DOGAMI Open-File Report Series

[Go back] | Publications Center | DOGAMI Home

Publication Preview

Open-File Report O-22-06, Earthquake and tsunami impact analysis for coastal Lane, Douglas, and Coos Counties, Oregon, by Jonathan C. Allan and Fletcher E. O’Brien; 124 p. report, including data tables and community-specific profiles.

WHAT'S IN THIS REPORT?
This report evaluates a Cascadia Subduction Zone earthquake (MW 9.0) and tsunami (M1, L1, and XXL1 scenarios) affecting coastal Lane, Douglas, and Coos counties, Oregon, to understand the degree of potential destruction, including building losses, debris generated, fatalities and injuries, and estimated numbers of the displaced populations. The goal is to help coastal communities prepare for this inevitable disaster.

PUBLICATION DOWNLOADS

• Report and appendix (124 p., 65 MB PDF), including an appendix of 42 tabloid-page-size community-specific profiles for Florence, Dunes City, Siltcoos River Campgrounds, Reedsport, Winchester Bay, Umpqua South Jetty, Lakeside, Coos Bay, North Bend, Barview, Charleston, Sunset Bay State Park, Bullards Beach State Park, and Bandon.
• Report only (80 p., 3.1 MB PDF)
• Appendix A community profile sheets only (44 p., 62 MB PDF)
• Spreadsheets only (Three Excel spreadsheets, 163 KB .zip file) - contains all the data that are the basis for this report’s tables and figures
• Publication bundle (report, appendix, spreadsheets; 61 MB .zip file)
  
  

EXECUTIVE SUMMARY

This report provides an evaluation of the potential impacts of a Cascadia earthquake and accompanying tsunami in coastal Lane, Douglas, and Coos counties. The analyses presented here include an assessment of the number of people, businesses, and critical facilities located in three Cascadia tsunami inundation zones (M1, L1, and XXL1). XXL1 represents the maximum-considered inundation scenario given our knowledge of the Cascadia Subduction Zone (CSZ). Large (L1) and Medium (M1) tsunami zones reflect smaller earthquake and tsunami scenarios that are more likely to occur than XXL1. L1 captures 95% of the uncertainty in tsunami modeling (there is a ~5% chance that the tsunami could exceed the L1 tsunami zone), whereas the M1 scenario captures 78% of the uncertainty (there is a ~22% chance that the tsunami could exceed the M1 tsunami zone).

A major focus of this study is to provide improved estimates of local population demographics in each community to better understand evacuation challenges that could affect different population groups, as well as socioeconomic impacts associated with a CSZ earthquake and tsunami. The results and analyses presented here reflect a comprehensive effort to document the likely effects the next great earthquake and tsunami will have on all three counties.

We used previously developed physical models of a CSZ earthquake and tsunami, “Beat the Wave” tsunami evacuation modeling, and the recently published Federal Emergency Management Agency (FEMA) Hazus Tsunami Model to develop standardized damage loss estimates for each community, as well as estimates of injuries, fatalities, and displaced population. From the building damage losses, we estimated the amount of debris generated. Our population model improves upon previous studies by providing spatially detailed estimates of permanent and temporary populations — the latter quantifying numbers of visitors, which vary widely throughout the calendar year. The tsunami injury and fatality modeling evaluates a nighttime (2 AM) evacuation scenario, which assumes people are in their homes/hotels/campgrounds at the time of the event (as opposed to on the beach or walking around town). We also maximize visitor occupancy by assuming all hotels/second homes/campgrounds are at capacity, to fully quantify potential impacts to permanent and temporary residents. Our major findings include the following:

• Total populations in coastal Lane, Douglas, and Coos counties that are within a tsunami zone are summarized below:



	Permanent population (M1-XXL1)	Permanent+temporary population (M1-XXL1)
Lane	550-1,870	2,600-6,040
Douglas	1,050-1,970	3,360-5,430
Coos	1,330-10,340	4,970-20,850



• The fraction of permanent residents within the three tsunami zones varies considerably between communities. These variations reflect contrasting patterns in the general shape and elevation of the county coastlines, whether it is open coast versus up an estuary; inundation extents; and the distribution of permanent residents within the communities. Notable observations:
  • Siltcoos, Sunset Bay, and Bullards Beach campgrounds are 100% inundated in all three scenarios
  • For the M1 scenario, communities with the largest number of people in the tsunami zone include Charleston (32%), Winchester Bay (54%), and Umpqua South Jetty (49%).
  • Winchester Bay is mostly located in the M1 tsunami zone and is 100% within the L1 and XXL1 tsunami zones.
  • Barview, Charleston, and Bandon each have relatively large numbers of people located in the XXL1 tsunami zone — 73%, 52%, and 68% respectively.
  • Florence, Dunes City, Lakeside, North Bend and Coos Bay have relatively few people in the various tsunami zones.
    
    
• All 17 communities and parks distributed along the Lane, Douglas and Coos coastlines can experience significant influxes of visitors, well exceeding their local resident populations. Of note, the population of Florence can swell by ~420% to 300% (M1 to XXL1), Winchester Bay by ~1,360% to 950%, and Bandon by ~210% to 225%. The popularity of these communities as centers of tourism present challenges associated with preparing such a large transient population for a CSZ earthquake and tsunami.
• An understanding of how population demographics are geographically distributed within each tsunami zone can provide an insight into those communities that may experience evacuation challenges. We use people over 65 years of age as a proxy for those who may experience increased evacuation difficulty (reduced evacuation travel speeds). Numbers of people over 65 years of age within a particular tsunami zone is summarized below:



	% of population ≥65 years	Number of ≥65 within M1	Number of ≥65 within L1	Number of ≥65 within XXL1
Lane	35% (M1 & L1), 34% (XXL1)	189	324	715
Douglas	34% (M1 to XXL1)	325	526	617
Coos	33% (M1), 31% (L1), 28% (XXL1)	436	914	2,749



• At the community level, Florence, Winchester Bay, and Bandon each have a large proportion (41%) of their resident population ≥65 in the XXL1 tsunami zone.


• The number of buildings located in a tsunami zone is a useful metric for determining exposure to the tsunami hazard. Building counts in the tsunami zones are particularly large in Barview, Bandon, Coos Bay, and to a lesser degree Florence. Interestingly, the largest single number of buildings fall within the “other” category (~2,102) in unincorporated Coos County, reflecting residential buildings established along the open coast outside of community boundaries, as well as along the shores of the Coos and Coquille estuaries. Communities with particularly high exposure to the tsunami hazard include:



% buildings inside the tsunami zone
	M1	L1	XXL1
Winchester Bay	56%	98%	98%
Charleston	58%	59%	70%
Barview	6%	17%	76%



• Building damage caused by earthquake shaking in the three coastal counties is estimated to be:
• Lane County: $1.23 billion
• Douglas County: $420 million
• Coos County: $4.42 billion

• An M1 event could yield damage levels that range from ~10% at Dunes City to ~90% at Bandon and Charleston. Damage caused by the XXL1 tsunami reveals destruction levels of >70% in most coastal communities, including Florence, Reedsport, Winchester Bay, Coos Bay, North Bend, Barview, Charleston, and Bandon; complete destruction occurs at the Siltcoos, Bullards Beach and Sunset Bay campgrounds. These findings can be attributed to the powerful hydraulic forces associated with the tsunami, and the prevalence of light-frame construction material (i.e., wood frame) on the Oregon Coast.
• Combined earthquake and tsunami damage losses for each tsunami zone and scenario are estimated to be significant:



	M1	L1	XXL1
Lane	$1.25 billion	$1.27 billion	$1.36 billion
Douglas	$440 million	$464 million	$530 million
Coos	$4.52 billion	$4.62 billion	$5.14 billion


These estimates reflect community-wide losses associated with the earthquake, combined with destruction caused by the tsunami. Note that these estimates exclude building content losses, such that the numbers may be viewed as minimum estimates.


• The destruction of buildings in coastal Lane, Douglas, and Coos counties is expected to generate substantial debris:



	M1	L1	XXL1
Lane	40,000 tons	50,500 tons	108,000 tons
Douglas	71,200 tons	106,000 tons	149,000 tons
Coos	191,300 tons	358,000 tons	785,000 tons


This equates to ~4,000 dump trucks for M1 in Lane County to as much as 78,500 dump trucks for an XXL1 event in Coos County (assuming dump truck capacity of ~10 yd3). These estimates are almost certainly on the low end, as they do not include debris associated with content from buildings (personal items, business equipment, furniture etc.), road rip-ups, vehicles, and vegetation.


• Modeled tsunami casualties (injuries and fatalities) vary widely between communities. This is due to many factors, but the most important is the relative distance to high ground. We estimate that, combined, countywide fatalities from the tsunami could reflect the following:



	M1	L1	XXL1
Lane	20	50	200
Douglas	610	1,180	1,380
Coos	440	1,070	5,290



• Low casualties associated with the M1 scenario in Lane County is indicative of the fact that high ground is located close to the population centers, allowing for quick access to high ground, or the tsunami simply was not large enough to reach them.
• For the XXL1 tsunami scenario — the largest-considered — the potential for significant fatalities is apparent for Bandon (~1,900), the “other” category in Coos County (~1,400), Winchester Bay (~1,200), and Barview (~980). Overall, the bulk of the fatalities (>61%) are likely to be from the temporary visitor population.
• High casualties associated with the temporary visitor population is predicated on the assumption that these facilities are at 100% occupancy.
• Several additional sites with the potential for large visitor fatalities include Siltcoos River Campground, Umpqua South Jetty, Sunset Bay campground and Bullards Beach campground. Fatalities in these areas are due to a combination of early wave arrivals and long travel distances required to reach high ground.
• These results demonstrate a need to evaluate alternative forms of high ground (e.g., vertical evacuation structures) and/or evaluate retrofitting bridges (e.g., Winchester Bay) to withstand the earthquake shaking, thereby allowing for faster evacuation.


• Following the earthquake and accompanying tsunami, communities will have to deal with many hundreds to potentially thousands of displaced people requiring immediate short-term shelter and care (for days to a few weeks), after which many people are likely to be evacuated from the coast. Hazus modeling indicates that the number of displaced people is significantly higher in the XXL1 scenario (~25,400) compared to the M1 scenario (~9,800). We expect large numbers of displaced people to severely challenge the following communities: Florence, Reedsport, Coos Bay, North Bend, Barview and Bandon. Furthermore, an estimated 4,800 people outside of community urban growth boundaries (UGB) and unincorporated boundaries will require shelter and care.
• Compared to fatalities, injuries from the earthquake were found to be moderately low. Overall, the number of critically injured (requiring hospitalization) as a result of the earthquake is on the order of:
• Lane County: 150
• Douglas County: 30
• Coos County: 380
• Injuries caused by the tsunami are expected to be on the order of:



	M1	L1	XXL1
Lane	~20	5	90
Douglas	~590	40	80
Coos	~180	350	1,700

Although each community in coastal Lane, Douglas, and Coos counties has unique circumstances and challenges, our results unequivocally demonstrate that in every community, injuries and fatalities from a tsunami can be minimized if people evacuate on foot toward safety as soon as possible and travel as fast as possible.