Elsevier

Ocean Modelling

Volumes 52–53, August 2012, Pages 54-68
Ocean Modelling

Tropical cyclone inundation potential on the Hawaiian Islands of Oahu and Kauai

https://doi.org/10.1016/j.ocemod.2012.04.009Get rights and content

Abstract

The lack of a continental shelf in steep volcanic islands leads to significant changes in tropical cyclone inundation potential, with wave setup and runup increasing in importance and wind driven surge decreasing when compared to more gently-sloped mainland regions. This is illustrated through high resolution modeling of waves, surge, and runup on the Hawaiian Islands of Oahu and Kauai. A series of hurricane waves and water levels were computed using the SWAN + ADCIRC models for a suite of 643 synthetic storm scenarios, while local wave runup was evaluated along a series of 1D transects using the phase-resolving model Bouss1D. Waves are found to be an extremely important component of the inundation, both from breaking wave forced increases in storm surge and also from wave runup over the relatively steep topography. This is clear in comparisons with debris lines left by Hurricane Iniki on the Island of Kauai, where runup penetration is much greater than still water inundation in most instances. The difference between steeply-sloping and gently-sloping topographies was demonstrated by recomputing Iniki with the same landfall location as Hurricane Katrina in Louisiana. Surge was greatly increased for the mild-slope Iniki-in-Louisiana case, while pure wind surge for Iniki-in-Kauai was very small.

For the entire suite of storms, maxima on Kauai show predicted inundation largely confined to a narrow coastal strip, with few locations showing more than a few hundred meters of flooding from the shoreline. As expected, maximum flooded areas for the 643 storms were somewhat greater than the Iniki inundation.

Oahu has significantly more low-lying land compared to Kauai, and consequently hypothetical tropical cyclone landfalls show much more widespread inundation. Under direct impact scenarios, there is the potential for much of Honolulu and most of Waikiki to be inundated, with both still water surge and wave runup contributing. Other regions of Oahu show inundation confined to a more narrow coastal strip, although there is still much infrastructure at risk.

Even for very strong storms in Oahu and Kauai, maximum still water surge is relatively small, and does not exceed 3 m in any storm modeled. In contrast, hurricane waves several kilometers from shore regularly exceed 10 m due to the lack of a continental shelf.

Highlights

► Steep sided volcanic islands show much lower surge potential and much greater runup potential than mainland locations. ► Large portions of southern Oahu may be inundated in a direct hurricane landfall. ► Inundation for the Island of Kauai is largely confined to a narrow coastal strip. ► Very large waves will propagate within a few kilometers of shore during hurricanes in the Hawaiian Islands.

Introduction

Landfalling tropical cyclones are a major source of damaging winds, waves and surge, causing hundreds of billions of dollars in damage to the United States over the past decade (Blake and Landsea, 2011). In the continental United States, hurricanes tend to make landfall in regions with a relatively broad, gently sloping continental shelf that is tens to hundreds of kilometers wide (Hurricane Katrina, 2005; Hurricane Ike, 2008; Hurricane Irene, 2011, nhc.noaa.gov). However, oceanic volcanic islands do not have continental shelves and water depths of hundreds to thousands of meters may be within a few kilometers of shore. For this reason, the important processes in inundation also change. For steady-state conditions, this may be shown by the one-dimensional steady-state storm surge balance (e.g. Dean and Darymple, 1991)gηx=τsρ(h+η)-1ρ(h+η)Sxxx-1ρPaxwhere η and h are the water surface elevation and depth, respectively, g is gravitational acceleration, ρ is the water density, τs is the wind surface stress, Sxx is the wave radiation stress, and Pa is the atmospheric pressure. Other terms will appear for two-dimensional, unsteady cases and when considering three-dimensional motions, but the basic cross-shore force balance is largely the same. For a wide continental shelf, surge is dominated by the integration of surface wind stresses over a long distance, with atmospheric pressure variations, wave setup, and wave modification to surface drag (e.g. Mastenbroek et al., 1993) also playing roles. However, as continental shelves shrink but the storm size remains constant, the contribution from wind stresses integrated over the shelf becomes much smaller because the integration distance is much shorter. In contrast, the pressure term remains the same, as it has no depth dependence. The wave setup contribution will increase in magnitude, as its value at the shoreline is proportional to the breaking wave height (Dean and Darymple, 1991), which can be very large in regions with deep water close to shore. Waves may partially reflect from steep-fronted reefs, while wind-wave reflection is negligible on gentle continental shelves. Thus, surge and inundation response is expected to be much different from that in mainland regions.

A further difference lies at the immediate shoreline. In very flat regions subject to surge, waves may be relatively small at the shoreline and thus wave runup is not very important. However, in regions like the Hawaiian Islands where large waves can propagate close to shore and shorelines may be relatively steep, wave runup can greatly increase storm inundation over still water surge (e.g., Cheung et al., 2003). Thus it becomes essential to include wave runup in all estimates of storm inundation for steep volcanic islands.

The Hawaiian Islands are classic examples of islands with no continental shelf. While they typically experience very mild weather, historical hurricanes have produced severe consequences. By far the most devastating storm in recent years was in 1992, when Hurricane Iniki made a direct landfall on the island of Kauai as a Category 4 storm on the Saffir-Simpson scale, with a central pressure of 945 millibars (mbar) and 140 mph (63 m/s) winds. Impacts were severe, with $1.8 billion damage, over 1400 homes completely destroyed and 10 times that number damaged in a relatively sparsely populated island (US Department of Commerce, 1993, Post et al., 1993). If a storm like Iniki had instead made landfall in Honolulu, impacts would have been even greater. Kauai was further impacted by Hurricane Iwa in 1982, which passed by as a strong category 1 storm. Other notable storms include Hurricane Dot (1959), which passed just off the coast of Hawaii, Oahu, and Kauai; Hurricane Nina (1957), which passed offshore of Kauai; and the Kohala cyclone of 1871, which caused great damage on the islands of Hawaii and Maui (Central Pacific Hurricane Center, http://www.prh.noaa.gov/hnl/cphc/); other hurricanes are known from the historical record. This infrequency combined with the large potential severity means that hurricanes in the Hawaiian Islands behave more like tsunamis in return frequency than typical hurricane-prone regions. Because of this, true return periods are difficult to obtain. Thus, it is useful to examine the inundation caused by different hurricane scenarios to determine potential areas of danger in the event of an oncoming landfall, and for determining evacuation routes, emergency shelter locations, and other longer term planning.

Section snippets

Inundation modeling on Oahu and Kauai

As part of the United States Army Corps of Engineers Surge and Wave Island Modeling Studies (SWIMS), inundation was computed for a suite of 643 possible hurricanes in the vicinity of Oahu and Kauai. Two sets of models were used: the first suite was the SWAN + ADCIRC wave and circulation models (Zijlema, 2010, Dietrich et al., 2011), and the second was a 1D phase-resolving Boussinesq surf zone model used to compute wave group inundation (Demirbilek et al., 2009).

Hurricane Iniki comparisons

Fig. 1 shows the track of Hurricane Iniki, and its landfall on the southern shore of Kauai. Iniki is particularly important because it is the only major hurricane with measurements for comparison in the Hawaiian Islands. In many parts of Kauai, a clear debris line was visible and was employed by Fletcher et al., 1995, Cheung et al., 2003 to develop runup inundation maps. Here, we use these runup inundation data for several purposes: both to compare to the modeled results, and as a means of

Inundation over different regions of Oahu and Kauai

Although they can be devastating, hurricane landfalls in the Hawaiian Islands are infrequent enough that reliable return period statistics are difficult to develop. However, potential scenarios may still be developed based on historical tracks and a knowledge of weather patterns. Using input from the Central Pacific Hurricane Center (http://www.prh.noaa.gov/hnl/cphc/), a suite of scenarios was developed and SWAN + ADCIRC and Bouss1D model runs were performed for 643 storms in the vicinity of Oahu

Wave height and water level envelopes

Because it is often useful to have an overall estimate of inundation for a given storm strength and general location, islands were divided into regions as defined in Fig. 1 (taken from http://www.hawaii-guide.com/), still water surge and significant wave heights in these regions were extracted for all storms, and results were separated according to storm characteristics (central pressure and landfall location). Next, for each storm in the database, all wet nodes on what is normally dry land

Conclusions

The lack of a continental shelf and relatively steep slopes of the Hawaiian Islands are a double-edged sword for storm inundation. Still water surge levels for a given storm will be greatly reduced over mainland locations, and for our computations do not exceed 3 m in either Oahu of Kauai for any of the 643 synthetic storm scenarios tested. The example of Iniki making landfall in Kauai compared to if it had made landfall in southern Louisiana demonstrates the large reduction in surge from 4.8 m

Acknowledgements

This work was funded under the Surge and Wave Island Modeling Studies under the Field Data Collection Program of the US Army Corps of Engineers. Permission to publish was granted by the Chief of Engineers.

References (24)

  • Blake, E.S., Landsea, C.W., 2011. The deadliest, costliest, and most intense United States tropical cyclones from 1851...
  • S. Bunya et al.

    A high-resolution coupled riverine flow, tide, wind, wind wave, and storm surge model for southern Louisiana and Mississippi. Part I: Model development and validation

    Month. Weather Rev.

    (2010)
  • P.C. Caldwell

    Validity of north shore, Oahu, Hawaiian Islands surf observations

    J. Coastal Res.

    (2005)
  • L. Cavaleri et al.

    Wind wave prediction in shallow water: theory and applications

    J. Geophys. Res.

    (1981)
  • K.F. Cheung et al.

    Modeling of storm-induced coastal flooding for emergency management

    Ocean Eng.

    (2003)
  • K.F. Cheung et al.

    Modeling of 500-year tsunamis for probabilistic design of coastal infrastructure in the Pacific Northwest

    Coast. Eng.

    (2011)
  • R.G. Dean et al.

    Water Wave Mechanics for Engineers and Scientists

    (1991)
  • J.C. Dietrich et al.

    A high-resolution coupled riverine flow, tide, wind, wind wave, and storm surge model for southern Louisiana and Mississippi. Part II: Synoptic description and analysis of Hurricanes Katrina and Rita

    Month. Weather Rev.

    (2010)
  • J.C. Dietrich et al.

    Modeling hurricane waves and storm surge using integrally-coupled, scalable computations

    Coast. Eng.

    (2011)
  • Demirbilek, Z., Nwogu, O.G., Ward, D.L., Sánchez, A., 2009. Wave transformation over reefs: evaluation of...
  • C.H. Fletcher et al.

    Marine flooding on the coast of Kaua’i during Hurricane Iniki: hindcasting inundation components and delineating washover

    J. Coastal Res.

    (1995)
  • K.O. Kim et al.

    Coupled process-based cyclone surge simulation for the Bay of Bengal

    Ocean Model.

    (2008)
  • Cited by (94)

    View all citing articles on Scopus
    View full text