Estimation analysis and mapping of coastal overwash on Kaua'i during the landfall of Hurricane Iniki

[From the U.S. Government Printing Office, www.gpo.gov]

QC
944
H3
E*88
1993

7 ESMA'nONANALysis
7 AND MAPPING OF
7 COASTAL OVERWASH ONKAuNi
DURING THE LANDFALL OF
7 HuRRIcANE INim

@-v

f-@ il; W.-

C.H. Fletcher, I.T. Stewart,
T.A. Schroeder, G.M. Barnes,
B.M. Richmond, HA. Krock,
and D. Neill
7

U.S. DEPARTMENT OF COMMERCE NOAA
COASTAL SERVICES CENTER
2234 SOUTH HOBSON AVENUE
CHARLESTON, SC 29405-2413
ESTIMATION ANALYSIS
AND MAPPING OF
COASTAL OVERWASH ON KAUA'I I
DURING THE LANDFALL OF
HURRICANE INIKI

C.H. Fletcher, I.T. Stewart
Geology and Geophysics, SOEST, University of Hawai'i,
Honolulu, HI 96822
T.A. Schroeder, G.M. Barnes
Meteorology, SOEST, University of Hawai'i, Honolulu, HI 96822
B.M. Richmond,
US Geological Survey, 345 Middlefield Rd., Menlo Park, CA 94025
H.-J. Krock, D. Neill
Ocean Engineering, SOEST, University of Hawai'i, Honolulu, HI 96822

Prepared for:

Hawaii Coastal Zone Management Program
Office of State Planning
Office of the Governor

National Coastal Geology Program
United States Geological Survey
Department of the Interior

National Science Foundation

October, 1993

A publication of the Hawaii Office of State Planning pursuant to National Oceanic and
Atmospheric Administration Award No. NA36OZ0000-01.

Property of CSC Library

COVER: The marine overwash of Hurricane Iniki on the southern coast of Kauai left a line
of debris behind this condominimum complex near Kukuiula Harbor. The lower floors,
and the foundation, were damaged by marine floodwaters. The upper floors and roof were
stroyed by wind. Beach erosion exposed soff layers that caused coastal turbidity for
several weeks following the storm.

Coastal Overwash During Hurricane Iniki

COASTAL, OVERWASH ON KAUA'i DURING HURRICANE INIKI

TABLE OF CONTENTS
List of Figures .............................................................................................3
Executive Summary ....................................................................................4
Introduction ..................................................................................................6
Storm History ..............................................................................................9
Tuesday, September 8 ............................................................................9
Wednesday, September 9 ....................................................................... I I
Thursday, September 10 ......................................................................... 12
Friday, September 11 ............................................................................. 13
Marine Conditions ....................................................................................... 16
Initial Set-up .....o....................o ................................................................ 19
Pressure Set-up ....................................................................................... 19
Estimation of RMW and Vemax ......................................................... 20
Wave Set-up .......................o ................................................................... 26
Wind and Bottom Stress Set-up ............................................................. 28
Wave Run-up ....................................................................o .................... 29
Discussion ..............................................................................o ............... 30
Mapping the Overwash ........................................................o ...................... 33
Methodology .......................................................................................... 33
Overwash Conditions ............................................................. 34
Kekaha ...............................................o ................................................... 36
Waimea .............o.................................................................................... 40
Hanapepe ................................................................................................ 42
Kukuiula to Keoniloa Bay ...................................................................... 44
Kapaa .....................................o.....................o ......................................... 50
Other Areas ............................................................................................ 52
Hanalei Bay ....................................................................................... 52
Na Pali Coast .......................o ............................................................. 53
Barking Sands ........................................................o .......................... 53
Pakala Village ..............................................................................o .... 53
Poipu to Nawiliwili ........................................................................... 53
Nawiliwili to Wailua ......................................................................... 54
Kapaa to Princeville .......................................................................... 55
Environmental Response ............................................................................. 56
Beach Erosion .....................................................................o .................. 57
Poipu Offshore Survey ........................................................................... 59
Coral Impacts .................................................................................... 59
Debris Build-up ................................................................................. 62
Field Observations ............................................................................ 62
Recommendations.. ........................................................................... 63
References ................................................................................................... 64

Coastal Overwash During Hurricane Iniki

COASTAL OVERWASH ON KAUA'I
DURING HURRICANE INnu

LIST OF FIGURES
Figure 1. Overwashed and wind-blown sand ..............................................5
Figure 2. An 80-unit condominium complex ..............................................7
Figure 3. The Bull Shed Restaurant ............................................................8
Figure 4. Houses in the Kukuiula Harbor area ............................................ 10
Figure 5. Path of Hurricane Iniki ................................................................. 11
Figure 6. Map of Kaua'i ............................................................................... 14
Figure 7. Tide gauge records ....................................................................... 18
Figure 8. Radial profiles of wind speed ...................................................... 21
Figure 9. NOAA Hurricane Research Division ........................................... 22
Figure 10. Overwash on an open field, Waipouli ....................................... 32
Figure 11. 'Me overwash debris line ........................................................... 35
Figure 12. Overwash map of Kekaha .......................................................... 37
Figure 13. The end of the Kekaha revetment .............................................. 38
Figure 14. Overwash was channeled down coastal avenues ....................... 39
Figure 15. Overwash map of Waimea ........................I ................................. 41
Figure 16. Overwash damage to a chain-link fence .................................... 42
Figure 17. Overwash map of Hanapepe ...................................................... 43
Figure 18. Overwash debris line near Kukuiula Bay .................................. 45
Figure 19. Overwash map between Kukuiula Bay and Keoniloa Bay ........ 46
Figure 20. Coastal damage between Kukuiula Bay and Keoniloa Bay ....... 48
Figure 21. Coastal damage at Poipu Beach Park ......................................... 49
Figure 22. Overwash map of Kapaa coast ................................................... 51
Figure 23. An overwash fan ........................................................................ 54
Figure 24. The Na Pali Coast suffered little damage .................................. 55
Figure 25. Erosion at Waipouli beach. Accretion at Mahaulepu ................ 58

Coastal Overwash Duzing Hunicane Iniki

-EXECUTIVE SUMMARY-

Hurricane Iniki was a minimal Category 4 subtropical cyclone that
crossed Kaua'i September 11, 1992. The eye made landfall at Kaumakani at
approximately 3:20 pra and departed the north shore by 4:00 pm. The loss of
life was miraculously minimized. Economic losses from high winds and
marine overwash rank the storm as the fifth costliest insured catastrophe in
U.S. history and the worst recorded storm in Hawaiian history.
This is a study of the marine overwash of Kaua'i during Iniki. The
goals of the study are to provide a numerically estimated description of the
meteorologic and oceanographic environment that contributed to the
overwash, to map the overwash line at significant points around the island,
and to describe impacts to the offshore environment of the south shore. The
research was supported by the Hawaii Coastal Zone Management Program
pursuant to N.O.A.A. Award No. NA360ZOO22-01, the National Coastal
Geology Program of the U.S. Geological Survey, and the National Science
Foundation.
Analysis of the initial set-up and astronomical tide, the pressure set-
up, wind and wave set-up, and wave run-up provides a description of
overwash components. These are compared to observed overwash elevations
at two areas on the southern Kaua'i coast, near Kukuiula Bay, and on the
Poipu-Koloa coast. Our estimations match the field observations.

Location Initial S Sww ASw Ruv Total Obs d
Kukuiula 1.8ft 1.3?ft 5.3 ft 0.2 ft 9.3 ft -18 ft 17.2 7
Koloa-Po ipu 1.8 ft 1.21 ft 5.3 ft 0.2 ft 14 ft 22.5 ft 19.7-27.8 ft

Coastal Overwash During Hurricane Iniki

Overwash maps are provided and discussed for five localities:
Kekaha, Waimea, Hanapepe, Kukuiula to Keoniloa Bay, and Wailua to
Kapaa. A continuous overwash line along the south and east coasts of Kaua'i
is archived in. the State of Hawai'i Geographic Information System. GIS
basernaps for the five regions displayed here depict the FEMA FIRM V-
Zone and A-Zone, as well as the 1986 GIS shoreline, and Iniki overwash.
Diver surveys between Kukuiula to Makauena Point determined that
no significant structural-debris accumulation is found in the marine
environment. Comparisons of coral-distribution data spanning 20 yr indicate
a one-third reduction in total coral coverage along the 30 ft depth contour.
Along the 70 ft contour, total coverage has not changed, but previously
dominant encrusting genera have been replaced by genus Pocillorpora.

0(41

lFigure 1. Overwashed and wind-blown sand on the coast road at.Kekaha.

Coastal Overwash During Hurricane Iniki

-INTRODUCTION-

Shortly after 3:00 PM (HST) on September 11, 1992, Hurricane Iniki
made landfall on the south shore of the island of Kaua'i. The eye passed the
Kaua'i coast near Kaumakani (-5 km west of Port Allen) with devastating
consequences. Sustained winds of about 130 mph, gusting to 160 mph, and
extensive coastal flooding commonly exceeding 10 ft in elevation, destroyed
or damaged 14,350 houses. Of these, 1,421 were destroyed, and another
5,152 suffered major damage. Statewide, an additional 607 houses sustained
damage or were destroyed. Six deaths are attributed to the storm, and over
one thousand injuries reported. A year later, the Property Claims Services
Division of the American Insurance Services Group, and the Insurance
Information Institute, ranked the $1.6 billion (1990 dollars) in losses from
Iniki as the fifth most costly insured catastrophe in U.S. history.
The principal goal of this report is to provide maps of the overwash
pattern on Kaua'i in and around the population centers of Kekaha, Waimea,
Hanapepe, Kukuiula to Poipu, and Kapaa. We also describe environmental
and meteorological factors that influenced overwash impacts in the coastal
zone of Kaua'i. Principal attention is given to the south shore of the island, in
the right forward quadrant of the storm as it made landfall. It was here that
overwash was most intense. Using field investigations, marine surveys, and
aerial photograph analysis, the overwash zone was mapped around the
island, and digitally recorded in the Office of State Planning Geographic
Information System. We also report on weather and marine conditions
during the overwash and discuss impacts to the offshore coral community.
Separate reports on various aspects of the Iniki tragedy are emerging
from various concerned economic and scientific sectors of our society. Many

Coastal Overwash During Hurricane Iniki

A
LIN,

Mont

AW 4
-#AW

40
A .4@e"

-40

" rla@

Figure 2. An 80 unit condominium complex, lower floors destroyed by
overwash, upper floors destroyed by wind. Note the overwash debris line in
@ 7-fw

the background. East of Kukuiula Bay, Kaua'i (see cover).

Coastal Overwash During Hurricane Iniki

of these have overlapping coverage, and, we have found, to some degree
report conflicting details. To the extent possible here, we utilize the most
reliable and reasonable of these in our o w*n. descriptions. Our approach is to
estimate, through field data and simple numerical calculations, the
meteorological and oceanographic processes that both mitigated and
exacerbated the marine overwash of Kaua'i during Hurricane Iniki.
At this writing, one year later, Kaua'i still struggles to recover her
lifes tyle and economy. It our sincere hope that the contributions in this
report provide some lessons that may help mitigate future levels of tragedy.

J
M U@l

ago,

[FiVure 3. The Bull Shed Restaurant, Kapaa. Overwash reached >2 m.
@t L-1

Coastal Overwash During Hurricane Iniki

-STORm HiSTORY-

Iniki was first identified as Tropical Depression (TD) 18-E
approximately 2690 km southwest of Baja California over the warm waters
of the eastern Pacific near , 120N 135OW on September 5, 1992.
Meteorologists at the National Weather Service in Silver Spring, Maryland,
and at the Central Pacific Hurricane Center in Honolulu, speculate that the
storm may have originated in August as a mid-tropospheric tropical wave
off the coast of Africa. An area of disturbed weather was tracked by the
National Hurricane Center across Central America for several days when it
moved into the Pacific on August 28. TD 18-E crossed 140OW and moved
into the Central Pacific on the morning of September 6 heading west-
northwest at 7 m/sec with maximum sustained winds of 15 m/sec. Initially
expected to dissipate, TD 18-E did weaken under unfavorable wind shear,
but by mid-day on September 7 the system became imbedded in a deep
easterly flow along the southern edge of the seasonal subtropical high-
pressure ridge extending from 40ON to 450N latitude between 130OW and
170OW longitude. By 5 pm that day, TI) 18-E was upgraded to Tropical
Storm Iniki moving west at 4.5 m/sec with maximum winds of 18 m/sec.
Tuesday, September 8 - Iniki slowly intensified as it moved west
with maximum sustained winds of 20 m/sec. Forecast to head west-
northwest, Iniki continued west at 7 to 7.5 m/sec throughout the day as
winds strengthened to over 30 m/sec with gusts to 38 m/sec by 5 pm. At 11
pm, Tropical Storm Iniki was upgraded to Hurricane Iniki at a position of
13.20N 152.10W moving still west with maximum sustained winds of 34
m1sec, gusting to 40 m/sec. With a forward translation speed of 7 m1sec,
Iniki's central pressure slipped to 992 mb as the hurricane assumed a new

Coastal. Overwash During Hurricane Iniki

4ft@

44k,

4W .Vqk

Ail

FIGURE 4. Houses in the Kukuiula Bay area, on the south coast of Kaua'i,
were subjected to gusts reaching 64 m/sec and overwash reaching elevations
4.5 to 5.2 m. Houses were swept from their foundations and carried inland as
much as 100 m by the inundation. Those buildings that remained in place
were battered by boulders and debris carried in the raging sea, and flooded
through their lower levels. Simultaneously, the wind ravaged the uppermost
floors. Demolished structures along the waterline, such as seawalls , poured
concrete lanai's, and paving stone acted as battering rams when carried by
e overwash. Several homes sited on bluffs near. Koloa landing were
damaged @y @reaking waves reportek exceeding 8 m in height.

10.

Coastal Overwash During Hurricane Iniki

9-13-92 5:00am
65/80 mph
Downgraded to
tropical depression P T H 0 F,

9-12-92 5:00prn National Weather Service
75/90 mph 990 mb Pacific Region
300- Extra-tropical in 17 'hours

P a c i f i c
(9-11-92) 11:00prn 0 c e a n
115/145mph 960mb
Storm weakening
0 (9-11-92) 5:00pm 130/160mph 945mb
Conditions on Kaua'i improve
(9-11-92) 3:00pm
130/160mph 945mb Kaua'i
Landfall imminent @%O'ahu
200 - (9-11-92) 11:00 am Maui Hawai'i
145/175mph 940mb
Maximum strength 0@
10-92) 1 1:00prn
0/155mph 947mb
(9-10-92) 5:00am auai on alert (9-8-92) 11:00prn
115/145mph 960mb 75/90mph 992mb
Turns sharply northward Upgraded to hurricane (9-7-92) 5:00pm -
Iniki= "piercing wind" 40mph 1002mb
(9-9-92) 5:00prn Tropical storm
100/120mph 980mb
Rapidly gaining strength
High surf advisory for all islands

30'l I I I I I
N 160OW 1500 1400

Figure 5. Path of Hurricane Iniki, and significant episodes in the evolution
of the storm (National Weather Service).
heading to the west-northwest by 5 am, the next day.
Wednesday, September 9 - At 6 am, high surf advisories for the
south-facing shores of the Big Island were issued. Predictions called for
waves between 1.2 and 2.5 rn high. When the swell eventually arrived, 11

Coastal Overwash During Hurricane Iniki

hours later, it measured 2.5 to 3.5 m. By I I am, sustained winds had
strengthened to 40 m/sec and gusts exceeded 49 m/sec. Expected to
strengthen still further, the system continued tracking west-northwest and by
late afternoon was approximately 484 km south of South Point, Hawail
hai was now heading west-northwest at 6.3 m/sec with maximum sustained
winds of 45 m/sec, and gusts to 54 m/sec.
At 5 pm, a high-surf advisory was issued for the south-facing shores
of all Hawaiian islands. By 11 pm on September 9, a USAF reconnaissance
plane found a "small intense hurricane with gale winds out 240 km in the
north semicircle." At this point Iniki was gusting to 56 m/sec, and central
pressure had dropped to 980 mb.
Thursday, September 10 - Iniki increased in intensity (sustained 52
m/sec, gust 65 m/sec, 951 mb) but did not change from its heading of a day
earlier. This situation was similar to previous hurricane systems that had
ventured near Hawai'i. Had Iniki continued on its path toward the western
edge of the Central Pacific subtropical high pressure ridge at 145oW, it
would have never approached the State any closer than its earlier position
south of the Big Island. However, further west, a large upper low pressure
system and cold trough had moved southward along and just east of the
dateline. A low pressure area associated with this movement developed near
30ON just east of the trough. T'his was destined to turn Iniki to a more
northerly track. The position of the storm at the time it turned and the path it
followed would determine the level of threat to the islands of 0'ahu and
Kaua'i.
During the morning hours of September 10, Iniki, located 675 km
south-southwest of Honolulu, ominously slowed its forward motion to 5
m/sec and maintained its earlier intensity. Throughout the afternoon,
however, conditions changed as the storm grew in strength and moved on a
new heading to the north-northwest. By 5 pm on September 10, Iniki turned
to the northwest. Having continued to grow in intensity with top winds of 56
m/sec, gusting to 69 m/sec, the center was now located near 160N and
160oW, 645 km south of the city of Lihue on KauaT
A hurricane watch was issued at 5 pm for islands from Kaua'i to
French Frigate Shoals. Reconnaissance plane and satellite pictures indicated
a strengthening of the storm, and forecasters predicted the storm would
follow a northwest track at about 4 m/sec. A National Weather Service news

12.

Coastal Overwash During Hurricane Iniki

release at 5 prn forecast a slow weakening of the storm as it turned on a more
northerly track, eventually passing 320 km west of Kaua'i within 48 hours.
Surf was reported subsiding on the Big Island, and building to 2.5 to 3.5 m
elsewhere. The south shore of Kaua'i was warned of sustained, rough surf
that would damage shoreline property by Friday. By 8:30 pm the aler 't was
elevated. A Hurricane Warning was issued for Kaua'i and Ni'ihau, a Tropical
Storm Warning for 0'ahu, and a Tropical Storm Watch for Maui Co.

"Iniki has turned on a more northerly track. Eye forecast to move close to
Kaua'i and Ni'ihau. Position 16.80N and 159.50W with max. winds 125
mph and gusts to 155 mph. Iniki moving NNW 14 mph."
National Weather Service Release, 8:30 pm HST Thu. Sept. 10, 1992

Within two and a half hours the alert status was elevated again. This
time 0'ahu was upgraded to a Hurricane Warning.

"Hurricane Warning issued for 0'ahu ... Upgraded from
Tropical Storm Warning. Hurricane warning continued for Kaua'i and
Ni'ihau. Hurricane Watch continued for islands west of Kaua'i to French
Frigate Shoals. Position 17.50N, 160.OOW moving north 15 mph with max
winds 125 mph and gusts to 155 mph. Tropical Storm Watch remains in
effect for Maui Co."

National Weather Service Release, 11:00 prn HST Thu. SepL 10, 1992

Friday, September 11 - Throughout the night of Sept. 10, and into
the early morning of Friday, Sept. 11 Iniki accelerated and strengthened as it
continued to turn to the north. All warnings and watches remained
unchanged as Iniki marched north at 6 m/sec with maximum sustained winds
of 65 m/sec, and gusts to 78 m/sec. By dawn on Sept. 11, residents of 0'ahu
and Kaua'i, who had been awakened earlier by several punctuated blasts on
the public siren system, were securing their homes and listening to
continuous media coverage on the progress and predictions of Hurricane
Iniki. "Life-threatening" surf of 9 ni or more was predicted for Kaua'i and
Ni'ihau by the evening. Building surf and waves of 3 to 6 m were forecast
for the south and west-facing shores of 0'ahu. Torrential rains and major
flooding were expected on Ni'ihau and Kaua'i by nightfall.

13.

Coastal Overwash Dtifing Hunicane Iniki

Departure
-4:00 prn HST
Haena Princeville

Kilauea

Makaha Ridge Kokee

Kapaa
Map 5
Wailua River

Barking
Sands
Lihue

Kekaha Waimea
Nawiliwili
Map 1 Kalaheo Harbor
Map
Kaumakani.j
Landfall Hanapepe port Poipu
-3:20 prn HST Allen Kuku la oloa Makahuena
t ap 3 LandingI Point
Map 4
10

0 lorni

0 10km

FFigure 6. Mal) of Kaua'i, and path of Hurricane Iniki.
By late morning, 11 am, Iniki had reached maximum recorded
strength due west of Maui and 210 krn south-southwest of Lihue.
Reconnaissance aircraft sent back reports of sustained maximum winds of 65
m/sec gusting to 78 m/sec, and a central pressure of 938 mb, the lowest ever
recorded in a Central Pacific Hurricane. The eye diameter was only about 33
kin, a small tight center with intense energy. On Kaua'i, the first waves were
reported crossing the coastal highway at Kekaha and Poipu. Winds had built

14.

Coastal Overwash During Hurricane Iniki

to 18 to 22 m/sec between Kekaha and Kalaheo. On O'ahu, the Civil Defense
ordered that all persons residing within "300 ft" of all shores evacuate to
higher ground. Flash flooding was forecast for 0'ahu, Kaua'i, and Ni'ihau.
Iniki moved northward toward Kaua'i throughout the day. A series of
slight wobbles in the movement of the eye led to short-term speculation. that
the storm had turned to the east and was on a path for the Kaua'i Channel,
between the Islands of O'ahu and KauaT These reports came through the
media at about the same time that power was being lost to much of O'ahu,
and the Honolulu waterfront was being overwashed in places to an elevation
of nearly 2 m. As power failed, many residents of O'ahu spent the afternoon
fearing that Iniki would pass close enough to cause major damage to the
entire island. However, by 3 prn Iniki was 60 km southwest of Lihue and
moving north, poised to strike the coast of Kaua'i.
As Iniki made landfall west of Port Allen on Kaua'i shortly after 3 pm,
the eye was only 18.6 km wide. A small eye diameter normally indicates that
strongest winds are located near the eyewall (Trapp, 1993). Unfortunately'.
no usable wind data is available from near the eye. The NOS tide station at
Port Allen did record a low pressure of 960 mb at 3 pm, and 968 mb at 4 pm.
Extrapolation of the hourly pressure record suggests a minimum pressure of
about 952 mb at Port Allen (NWS, 1993). Twenty kilometers to the east, at
Makahuena Pt., an anemometer recorded sustained winds at 36.3 m/sec
before it failed. A peak gust of 64 m/sec was later extracted from the digital
data recorder at Makahuena. At Lihue, 32 km east of the eye as it passed
over the island, sustained winds of 43.5 m/sec (3:52 pm) and a peak gust of
57.8 m/sec: (3:02 pm) were recorded.
By 5 pm, having left the island along a path between Haena and
Wainiha, Iniki was already 80 km north of Kaua'i, and the Hurricane
Warning for O'ahu was downgraded to a Tropical Storm Warning. In
confusion, some 'officials concluded that this new warning heralded the
approach of a second storm. All warnings for the State were either cancelled
or downgraded by 11 pm as Iniki sped on to the north. High Surf Advisories
were maintained the following day for south and west-facing shores of all
islands. September 12 dawned with clear skies as citizens of Kaua'i faced the
task of recovering from the worst recorded catastrophe in the State's history.

15.

Coastal Overwash During Hurricane Iniki

MARINE CONDITIONS -

Coastal overwash results from the movement of a hurricane across the
ocean surface when it encounters land. Flooding of the coast is the product
of a number of meteorological and oceanographic factors related.to:

-Initial set-up of the water from either an early arriving long wave, or
local sea-level anomalies, plus phasing of the astronomical tide
-Pressure set-up from decreased atmospheric weight on the water column
-Wind set-up from the shear of the winds across the water surface
-Wave set-up effects inside the breaker zone
-Wave run-up controlled by bathymetry, topography and surface
roughness
-Shape of the coastline, and bathymetric morphology

Other factors adding to overwash, such as rainfall and the Coriolis
effect are usually considered negligible on open ocean coasts lacking run-off
collection basins such as estuaries and lagoons, or shoreline features
perpendicular to the mass movement of deflected water.
Typically the region in the right front quadrant of a moving storm
experiences the greatest wind shear because the speed of the storm adds to
the actual wind velocity. When a storm crosses a shoreline, or moves
adjacent, the combined effects of storm surge and high waves can cause
flooding, called overwash. Overwash can lead to coastal erosion, loss of soil
fertility and standing-water quality by salinization, damage to buildings and
transportation networks, interruption of power and communication grids,

16.

Coastal Overwash During Hurricane Iniki

bodily injury and loss of life, and crop destruction. Numerous secondary
effects may result from overwash, including landsliding, excessive coastal
turbidity with deleterious effects on the coastal ecosystem, and even elevated
seismic risk in seismically active regions associated with the dramatic
pressure changes. The decrease in the weight of air in a hurricane can.be as
much as 2-3 million tonnes per kM2of land over a matter of hours (Bryant,
1991). Overwash on the order of 8-10 m in height can lead to an increase of
pressure on the earth's surface of 5-7 million tonnes per kM2. Areas such as
the south shore of the Big Island may be under enhanced seismic risk during
landfall of a hurricane.
In the following estimation analysis we attempt to unravel the various
components that additively comprise the rise in sea-level that led to flooding
of the Kaua'i shore in the region between Kukuiula and Poipu. More detailed
descriptions are provided in Agrawal (1993) and, based on his analysis,
Bretschneider et al. (1993). We also refer the reader to the excellent
modeling of Sea Engineering and Bretschneider (1986). Their worst case
scenario predicts the overwash experienced on this coast during Iniki. A
report currently being generated by the U.S. Army Corps of Engineers also
tackles this problem (Yamamoto and Sullivan, 1993). These references
provide the reader with more detailed descriptions of the numerical
approach.
Our assumptions simplify our findings. We use a maximum deep
water wave height, maximum wind speed, and we reduce the complicated
coastal zone experiencing overwash to a uniform, rough slope extending
approximately from the wave break point to the upland 6 m contour. Our
analysis is only applied in two locations, east of Kukuiula Harbor, and
between Koloa Landing and Poipu Beach Park. It is not our purpose to
produce a new numerical model of overwash, rather it is our goal to
understand the relative components of overwash. In so 'much as our final
estimations of overwash are close to observed values, our analysis has
achieved its goal.

17.

Coastal Overwash During Hurricane Iniki

PORT ALLEN - HURRICANE INIKI
AS RECORDED AT NOS TIDE STATION
1020
OR]) M GH. FID] 1: 6@. FT A'L
f" 6 IM0
L C
5
L
4 990
0 0

3 980

2 970

960

0 950
aaw (pw mw i@ i@ m@ mw mw uw i&w 1@ um c@
TBM (UTC) 11-12 SEFIF 1992

NAWILIWILI - HURRICANE INIKI
AS RECORDED AT NOS TIDE STATION
7

R EU TWE: 5.8FMFJ T

0 3

2 A

0
lz,m L
@m 019 L'rW IM M9 MW M.00 D&P, MOD ZM =59 MM OLM MM ffiM 000 ftW mw am 07:9 Gom Um I" 11M
TIME IN UTC (11-12 SEPT 1992)
AQUATRAK SENSOR - BACK-LJP PRESSURE SENSOR
AQUAMARSENSOR REACRFDBL4NZMVG ZONEATABOUY5.0 FEET

Figure 7. Tide gauge records from Port ARen and Nawildwili duflng In
I xe @AA

18.

Co:astal Overwash During Hurricane Iniki

INITIAL SET-UP

If a decrease in atmospheric pressure should suddenly be applied to
the ocean surface, a wave will be generated with a celerity that is depth
dependent. Over the open sea the velocity of propagation of such long waves
is 180 to 270 m/sec and thus the storm cannot keep pace. This long wave
arrives at the coast as a "forerunner", and may raise the still-water level as
much as 0.5 rn or more. Sea Engineering and Bretschneider (1986) argue
that the forerunner should not be treated as an explicit term because it is
accounted for in the wave set-up and pressure set-up. In Agrawal (1993) it is
treated with the astronomical tide. Values for initial set-up on the Texas
coast have been found to be 0.6-0.75 m (Marinos and Woodard, 1968).
Initial set-up values on the Atlantic coast are typically lower than either Gulf
coast or Pacific values.
Because the forerunner is poorly understood, and cannot be accurately
predicted, we will express it imbedded within the astronomical tide
component of the sea-level. Based on the available tidal hydrograph data
(Fig. 4), Bretschneider et al. (1993) estimate a value of 0.55 m above mllw
(local Hawaiian datum) for the combined tide and forerunner at Port Allen at
the landfall of Hurricane Iniki. In the hurricane vulnerability study of Poipu,
Kaua'i, Sea Engineering and Bretschneider (1986) use a value of 0.52 m (1.7
ft) for their scenario hurricanes because of the frequency of occurrence of
this tide level. Landfall of the eye of Hurricane Iniki unfortunately coincided
with the peak diurnal astronomical tide for the day, thus the resulting
overwash was enhanced by the phased. tide and surge. As Iniki crossed the
coast of Kaua'i, then, the initial water-level was at or near a maximum, and
damage was exacerbated.

PRESSURE SET-UP

The water-level increase, or set-up (Sp), due to a pressure variation
(Ap) on the surface of the ocean is a function of the pressure drop from the
periphery to a point within the hurricane (Ap/p,,g), where p,, is the density
of seawater (1025 kg/M3). Myers (1954) developed an empirical relation for
the surface pressure distribution based on an analysis of historical data.

19.

Coastal Overwash During Hurricane Iniki

Pa-Pr @ [(pa-p,) 100]( 1-e -Rmw/r)

Where Pa is the ambient pressure (assumed to be 1013.3 mb), p, is the
pressure at any distance r from the center, and pc is the central pressure of
the storm (945 mb). RMW is the radius of maximum sustained wind, usually
somewhere in the eyewall. Thus, the set-up becomes,

SP [(Pa-Pr)(l00)APswg)1 (2)

[(Pa-Pc)(l00)APswg)1 (I- e -RMwIr) (3)

Using (2), the set-up under the center of Iniki, where Pr= 945 mb, becomes
0.68 m (2.2 ft). However, to calculate the pressure set-up outside the center
of the storm requires an estimate of the distance of maximum sustained
winds, RMW, from the center.

Estimation of RMW and Vemax

The clear eye radius is approximately 9.3 krn (reported diameter, 10
nautical miles at sea prior to landfall; Trapp, 1993). Maximum winds, Vema,
will be located an unknown distance beyond the eye, within the wall at the
right frontal quadrant of the storm.
An approximation Of Vema,, and its location, RMW, begins with a
theoretical consideration of the cyclostrophic. structure of a symmetrical
vortex. The ratio of the square of maximum velocity to the radius of
maximum winds (Vgma.,2/RMW) is estimated by the inverse of atmospheric
density (1/1.1 kg/m3) and the pressure gradient across some representative
distance (Ap/Ar). Dropsonde data near Kaumakani records a central pressure
of 945 mb, while 5 krn away at Port Allen, the extrapolated pressure record
suggests a minimum of 952 mb (NWS, 1993). If we choose a distance of 12
km for the radius of maximum winds, an estimate of the sustained wind
around the vortex center is given by

20.

rpm
Coastal Overwash During Hurricane Iniki

(VOmax2)/(RMW) = G/Paun) (AP/Ar) (4)
VOmax = 1) (952-945) (100)) 12000] 0.5 (5)
5000
r = 40.98 m/sec

PR 80 +200
U
MOO GMT) (0640 GMT)'
70 - -+too
U
r 60 - - 0
3>
C:)
'j; 50 - - -1100 C-
C:

Lw'j 4 0 - -200

3>
r

R- 30 - -300

20 - HURRICANE ANITA Vorl= CONST - -400
SEPT. 2, 1977
ALTITUDE 30Q M X=.5 - -500
to-. REL. WINDS ........ X z.6
D VALUE
01 1 - I I 1 1 1-600
150 too 50 0 50 too 1150
RADIAL DISTANCE (KM)

Figure 8. Radial profiles of wind speed (m/sec) and pressure in Hurricane
Anita. Note increased velocity to right of eye. Also shown are plots of the
Modified Rankine Vortex model, V@ r W for x--0.5 and 0.6 (Anthes, 1982).
This is the stationary velocity. Winds in the right frontal quadrant
would have an additional translation speed (Fig. 8). There is some dispute
about the actual speed of Iniki as in crossed the south coast of Kaua'i.
According to the Iniki Disaster Survey Rep ort (NWS, 1992) the storm had a
forward speed of "30 mph" (about 13.4 m/sec). However, in its 3 pni HST

21.

Coastal Overwash During Hurricane Iniki

news release, prior to landfall, the National Weather Service citing
reconnaissance aircraft'data reported that "Hurricane Iniki was speeding
up(,) moving north at 21 mph" (about 9.4 m/sec). Thus, the effective
maximum sustained velocity, Ve,,., was in the range of 50.4 m/sec to 54-38
rn/sec (52.4�2 rn/sec, for an average translation speed of 11.4 m/sec).

NOAA Hurricane Research Division Surface Wind Analysis
Hurricane Iniki 12 Sept, 1992 0006 UTC
based on data between 1 10OUTC 11 Sept.-0600UTC 12 Sept.
empirical adjustment of Air Force recon winds from 10,000 ft.
ON
30
for over-
2 3. 5 -P'reli hary
-expos u re

3. 0

-17

0-
. . .... -70 -

1 0

:::20 5

161.0 160.5 160.0 159.5 159.0 158.5
Maximum 1, min sustained surface winds 10 kt interval

............ ..... ........

Figure 9. NOAA Hurricane Research Div. surface winds at landfall using
empirically adjusted reconnaissance aircraft data (Powell, pers. comm.).

22.

Coastal Overwash Duzing Hunicane Iniki

An independent estimation of maximum winds can be performed by
several means. Reconnaissance aircraft reported observations of maximum
sustained flight-level winds of 65.5 m/sec (127 knots, 146 mph) south of
Kaua'i. Powell (1987), in an analysis of the wind structure of Hurricane
Alicia (1983) at landfall, suggests a conversion ratio of 0.78 for estimating
surface winds from reconnaissance flight-level data. This suggests a value of
51.1 m/sec (99 knots, 114 mph) for maximum sustained winds at or near
landfall. Figure 9 shows the preliminary surface wind analysis using
empirically adjusted reconnaissance aircraft data (Powell, pers. comm.). This
is a close match to the estimate of'Vem,,., given above.
An alternate method utilizes the Atlantic Empirical Equation, which
relates the difference of ambient and observed central pressure to the
velocity of maximum wind (Vemax=14 [1013.3 mb - 945 mb]0.5) based on
historical observations in the North Atlantic basin. This provides a value of
59.6 m/sec (115.7 knots, 133 mph).
Our estimation of 52.4 �2 m/sec for Vom. is based upon a number of
stated assumptions, each of which may legitimately be challenged.
Nonetheless, the value is a reasonable one, and each of the assumptions, by
implication, are therefore within the range of possibility. In addition, our
purpose in estimating Vomax is towards the goal of increasing our
understanding of the factors influencing marine overwash, and not motivated
by pure meteorological curiosity alone.
A number of variables in this evaluation can be modified to allow for
a range of Von,,, estimations. Tle extrapolation of the pressure trace at Port
Allen to a value of 952 mb should not be accepted without qualification. It is
not a directly measured value, nor is it theoretically derived. In fact, the
assumptions upon which it is based were not reported. It is, rather, a "best
guess" of what a continuous record might have provided. In addition,
equation (3) extrapolates the pressure slope between the center and Port
Allen as representative of the entire pressure field on the right side. Is this
appropriate? Unfortunately, no other data to constrain the pressure field is
available within the RMW. Likewise, the choice of 12 kin for the radius of
maximum winds is a "best guess", arrived at iteratively, and should not be
accepted uncritically. The Atlantic Empirical Equation suggests a Vomax of
59.6 m/sec. Hindcasting with this value using (3) suggests that the radius of
maximum winds is greater, on the order of 15 km for a translation speed of

23.

Coastal Overwash During Hurricane Iniki

13.4 m/sec. This is also an acceptable result. So the derivation provided here
is not a unique solution. Lastly, as we have stated, the speed of the storm is
under question. The NWS (1993) estimates that Rai crossed Kaua'i between
3:20 and 4:00 pm on Sept. 11, 1992. Covering a distance of 22.2 miles in 40
r minutes equates to about 30 mph, giving some credence to the higher
number. Nonetheless, we choose to take an average of the two reported
r values.
An additional test for our assumptions is provided by the data
available from Makahuena Point. A maximum gust value of 124 knots was
extracted from the digital recorder there. Powell's (1987) analysis of the
landfall of Hurricane Alicia provides a useful gust factor of 1.64, suggesting
that maximum sustained winds at Makahuena Pt. were about 75 knots (39
m/sec; 87 mph). By calibrating the Modified Rankine Vortex Model
(Anthes, 1982) with Vemax, we estimate the maximum sustained winds.

Ve r (.5-.6) = Iniki Constant = (40.98) (12,000)0.55 (6)
= 7179.95 m2/sec
Vo = 7179.95/ r (.55) (7)

Makahuena Pt. is a distance (r) of about 20 krn from the center.
Equation (7) predicts a Ve of 30.9 m/sec at this location, which becomes
42.3�2 m/sec for a translation speed of 11.4�2 m/sec. Thus, estimates of Vo
at Makahuena Pt. using Powell's gust factor (39 m/sec) and the calibrated
vortex model (42.3�2 m/sec) correlate to within 8%, a reasonable fit.

24.

Coastal Overwash During Hurricane Iniki

Let us return to a consideration of the hydrostatic pressure set-up (SP)
at points away from the center of hai. Using a value of 12,000 m for RMW,
with Myers' (1954) pressure distribution relation, we can estimate the
pressure set-up at two representative locations where extensive overwash
occurred in the region from Kukuiula Harbor to Poipu.

Location Observed Overwash 1: (km)
(1) Kukuiula 5.2 m (17.2 ft) 16.5
(2) Koloa-Poipu - 6-8.5 m (19.7-27.8 ft) 19

SP = 1(pa-PAl00)/(pswg)] (1- e -Rmw1r) (3)
S1 = [0.68] (0.52) = 0.35 m (1.2 ft)
S3 = [0.681 (0.47) = 0.32 m (1.1 ft)

This is a time-dependent result however, and it is incomplete until we
include a factor for the depth- and time-relative resonance of the hydrostatic
set-up by a long-wave generated with the moving storm. These waves reach
greatest amplitude when the speed of the storm equals the speed of a shallow
water wave [(gd)0-5], and when the storm has travelled at the required speed
for sufficient time for the wave to completely develop. A factor for the
influence of this wave is used to correct the pressure set-up calculated by (3).
The critical water depths for hurricane speeds of 10, 20, and 30 knots are 3,
11, and 26 m respectively (Sorensen, 1978). Ewing, et al. (1954) report on a
squall moving over the appropriate depth of southern Lake Michigan for 0.5
hr. The hydrostatic set-up (0.03 m) was amplified by water column
resonance and a 1-2 m surge was observed in the path of the squall-
generated wave at Michigan City, Indiana.
Based on a laboratory study, Abraham (1964) suggests the long-wave
surge can approach twice the value given by (3) if a storm moves over water
of between 0.75-1.25 times the critical depth for a period of 1 hr or more,
and triple that value for a depth range of 0.9-1.1 times the critical depth for
the same period. hai moved at speeds between 8.9-17.6 m/sec (15-30 mph)
in the period prior to landfall. The critical resonance depth can be calculated
from [dc, = V2/g]. Thus 8.1 m < dcr > 31.6 m, and 0.75 dcrmin = 6 m, while
1.25 damax = 39.5 m. According to Abraham (1964) if the hurricane remains

25.

Coastal Overwash During Hurricane Iniki

in the depth range 6-39.5 in for an hour then resonance amplification
becomes a factor of 2. The bathymetry of the southern Kaua'i coast is steep,
and these two depths are an average of 2-2.5 km apart. At an average
translation speed of 11.4 m/sec, Iniki would only have remained in this depth
interval for 3-4 minutes. Since the resonance factor of 2 requires Iniki to
have travelled this depth range for an hour, the resonance factor must be
significantly less. But because Iniki was moving, the value must be greater
than 1. We will use a value of 1. 1 for the resonance factor.
Agrawal (1993) defines a resonance factor, R, that is dependent upon
wave celerity and storm translation speed, [(gd)0-5]2/([(gd)0.5]2 -V?). He
chooses a value of R=1.2, but his criteria are not made clear.
Our values of resonance enhanced, pressure set-up are:

Location Hydrostatic Combined (Sp)
(1) Kukuiula 1.2(l.1) 1.32 ft
(2) Poipu 1.1 (1.1) 1.21 ft

WAVE SET-UP

Waves generated by Iniki winds resulted in mass transport into the
shorezone that elevated the water level, called wave set-up. In small-
amplitude wave theory, water-particle motion in the water column is
described with closed orbits, and therefore no net transport of water mass
occurs as waves advance to the coast. In reality, measurements of water level
inshore of the point where waves break reveals that sea levels increase with
larger waves. Tbus, some mass transfer of water into the coastal zone must
occur as a function of wave characteristics. In real waves, there is a small
forward mass transport as water particles advance slightly with each orbit.
The velocity of transport increases as wave steepness and wave celerity
(speed) increase, and as the relative water depth decreases. As mass
accumulates in the breaker zone, currents are generated by the return flow
which may move quasi-uniformly along the bottom, or as rip currents spaced
periodically along the coast, and as longshore currents.moving down a
pressure gradient towards areas of lower set-up or offshore flow.

26.

Coastal Overwash During Hurricane Iniki

Laboratory and theoretical studies (USACE, 1984) suggest that wave
set-up (S,,) can be estimated by
S"'= 0.19 [1-2.82 (Hb/gT2)0.5] Hb (8)

Where Hb is the breaker height in the surf zone, T is wave period. Typically
wave set-up will amount to about 15% of the breaker height. In this analysis
we neglect the influence of wave-set-down. Breaking wave heights of about
9 m were reported in the region of the Poipu coast at the height of the
hurricane. Calculations using solitary wave theory in Bretschneider et al.
(1993) suggest that wave heights prior to breaking at Kewalo Basin, O'ahu
were in the range 7-8.9 m with periods of 14 sec. Offshore Buoy 51002,
located about 280 miles south-southest of Honolulu, recorded a significant
wave height of 6 m, with a most probable maximum deep-water height of
nearly 11 m. (NOS Buoy Data). However, Buoy 51002 is located somewhat
off the path of Iniki and we can conservatively estimate deep-water wave
heights (Hd) of about 12.2 m (40 ft) for positions along the hurricane track.
Agrawal (1993) on the same basis assumes that Hd=45 ft. Equation (9)
derives the relationship between breaker height and deep-water height,

Hb=Hd/3.3(Hd/Ld)0-33 (9)

A deep-water wave length (Ld) can be calculated using (Ld=[g/2n]T2).
Assuming a wave period of 14 sec, Ld= 306 m (1004 ft). Equation (9)
predicts a breaking wave height of 10.7 rn (35 ft). Thus, (8) provides an
estimate of the wave set-up, S,,w, of about 1.6 m (5.3 ft). Pressure set-up and
wave set-up are additive:

Location Initial SP SWW
(1) Kukuiula 1.8 ft 1.32 ft 5.3 ft.
(2) Koloa-Poipu 1.8 ft 1.21 ft 5.3 ft.

27.

CoastaL Overwash During Hurricane Iniki

WIND AND BOTTOM STRESS SET-UP

A stress (ts) is generated as the wind blows across the water surface
that is a function of air density (Pa), wind velocity (U m/sec), and drag (Cd),

Ts=CdPaU2 (10)

Values Of Cd typically vary from 1.5x10-3 for light winds to 2.4x10-3 for
strong winds. Equation (10) is often written relative to water density (p),

Ts= k pU2 (11)

where k =(1.19xl0-3)Cd. It has been shown (Van Dorn, 1953) that,

k=1.21x10-6 + 2.25x10-6 (1-[7.2/U])2

Using our estimation of 52.4�2 m/sec for Vomax, k becomes 3x10-6.
The surface wind stress leads to development of bottom currents that
generate a bottom stress (Tb). Saville (1952) found that Ts/Tb=(3.3x 10-6)pU2.
Assuming that k =3.0X10-6, he suggested that Tb/Tb=0.1.This can be used to
determine a combined surface and bottom stress coefficient K, where
(Ts+Tb=KpU2). Thus, K becomes 3.3x10-6. Across a segment of seafloor
normal to the coast of length AX, the depth changes from d to d+^Sw, where
Sw is the set-up due to wind and bottom stresses acting over the length ^X.
Because the water surface slope (^Sw/^X)is negligible,

Ts^X+Tb^X+0.5gpd2-0.5gp(d+^Sw)2=0

^SW=d[([(2KU2^X)/gd2]+1)0.5-1] (12)

Equation (12) is applied sequential.ly along the bottom profile offshore of the
area of interest. Flat sections of bottom of constant depth, d, and length ^X
are used to calculate ^S,. The cumulative sum of preceeding set-up values
are added to (d) along the profile in the shoreward direction. The highest
^Sw then becomes the wind and bottom stress component.

28.

Coastal Overwash During Hurricane Iniki

Agrawal (1993) provides data on the bathymetry offshore of the
Koloa-Poipu region:

Koloa-Poipu Region
U d (m) AX (in) AS,,
5.5 361 0.06 in
PR 18.3 626 0.03 in
U 91.4 2022 0.02 rn
183 2359 0.01 M
U ASwmax 0.06 m (0.2 ft)

r
U In this way the combined wind stress and bottom stress set-up under Vema,,=
52.4 m/sec becomes about 0.2 ft for the Poipu coast. We win assume the
r same value for the Kukuiula coast.

Location Initial SP SWW ASW
U (1) Kukuiula 1.8ft 1.32 ft 5.3 ft 0.2 ft
r (2) Koloa-Poipu 1.8 ft 1.21 ft 5.3 ft 0.2 ft
r WAVERUN-UP

U The extent of wave run-up on a slope is a function of permeability or
roughness, bathymetric and onshore slope, coastline configuration, breaking
wave height and wave period. In this estimation analysis, we neglect
coastline configuration by assuming the bathymetry consists of straight and
r parallel contours, a common assumption. Although it is not possible to
theoretically predict run-up because of the large number of variables, it is
possible to estimate run-up from a few simple parameters following the
guidelines of some laboratory studies (USACE, 1984).
The following conditions apply: Hd= 12.2 in, T=14 sec, and the slope
(tanO) from the break point to the 6 in onshore contour at Kukuiula is
approximately 0.02, and at Koloa-Poipu approximately 0.03. This approach
also neglects the influence of the water column between the break point and
the shoreline, and treats the run-up as a simple function of slope and

29.

Coastal Overwash During Hurricane Iniki

permeability between the break point and the upland. Equation (13) is a
modified form of Hunt's (1959) empirical formula relating onshore slope,
breaker height and wave period to significant vertical wave run-up
[Ruv=T(gHb)0-5tan0]. Horizontal excursion can be calculated with
r Ruh=Tr(gHb)0-5, where r is a roughness factor.
We have estimated a breaker height (Hb) Of 11 m. Agrawal (1993)
calculates a probability parameter (1.5), using wave-energy analysis applied
r to a modified -Hunt formula. This factor converts from significant wave run-
p to maximum run-up, with a 2% exceedance, under given hydraulic
conditions. Agrawal (1993) and Sea Engineering and Bretschneider (1986)
r u
PR assume a friction value of r=0.85 to estimate the effects of slope
U permeability and roughness which is a value used to describe grassy slopes.
However, the shoreline at both areas of interest is composed of basalt
r boulders, rocky outcrops, seawalls, and various structures. Accordingly we
will use r--0.65.
r Ruv= (1.5)(0.65) T[Hb g]0.5 tanO (13)

Thus, for Kukuiula Ru=2.8 m (9.3 ft), and for the steeper, Koloa-Poipu coast
Ru=4.25 m (14 ft).

DiscusSION

Location Initial Sp Sww ASw Ruv Total Observed
Kukuiula 1.8ft 1.32 ft 5.3 ft 0.2 ft 9.3 ft -18 ft 17.2 ft
Koloa-Poipu 1.8 ft 1.21 ft 5.3 ft 0.2 ft 14 ft -22.5 ft 19.7-27.8 ft

Our estimations provide a picture of the relative contribution of the
various components comprising the overwash of the south shore of Kauai.
Clearly, from the point of view of damage in the coastal zone, the most
important aspects of the overwash are wave induced set-up and wave run-up.
Natural factors that mitigate the wave hazard are a gentle offshore slope,
natural roughness of the coastal zone, and the presence of offshore
topographic barriers such as a barrier reef and submerged headlands and

30.

Coastal Overwash During Hurricane Iniki

rocky outcrops. The wave hazard is exacerbated by a steep offshore slope, a
featureless bathymetry, initial water set-up, and the pressure and wind stress
set-up. Features that mitigate the wave environment under incident
hurricane-force winds should be considered in coastal zone management
decisions.
The shape of the coastline is an important factor. Any sector of the
shoreline that is V-shaped, including submarine topographic channels, will
lead to lateral compression of the storm-surge wave, increasing the wave
height. The size of such a coastal basin can also lead to resonance if it has a
shape that matches the period of any entering wave.
The V-shaped shoreline at Koloa Landing is probably responsible for
the highest level of overwash resulting from Hurricane Iniki. Located
immediately to the east of Nahumaalo Point, overwash into this narrow
embayment, and up the boat ramp and the adjacent Waikomo Stream Valley
reached an elevation of over 22 ft. Immediately east of the boat ramp an
overwash debris line behind a house measured over 26 ft above mean sea
level. Offshore of Koloa Landing, to the immediate west and east, are two
submerged rocky platforms, and between them a low-lying coralline-algal
surface with a relative relief of about 15 ft. This submerged valley, leading
directly to the Koloa Landing area must have channeled and heightened the
storm surge and allowed deeper penetration shoreward of the relevant wave
characteristics.
It is clear that careful evaluation of offshore morphology and slope
should be a consideration in coastal management decisions with regard to
development planning and zoning on shorelines with high hazard ratings.

31.

Coastal Overwash During Hurricane Iniki

-Ir

L
A

Figure 10. Top, overwash on an open field at Waipouli near Kapaa. Bottom,
200m north of the top photo, roadways and houses suffered severe damage.
Overwash causes little concern where development is adequately set-back
jLrom the shore.
04,
IAMO

32.

Coastal Overwash During Hurricane Iniki

MAPPING THE OVERWASH

METHODOLOGY

On Tuesday, September 15, 1992 a research team composed of
personnel from the University of Hawai'i, and the U.S. Geological Survey
Center for Coastal Studies, conducted an aerial reconnaissance of the coastal
zone of Kaua'i to assess the level of marine inundation and damage to the
littoral environment. Videotapes of the entire coast were acquired to assist in
later mapping exercises at an average elevation of 500 ft and a flight speed
of 80 knots. Three days of ground-based surveying followed to identify the
extent, and characteristics of overwash at Kekaha, Waimea, Hanapepe,
Kukuiula to Kawelikoa Point, Wailua to Kapaa, Princeville, and Hanalei.
Personnel from the University of Hawai'i conducted another three weeks of
ground-based investigations at the same locations that autumn.
In February, 1993 a team from the University of Hawai'i conducted a
series of marine surveys between Kukuiula Bay and Makahuena Point to
determine the level of damage to the coral community, and the extent of
debris coverage in the depth range 90 ft to the shore. The results of this study
are described in a later section of this report, and in Krock and Neill (1993).
Aerial photographs at an enlarged scale of 1:1200, and high altitude
NASA infra-red aerial photography provided coverage of all coastal
population centers on Kaua'i. These were used in conjunction with field
observations, and with survey data from the U.S. Army Corps of Engineers
(Pacific Ocean Division), to construct maps of the continuous overwash line
from west of Kekaha along the south and east shore of Kaua'i, to north of

33.

Coastal Overwash During Hurricane Iniki

Kapaa. The overwash line was drafted from field data and aerial
photographic interpretation onto digital basemaps supplied by the Office of
State Planning. The overwash line was digitized on the Universal
Transmercator Zone 4 Coordinate System and entered into the State of
Hawai'i, Office of State Planning Geographic Information System under the
file heading INIKLOWASH.
In the following sections we present and discuss maps showing the
digitized hai overwash line, the 1986 shoreline, the FEMA Flood Hazard
Zone V ("subject to inundation by the 100-year flood with the additional
hazards associated with storm waves"; FEMA, 1987), and the FEMA Flood
Hazard Zone A ("subject to inudation by the 100-year flood"; FEMA, 1987).
These maps are State GIS products. Because of intervening shoreline
movement between determination of the FIRM lines, and the overwash line,
and differences in the FEMA shoreline and the State shoreline, there is some
significant positioning inaccuracy in these maps. There are instances (see
Waimea) where the FEMA FIRM lines wander slightly offshore. These
obvious errors have been retained in the maps as an example of the inherent
inaccuracies in this type of mapping technique in a dynamic environment.
We estimate a positioning accuracy of +50 m.

OVERWASH CONDITIONS

A total storm-surge height of 1.8 m above mllw is recorded at near
peak high tide at Port Allen, Kaua'i located 5 km southeast of eye landfall at
Kaumakani. At Nawiliwili Harbor, 30 km northeast of Kaumakani, a storm-
surge height of 1.7 m (ml1w) is recorded. Moored wave buoys located 355
kin southwest of Honolulu and 452 km south-southeast of Honolulu,
recorded significant wave heights of 5.5 m and 6.0 m (10.97 m most
probable maximum wave height), resp. Winds at Barking Sands Naval
Facility, 22 km left of the eye, reached a maximum sustained speed of 31
m/sec (69 mph) with gusts to 45 m/sec (100 mph). At Lihue, 32 km right of
the eye, sustained winds of 43.5 m/sec (97 mph) with gusts to 57.8 m/sec
(129 mph) were recorded. At Makahuena Point, 20 km right of landfall, a

34.

Coastal Overwash During Hurricane Iniki

Po "f

jw,

17@
iji.i

Or,

7, Figure 11. The overwash debris line east of Kukuiula Bay, south shore of
Kaual Overwash reached elevations between 4 and 5 rn (mllw) along this
coast. The debris line is composed of irrigation piping, housing debris, and
L
00'"'*

,[various structural elements from the coastal developments near the water. A

35.

Coastal Overwash During Hurricane Iniki

maximum gust value of 124 knots was extracted from the digital recorder,
and we have estimated that maximum sustained winds there were between
39-42 m/sec (87-94 mph).
Overwash excursion and elevation are primarily a function of relative
position and orientation to the storm core, and secondarily a function of local
shoreline geometry, topography and bathymetry, and marine conditions.
Maximum overwash excursions (>200 m inland) and elevations (-8.2 m
msl) are not coincident, apparently this is the influence of secondary forcing
parameters related to bathymetry and topography.
As we have discussed, Iniki traveled north-northeast across Kauai,
exiting east of Haena on the north shore. The south coast experienced the
highest overwash, and broad patterns of overwash elevation and inland
excursion are relative to Kaumakani.

KEKAHA

Ten kilometers left of the eyewall, in Kekaha, maximum debris line
and still-water level heights averaged 3.4 m (msl). Damage from overwash,
while relatively light with respect to other areas of the island, was related to
high velocity overwash flooding at the first and second line of houses in the
vicinity of Oomano Point. Shore-normal streets acted as conduits for
channelizing overwash, but these flows generally remained confined to the
paved roads and did not cause significant damage among dwellings away
from the water front. Although the revetment along Kaumualii Highway
(coastal road) was overtopped by the flooding, apparently the wall was a
significant factor in mitigating overwash damage to much of Kekaha west of
Oomano Point. The revetment, built to halt shoreline retreat and chronic
erosion resulting from the updrift (east) construction of Kikiaola Harbor,
effectively reduced both the volume and velocity of the overwash in the
western and central portions of Kekaha.
The history of shoreline erosion was an exacerbating factor in Kekaha,
as in many other overwash areas on Kaua'i, and identifies chronic erosion as
one common preconditioning agent that tends to maximize flood damage.

36.

Coastal Overwash During Hurricane Iniki

E!
CU

ctj

ebb

...... .. .
. . .. ... ......

.... ..... .

=Fiure 12. Overwash map of Kekaha

37.

Coastal Overwash During Hurricane Iniki

Abundant aeolian transport occurred in central and western Kekaha.
r", Beach sands there are coarse-grained carbonate, whereas in east Kekaha,
IL they are fine-grained volcaniclastics with abundant heavy minerals. Wind
transport increased the amount of sand deposition occurring on the highway
and immediately inland, requiring special efforts to clear roads for vehicular
traffic.
Overwash excursion exceeded 200 rn at Oomano Point. The overwash
L line takes a significant shift landward, moving from approximately one half
of a city block landward of the highway to a full block and a half landward,
exactly where the protection afforded by the revetment ends.
Figure 5 demonstrates that throughout the region overwash exceeded
the FEMA V-Zone prediction of 100-year flooding with hazardous wave
action. The overwash line falls within the 100-year flood zone, and had all
7 buildings been elevated to the suggested AE and/or VE elevations, the level
of damage would have been significantly reduced. As noted in the FEMA
7

ilk-

oar-
L

Figure 13. The end of the Kekaha revetinent, and the erosional offset in the
shoreline caused by Kikiaola Harbor in the distance.

38.

7
Coastal Overwash.During Hurricane Iniki

or
0
t
AW_ A
OR
4:1
4
X - .@7,
.1:5r;

Figu e 14. Overwash was channeled down coastal avenues, although
7 velocity and total water volume was mitigated by the Kekaha revetment.
L_ Building Performance Assessment Report (1992), homes that were elevated,
even as little as 0.6-1 m above grade, often were spared flooding and
7 sustained little flood damage. As also noted in that report, the level of flood
damage prevented, reinforces the importance of properly elevating a
structure to suggested elevations within flood-hazard areas.
The coast between Oomano Point and Kikiaola Harbor has undergone
retreat at a rates of 0.15-0.6 m/yr (MOESE, 1992). Construction of the
harbor in 1959 interrupted the net westward littoral drift and when erosion
threatened the coast highway, the Kekaha revetment was built. Erosion
continues west of the harbor, threatening the onshore stability of the west
jetty which was flanked by overwash during Iniki. Flooding skirted, and
partially eroded a dune system behind the west jetty and penetrated a local
cemetary. Damage was minimal.
Between the harbor and the neighboring coastal town of Waimea, the
shoreline is largely undeveloped and no significant structural damage was

39.

Coastal Overwash Dufing Hunicane InW

sustained other than to some lands under agricultural use, and the native
vegetation. Overwash fans, extending landward 50 to 150 m, are noted in the
aerial photography demonstrating that a large volume of sediment was
removed from the beach and deposited beyond the active littoral zone. Much
of this may be returned to the beach and littoral dunes by the tradewinds.
Possible salinization of the worked soil may have occurred. The
geomorphology of the backshore region is uniform, and the overwash line
shows no significant features other than as reflected by variations in run-up
and wave energy. Overwash elevations are approximately 1.8-2.4 m (msl).

WAIMEA

In Waimea, maximum overwash heights averaged 2.6 m (msl).
Landward excursion of the flooding was generally confined to the back
beach and immediate areas (10-30 m). Marine flooding damaged the lower
level of the first row of houses on the west bank of the Waimea River mouth.
The highest measured overwash in Waimea occurred near the river,
-2.5-2.8 in. Debris lines located below the top of the embankment revetment
at the river mouth suggest that only minor overtopping occurred. Thus,
flooding onto the low-lying coastal plain adjacent to the river was
predominately marine, and did not contain any notable contribution from
riverine discharge. It has been noted elsewhere that rainfall during Hurricane
Iniki was generally below expected levels, and little actual river flooding
occurred on Kaua'i. Precipitation at Lihue was approximately 2.3 cm, and 3
cm at Princeville (Trapp, 1993). 'Me notable increase in marine excursion
near the river was the result of the generally lower elevation around the
mouth, rather than actual fluvial flooding. The east bank of the river is a high
basalt exposure several meters in elevation and marine overwash was
confined to a gravel beach located at the base of the headland.

40.

Coastal Overwash During Hurricane Iniki

ii li f f

0
ii@ @j:

. ...........

......... . .

iR
4,

O!A M, zii@

FFiaure 15. Overwash map of Waimea.

41.

Coastal Overwash During Hurricane Iniki

HANAPEPE

Hanapepe, 4.5 km right of landfall at Kaumakani, experienced
maximum overwash heights exceeding 2.8 m (msl). The most extensively
flooded area was on the low-lying fluvial floodplain east of the mouth of the
Hanapepe River. Residents report Iniki overwash penetrated several hundred
meters inland along this lowland, to the edge of the coastal highway (Rt. 50)
where it crosses the river. The bridge there is a critical transportation link for

e 16. Overwash damage to a chain-link fence, Hanapepe Coastal

42.

EA C73 C:A E@A CA EA CA EA E:A CA EA CA F-I CA C@

. ..... ........ . ............... .......... pe Rive
CID

4 ...... ........
CA

. ... ..... ... .......

CD
PORT
ALLEN

Salt Pond..
Park

....... ...

EMA2
I. Ittzoll
....... . . ....
EMA04000

Coastal Overwash During Hurricane Iniki

communities in west Kaua'i and a determination should be made as to
whether the bridge was at any point threatened by the overwash. This region
falls within the FEMA 100-yr flood Zone A, but beyond their determination
for the marine hazard Zone V.
Extensive damage to the coastal park in Hanapepe Bay occurred. A
revetment was one source of large boulders that were carried in suspension
by the overwash and destroyed park facilities including a bathing house and
tennis court fencing. Soil erosion, evidenced by scarping of terrigenous
deposits behind the revetment, led to high turbidity levels in the Bay.

KUKUIULA To KEONILOA BAY

At Kukuiula, 16-17 km right of the eye at landfall, maximum
overwash around the Bay, and in the fields to the east, measured between
3.8-4.9 in (msl). Because of the low relief, the gentle onshore gradient, and
the absence of offshore barriers, this sector experienced a high degree of
inland penetration by marine waters. The overwash passed through a line of
coastal dwellings, across a road and drainage ditch and carried debris deep
into the adjacent agricultural fields. Aerial photos of the resulting debris line,
in places over 250 in inland, have been widely published (Fig. 11). These
provide telling testimony to the power and destructive potential of overwash
in the region.
This same reach of coast was devastated by Hurricane Iwa in 1982,
and was the subject of numerical modeling under contract to the U.S. Army
Corps of Engineers in 1986 (Sea Engineering and Bretschneider, 1986). That
work reports on scenario models that predict the overwash expected during
an Iniki-type storm. One lesson learned by this experience is that numerical
models of hurricane overwash have been developed to a high level of
reliability, and should be incorporated in future planning along the Hawaiian
coastline. Another lesson learned through the combined overwash under Iwa
and Iniki, is that the coastal developments around and east of Kukuiula Bay
are vulnerable to marine flooding and massive damage. Despite the obvious
consequences, rebuilding activity proceeds unabated at this writing along

44.

Coastal Overwash During Hurricane Iniki

Apt#
7

Figure 18. Overwash debris line near Kukuiula Bay.

45.

Coastal Overwash During Hurricane Iniki

...... .....

............

i Kft

01894S 0,
..... ........

Uj
.. . ... .. ..

......... ...... .

=Fiure 19. Overwash map between Kukuiula Bay and Keonfloa Bay-

46.

Coastal Overwash During Hurricane Iniki

this twice-devatated coast.
The overwash line runs well inland of coastal developments between
Kukuiula Bay and Nahumaalo Point. Single family dwellings, multi-unit
condominium structures, and resort hotels were all subjected to high-velocity
marine flow. A double debris line in places (Fig. 18), suggests that the
overwash occurred in waves and that structures were subjected to multiple
episodes of flooding in a short space of time. Dry stream-beds and former
natural drainage valleys contain evidence that marine penetration of inland
regions was heightened by locally lower elevations. Several low-lying
regions, some over 100 rn from the coast, contained extensive ponded sea
water and were clogged with housing and pavement debris. Local pocket
beaches along this coast experienced severe erosion, and the resulting
exposure of underlying soil led to high levels of coastal turbidity that
persisted for over a month after the hurricane. Beach accretion has been slow
to develop and at this writing, a year later, many sections of beach remain in
an eroded state and have not fully recovered from the effects of Iniki.
Although we cannot determine if there was pre-existing ecosytern
degradation, the prolonged turbidity in the Kukuiula to Poipu coastal zone
may be responsible for some shallow-water coral mortality and algal growth.
Because of the intensity and extent of overwash in this sector a
number of structures sustained damage from high-velocity flooding,
breaking wave forces, foundation scour and undermining, and battering by
suspended debris in the water column. There are field reports of hydraulic
buckeling of cement slab-type foundations, and examples of heightened
damage from large boulders and paving stone literally battering down walls.
At Poipu Beach Park coconut trees bear impact scars to a height of 1.8 in
above ground level. Smaller structures were frequently swept off their
foundations and either floated inland, or experienced destruction of the
lower floors while upper floors and roofs survived to be deposited some
distance inland. Residents at Nahumaalo Point report that wave spray against
the headland overtopped three-story structures, and had sufficient velocity to
break windows and damage roofing materials.
The highest elevation of overwash evidence, a debris line behind a
small dwelling, was discovered by surveyors at the U.S. Army Corps of
Engineers. Tlis was in the area immediately east of Koloa Landing. As
previously mentioned, the configuration of the coast and bathymetry there

47.

Coastal Overwash During Hurricane Inild

-'S

Figure 20. Coastal damage between Kukuiula Bay and Keoniloa Bay
included beach erosion, coastal turbidity, structural damage from high-
velocity overwash, and salinization of terrestrial envirom-nents in the coastal
zone. Note overwash ponding, and the debris line behind this hotel.

48.

Coastal Overwash During Hurricane Iniki

apparently led to maximum run-up and breaking wave heights, and
amplification of stonn surge components. Debris on. the road at the head of
the boat ramp at Koloa measured 7 m in elevation. While to the east the
maximum debris line was measured at over 8 m. Nearby, at an elevation of
nearly 6 m was an overwash debris line consisting of stone blocks weighing
several hundred pounds, and large household appliances.
Toward the Poipu coast the overwash line meanders inland behind
townhouse and condominium developments, and the Sheraton Resort Hotel,
keeping in the elevation range between 3.5 and about 5 m. Large ponds of
brackish water mark upland areas where return flow collected in local basins
(Fig. 20). In the parking lot behind the Waiohai Resort, automobiles were
actually stacked atop one another by the overwash. The much of the
subdivision landward of Poipu Beach Park and Brenecke's Beach, a popular
recreation site, was overwashed. A still-water mark in a house on Pane Rd.
measured 5.6 m. The small beach in front of the seawall at Brenecke's,
originally destroyed by Hurricane Iwa and having only returned in recent
years, was completely eroded by Iniki. One year later it has not returned.
F------

UFA

6W@

.liq MOR

=Fiiare 21. Coastal damage at Poipu Beach Park.
7----

49.

Coastal Overwash During Hurricane Iniki

The seawall at this locality was damaged by Iniki, and this is an excellent
opportunity to remove it completely in order to encourage beach recovery
and long-term stability.
On Makahuena Point, a headland with a coastal elevation exceeding 6
m (msl), the development is sufficiently set-back, and the protection
afforded by the cliffs sufficiently effective, that little overwash damage
occurred. That portion of the headland composed of carbonate sandstone
(aeolianite) did experience some hydraulic undermining from wave action.
This was also noted across Keoniloa Bay at Makawehi Point, another
limestone headland. While no slumping was observed on basaltic coasts,
there was overhang collapse and slumping along many sectors of limestone
shore, demonstrating that a hazard is associated with development on these
sites. '
The beach at Keoniloa Bay was temporarily eroded, and beachrock
exposed. Accretion, and the construction of a partially buried, armored
overwash barrier, has since taken place. Overwash in the region of the Grand
Hyatt Hotel exceeded 4 m, but damage was min imized by the wide set-back
there. Unconsolidated coastal dunes in this area were cut and channelized to
a depth of over 2 rn where overwash apparently forced breaching of the
backshore area. To the north, Mahaulepu Beach was overwashed to an
elevation of 3 to 4 m, but nonetheless experienced massive lateral beach
accretion exceeding 10 m because of a large sand supply available from the
adjacent dune system there.

KAPAA

The highest overwash on the east coast of Kauai, in and near the town
of Kapaa, was over 3 m, but averaged approximately 2.5-2.8 m. Because the
Kapaa coast is oriented north-south, the winds of Iniki did not blow directly
onshore for sustained periods and the additional translation speed of the
storm did not exacerbate the overwash. The coastal orientation was oblique
to most of the surge components. Also, waves generated by Iniki had to
refract around the south coast of the island before striking the Kapaa

50.

Coastal Overwash During Hurricane Iniki

shoreline. As a result, breaking wave heights were in the range 3-5 m. Kapaa
generally did not experience extreme overwash. Nonetheless, there were
pockets of extensive damage in the first row of houses on coasts with a
history of chronic erosion. One of these, the Waipouli neighborhood,
suffered overwash damage as a result of crowded shoreline development,
low-grade housing lacking the minimum elevation suggested by FEMA
flood maps, and a condition of pre-existing erosion.

.. ........

n
Rd
a

KA@

Ma
NIA N@

... ..... .:::.

un
h on
Wan

=Fjure 22. Overwash map of Kapaa coast.

51.

Coastal Overwash During Hurricane Iniki

Frontage properties in Papaloa, Waipouli, and Kapaa Town suffered
damage from overwash, flooding, road washouts, and bank failure under
wave assault. Coastal erosion was prevalent and beachrock exposure was a
frequent tell-tale that entire beaches had been eroded. 71is was especially
true where shorelines were armored, or soil and rock cliffs marke 'd the
landward extent of the beaches. The region between Waikaea Canal, which
traps littoral sediment from the south, and Moikeha Canal, which traps
littoral sediment from the north, has experienced chronic erosion for over a
decade. Major portions of this coast have lost their beaches and thus had no
natural buffer to absorb wave energy and dissipate run-up. Fortunately much
of the development in this area is set-back behind Kapaa Beach Park, and
while overwash reached significant elevations (3.4 m) and landward
excursion was on the order of 25 to 50 m, structural damage was limited.
South of Waikaea Canal lies yet another jettied canal at Waipouli. This, like
the others, is unfortunately responsible for effectively trapping the littoral
transfer of sediment. A condition of sediment starvation along nearly 2.5 km
of coast is responsible for exacerbating what probably would have been a
significantly less damaging overwash environment. Removal, or at the least
truncation and lowering of these jetties would not only improve coastal
sediment flux and lead to reduced beach erosion, possibly beach accretion,
but more robust sediment storage along the coast would provide some
mitigation against future overwash damage.

OTHERAREAS

HANALEI BAY: Overwash averaged 1-1.5 m and breaking wave heights
were negligable because first winds (as storm approached) blew westerly
and offshore, and second winds (as storm retreated) reversed to the east,
while remaining offshore. Wind damage to Hanalei Town was extensive,
and housing debris was noted on the seafloor at offshore sites in Hanalei
Bay. No river flooding occurred, and the only coastal impact from the storm
was post-event turbidity from deforested drainage areas. No observable reef
damage resulted from the storm.

52.

Coastal Overwash During Hurricane Iniki

NA PALI COAST: Although the media initially reported that the Na Pali
Coast, on the north shore of Kaua'i had "geologically aged thousands of
years", in fact there was little environmental damage to the coastal zone. Our
reconnaissance of this coast, by helicopter, revealed little evidence of beach
erosion, no slumping of the headlands, and no evidence of alteration to the
unique coastal architecture of caves and arches.
Turbidity of coastal waters was noted adjacent to streams draining
uplands that had experienced deforestation and uprooting in runoff areas.
But this was minor in comparison to the extensive turbidity of the Kukuiula-
Poipu coast. We suspect that the relatively light rainfall associated with Iniki
was an impotant reason that environmental damage was reduced in the Na
Pali coastal zone. Another mitigating factor was the direction of storm
movement and predominate winds that placed this sector in the lee of
maximum sustained winds.

BARKING SANDS: Visits to Barking Sands recreational area and the
surrounding dune fields revealed no environmental damage. Comparisons to
pre-Iniki aerial photography showed no significant beach erosion, and some
isolated cases of new dune formation.

PAKALA VELLAGE: Pakala Village coast was overwashed to an elevation of
over 3 m and the first line of houses experienced flooding. There was also
flooding in the cane field to the immediate east of town behind a small
coastal bluff as evidenced by a line of dead plants. A history of erosion, and
poorly engineered shoreline protection structures exacerbated the impact of
the storm. No data is available regarding historical erosion rates at the site.
The beach has disappeared along much of the fronting coast at Pakala, and
trees are toppling into the water. Apparently erosion has been chronic and
rapid.

POIPU To NAWILIWELI Local beach erosion occured at Gilliens Beach, a
local popular recreational site. As mentioned earlier, accretion was also
noted in the area where beaches are backed by large dune fields. Cultural
artifacts were reported exposed by the erosion at Gilliens, and strewn by the
waves. Maximum overwash measured over 4.5 m in elevation. Carbonate

53.

Coastal Overwash During Hurricane Inild

aeolian headlands and cliffs were hydraulically excavated to the point of
overhang collapse at several localities.

NAWELBVILI TO WAELUA: Nawiliwili Harbor sustained damage that was
noted from aerial reconnaissance only. A large sailboat was washed
broadside against the rock jetty, and several vessels sustained damage from
wave surge in the harbor. Of those vessels that were hauled out, collapsed
blocks apparently resulted in damage as they fell against the pavement.

Figure 23. An overwash fan and dead sugar cane near Pakala Village.

54.

Coastal Overwash During Hurricane Iniki

Local beach erosion was noted along the ocean coast. Several incidents of
coastal turbidity were observed, and the Wailua Golf Course revetment
failed at a site that has suffered chronic erosion adjacent to a break in the
fringing reef.

KAPAA TO PRINCEVILI : The natural coastal environments along this sector
apparently sustained little environmental damage. There was widespread and
massive defoliation of the upland forests that led to fears of stream mouth
debris build-up, a hazard that can produce flash flooding and further
environmental damage. Many of the beaches in this sector are boulder ramps
and suffered little under the high waves. Occasional overwash was noted
from the air, to elevations of approximately 1 rn or less. Rarely was the
back-beach envirorunent overwashed.

=Fiure 24. The Na Pali Coast suffered little coastal zone damage.

55.

Coastal Overwash During Hurricane Iniki

-ENVIRONMENTAL RESPONSE-

Environmental response to the overwash varied. Beaches suffered the
greatest erosion on stabilized shorelines where steep, energy-reflective faces
interfered with overwash. In unstabilized regions both erosion and accretion
occurred. Beach erosion exposed ferruginous soil beds that leached silt into
coastal waters, and created sustained turbidity following the storm. Surveys
of coral response in the region of greatest overwash (Kukuiula to Poipu)
revealed slight mechanical damage to the sparse Porites sp. and Pocillopora
sp. community in the nearshore region, and negligible damage in deeper
offshore waters. Heightened algal growth, and coral mortality in waters
shallower than 5 m may be a response to post-storm turbidity.
A short list of environmental effects in the coastal zone includes:
defoliation of standing forests leading to mortality and debris build-up; soil
erosion from highland regions and coastal settings; overwash mortality to
coastal vegetation and agriculture; local soil and water salinization; impact
damage and uprooting of trees under intense overwash; beach erosion and
accretion; coastal dune channelization and erosion, some dune formation and
migration; coastal soil exposure by beach erosion and leaching of soil into
coastal waters causing turbidity and siltation; some coral abrasion; possible
coral mortality by turbidity in the shallow zone between Kukuiula and
Makahuena Point; structural debris in the nearshore zone; boulder-strewn
fields across the coastal upland under intense overwash; undercutting of
carbonate headlands; debris build-up by overwash.
In the remaining sections of this study, we report briefly on beach
erosion, Poipu coral coverage, and offshore debris build-up.

56.

Coastal Overwash Dufing Hw-ricane Iniki

BEACHEROSION

While many beaches sustained little erosion during Iniki, and cases of
accretion were noted, the more common response of impacted beaches was
to erode. We have made a qualitative assessment of beach erosion by
measuring beach width on aerial photographs of the coast between Kukuiula
Bay to Makahuena Point, and between the Wailua River mouth to north
Kapaa Beach Park. Black and white photographs taken 7/28/87 at a scale of
1 "=500', were compared to a set of color photos at the same scale taken on
9/18/92. No photogrammetric, or tidal corrections have been applied in this
assessment. Our criteria for erosion were simply the recognition of minor
subaerial beach narrowing (ca. <15 m change over the period 1987-1992), or
significant subaerial narrowing or disappearance (ca. > 15 rn change over the
period 1987-1992). We cannot deten-nine when the narrowing may have
occurred, but we feel there is some likelyhood that Iniki played a role on
those beaches where significant narrowing was noted over the 5 year period.

Kukuiula-Makahuena Point Wailua River-Kapaa Beach Park
Kukuiula Bay beach (min.) Wailua River State Park beach (sig.)
East of Ekaha Point (sig.) Wailua Bay beach (sig.)
Hoai Bay (min.) Kapaa Beach Park north end (sig.)
Kaheka beach (sig.) South of Waipouli (min.)
Kihouna Point, E&W (sig.) Waipouli coast (sig.)
Keoniloa Bay beach, middle Waipouli Beach Park (sig.)
portion (sig.) North of Moikeha Canal (sig.)
Poipu Beach Park (sig.) North of Waikea Canal (sig.)
Brennecke's beach (sig.) South of Moikeha Canal (sig.)
Total length: 855+100 m Total length: 3045+100 m

The total measured length of significantly-eroded beach between
Kukuiula Bay and Makahuena Point is 855+100 m. Between the Wailua
River and north Kapaa Beach Park it is 3045+100 m. These lengths include
beaches that have either disappeared or significantly narrowed over the
period 7/28/87 to 9/18/92.

57.

Coastal Overwash During Hurricane Iniki

Z@:

M-

1IFigure 25. Top, erosion at Waipouli beach. Bottom, accretion at Mahaulepuj
'IT

58.

Coastal Overwash During Hurricane Iniki

POIPU OFFSHORE SURVEY

Coral Impacts - Divers surveyed coral coverage at 16 stations along the 30
ft and 70 ft depth contours in the region between Makahuena Point and
Kukuiula Bay (Krock and Neill, 1993). These sites were chosen to duplicate
stations used in a similar survey conducted in 1972 as part of a water quality
management study for the county government (Sunn, Low, Tom & Hara,
Inc., 1973). Sites were spaced approximately at 0.5 mile intervals along each
contour. Sampling procedure consisted of photographing 2 to 3
representative bottom locations at each station with a reference scale and
station identification number. The percentage of coral coverage in the
photographs was determined by planimeter. The coral were grouped into two
classes: 1.) those that present little cross-sectional profile to the nearshore
hydrodynamic environment, consisting principally of the encrusting genera
Porites and Montipora; 2.) and those with a larger cross-sectional profile,
head coral, principally the genus Pocillapora.
Dr. Ralph Bowers, the biologist responsible for the 1972 survey,
revisited the region following Hurricane Iwa and reported extensive damage
to the coral community along both depth contours. However, his
observations were not quantitative, and do not serve as a basis for numerical
comparison (Krock and Neill, 1992).
Data in the following tables is from Krock and Neill (1992), and serve
to illustrate several points. Over the period of observation there has been
approximately a one-third reduction in total coral coverage on the 30 ft
contour. This is supported by diver observations of abundant broken
Pocillapora , and abraded and broken encrusting corals at this depth. The
greatest damage was observed where unconsolidated boulder pavements are
present. The 30 ft contour was the approximate depth of initial wave
breaking during Iniki, and it is likely that this high-energy environment
resulted in the entrainment, and saltation of many of the clasts.
The data suggests that there has been little overall change in coral
coverage at the 70 ft contour. The previously dominant encrusting coral at
this depth has, however been replaced by higher profile Pocillapora. The
1972 survey showed an encrusting coral and head coral mean coverage of

59.

Coastal Overwash During Hurricane Iniki

about 4% and 0.5% respectively. By 1993, the relative coverage had
changed to approximately 0.9% and 5% for the encrusting and head coral
respectively. Although damage was noted at this depth following Hurricane
Iwa (Bowers, pers. comm.), divers in the 1993 survey found very little
evidence of damage despite the presence of unconsolidated boulder
pavements.
If high levels of damage were sustained at 70 ft during Hurricane Iwa,
it is likely that the current abundance of Pocillapora represents
recolonization of the substrate. Although tropical storms had passed nearby,
prior to Iwa, Kauai had not sustained a direct hurricane hit since Dot in
1959. The intervening 22 yr may have been enough to allow sustained coral
colonization that was especiaRy vulnerable to damage. Sport diver reports
from 0'ahu suggest that Hurricane Iwa caused more damage than Iniki.
Possibly this is a result of the relative length of quiet periods between major
storms. It is unclear if the benthic environment during Iwa was significantly
more energetic than during Iniki, but there is no hydrodynamic argument to
support this conclusion.

Table L
Percent coral coverage for nearshore Kauai stations

Station % Encrusting Coral % Head Coral Approx. Depth
Number 1993 (1972) 1993 (1972) (ft)
81 1.8 (3.2) 2.7 (6.2) 30
82 sand (1.0) sand (0.4) 70
83 2.5 (1.2) 4.9 (1.3) 30
84 3.0 (11.0) 3.6 (0.0) 70
85 5.0 (4.7) 3.4 (4.3) 30
86 0.2 (1.7) 0.0 (0.0) 70
87 1.2 (7.3) 4.1 (4.3) 30
88 sand (sand) sand (sand) 70
89 0.7 (6.0) 10.1 (8.0) 30
90 0.1 (5.0) 6.5 (0.3) 70
91 sand (1.5) sand (7.0) 30
92 0.0 (sand) 7.2 (sand) 70
93 0.3 (1.0) 0.0 (8.0) 30
94 0.1 (2.0) 7.0 (0.8) 70
Note: 1972 Data is in Parenthesis

60.

Coastal Overwash During Hurricane Iniki

Alternatively, the abundance of damaging clasts at the 70 ft level may
have declined in the interim. Coral stress levels prior to Hurricane Iwa must
also be considered in evaluating the survey results. It is possible that the
benthic substrate in 1982 consisted of a large number of available clasts that
accumulated over the relatively long period of quiesence since earlier
storms. Increased coral damage during Iwa may have occurred as a result
and the responsible materials shifted to an offshore location. There is a ledge
very near the. 70 ft level (at approximately 80 ft) that would serve as an
effective barrier against onshore return of these clasts, under fairweather

waves.

Table 11.
Patchiness of coral growth
Percent of area with hard substrate
Arithmetric mean and variance over mean

Station Stations % Encrusting Coral % Head Coral
Number/ w/ Hard 1993 Mean Var/Mean 1993 Mean Var/Mean
Depth Subst 1972 Mean Var/Mean 1972 Mean Var/Mean
89-83 6 1.92 1.51 4.2 2.22
30 ft (7) (3.22) (1.56) (6.57) (0.96)
82-94 5 0.86 1.61 4.86 1.56
70 ft (5) (3.96) (3.67) (0.49) (0.35)]
Note: 1972 Data is in Parenthesis

Table Ill.
Estimate of gross coral coverage
Percent of area with hard substrate
Arithmetric mean and 95% confidence interval

Station Stations % Encrusting Coral % Head Coral
Number/ w/ Hard 1993 Mean Conf. Interv 1993 Mean Conf. Interv
Depth Subst 1972 Mean Conf. Interv 1972 Mean Conf. Interv

89-83 6 1.92 0.14<m<3.70 4.2 1.00<m<7.40
30ft (7) (3.22) (1.35<<4.58) (6.57) (3.94<<8.97)
82-94 5 0.86 0.00<m<2.31 4.86 1.44<m<8.28
70 ft (5) (3.96) (1.06<<8.46) (0.49) (.001<<1.63)

Note: 1972 Data is in Parenthesis

61.

CoasW Overwash Dining Hurficane InM

Debris Build-up - Surveys of nearshore waters in the region revealed only
limited distribution of hurricane debris. Collecting principally in channel
areas, items such as lanai furniture, acoustic ceiling tiles, plastic pipes and
sheeting, guttering, occasional small items from kitchens, and plastic siding
were noted on an individual basis. Reports from the local sport diving
community identified specific clean-up activities had been conducted on a
volunteer basis following Iniki. Local divers reported having removed
several large items such as bicycles, a refrigerator, micro-wave ovens, bed
springs assemblies, and some household furnishings. The same individuals
reported that debris was no longer a concern in this region, or in any other
area they had heard of.

Field Observations - related to the diving survey are as follows:
a. There is no reef between Makahuena Point and Spouting Hom.
b. There is instead a sandy or varying volcanic bedrock seafloor consisting
of: outcropping lava flows that form benches and overhangs; bouldery basalt
pavements with diameters ranging from 0.5m to 2m; polygonal basalt
pavements, usually in association with boulder fields, that are of low relief
and smooth texture; sandy seafloor with uniformly fine to medium sand that
is well-sorted and of mixed carbonate-volcanogenic mineralogy with a
minor silt/organic component.
c. The sandy bottom had no coral growth.
d. The volcanic hardbottom had sporatic coral development with
Pocillapora as the dominate genus and Porites sub-dominate, both genera
grow in this area as separate and usually isolated patches approx. 5 cm to 30
cm in diameter.
e. Because it presents a prominate vertical profile, Pocillapora sp. displayed
varying degrees of abrasion, breakage, and scour generally increasing in
frequency in the shallow depths, and increasing where a bouldery bottom
existed, several instances of total abrasion were noted where an that was left
was a white scar, a former foothold, on a basalt substrate.
f. Porites sp. displayed a lesser degree of damage due to mechanical action.
g. Debris was absent to sparse in all sectors of the survey, there was a clear
increase in the amount of debris in the shallow water zone, there were no
areas of debris build-up, debris components were isolated and widespread
consisting of pvc piping of varying length and diameter, aluminum gutter,

62.

Coastal Overwash During Hurricane Iniki

kitchen storage racks, ceiling acoustic particle board, lightweight lanai
funishings, some textile material, occasional plastic sheeting, and other
unidentifiable components, a clean-up effort by the local population has
apparently been successful in removing the majority of large and abundant
debris.
h. A common debris element was fishing line wrapped around Pocillapora
sp., which did appear to cause damage to the coral, this would seem to be
unrelated to the hurricane.
i. There was notable algal growth in the nearshore zone on every exposed
surface save living coral, a significant percentage (1-5%?) of the Porites sp.
and Pocillapora sp. poulation were dead and covered with algae, some
(much?) of this may be due to month-long water column turbidity following
the hurricane, turbidity developed because beach erosion exposed large
amounts soil to wave action, silt eroded into the water for several weeks
following the storm, the dramatic decrease in photic penetration may have
impacted the viabilitiy of nearshore corals and resulted in the algal growth.
j. Algal growth in the nearshore (0 ft to 20 ft) was the dominate biological
component of the substrate.
k. There are bathymetric highs in the coastal zone that appear to be remnant
bedrock outcrops from shoreline retreat, these often have crests that are intertidal
and that extend 100m or more offshore, it is likely that these bathymetric highs,
and the accompanying adjacent lows (total relief of approx. 20 ft) influenced the
elevation and excursion distance of the hurricane overwash by modifying the
wave height at the shoreline, by increasing the immediate roughness coefficient
and decreasing run-up, and by modifying the energy level and set-up.

Recomendations -
1.) No further clean-up is necessary.
2.) Continued monitoring of coral coverage and algal growth at 3-yr intervals
would provide anexcellent data-set on this ecosystem in the eventuality of
another storm.
3.) The algal growth should be monitored and surveyed on a 6-month interval,
sources of nutrification should be identified and monitored.

63.

Coastal Overwash During Hurricane InW

References
Abraham, G. (1964) "Hurricane Storm Surge Considered as a Resonance
Phenomenon", Waterloopkundig Laboratorium, Delft, Netherlands
pp.585-602.
Agrawal, J. D. (1993) "Calculation of Storm Surge for Hurricane Iniki,"
Department of Ocean Engineering, University of Hawaii at Manoa,
Honolulu, pp.81.
Anthes, R. A. (1982) "Tropical Cyclones, Their Evolution, Structure and
Effects", American Meteorological Society, Boston, MA, pp. 208.
Bretschneider, C. L., J. D. Agrawal, N. C. Cannella, and K. A. Wohlmut
(1993) "Hurricane Iniki, a Hindcast, Winds, Windwaves, Swells, Storm
Surge, and Wave Run-Up," Proceedings of the American Society of Civil
Engineers, Hurricanes of 1992 Technical Conference, Miami, Dec 1-3,
1993.
Bryant, E. A. (1991) "Natural Hazards", Cambridge University Press,
Cambridge, pp 285.
Ewing, M. F., F. Press and W.L. Donn (1954) "An Explanation of the Lake
Michigan Surge of 26 June, 1954, " Science pp. 684-686.
Federal Emergengy Management Assessment Team (FEMA) (1987)
"Questions and Answers on the National Flood Insurance Program"
Pamphlet.
Federal Emergengy Management Assessment Team (FEMA) (1992)
"Hazard Mitigation Report, Hurricane Iniki", Report FEMA-961-DR-HI,
Hawaii, pp.43.
Federal Emergengy Management Assessment Team (FEMA) (1992)
" Preliminary Report in Response to Hurricane Iniki, Kauai County,
Hawaii", Building Performance Assessment Team, Federal Emergency
Management Agency, Federal Insurance Administration, Hawaii, pp.72.
Hensley, J. M., (1992) "Offshore Buoy Data", U.S. Department of
Commerce, National Oceanic and Atmospheric Administration, Stennis
Space Center, Miss.
Hunt, I. A., (1959) "Design of Seawalls and Breakwaters," Proceedings,
American society of Civil Engineers, ASCE, Waterways and Harbors
Division, Vol. 85, No, WW3.
Krock, H. J., and D. Neill (1993) "Hurricane Iniki Effects on Coral Sand and

64.

Coastal Overwash During Hurricane Iniki

Beach Processes on Kauai," Proceedings of the American Society of Civil
Engineers, Hurricanes of 1992 Technical Conference, Miami, Dec 1-3,
1993.
Marinos, G. and J. W. Woodward (1968), "Estimation of Hurricane Surge
Hydrographs," Journal Waterways and Harbors Divsion, American
Society of Civil Engineers, May, pp. 189-216.
Myers, V.A. (1954), " Characteristics of U.S. Hurricanes Pertinent to Levee
Design for Lake Okeechobee, Florida," H.R. Report 32, U.S. Weather
Bureau.
National Weather Service (1992) "Hurricane Iniki, September 6-13, 1992",
Natural Disaster Survey Report, U.S. Department of Commerce, National
Oceanic and Atmospheric Administration, Silver Spring, Md.
Saville, T. (1952), "Wind Set-Up and Waves in Shallow Water", Tech.
Memo. 27, U.S. Army Beach Erosion Board, Washington, D.C., 36p.
Sea Engineering and C.L. Bretschneider (1986) "Hurricane vulnerability
Study for Kauai, Poipu and Vicinity, Storm Wave Run-up and
Inundation", U.S. Army Engineer Division, Fot Shafter, Hawaii.
Sorensen, R. M. (1978), "Basic Coastal Engineering," John Wiley & Sons,
New York.
Sunn, Low, Tom &Hara, Inc.(1973) "County of Kauai Water Quality
Management Plan", Honolulu, Hawaii.
Trapp, G. (1993), "Hurricane Iniki: The Event and its Impact," Preliminary
Report, Draft Copy.
U.S. Army Corps of Engineers (1984) "Shore Protection Manual" Vol. I &
H, Coastal Engineering Research Center, Department of the Army,
Vicksburg, Miss., pp. 893.
Van Dom, W. G. (1953), "Wind Stress on an Artificial Pond", Journal of
Marine Research, Vol. 12, pp. 249-276.
Yamamoto, S. H., and S. P. Sullivan (1993) "Measurement and Hindcasting
of Hurricane Iniki Coastal Inundation", Proceedings of American Society
of Civil Engineers - Hurricanes of 1992", Technical Conference, Miami,
Dec 1-3, 1993.

65.

mop--"

Sam

State of Hawai'i Office of State Planning
Coastal Zone Management Program

U.S. Geological Survey
National Coastal Geology Program

National Science Foundation
V

W'ww

Y' - @-At

3 6668 00000 8385