a second generation of tsunami inundation maps for the state of california

A Second Generation of Tsunami Inundation Maps for the State of California

A. BARBEROPOULOU,1 J. C. BORRERO,1,2 B. USLU,3 M. R. LEGG,4 and C. E. SYNOLAKIS1,5

Abstract—A new generation of tsunami inundation maps is

now available for 20 coastal counties in California. These maps

represent an improvement over previous efforts, as they are based

on the most recent descriptions of potential tsunami sources, apply

recently updated numerical modeling techniques, and cover pre-

viously unmapped regions of the State. Since the maps are based on

deterministic rather than probabilistic modeling, they are only

intended for emergency preparedness and evacuation planning and

are not to be used in engineering siting studies. The California

maps cover a greater coastal area than any other US State. To be

helpful, the maps need to be integrated into a consistent statewide

hazard-planning framework. Indeed, tsunami preparedness in Cal-

ifornia was tested on several occasions over the past 5 years, i.e.,

during the 14 June 2005 event, about 90 miles SW of Crescent

City, the 15 November 2006 Kuril Islands, and the 27 February

2010 Chile earthquake. We discuss briefly the State’s response as

these events unfolded.

Key words: California, Tsunami, inundation, inundation

maps, mitigation, coastal, evacuation, natural hazards.

1. Introduction

Since the early 1990s, steady progress has been

made in understanding and mitigating tsunami haz-

ards in California. In response to tsunami events that

have highlighted previously unrecognized threats,

scientists and federal and state policy-makers have

undertaken a comprehensive effort to produce evac-

uation maps showing areas of possible tsunami

inundation. Advances in tsunami modeling, source

characterization, computational resources, and social

science studies have allowed better identification of

high-risk areas and have made this information

available to the public.

Prior to the early 1990s, California was thought to

be primarily vulnerable to teletsunamis caused by

large earthquakes in the subduction zones around the

Pacific Rim (HOUSTON and GARCIA, 1978; MCCULLOCH,

1985). The State was generally thought to be at lesser

risk from local or regional tsunamis. However,

identification of the Cascadia Subduction Zone (CSZ)

as a tsunami source (CLARKE and CARVER, 1992) and

the 1992 Cape Mendocino earthquake and tsunami

demonstrated otherwise (MCCARTHY and BERNARD,

1993; SYNOLAKIS et al., 1997; GONZALEZ et al., 1995).

Post-tsunami field surveys (SYNOLAKIS and OKAL,

2005; SYNOLAKIS and KONG, 2006) have provided a

wealth of field observations that have increased

awareness of tsunami hazards (SYNOLAKIS et al., 2002;

BARDET et al., 2003). Three recent events have further

reinforced the need for continued vigilance (e.g.,

SYNOLAKIS and BERNARD, 2006). Although the first of

these, the June 14, 2005 earthquake offshore of

Northern California, did not generate a significant

tsunami, it presented a challenging tsunami-warning

situation and highlighted shortcomings in the existing

response system (GIBBONS, 2005; MURPHY, 2005). The

November 15, 2006 earthquake in the Kuril Islands

generated a teletsunami that caused US $10 million

damage to Crescent City Harbor (DENGLER, 2008;

DENGLER et al., 2009). The Chilean earthquake of

February 27, 2010 produced a similar teletsunami in

California, which nonetheless caused substantial

currents in California ports and damage estimated at

US $10 million.

1 Tsunami Research Center, University of Southern California,

3620 S. Vermont Avenue, Los Angeles, CA 90089, USA.

E-mail: [email protected]; [email protected] ASR Limited, Marine Consulting and Research, 1 Wainui

Road, Raglan 3225, New Zealand.3 Pacific Marine Environmental Laboratory, NOAA, 7600

Sand Point Way NE, Seattle, WA 98115, USA. E-mail:

[email protected] Legg Geophysical, 16541 Gothard St # 107, Huntington

Beach, CA 92647-4472, USA. E-mail: [email protected] Laboratory of Natural Hazards, Technical University of

Crete, 73100 Chanea, Greece.

Pure Appl. Geophys. 168 (2011), 2133–2146

� 2011 Springer Basel AG

DOI 10.1007/s00024-011-0293-3 Pure and Applied Geophysics

With more than 2,000 km of coastline, California

has a large at-risk population and several economically

vital ports and harbors. The vulnerability is high, and,

to increase resilience, advances in tsunami planning,

hazard mitigation, and warning need to be utilized.

California faces additional challenges such as a very

short historical record of tsunamis and a long history of

geologic studies that have focused on inland faults,

while generally ignoring offshore fault zones

(SYNOLAKIS et al., 1997). Several of these fault zones lie

only a few kilometers from the coast, yet seismometer

coverage of offshore areas is inadequate, leading to

inaccurate determination of epicenter locations and

focal mechanisms (LEGG and BARBEROPOULOU, 2007).

Not surprisingly, the frequency of occurrence for

locally generated tsunamis is not much better known

today than it was in the early 1980s.

The tsunami hazard in Northern California is

dominated by the CSZ and can be inferred from

GONZALEZ et al. (1995). Further south, the 1927 Point

Arguello earthquake (Ms 7.3) is to date the best-

documented locally generated tsunami in Southern

California (SATAKE and SOMERVILLE, 1992; LANDER

et al., 1993); however, its occurrence in a sparsely

populated area minimized the impact. Prior to 1927,

historical accounts of 19th century events describe

waves that would be damaging today, due to

increased coastal populations and higher building

density, particularly in the Santa Barbara Channel

region (MCCULLOCH, 1985; LANDER et al., 1993;

BORRERO et al., 2001) and in San Diego Bay (LEGG

and KENNEDY, 1979). Confirmation of prehistoric

tsunami deposits in the Southern California coastal

area remains to be established, although candidate

sites have been identified (e.g., G. KUHN, pers. com-

munication, 2003; GOYA, 2005).

2. Previous Efforts to Quantify Tsunami Hazards

in California

HOUSTON and GARCIA (1974) and HOUSTON and

GARCIA (1978) used a combination of finite-difference

computations, one-dimensional analytical results

for periodic waves, and finite-element models to

estimate offshore wave heights caused by far-field

hypothetical scenarios designed to resemble the 1960

Chile and 1964 Aleutian Island earthquakes. Their

numerical computations, while groundbreaking at the

time (SYNOLAKIS et al., 1997), did not include inun-

dation. Their 100- and 500-year recurrence rates for

offshore tsunami heights were used in the Flood

Insurance Rate Maps (FIRM) for the National Flood

Insurance program. MCCULLOCH (1985) considered

far-field and near-field tsunami hazards as part of an

evaluation of the earthquake hazard in the Los

Angeles region. However, due to an arithmetic error,

his estimate of tsunami wave heights caused by near-

shore submarine landslides was deficient by nearly

two orders of magnitude (BORRERO, 2002), leading to

the earlier significant underestimation of this hazard.

MCCARTHY and BERNARD (1993) qualitatively

assessed the tsunami hazard for the entire state fol-

lowing the 1992 Cape Mendocino earthquake

(OPPENHEIMER et al., 1993). They considered local and

regional sources and offshore slumping and were the

first to consider the tsunami hazard from earthquakes

along the CSZ. They proposed that the coastal areas

north of Cape Mendocino and south of the Monterey

Peninsula to Palos Verdes (Los Angeles County) are

high risk, and the remaining coastlines ‘‘moderate’’

risk.

The first quantitative study of tsunami inundation

due to a near-source event was that of BERNARD et al.

(1994). This study was motivated during discussions

between the Federal Emergency Management Agency

(FEMA) and the scientific community in the aftermath

of the 1992 Cape Mendocino earthquake and tsunami.

While BERNARD et al. (1994) presented careful seis-

mological inferences for possible CSZ events, the

characterization of the wave itself was less rigorous.

The tsunami from the Mw 8.4 tremor was modeled as a

sinusoid with 10 m amplitude and 33.3 min period at

50 m depth. The model used was untested in terms of

the veracity of inundation predictions, and likely it

stopped computations at some threshold location. The

results were published as part of an earthquake plan-

ning scenario produced by the California Division of

Mines and Geology, now known as the California

Geological Survey (CGS) (TOPPOZADA et al., 1995).

Since then, the Redwood Coast Tsunami Working

Group (RCTWG), a consortium of private, academic,

and government scientists and planners, has developed

a more detailed set of tsunami hazard maps for the

2134 A. Barberopoulou et al. Pure Appl. Geophys.

same area in Northern California (e.g., Fig. 1;

DENGLER, 2006).

In 1995, the US Congress created the National

Tsunami Hazard Mitigation Program (NTHMP) and

directed the National Oceanic and Atmospheric

Administration (NOAA) to form working groups at

the federal and state levels charged with addressing

tsunami hazard assessment, mitigation, and response

issues (BERNARD and GONZALEZ, 1994). In 1996, this

group submitted a report recommending the prepa-

ration of tsunami inundation maps for the five Pacific

states of Alaska, Washington, Oregon, California,

and Hawaii (EISNER et al., 2001). By 1997, a series of

local tsunamis (Nicaragua, 1992; Flores, Indonesia,

1992; Okushiri, Japan, 1993; East Java, Indonesia,

1994; Mindoro, Philippines, 1994; Shitokan, Russia,

1994; Manzanillo, Mexico, 1995; Biak, Indonesia,

1996; Sulawesi, Indonesia, 1996; Chimbote, Peru,

1996) further focused attention on hazards associated

with locally generated tsunamis. Based on MCCARTHY

and BERNARD (1993), SYNOLAKIS et al. (1997) con-

sidered a possible rupture of the San Clemente Fault

in Southern California and found that, using inunda-

tion models that had just started to become available

(TITOV and SYNOLAKIS, 1998), runup predictions ran-

ged up to two times greater than results obtained

using 1980s-type threshold computations (see

SYNOLAKIS and KANOGLU, 2009, for definitions). In the

immediate aftermath of the 1998 Papua New Guinea

landslide tsunami, a series of workshops combined

with a reassessment of the tsunami hazard led the

California Office of Emergency Services [now known

as the California Emergency Management Agency

(CalEMA)] to collaborate with the University of

Southern California-Tsunami Research Center (USC-

TRC) for the development of a first generation of

inundation maps for the State’s coastlines using funds

provided by the NTHMP and the California Seismic

Safety Commission.

Because the HOUSTON and GARCIA (1974) study

considered only distant events, this early effort

focused on locally generated tsunamis. Limited

resources and a large at-risk population spread along

2,000 km of coast meant that difficult decisions had

Figure 1Example of an inundation and evacuation map for Crescent City, CA produced by the Redwood Coast Tsunami Working Group

Vol. 168, (2011) A Second Generation of Tsunami Inundation Maps 2135

to be made regarding mapping priorities and meth-

odology. The first-generation maps were produced

with priority given to coastal areas of high population

density, numerical grid resolution of 250 m (com-

pared with an average 1.7–2.0 km spacing used by

HOUSTON and GARCIA, 1974), and emphasis on near-

shore earthquake and landslide sources over distant

events. The State’s strategy is discussed in EISNER

et al. (2001) and BORRERO et al. (2003), while the

methodology and computational details are discussed

in BORRERO (2002).

Based on feedback from the emergency manage-

ment community, these maps reflected a single runup

value for each modeled region, derived from com-

putations of wave evolution and inundation from

extreme but plausible tsunamigenic scenarios. The

inundation line on each map corresponded to the

highest modeled tsunami runup along the modeled

coast, among the scenarios. As they were completed,

the first-generation maps were ‘‘ground-truthed’’

through field surveys conducted by the USC-TRC

team and State officials. The purpose of these field

surveys was to compare the model results to physical

reality and to familiarize both parties with areas of

particular vulnerability.

The California practice differs from that in other

Pacific states. For example, the Oregon Department

of Mineral and Industry’s tsunami inundation maps

included two lines, one for a locally generated tsu-

nami from the CSZ, and another for far-field

tsunamis, leaving the final choice of which map’s

projection to use with emergency managers. The

reasoning behind California’s single inundation line

was to facilitate communication to the public with a

clear message: if you are at a location situated below

the inundation line, you should evacuate to high

ground in the event a tsunami warning is issued or if

you feel an earthquake lasting longer than 30 s, or if

you see any unusual water motions near shore.

Another difference was that the computational suite

used by California was the only one benchmarked

with both laboratory and field data. At the time of this

writing, the five Pacific states are planning a code

benchmarking workshop, and it remains to be seen

which state models comply with the recently adopted

standards and guidelines for numerical models (SYN-

OLAKIS et al., 2008).

While groundbreaking as compared with maps

available in other states (for example, maps in Hawaii

until recently relied on 1970s-type modeling with

one-dimensional propagation results), California’s

early mapping effort had limitations. First, the maps

were based only on computations of tsunami evolu-

tion from local sources. Second, the relatively coarse

grid resolution of 250 m was difficult to transfer to

the onshore land surface. At the time, it was difficult

to register on-land topographic grids and offshore

bathymetric grids to the same datum. Third, the

inclusion of submarine landslide scenarios—events

that are plausible but with unknown return periods—

resulted in inundation distances that would make an

orderly evacuation nearly impossible, particularly on

a summer afternoon when there may be hundreds of

thousands in beach areas. Last, while the maps cov-

ered all densely populated areas, vast areas of the

State remained unmapped. These constraints moti-

vated the development of a new set of inundation

maps for California.

For the record, the first study of probabilistic

hazard assessment since HOUSTON and GARCIA (1974)

is that of USLU (2008), where both time-dependent

and time-independent methods were used to assess

the tsunami risk. In the latter, slip rates were obtained

from global positioning system (GPS) measurements

of tectonic motions and then used as a basis to esti-

mate the return period of possible earthquakes. The

resulting tsunami heights from these earthquakes

were computed at selected locations offshore ports

and harbors in California, and then return periods

were estimated to provide a total probability of

exceedance of given wave heights. In the time-inde-

pendent method, return periods of candidate

earthquakes were estimated using Gutenberg–Rich-

ter-type relationships. For Crescent City Harbor, a

full inundation computation was performed so that

recurrence estimates were provided at the location of

the tide gage, for comparison with earlier events.

The USLU (2008) analysis found that San Fran-

cisco Bay and Central California are most sensitive to

tsunamis originating from the Alaska and Aleutian

Subduction Zone (AASZ). Using an earthquake

source of magnitude comparable to the 1964 Great

Alaska Earthquake located on the central AASZ

resulted in twice the wave height as experienced in


1964 in San Francisco Bay. Detailed inundation

modeling is currently underway to evaluate whether

this doubling of offshore heights results in double the

flooding distances. Both time-dependent and time-

independent methods showed that Central California

and San Francisco Bay receive more frequent tsuna-

mis from the AASZ, while Southern California is

mostly impacted by tsunamis generated on Chile and

Central American Subduction Zone, as well as the

AASZ.

3. New Tsunami Inundation Maps

On December 26, 2004, the second largest

instrumentally recorded earthquake in history (second

only to the 1960 Chilean Earthquake, Mw = 9.5)

occurred off of the west coast of Sumatra, Indonesia

(i.e. GEIST et al., 2006; OKAL and STEIN, 2005). The

aftermath of the subsequent tsunami prompted gov-

ernments worldwide to reassess coastal vulnerability.

The US Congress supplemented funding in 2005 and

2006 to support the expansion of the tsunami warning

system, additional Deep-Ocean Assessment Report-

ing of Tsunamis (DART) tsunameters in the Pacific

the development of the infrastructure for producing

real-time forecasts (WEI et al., 2008), and accelerated

production of inundation maps for at-risk or previ-

ously unmapped areas.

In 2006, CalEMA responded by requiring that

updated inundation maps be developed including

distant and local sources. Hence, the most recent

mapping collaboration between USC-TRC, CalEMA,

and CGS included 35 separate grid areas in 20

counties (BARBEROPOULOU et al., 2009; Fig. 2a, b, c;

Tables 1, 2, 3), by far the largest coastal area of any

US state. These maps cover the most significant ports,

harbors, coastal urban centers, and popular recrea-

tional areas in California. The tsunami modeling took

advantage of updated bathymetric and topographic

datasets, and included additional tsunami source

information and descriptions published or identified

since 1998 (i.e., BARDET et al., 2003; LEGG et al.,

2004; BORRERO et al., 2004).

As with the first-generation maps, computations

of wave evolution and inundation were undertaken

234˚ 236˚ 238˚ 240˚

36˚

38˚

40˚

42˚

237.5 238

37.5

38

SanFrancisco Bay

RCF

SGF

0 10 20

km

HF

OREGON

Morro Bay

Crescent City

Cape Mendocino

1

2

3

4

5

67

8

9

10

11

1213

14

15

16

Monterey Bay

Point Arguello

CALIFORNIACSZ

PRF

50010002000

3000

4000

239˚ 240˚ 241˚ 242˚ 243˚32˚

33˚

34˚

CALIFORNIA

Cata.

Los Angeles

Long Beach

SanPedro Bay

SanDiego

Malibu

Santa Barbara

SN

Newport Beach

16 17 18

19

20

LKCAT

SCI

Santa B

a

rbara Channel

500

1000

20003000

Santa Monica Bay

SMT/C

BT

CIT

30MB

GS

NIFPVS

S.Clem. Is.

PointConcepcion

ACD

MCF

CB

ChannelIslands

a

b c

Figure 2Locations of geographical features and local tsunami source regions mentioned in the text. Numbers correspond to the county names listed in

Table 2. a Northern California: CSZ Cascadia Subduction Zone, PRF Point Reyes Fault, (inset), RCF/HF Rodgers Creek Fault/Hayward Fault

Stepover in San Pablo Bay, SGF San Gregorio Fault. Black star is the approximate epicenter of the 1927 Point Arguello earthquake.

b Southern California: CIT Channel Islands Thrust Fault, GS Gaviota and Goleta Slide features, ACD Anacapa-Dume Fault, MCF Malibu

Coast Fault, PVS Palos Verdes Slide, NIF Newport Inglewood Fault, CAT Catalina Island Restraining Bend, LK Lasuen Knoll Restraining

Bend, SMB/CBT San Mateo and Carlsbad Thrust Faults, CB Coronado Bank Fault and Restraining Bend, SCI San Clemente Island, 30 MB

Thirty Mile Bank Fault and Restraining Bend. c Teletsunami sources from the Pacific Rim used in the tsunami inundation study


with MOST (TITOV and GONZALEZ, 1997). Originally

developed at the USC-TRC (TITOV and SYNOLAKIS,

1998) prior to its adoption by NOAA, MOST solves

the nonlinear two-dimensional shallow-water wave

equations and allows for calculations of the wave

evolution over complex bathymetry and overland

flooding. Inundation computations are notoriously

temperamental, as solving the overland flow involves

nontrivial interactions of the incident wave with the

shoreline (LIU et al., 1991). MOST has been validated

repeatedly at benchmark workshops funded by the

National Science Foundation (YEH et al., 1996; LIU

et al., 2005; LIU et al., 2008), through the steps out-

lined in SYNOLAKIS et al. (2008). MOST is the

computational suite used for the official tsunami

forecasts in the USA. It was most recently used to

accurately predict the far-field wave evolution of the

2010 Chilean tsunami, as well as the extent of

flooding in Hawaii and California, faster than real

time.

In terms of maps, most of California’s coastline

was modeled with grid resolution of 3 arcsec,

(approximately 75–90 m). The three major ports, Los

Angeles/Long Beach, San Francisco, and San Diego,

were modeled with 1 arcsec resolution (*25–30 m).

The numerical grids were adjusted to mean high

water (MHW) conditions to represent a conservative

sea level for tsunami inundation. We note that only

maximum inundation was computed in the ports;

computations for currents and assessment of the

impact of harbor resonance possible in narrow

channels need even higher resolution and far longer

computational times.

Both local (BORRERO et al., 2001; BORRERO et al.,

2004; USLU et al., 2007) and far-field (OKAL et al.,

2003; DENGLER et al., 2008) tsunami sources were

considered in this effort. Local tsunami sources

included offshore reverse and thrust faults, major

restraining bends on strike-slip fault zones, and large

submarine landslides capable of significant seafloor

displacement and tsunami generation. Far-field tsu-

nami sources included historical great subduction

zone events such as the 1700 Cascadia, 1952 Kurile

Island, 1960 Chile, and 1964 Alaska earthquakes, as

well as a suite of plausible large-scale events occur-

ring around the Pacific Rim.

To create overland flooding projections, the

computed inundation from each of the sources in a

given region was combined to create a dataset

showing the area of potential flooding from all

sources considered. The location of the final inun-

dation line was determined with the benefit of a

ground-truth survey to check the preliminary pre-

dictions against a high-resolution (3–10 m)

topographic dataset and compare them with the actual

coastal conditions in terms of sand dunes, river

mouths, wetlands, and coastal structures that may

interfere with the advancing tsunami. These surveys

were coordinated with local emergency managers,

and considered historic inundation information, when

available (Fig. 3). As opposed to the first generation

Table 1

Some facts about the California tsunami inundation maps

Number of numerical grids used in the study 35

Numerical grid resolutiona *90 m

Total number of distant sources 12

Total number of local sources 23

Total number of local earthquake sources 16

Total number of landslide sources 7

Number of maps produced (quadrangle USGS format) 130

Approximate cost per map (US $) $3,000

a Port areas of Los Angeles/Long Beach and San Diego were

modeled at *25 m resolution

Table 2

California coastal counties

1 Del Norte

2 Humboldt

3 Mendocino

4 Sonoma

5 Marin

6 Napa

7 Solano

8 San Francisco

9 Contra Costa

10 San Mateo

11 Alameda

12 Santa Cruz

13 Santa Clara

14 Monterey

15 San Luis Obispo

16 Santa Barbara

17 Ventura

18 Los Angeles

19 Orange

20 San Diego

Numbers correspond to the regions indicated in Fig. 2a, b


of maps (whose inundation line represented one

global maximum over the entire map), each point on

the second-generation maps reflected the extreme

inundation at that particular point, among different

scenarios at that point, therefore a local maximum.

California completed its second generation of

inundation maps, as planned, by the fifth anniversary

of the 2004 Sumatra tsunami. They were presented at

a joint CalEMA, USGS, and USC-TRC press con-

ference at the 2009 Fall Meeting of the American

Geophysical Union. These maps are a significant

improvement over the earlier generation, as they are

of higher resolution, are based on more recent geo-

logic investigations, and cover areas not previously

mapped. Once implemented in local tsunami hazard

mitigation and response planning, the maps will

provide an additional level of confidence for visitors

and residents of coastal locales.

4. Discussion of the Second-Generation Maps

4.1. Deterministic versus Probabilistic

For these maps, a deterministic approach was

used. Probabilistic maps are more appropriate for

land use planning and zoning, but deterministic maps

are more common for evacuation planning. The latter

are less expensive to produce than the former, and

often there is simply not enough information to

warrant the production of probabilistic maps, even if

resources exist. This happens when there are insuf-

ficient historical records or geological and

seismological data available for meaningful estimates

of tsunami recurrence rates.

While specific scenarios used to develop deter-

ministic maps do, in general, have quantified return

periods, for many scenarios used in this study, the

return periods are poorly known, especially for

submarine landslide sources. For example, the repeat

time of the Palos Verdes debris avalanche, a feature

believed to be capable of generating a significant

tsunami off the Ports of Los Angeles/Long Beach,

has been estimated to 3,000–7,500 years (e.g., LEE

et al., 2009), with the last event estimated to have

taken place over 2,000 years ago.

The inundation mapping effort discussed here

included submarine landslide sources placed at various

likely locations along the coast. Indeed, these submarine

landslide sources dominated the runup predictions for

most cases in Southern California. The relatively long

expected recurrence frequencies for these large coastal

submarine landslides—up to 10,000 years—are roughly

one order of magnitude lower than the frequency of large

(M [ 7.0) offshore earthquakes on active offshore faults

with recurrence rates of 150–3,500 years (150 years for

the offshore San Andreas in Northern California,

3,500 years for the less active faults along the coast

such as the Newport-Inglewood-Rose Canyon Fault

Zone). Reconciling this disparity is an issue that will be

revisited in future studies.

4.2. Numerical Grid Resolution

California, with its complex offshore and near-

shore bathymetry, offers a challenging hydrodynamic

setting for numerical simulations. While flat and

Table 3

Sources used in the tsunami inundation modeling

Distant sources

Cascadia Subduction Zonea—CSZ

Alaska Aleutian Subduction Zone—AASZ (4 sources)

Kuril Japan Trench (4)

Nazca/South America Subduction Zone (2 sources)

Mariana Izu-Bonin Trench

Local earthquake sources

Cascadia Subduction Zonea—CSZ (6 sources)

Santa Monica Bay

Santa Barbara Channel (2 sources)

Catalina Island Restraining Bend

Newport Inglewood Fault

Lasuen Knoll

San Mateo Thrust

1927 Point Arguello Earthquake

San Gregorio Fault

Rodgers Creek/Hayward Fault Stepover

Point Reyes Fault

Carlsbad Thrust Fault

Coronado Bank Fault

San Clemente Fault

Local landslide sources

Palos Verdes Slide (2 sources)

Goleta/Gaviota Slide (2 sources)

Coronado Canyon

Monterey Canyon

Thirty Mile Bank Fault

a The CSZ is a local source for Northern California and a distant

source for Central and Southern California


planar beaches can be realistically modeled using

90 m numerical grids, more complex ports and harbors

require higher-resolution grids to resolve complex

wave dynamics and possible seiching (i.e. BARBERO-

POULOU, 2008). For example, in the Ports of Los

Angeles and Long Beach, the breakwater is not well

represented with a 90 m grid. Using a 25–30 m grid,

the breakwater is better resolved, resulting in more

accurate hydrodynamic simulations. Furthermore, as

well documented in previous inundation modeling

studies, coarse numerical grid resolution leads to

underprediction of tsunami inundation and runup

(TITOV and SYNOLAKIS, 1998; TITOV and GONZALEZ,

1997). The issue of modeled inundation approaching

its physical asymptotic value with increasing grid

resolution has been addressed by TITOV and SYNOLAKIS

(1998) and PEDERSEN (2008) and is illustrated in Fig. 4,

which compares inundation projections in San Diego

Bay using 3-arcsec and 1-arcsec resolution grids.

The issue of the appropriate grid resolution for

numerical prediction of tsunami inundation in com-

plex bays and harbors was raised as recently as April

2009, during a meeting of the NTHMP Mapping and

Modeling Subcommittee (MMS). Minimum require-

ments were placed for the inundation map production,

wherever the quality of available data permits. As

SYNOLAKIS et al. (2002) have argued, while higher

numerical resolution may improve computational

accuracy, it does not improve physical realism if not

based on equivalent or higher resolution in measured

bathymetry and topography. If one interpolates

numerically in a given coarse grid, features present

in the prototype landforms may be missed and the

fine-resolution computation may provide a false sense

of confidence. In California, where most of the

inundation maps relied on a 90 m resolution grid,

field surveys attempted to improve the maps by

carefully checking the actual terrain against the digital

elevation model used in the computations. This

approach has merit in terms of finding out if on-land

features are reflected in the topographic grids, but it

does not improve computational accuracy.

4.3. Bathymetry and Topography Data

The bathymetric and topographic data used in

California’s maps originate from the National

Geophysical Data Center (NGDC). The NGDC is

the primary developer of registered bathymetry and

topography datasets in the USA. Grids are produced

at cost—a grid with 30 m resolution for a 40 km

length of coastline is valued at approximately

US $45,000. The NGDC produces datasets in a

format readily usable by MOST, thus allowing for

considerable time-saving in the production of maps.

Grids are subsectioned as appropriate for site-specific

scenario modeling.

4.4. Tsunami Sources

Initial conditions for local and distant sources

were implemented differently in the computations.

For distant events, we relied on the National Oceanic

and Atmospheric Administration Pacific Marine

Environmental Laboratory (NOAA–PMEL) database,

which subdivides the major subduction zones of the

Pacific Rim into discrete 50-km-wide by 100-km-

long fault segments. The rupture along each segment

is modeled as a low-angle, pure thrust mechanism

with a uniform 1 m of co-seismic slip, equivalent to

an earthquake of Mw = 7.5. The resulting sea-floor

deformation is then used as an initial condition for

MOST, and the numerical results stored in a database.

More complex rupture scenarios are created through

linear superposition and scaling of these individual

scenarios into larger events of the desired earthquake

magnitude. The resulting composite wave field is

used to calculate boundary conditions for subsequent

computations in the high-resolution grids for the

region of interest. In this manner, it is not necessary

to compute the wave evolution over large transoce-

anic distances for every scenario, since it is supplied

by the database, and the computational effort is

limited to solving the nonlinear near-shore evolution

problem. It is possibly for this reason, that the cost of

the California maps is the lowest among the five

Pacific states.

Note that, in this methodology of unit sources,

geometric parameters describing each unit fault

segment cannot be modified. However, the NOAA/

PMEL database is expanding and now covers every

major subduction and faulting zone in the Pacific,

Atlantic, and Indian Oceans, and most known earth-

quakes can be described through combination of


117°22'30"W

117°22'30"W

33°15'0"N

33°7'30"N

33°7'30"N

117°22'30"W

117°22'30"W

33°15'0"N

33°7'30"N

33°7'30"N

TSUNAMI INUNDATION MAPFOR EMERGENCY PLANNING

0.5 0 0.5 10.25

Miles

SCALE 1:24,000

1,000 0 1,000 2,000 3,000 4,000 5,000500

Feet

0.5 0 0.5 10.25

Kilometers

State of CaliforniaCounty of San Diego

Initial tsunami modeling was performed by the University of Southern California (USC)Tsunami Research Center funded through the California Emergency Management Agency(CalEMA) by the National Tsunami Hazard Mitigation Program. The tsunami modelingprocess utilized the MOST (Method of Splitting Tsunamis) computational program(Version 0), which allows for wave evolution over a variable bathymetry and topographyused for the inundation mapping (Titov and Gonzalez, 1997; Titov and Synolakis, 1998).

The bathymetric/topographic data that were used in the tsunami models consist of aseries of nested grids. Near-shore grids with a 3 arc-second (75- to 90-meters)resolution or higher, were adjusted to “Mean High Water” sea-level conditions,representing a conservative sea level for the intended use of the tsunami modelingand mapping.

A suite of tsunami source events was selected for modeling, representing realisticlocal and distant earthquakes and hypothetical extreme undersea, near-shore landslides(Table 1). Local tsunami sources that were considered include offshore reverse-thrustfaults, restraining bends on strike-slip fault zones and large submarine landslidescapable of significant seafloor displacement and tsunami generation. Distant tsunamisources that were considered include great subduction zone events that are known tohave occurred historically (1960 Chile and 1964 Alaska earthquakes) and others whichcan occur around the Pacific Ocean “Ring of Fire.”

In order to enhance the result from the 75- to 90-meter inundation grid data, a methodwas developed utilizing higher-resolution digital topographic data (3- to 10-metersresolution) that better defines the location of the maximum inundation line (U.S.Geological Survey, 1993; Intermap, 2003; NOAA, 2004). The location of the enhancedinundation line was determined by using digital imagery and terrain data on a GISplatform with consideration given to historic inundation information (Lander, et al.,1993). This information was verified, where possible, by field work coordinated withlocal county personnel.

The accuracy of the inundation line shown on these maps is subject to limitations inthe accuracy and completeness of available terrain and tsunami source information, andthe current understanding of tsunami generation and propagation phenomena as expressedin the models. Thus, although an attempt has been made to identify a credible upperbound to inundation at any location along the coastline, it remains possible that actualinundation could be greater in a major tsunami event.

This map does not represent inundation from a single scenario event. It was created bycombining inundation results for an ensemble of source events affecting a given region(Table 1). For this reason, all of the inundation region in a particular area will not likelybe inundated during a single tsunami event.

Tsunami Inundation Line

Tsunami Inundation Area

MAP EXPLANATIONMETHOD OF PREPARATION

References:

Intermap Technologies, Inc., 2003, Intermap product handbook and quick start guide:Intermap NEXTmap document on 5-meter resolution data, 112 p.

Lander, J.F., Lockridge, P.A., and Kozuch, M.J., 1993, Tsunamis Affecting the West Coastof the United States 1806-1992: National Geophysical Data Center Key to GeophysicalRecord Documentation No. 29, NOAA, NESDIS, NGDC, 242 p.

National Atmospheric and Oceanic Administration (NOAA), 2004, InterferometricSynthetic Aperture Radar (IfSAR) Digital Elevation Models from GeoSAR platform (EarthData):3-meter resolution data.

Titov, V.V., and Gonzalez, F.I., 1997, Implementation and Testing of the Method of TsunamiSplitting (MOST): NOAA Technical Memorandum ERL PMEL – 112, 11 p.

Titov, V.V., and Synolakis, C.E., 1998, Numerical modeling of tidal wave runup:Journal of Waterways, Port, Coastal and Ocean Engineering, ASCE, 124 (4), pp 157-171.

U.S. Geological Survey, 1993, Digital Elevation Models: National Mapping Program,Technical Instructions, Data Users Guide 5, 48 p.

California Emergency Management AgencyCalifornia Geological SurveyUniversity of Southern California

Tsunami Inundation Map for Emergency PlanningOceanside Quadrangle/San Luis Rey Quadrangle

State of California ~ County of San Diego

OCEANSIDE QUADRANGLESAN LUIS REY QUADRANGLE

June 1, 2009This tsunami inundation map was prepared to assist cities and counties in identifyingtheir tsunami hazard. It is intended for local jurisdictional, coastal evacuationplanning uses only. This map, and the information presented herein, is not a legaldocument and does not meet disclosure requirements for real estate transactionsnor for any other regulatory purpose.

The inundation map has been compiled with best currently available scientificinformation. The inundation line represents the maximum considered tsunami runupfrom a number of extreme, yet realistic, tsunami sources. Tsunamis are rare events;due to a lack of known occurrences in the historical record, this map includes noinformation about the probability of any tsunami affecting any area within a specificperiod of time.

Please refer to the following websites for additional information on the constructionand/or intended use of the tsunami inundation map:

State of California Emergency Management Agency, Earthquake and Tsunami Program:http://www.oes.ca.gov/WebPage/oeswebsite.nsf/Content/B1EC51BA215931768825741F005E8D80?OpenDocument

University of Southern California – Tsunami Research Center:http://www.usc.edu/dept/tsunamis/2005/index.php

State of California Geological Survey Tsunami Information:http://www.conservation.ca.gov/cgs/geologic_hazards/Tsunami/index.htm

National Oceanic and Atmospheric Agency Center for Tsunami Research (MOST model):http://nctr.pmel.noaa.gov/time/background/models.html

The California Emergency Management Agency (CalEMA), the University of SouthernCalifornia (USC), and the California Geological Survey (CGS) make no representationor warranties regarding the accuracy of this inundation map nor the data from whichthe map was derived. Neither the State of California nor USC shall be liable under anycircumstances for any direct, indirect, special, incidental or consequential damageswith respect to any claim by any user or any third party on account of or arising fromthe use of this map.

Topographic base maps prepared by U.S. Geological Survey as part of the 7.5-minuteQuadrangle Map Series (originally 1:24,000 scale). Tsunami inundation lineboundaries may reflect updated digital orthophotographic and topographic data thatcan differ significantly from contours shown on the base map.

PURPOSE OF THIS MAP

MAP BASE

DISCLAIMER

Table 1: Tsunami sources modeled for the San Diego County coastline.

Areas of Inundation Map Coverageand Sources UsedSources (M = moment magnitude used in modeled

event) DanaPoint

Oceanside San Diego

Carlsbad Thrust Fault X XCatalina Fault X X XCoronado Bank Fault XLasuen Knoll Fault X XSan Clemente Fault Bend Region XSan Clemente Island Fault XSan Mateo Thrust Fault X X

LocalSources

Coronado Canyon Landslide #1 XCascadia Subduction Zone #3 (M9.2) X XCentral Aleutians Subduction Zone#1(M8.9) X X XCentral Aleutians Subduction Zone#2(M8.9) X XCentral Aleutians Subduction Zone#3(M9.2) X X XChile North Subduction Zone (M9.4) X X1960 Chile Earthquake (M9.3) X X1952 Kamchatka Earthquake (M9.0) X1964 Alaska Earthquake (M9.2) X X XJapan Subduction Zone #2 (M8.8) X XKuril Islands Subduction Zone #2 (M8.8) X XKuril Islands Subduction Zone #3 (M8.8) X X

DistantSources

Kuril Islands Subduction Zone #4 (M8.8) X X

Figure 3Sample standard inundation maps for San Diego County (Oceanside quadrangle), California. The legend below the map contains the map

symbology, a brief explanation behind the making of the map, and basic references. It also contains the list of tsunami scenarios used and the

important disclaimer that the map is solely for the purpose of emergency and evacuation planning. Each county will use a similar map for

determining evacuation routes and locating tsunami signage


faults in the expanded database (GICA et al., 2008).

When complex local faulting and/or sources occur

within the computational areas, we devise more

realistic faulting scenarios that incorporate variable

slip distributions and alternate fault geometries, i.e.,

we do not use the database. Tsunami inundation

calculations in the Southern California Bight

(BORRERO et al., 2004; LEGG et al., 2004) and near-

field effects of ruptures along the CSZ (USLU, 2008)

were modeled from the epicenter to the maximum

inundation point, without using precomputed data.

4.5. Applicability and Use of the California Maps

The primary use of the new inundation maps is for

evacuation planning, increasing awareness, and pub-

lic education. The latter is believed the weakest link

in the chain. The maps are not appropriate for use in

calculating local effects at power plants, utilities or

vital coastal infrastructure. For example, when con-

sidering whether a near-shore structure will be

overtopped, the information on the maps is of little

use, particularly for marginal events, which may be

the most common. The overland flow depths of

smaller tsunamis may be of the same size as the

height of the structure, hence computations at high

resolution with at least two grid points in the tank are

needed for any meaningful inferences as to the

hazard. Hence, extreme caution needs to be exercised

when using the inundation maps for any engineering

study, which by definition involves analysis of risk at

the level of particular structures. For such needs, a

detailed site-specific study should be undertaken

using state-of-the-art numerical modeling techniques.

4.6. Recent Tsunami Events in California

Inundation maps are but one link in hazard

mitigation planning and emergency response. In an

actual event, State officials (hopefully with the

benefit of the knowledge of tsunami researchers)

decide whether to proceed with a local or regional

evacuation after a tsunami message (watch, advisory

or warning) is issued. Here, we will discuss briefly

how the system worked during three recent events.

The earthquake of 14 June 2005 (MW = 7.2)

occurred on the Gorda Plate near the Mendocino

Triple Junction offshore of Northern California

(OPPENHEIMER et al., 1993). This event was particularly

worrisome, since it occurred near the southern end of

the CSZ and immediately raised the possibility of a

destructive local tsunami. Fortunately, the generated

waves were only a few centimeters high, due primarily

to the strike-slip motion of the source. While strike-

slip faults are not usually thought of as tsunamigenic,

there are examples where strike-slip faults may have

triggered significant tsunamis (IMAMURA et al., 1995).

Also, there is speculation that restraining bends

associated with strike-slip faulting are possible tsu-

nami sources, particularly in Southern California

(LEGG et al., 2004). During the 2005 event, residents

of Crescent City were successfully warned about a

possible tsunami; however, this message failed to

reach residents in other areas on time (GIBBONS, 2005).

On November 15, 2006 an earthquake in the Kuril

Islands produced a tsunami that affected parts of

California (DENGLER et al., 2009). The distance from

the epicenter allowed time for officials to respond.

Once a relatively small tsunami was detected on

deep-ocean tsunameters, the warning was lifted for

the US West Coast, approximately 3.5 h after the

event (DENGLER et al., 2009). At the time, warning

center protocols and procedures did not allow for a

localized watch or warning to be issued for specific

areas of Northern California, such as Crescent City,

which, despite the small tsunami amplitude, remained

vulnerable to damaging currents and surges.

Figure 4Maximum water levels and overland flow depths in San Diego Bay

for a hypothetical Mw 7.6 earthquake on the Catalina Fault. a,

b Show that the extent of the inundated area land is greater when

higher-resolution grids are used. The figure on the left (a) uses a

1-arcsec (*30 m) grid while the figure on the (b) uses a 3-arcsec

(*90 m) grid


The lines of communication between local

experts, emergency management, and the warning

centers remained open, despite the cancelation of the

official warning. Hence, certain low-lying areas were

evacuated, despite the warning cancelation. The

tsunami arrived at Crescent City at 11:30 a.m. local

time and caused approximately US $10 million in

damage. Most of it was confined to the harbor area

and caused primarily by strong currents generated by

the rapid rise and fall of the water levels. It is

important to note that the damaging surges occurred

11 h after the predicted first wave arrival, by which

time many evacuees had returned to low-lying

waterfront areas. While local officials and experts

were prepared and handled the situation appropri-

ately, they were not aware of the possibility of the

delayed response in Crescent City, and the evacuation

order was lifted too soon (BARBEROPOULOU et al.,

2008).

The Chilean earthquake of 27 February 2010

generated a trans-Pacific tsunami that had small

impact in Japan and minimal effects in Hawaii and

the Marquesas Islands, as predicted by NOAA–

PMEL approximately 8 h before the first wave

arrival. While Hawaii opted for a full evacuation as

per their inundation maps, California, which had a

shorter time window, remained at the advisory level.

Several municipalities evacuated local beaches only,

others not at all. What is relevant is that the tsunami

did cause strong and damaging currents in San Diego,

Los Angeles, Long Beach, and Ventura Harbors, but

otherwise had even less impact than in Hawaii.

The tsunami tested the level of education among

managers and first-responders, such as firefighters,

coastguard, and police. In the ports, emergency boats

sailed offshore, to depths that were greater than

150 m, under the impression that this was the

threshold of safety. One co-author (C.E.S.) observed

the response from a Los Angeles County Fire

Department boat in the Port of Los Angeles through-

out the day of 28 February, and was astounded to find

that, once the expected ‘‘landfall’’ time had passed,

the flotilla of boats started heading back to the harbor.

There was no understanding that tsunamis are long-

duration events and that more often than not the

waves following the first one are larger. In fact, at

1:00 p.m., about 45 min after the forecast time for the

first arrival, another Fire Department vessel encoun-

tered strong currents while attempting to reenter the

harbor, which for a few moments, made the vessel

difficult to steer.

On the same morning (28 February), the US

Coastguard continuously broadcasted every 10 min

over emergency radio channels that 6–8 ft (2–2.5 m)

surges were expected in Southern California, but did

not specify where. Many assumed that this meant

walls of water, and having not noticed even a 1 ft

surge, they believed the tsunami was a ‘‘no show.’’

Later in the day, after receiving reports of strong

currents, the Coast Guard decided to stop the

boarding of passengers onto a cruise ship. This

caused a number of heated arguments, with people

questioning why boarding was being stopped at that

time, since the threat was believed to be long over.

Even emergency personnel had difficulty identifying

the rapid 2 ft drop in water level near their station as

being the tsunami; they instead assumed it was the

tide, not noticing that the water level change took

place in less than 30 min.

A follow-up meeting was organized on 4 March

2010 to discuss California’s response to the Chile

tsunami with first-responders and emergency manage-

ment. Key issues identified as problematic were

uncertainty as to the duration of the event, the number

of waves, and the confusion as to what the tsunami

forecasts referred to in terms of heights or amplitudes.

There was no question that, for port operations,

information about the sequence of waves is essential

to allow rescue operations to proceed ‘‘safely,’’ but

along the open coastlines a flooding forecast would

have been more useful than vague announcements

about tsunami heights or amplitudes, which can change

depending on the offshore location they refer to.

The three described earthquakes challenged

assumptions of what tsunamis look like upon reach-

ing the coast, illustrated the extended duration of the

hazard, and tested the readiness of California’s

tsunami response plan. It is quite clear that what

the science community believes is common knowl-

edge is not so for emergency management or the

general public, particularly in terms of the extended

duration of the hazard, which persists well beyond the

first wave arrival, and inferences about tsunami

‘‘heights’’ and ‘‘amplitudes.’’


It is here that inundation maps are very impor-

tant. The maps depict flooding zones, just as the

NOAA–PMEL forecasts are designed to do. What

emergency managers need to know is whether a full

or partial evacuation is warranted, and they can do

that with estimates of offshore heights, even when

there is ample time for decision-making. In the

recent event, Hawaii and California each had over

12 h to prepare. Ports need estimates of currents,

duration, and the possibility of seiche, in real time.

Inundation maps provide estimates of flooding and

can provide maximal currents and serve to plan

evacuations in advance, in terms of routes and

resources, particularly important for near-field

events, when there is a short warning window. By

2012, it is expected that in an actual emergency

there will be real-time flooding forecasts from

NOAA–PMEL including margins of error, and

evacuation decisions will be somewhat easier. Even

then, if emergency management, first-responders,

and the public do not become better informed as to

how to interpret forecasts, the useless evacuation in

Hawaii during the 2010 tsunami may be repeated. In

terms of California, one wonders what will happen

in heavily populated southern beaches after a

moderate to large local offshore earthquake.

5. Conclusions

A new set of inundation maps has been developed

for California. These maps cover the largest coastal

area of any US state. They are deterministic, and their

primary use is for evacuation planning. The release of

these maps offers an opportunity to develop a con-

sistent, statewide tsunami preparedness plan for all

coastal communities in California.

These new products are a result of a multiyear

collaboration between CalEMA, CGS, and USC-

TRC. Because planning for a natural disaster is an

ongoing process, the release of these maps is both a

milestone and a call for improvements in tsunami

preparedness.

For more information regarding tsunami hazards

in California and the new tsunami inundation maps

please visit the following websites:

New statewide inundation maps:

http://www.conservation.ca.gov/cgs/geologic_hazards/

Tsunami/Inundation_Maps/Pages/Statewide_Maps.aspx

University of Southern California—Tsunami

Research Center:

http://www.usc.edu/dept/tsunamis

Acknowledgments

We would like to thank the National Tsunami

Hazards Mitigation Program (NTHMP) for providing

the funds for developing the first and second gener-

ation of maps, and particularly our collaborators at

CalEMA (Jim Goltz, Kevin Miller, Johana Fenton)

and at CGS (Rick Wilson) for their contributions in

the implementation of the inundation into evacuation

maps. We thank Dick McCarthy and Jim Anderson

from the California Seismic Safety Commission for

many interesting suggestions and for providing travel

funds in transition periods. We thank the National

Science Foundation for providing grant support for

Burak Uslu’s PhD thesis.

REFERENCES

BARBEROPOULOU, A. (2008) ‘‘A seiche hazard study for Lake Union,

Seattle, Washington’’, Bulletin of the Seismological Society of

America 98 (4), 1837–1848.

BARBEROPOULOU A., DENGLER L., and USLU B. (2008), Lessons for

California from the November 2006 tsunami in Crescent City,

California, in Solutions to Coastal Disasters, Proc. ASCE,

192-203, doi:10.1061/40978(313)18.

BARBEROPOULOU, A., BORRERO J.C., USLU, B., KALLIGERIS N., GOLTZ

J.D., WILSON R.I. and SYNOLAKIS C.E. (2009), New maps of

California to improve tsunami preparedness, Cover article, EOS

Trans. AGU, 90 (16), 137-138.

BARDET, J.-P., SYNOLAKIS, C.E., DAVIES, H.L., IMAMURA, F. and

OKAL, E.A. (2003), Landslide tsunamis: Recent findings and

research directions, Pure and Applied Geophysics, 160, (10/11),

179–1809.

BERNARD, E.N., and GONZALEZ F.I. (1994), Tsunami Inundation

Modeling Workshop Report (November 16–18, 1993), NOAA

Tech. Memo, ERL PMEL-100 (PB94-143377), 139 pp.

BERNARD, E., MADER, C., CURTIS, G., SATAKE, K. (1994), Tsunami

inundation model study of Eureka and Crescent City, California,

NOAA Tech. Memo, ERL PMEL-103, 80 pp.

BORRERO, J.C. (2002), Tsunami Hazards in Southern California,

Ph.D. Thesis, University of Southern California, Los Angeles,

California, 262 pp.


http://www.conservation.ca.gov/cgs/geologic_hazards/Tsunami/Inundation_Maps/Pages/Statewide_Maps.aspx

http://www.conservation.ca.gov/cgs/geologic_hazards/Tsunami/Inundation_Maps/Pages/Statewide_Maps.aspx

http://www.usc.edu/dept/tsunamis

http://dx.doi.org/10.1061/40978(313)18

BORRERO J, YALCINER AC, KANOGLU U, TITOV V, MCCARTHY D,

SYNOLAKIS CE, (2003), Producing tsunami inundation maps: The

California experience in Submarine Landslides and Tsunamis,

NATO Science Series IV Earth and Environmental Sciences, 21,

315-326, IDS Number: BY68 ISBN: 1-4020-1348-5.

BORRERO J.C., LEGG M.R., SYNOLAKIS, C.E. (2004), Tsunami Sour-

ces in the Southern California Bight, Geophysical Research

Letters, 28 (4) 643-646.

BORRERO, J. C., DOLAN, J. F., and SYNOLAKIS, C. E. (2001), Tsunamis

within the eastern Santa Barbara Channel, Geophysical

Research Letters, 28, 643-646.

CLARKE, S. H., and CARVER, G. A. (1992), Late holocene tectonics

and paleo- seismicity, southern Cascadia Subduction Zone,

Science, 255, 188 – 192, doi:10.1126/science.255.5041.188.

DENGLER, L., USLU B., BARBEROPOULOU A., BORRERO J. and SYN-

OLAKIS C. (2008), The vulnerability of Crescent City, California

to tsunamis generated by earthquakes in the Kuril Islands region

of the northwestern Pacific Seismological Research Letters.

DENGLER, L. (2006), Cascadia earthquake and tsunami scenario for

California’s North Coast: Eos, Transactions, American Geo-

physical Union 87(52), Fall Meeting Supplement, Abstract

GC43A-03.

DENGLER, L., USLU B., BARBEROPOULOU, A., YIM S. C. and KELLY A.

(2009), The November 15, 2006 of Crescent City, Kuril Island

generated tsunami in Crescent City, California. Pure and appl.

Geophys., 166 (2009),37-53.

EISNER, R., BORRERO, C., SYNOLAKIS, C.E. (2001), Inundation Maps

for the State of California, International Tsunami Symposium,

ITS 2001. In: Proceedings of the International Tsunami Sympo-

sium 2001 (ITS 2001), NTHMP Review Session, R-2, Seattle,

WA, 7–10 August 2001, 67-81.

GEIST, E.L., TITOV, V.V., and Synolakis, C.E., 2006. Tsunami: wave

of change, Scientific American, 294, 56{63.

GIBBONS, H. (2005), Brief tsunami warning startles US West coast,

reveals strengths and weaknesses in tsunami preparedness, Sound

Waves monthly newsletter, United States Geolog. Survey, July

2005.

GICA, E., SpILLANE, M., TITOV, V., CHAMBERLIN, C., and NEWMAN,

J. (2008), Development of the forecast propagation database

for NOAA’s Short–term Inundation Forecast for Tsunamis

(SIFT). Tech. Memo. OAR PMEL–139 NTIS: PBB2008-

109391, NOAA/Pacific Marine Environmental Laboratory,

Seattle, WA.

GONZALEZ, F.I., BERNARD E.N., and SATAKE K. (1995), The Cape

Mendocino tsunami, 25 April 1992. In Tsunami: Progress in

Prediction, Disaster Prevention and Warning, Y. Tsuchiya and

N. Shuto (eds.), Series of Advances in Natural and Technological

Hazards Research, Kluwer Academic Publishers, 151–158.

GONZALEZ, F.I., SATAKE, K., BOSS, E.F., and MOFJELD, H.O. (1995),

Edge wave and non-trapped modes of the 25 April 1992 Cape

Mendocino tsunami. Pure and Appl. Geophys, 144 (3/4), 409–426.

GOYA, H. (2005), Stratigraphic and lithologic characteristics of

tsunami deposits and an evaluation of possible paleo-tsunami

deposits in southern California: M.S. Thesis, California State

University, Los Angeles, California.

HOUSTON, J. R. and GARCIA, A. W. (1974), Tsunami runup predic-

tions for southern California coastal communities, U. S. Army

Corps of Engineers, Waterways Experimental Station, 17 p.

HOUSTON, J.R. and GARCIA, A.W. (1978), Type-16 flood insurance

study: tsunami predictions for the west coast of the continental

United States, Waterways Experiment Station, Technical Report

H-78-26, Vicksburg, MS.

IMAMURA, F., SYNOLAKIS, C.E., GICA, E., TITOV, V., LISTANCO E. and

LEE, H.J. (1995), Field survey of the 1994 Mindoro Island,

Philippines tsunami, Pure and Applied Geophysics, doi:

10.1007/BF00874399, V.144,No 3-4.

LANDER,J.F, LOCKRIDGE, P.A., KOZUCH, M.J. (1993), Tsunamis

affecting the West Coast of the United States 1806-1992, NOAA

report.

LEE, H.J., GREENE, G.H., EDWARDS, B.D., FISCHER, M.A. and NOR-

MACK, W.R. (2009),Submarine landslides of the Southern

California Borderland, GSA special papers, V.454,251-269 doi:

10.1130/2009.2454(4.3).

LEGG, M.R., and KENNEDY, M.P. (1979), Faulting offshore San

Diego and northern Baja California. in Abbott, P.L., and W.J.

Elliott, eds., Earthquakes and other perils, San Diego region. San

Diego Assoc. of Geologists, 29-46.

LEGG, M.R., and BARBEROPOULOU, A., (2007), Complex faulting in

the Pacific-North America transform offshore southern Califor-

nia and implications for plate boundary tectonics and

tsunamigenesis: Eos, Transactions, American Geophysical

Union, Fall Meeting, Abstract.

LEGG, M.R., BORRERO, J.C., SYNOLAKIS, C.E. (2004), Tsunami

Hazards from the Catalina Fault Offshore Southern California,

Earthquake Spectra, Vol. 20, No. 3, p. 917 – 950.

LIU, P.L.-F., SYNOLAKIS, C.E. and YEH, H. (1991) Report on the

International Workshop on Long Wave Runup, Journal of Fluid

Mechanics, vol 229, 675-688.

LIU, P.L.-F., WU, T.-R., RAICHLEN, F., SYNOLAKIS, C.E., and BOR-

RERO J. (2005) Runup and rundown generated by three-

dimensional sliding masses, J. Fluid Mech., 536, 107–144.

LIU, P.L.-F., YEH, H. and SYNOLAKIS, C.E. (2008), Benchmark

Problems, in Advanced numerical models for simulating tsunami

waves and runup (ed. Liu, P.L-F.,Yeh H. and Synolakis C.E.),

World Scientific,, pp.3-42.

MCCARTHY, R.J., BERNARD, E.N. and LEGG M.R. (1993), The Cape

Mendocino Earthquake: A local tsunami wakeup call?, Pro-

ceedings of the Eighth Symposium on Coastal and Ocean

Management, Vol. III, 2812–2828.

MCCULLOCH, D. S. (1985), Evaluating tsunami potential, Evaluat-

ing Earthquake Hazards in the Los Angeles Region-An Earth

Science Perspective, USGS Prof. Paper 1360, 375-413.

MURPHY, K. (2005), West Coast Tsunami Warning June 14 2005,

After action report, Oregon Emergency Management, August

2005.

OKAL, E. A., STEIN, S. (2005), Seismology: Speed and size of the

Sumatra earthquake, Nature 434, 581–582, doi:10.1038/434581a.

OKAL, E.A., PLAFKER, G., SYNOLAKIS, C.E. and BORRERO, J.C.

(2003), Near-field survey of the 1946 Aleutian tsunami on Uni-

mak and Sanak Islands, Bull. Seismol. Soc. Amer., V. 93, No. 3,

pp. 1226 – 1234.

OPPENHEIMER, J. EATON 1, JAYKO, A., LISOWSKI, M., MARSHALL, G.,

MURRAY, M., SIMPSON, R., STEIN, R., BEROZA, G., MAGEE, M.,

CARVER, G., DENGLER, L., MCPHERSON, R., GEE, L., ROMANOWICZ,

B., GONZALEZ, F., LI, W. H., SATAKE, K., SOMERVILLE, P., and

VALENTINE D. (1993), The Cape Mendocino, California, Earth-

quakes of April 1992: Subduction at the Triple Junction, Science,

Vol. 261. no. 5120, 433 – 438.

PEDERSEN, G. (2008), A lagrangian model applied to runup prob-

lems, In Advanced numerical models for simulating tsunami


http://dx.doi.org/10.1126/science.255.5041.188

http://dx.doi.org/10.1007/BF00874399

http://dx.doi.org/10.1130/2009.2454(4.3

http://dx.doi.org/10.1038/434581a

waves and runup (ed. Liu, P.L-F.,Yeh H. and Synolakis C.E.),

World Scientific, 2008, pp. 311-318.

SATAKE, K. and SOMERVILLE, P.G. (1992), Location and size of the

1927 Lompoc, California, earthquake from tsunami data, Bull.

Seismol.Soc.Amer., V82,(4), 1710-1725.

SYNOLAKIS CE, MCCARTHY D, TITOV V.V. (1997), Evaluating the

tsunami risk in California and the World Ocean ‘97 –Conference

Proceedings, Editors Magoon OT et al., San Diego, California,

IDS Number BL35F, ISBN 0-7844-0297-3, 1225-1236.

SYNOLAKIS, C.E., and OKAL, E. A. (2005), 1992 – 2002: Perspective

on a decade of post-tsunami surveys, in Tsunamis: Case Studies

and Recent Developments, K. Satake (ed), Adv. Nat. Technol.

Hazards, 23, 1 – 30.

SYNOLAKIS, C.E., BARDET, J.-P., BORRERO, J.C., DAVIES, H.L., OKAL,

E.A., SILVER, E.A., SWEET, S. and TAPPIN D.R. (2002), The slump

origin of the 1998 Papua New Guinea tsunami, Proceedings of

the Royal Society (London), Ser. A, 458, 763-789.

SYNOLAKIS, C.E., and BERNARD, E.N. (2006), Tsunami science

before and beyond Boxing Day 2004. Philisophical Transactions,

364 (1845),, 2231–2265, doi:10.1098/rsta.2006.1824.

SYNOLAKIS, C.E., and KONG, L., 2006, Runup measurements of the

December 26, 2004 tsunami, Earthquake Spectra 22 (S3),

S67S91, doi:10.1193/1.2218371,6, 2006.

Synolakis, C.E., Bernard, E.N., Titov V.V., Kanoglu, U. and

Gonzalez, F. I. (2008), Validation and Verification of Tsunami

Numerical Models, Pure and Applied Geophysics, V. 165(11/12),

2197-2228, doi:10.1007/s00024-004-0427-y.

SYNOLAKIS, C.E.and KANOGLU, U., (2009) Tsunami modeling,

development of benchmarked models, The Sea (Bernard, E.N &

Robinson, A.R., eds), V. 15, Harvard University Press, Cam-

bridge, Massachussetts, ISBN 0674031733.

TITOV V.V. and GONZALEZ, F. I. (1997), Implementation and Testing

of the Method of Tsunami Splitting (MOST). NOAA Technical

Memorandum ERL PMEL – 112. November.

TITOV, V.V. and SYNOLAKIS, C.E. (1998), Numerical modeling of

tidal wave runup, Journal of Waterways, Port, Coastal and Ocean

Engineering, ASCE, 124 (4), 157-171.

TOPPOZADA, T., BORCHARDT, G., HAYDON, W., PETERSEN, M., OLSON,

R., LAGORIO, H., and ANVIK, T. (1995), Planning Scenario in

Humboldt and Del Norte counties, California, for a great

earthquake on the Cascadia subduction zone, California Division

of Mines and Geology Special Publication 115, 157 p.

USLU, B., BORRERO, J.C., DENGLER, L., and SYNOLAKIS, C.E. (2007),

Tsunami Inundation at Crescent City, California generated by

earthquakes on the Cascadia Subduction Zone. Geophysical

Research Letters, Vol. 34, L20601.

USLU, B. (2008) Deterministic and probabilistic tsunami studies in

California from near and farfield sources, Ph.D. Thesis, Uni-

versity of Southern California, Los Angeles, California, 208 pp,

publication number 3341809.

WEI, Y., BERNARD, E., TANG, L., WEISS, R., TITOV, V., MOORE, C.,

SPILLANE, M., HOPKINS, M., and KANOGLU U. (2008) Real-time

experimental forecast of the Peruvian tsunami of August 2007 for

U.S. coastlines. Geophys. Res. Lett., 35, L04609, doi:

10.1029/2007GL032250.

YEH, H., LIU, P., and SYNOLAKIS, C. (eds.). (1996) Long-wave runup

models. World Scientific, Singapore.

(Received January 4, 2010, revised October 16, 2010, accepted October 19, 2010, Published online February 27, 2011)


http://dx.doi.org/10.1098/rsta.2006.1824

http://dx.doi.org/10.1193/1.2218371,6

http://dx.doi.org/10.1007/s00024-004-0427-y

http://dx.doi.org/10.1029/2007GL032250

a second generation of tsunami inundation maps for the state of california

Documents