features like highways and canals
  • Good to model overland flooding

The numerical storm surge model most widely used in the United States is the National Weather Service SLOSH (Sea, Lake, and Overland Surges from Hurricanes) model. Developed more than 25 years ago, this computer model has served us well, but there have been major advancements in computer technology, GIS and GPS since that time. The SLOSH model tends to produce large uncertainty in the predicted flooded area because of its relatively coarse resolution. The flood zone can be delineated much more precisely by combining SLOSH and airborne LIDAR 30m DEMs.

The ADCIRC model was shown to perform well by comparing predictions to observed storm surge levels, especially at the shoreline.

A new 3-dimensional High Resolution Surge Model (HRSM) was developed by the IHRC. This represents a major step forward, providing much improved forecasts of storm surge flooding (see map). By using curvilinear-orthogonal grids, this model is very flexible and can incorporate the effect of convoluted shorelines and major features such as highways (which often serve as barriers to surges). This high resolution model exhibits good agreement between predicted and observed storm surge penetration and high water marks (see map and graph; attached CD contains animation). There is a need to fully develop this prototype HRSM model for all of Florida.

Resolution Surge Model (HRSM) Predicted vs. Observed Storm Surge for Hurricane Andrew

Resolution Surge Model (HRSM) Predicted vs. Observed Storm Surge for Hurricane Andrew

 

Water Marks

Water Marks

Click the following link to get the movie clip for Hurricane Andrew Surge

Hurricane Andrew Surge (avi format, 1.63 MB)

IHRC Website © 2017 All Rights Reserved