The Pacific Northwest Megaquake: Cascadia Worst-Case Scenarios

At approximately 9 PM on January 26, 1700, the ground along the entire Pacific Northwest coast heaved upward and then dropped. A rupture 1,100 kilometers long tore open along the Cascadia subduction zone — the interface between the Juan de Fuca plate and the North American plate — releasing energy equivalent to a magnitude 9.0 earthquake. The sea retreated from beaches that had never seen it behave that way. Then it came back. Waves 10 to 20 meters high swept inland across the low-lying coastal terrain of what is now Oregon and Washington, drowning the cedar forests, inundating the villages of the Chinook, Tillamook, and Huu-ay-aht peoples, and generating a trans-Pacific tsunami that flooded rice paddies in Japan and was documented with enough precision for Japanese scholars to eventually calculate the earthquake's exact date and approximate magnitude — and for modern scientists to use those Japanese records to precisely date the event 326 years after it happened.

No one alive today in the Pacific Northwest has felt a Cascadia megaquake. No seismograph has ever recorded one. The region's entire modern history — its cities, its ports, its Interstate bridges, its hospitals and schools and water treatment plants — was built during a 326-year window of seismic quiet that the geological record shows is unusual only in the sense that it eventually ends. It will end. The Cascadia subduction zone has produced great earthquakes on average every 200 to 500 years, and in the southern section every 200 to 300 years. The 30-year probability of a full-margin M9+ event is approximately 10–15%; of any Cascadia rupture M8.0+, approximately 37%. These are not remote possibilities — they are planning parameters for the worst natural disaster in the history of the United States, likely to occur within the lifetimes of millions of current residents.

This post covers the worst-case: a full-margin Cascadia rupture of the type that occurred in 1700. Not because partial ruptures are less real or less important — they are not — but because the full-margin scenario defines the upper envelope of consequence, and because understanding that envelope is the foundation of every preparedness and resilience investment the Pacific Northwest is currently making or failing to make.

The Cascadia Subduction Zone: Geometry and Behavior

The Cascadia subduction zone spans approximately 1,100 kilometers from the Mendocino Triple Junction in northern California — where the San Andreas Fault, the Gorda plate, and the North American plate meet — northward to the northern end of Vancouver Island and the Queen Charlotte Fault system at the southern tip of Alaska. Along this entire length, the small Juan de Fuca plate (and its southern extension, the Gorda plate) is converging with and subducting beneath North America at approximately 35–40 mm per year.

Unlike many subduction zones — the Japan Trench, for instance, which has a long history of instrumentally recorded great earthquakes — Cascadia has produced no great earthquake in the 170 years of European-American settlement, and no instrumentally recorded megathrust rupture at all. Until the 1980s, some geologists even argued that Cascadia might be incapable of great earthquakes — that the young, warm Juan de Fuca plate was too buoyant and the subduction too slow to generate the locked, seismically coupled interface required for an M8+ event. That view was demolished by the discovery of the paleoseismic record.

The Paleoseismic Record: Proving the Hazard

The scientific transformation of Cascadia from a debated hazard to a recognized catastrophic threat happened primarily in the 1980s and early 1990s, driven by three researchers and three independent lines of geological evidence: Brian Atwater's discovery of buried coastal marshes, Kenji Satake's Japanese tsunami analysis, and Alan Nelson's stratigraphic work on tsunami deposits at multiple coastal sites.

Atwater's key discovery, made at Willapa Bay in Washington state, was a sequence of buried soil horizons separated by sand layers — each buried soil representing a coastal marsh that had been suddenly dropped below sea level by the subsidence accompanying a great earthquake, then buried by the tsunami that followed. The pattern of sudden subsidence (a meter or more, occurring instantaneously) followed by gradual re-emergence over centuries is the characteristic geological signature of a megathrust earthquake cycle at a subduction zone: the land subsides when the locked fault finally ruptures, and gradually re-emerges as the fault re-locks and the crust rebounds elastically in the interseismic period.

Atwater found not one buried soil but many, stacked atop each other in sequence, each representing a separate megathrust event. The stratigraphic record at Willapa Bay and subsequently at dozens of other coastal sites from northern California to Vancouver Island has now documented approximately 40 great earthquakes over the past 10,000 years — an average recurrence interval of roughly 240 years for events affecting at least the southern half of the subduction zone, and roughly 500 years for full-margin events affecting the entire length from California to British Columbia.

Full-Margin vs. Southern-Section Ruptures

One of the most important findings from the paleoseismic record is that Cascadia ruptures come in two varieties with very different consequence profiles. Full-margin ruptures — like the 1700 event — involve simultaneous rupture of the entire 1,100-kilometer length of the subduction interface, producing M9.0 or larger earthquakes and generating tsunamis that affect the entire Pacific Northwest coast from Northern California to British Columbia. These events recur roughly every 500 years on average, with large variability (intervals ranging from roughly 200 to 1,000 years in the paleoseismic record).

Southern-section ruptures — involving only the portion of the Cascadia megathrust south of approximately 48°N latitude (roughly the Olympic Peninsula) — produce M8.0–8.7 events and generate tsunamis that affect primarily the Oregon coast and Northern California. These events recur more frequently — roughly every 200–250 years in the southern section — and the last one preceded the 1700 full-margin rupture by approximately 300 years. The more frequent recurrence of southern-section events means that the 30-year probability of any Cascadia rupture significant enough to be catastrophic is higher than the full-margin rupture probability alone — approximately 10–15% for a full-margin M9+ but roughly 37% for any Cascadia M8+ event.

The Ground Motion Scenario: What Shaking Looks Like

A full-margin Cascadia M9.0 would produce ground shaking qualitatively different from any earthquake experienced in the Pacific Northwest in modern times. The closest analogue felt by people currently living in the region was the 2001 M6.8 Nisqually earthquake — which lasted about 45 seconds and caused $2 billion in damage. A full-margin Cascadia event would shake the region for 4 to 6 minutes, with peak ground motions 3 to 10 times more intense than Nisqually across most of the urban corridor.

Inland Cities: Portland, Seattle, Victoria

Portland, Oregon — the largest city in the Pacific Northwest — sits approximately 120 kilometers from the Cascadia trench. Physics-based ground motion simulations for a full-margin Cascadia rupture project Modified Mercalli Intensity (MMI) of VIII (severe) to IX (violent) throughout the Portland metropolitan area, with shaking duration of 3–5 minutes. The Willamette Valley, in which Portland sits, is underlain by deep alluvial and lacustrine sediments that will amplify the long-period seismic energy characteristic of a distant great earthquake — particularly dangerous for tall buildings and large bridges whose natural periods match the dominant period of the amplified ground motion.

Seattle sits at similar distance (120–140 km) from the trench and faces the additional complication of the Seattle basin — a deep sedimentary basin beneath downtown and much of King County that amplifies long-period ground motion by factors of 3–5 and extends shaking duration well beyond the source duration. Seattle also faces a secondary hazard from the Seattle Fault — a shallow crustal fault directly beneath the city — which could be triggered by Coulomb stress transfer from the Cascadia megathrust rupture, producing a local M7.0+ secondary earthquake in the hours to days following the main event.

Ground Motions at Specific Distances

USGS Hazus models and physics-based simulations for the full-margin Cascadia scenario project the following representative ground motion intensities:

The Tsunami: The Catastrophic Second Event

Ground shaking will kill people in the Pacific Northwest in a full Cascadia rupture. But the earthquake's most lethal consequence — and the one that separates the Cascadia scenario from other American earthquake scenarios — is the tsunami. When the locked Cascadia megathrust ruptures, the seafloor above the rupture zone will lurch upward by several meters as the elastic strain accumulated over centuries is released. This sudden vertical displacement of thousands of square kilometers of seafloor displaces an enormous column of water — generating tsunami waves that propagate outward from the rupture zone in all directions, reaching the coasts of Japan and Hawaii in hours and reaching the Pacific Northwest coast in 15 to 40 minutes depending on location.

Wave Heights and Inundation

NOAA and Oregon Department of Geology and Mineral Industries (DOGAMI) have produced the most detailed tsunami inundation modeling for the Cascadia scenario. Their models, based on a range of rupture scenarios from southern-section-only to full-margin M9.1, project the following wave heights at coastal communities:

• Seaside and Gearhart, Oregon: 8–20 meters wave runup, with nearly complete inundation of the historically developed coastal flat extending inland 1–3 kilometers
• Cannon Beach, Oregon: 10–25 meters, with the flat areas behind the beach completely inundated to the base of the surrounding headlands
• Newport, Oregon: 8–15 meters at the coast, with Yaquina Bay experiencing surge and reverse flow up the estuary
• Astoria, Oregon: 6–12 meters along the Columbia River estuary, with potential for tsunami energy to propagate 40–80 kilometers upriver
• Westport, Washington: 10–20 meters, with essentially the entire developed area of the Grays Harbor spit inundated in the worst-case scenario
• Ocean Shores, Washington: 8–15 meters, inundating the Ocean Shores peninsula virtually completely

These wave heights, expressed as runup (the maximum elevation reached by the wave), do not mean that a 20-meter wall of water approaches the coast — the waves arrive as a series of rapid tide-like surges, with the first wave reaching the coast as a withdrawal followed by a rapid, turbulent inflow of fast-moving, debris-laden water. The current from the inflow typically reaches 5–10 meters per second — fast enough to sweep away vehicles, collapse most structures, and kill anyone caught in the inundation zone.

Coastal Community Isolation: Cutting Off the Oregon and Washington Coast

Beyond the immediate inundation, the tsunami and ground shaking will sever virtually every transportation link between the Pacific Northwest coast and its inland cities. All of the major highway routes crossing the Coast Range — US-26, US-30, OR-6, OR-22, OR-18, US-101 crossings — will be affected by slope failures, bridge damage, and road washouts from both ground shaking and tsunami inundation. Oregon's coastal Highway 101 itself runs directly through the tsunami inundation zone for much of its length — the road will cease to exist in many segments immediately after the event.

The consequence is that every surviving coastal community west of the Coast Range will be isolated — cut off from emergency medical care, emergency management, and resupply by land for days to weeks after the event. Communities that have prepared for this — that have pre-positioned emergency supplies, trained community emergency response teams, and built vertical evacuation structures — will be able to sustain their survivors until helicopter and maritime access allows resupply. Communities that have not will face a survival situation, not merely a disaster response.

The FEMA Cascadia Rising Scenario: The Full Consequence Picture

The Cascadia Rising exercise, conducted by FEMA Region X in 2016 with participation from Oregon and Washington emergency management agencies, the U.S. military, and dozens of local governments, used a full-margin M9.0 Cascadia scenario to test regional response capacity and identify planning gaps. The scenario inputs — developed from USGS ground motion models and DOGAMI tsunami inundation maps — produced the following headline findings:

• Approximately 13,000 deaths in the immediate period (first 72 hours), concentrated on the Oregon and Washington coasts from tsunami inundation and in inland communities from structural collapse
• Approximately 27,000 injuries requiring medical care, in a region whose hospital system will itself be damaged and partially non-functional
• Approximately 1 million people displaced from their residences — the largest mass displacement event in American history by a factor of three
• Nearly complete destruction of the coastal highway system and major damage to I-5 corridor bridges, with reconstruction timelines measured in years
• Three to six months before I-5 is fully operational; one to three years before all coastal communities are reconnected to inland areas by road
• Total economic losses in the range of $138–$270 billion (2016 dollars), with subsequent estimates using updated models suggesting losses above $300 billion

Infrastructure: The Cascading Failures That Shape Recovery

The full consequence picture of a Cascadia megaquake is determined less by the initial deaths and injuries — as horrific as those are — than by the infrastructure failures that govern how long the region remains non-functional and how quickly the 1 million displaced people can return to normal life. The critical systems are transportation, water, energy, and communications, and each faces Cascadia-specific vulnerabilities that are more severe than any scenario previously planned for in the United States.

The I-5 Corridor: America's West Coast Spine

Interstate 5 — the primary highway connecting Seattle, Portland, and the San Francisco Bay Area — crosses the Columbia River on bridges that have been assessed as seismically vulnerable and potentially non-functional after a Cascadia event. The I-5 Columbia River crossing between Portland, Oregon and Vancouver, Washington is a critical chokepoint: two parallel bridges carrying all northbound and southbound I-5 traffic, both built in 1917 (the original structure) and 1958, both lacking modern seismic isolation, and both crossing a river that will experience tsunami bore propagation up to 80 kilometers from the ocean after the megaquake. ODOT's seismic vulnerability assessment has identified these bridges as among the highest-priority retrofit needs in the Oregon transportation system, and preliminary engineering for seismic retrofit has been funded, but as of recent assessments the bridges remain vulnerable.

Fuel Supply Disruption

The Pacific Northwest's petroleum fuel supply is particularly vulnerable to Cascadia because the region's primary fuel terminal infrastructure is concentrated in a small area of Portland's northwest industrial waterfront — directly in the tsunami inundation zone for the Columbia River estuary. The tank farms, pipeline terminals, and marine loading facilities serving Oregon and Washington's fuel supply are collectively the most significant concentration of critical energy infrastructure at tsunami risk anywhere on the West Coast. A full-margin Cascadia rupture could disable the region's ability to receive and distribute petroleum fuel — gasoline, diesel, and jet fuel — for months, directly limiting emergency response, utility restoration, and reconstruction activities that depend on heavy equipment.

Water Systems

Portland's Bull Run watershed water system — serving approximately 1 million customers — draws water from reservoirs in the Mount Hood National Forest and transmits it to the city through a pipeline corridor that crosses multiple stream valleys with bridges and culverts that will be damaged by ground shaking. The system's primary pipeline parallel to the Columbia River Gorge runs through terrain subject to significant slope failure hazard in a major earthquake. Portland Water Bureau's seismic vulnerability assessment projects significant water system damage in a Cascadia event, with restoration taking weeks to months in the most affected portions of the distribution network.

The Coastal Communities: Survival, Not Response

For the 140,000 people living in Oregon's coastal tsunami hazard zone — the area expected to be inundated in a full-margin Cascadia rupture — the earthquake response framework is fundamentally different from that for inland communities. There will be no time for organized evacuation, no effective warning beyond the earthquake itself, and no help arriving from outside for days. The planning concept for coastal communities is not response — it is survival followed by self-sufficiency until external assistance can reach them.

Vertical Evacuation

Where horizontal evacuation to high ground is not possible within the available time — as is the case for flat, low-lying communities like Seaside, Ocean Shores, and the Grays Harbor peninsula — vertical evacuation structures provide the only alternative. Oregon and Washington have invested in designing and building tsunami vertical evacuation structures (TVESs): reinforced concrete buildings specifically designed to withstand tsunami forces and wave overtopping, located close enough to at-risk populations to be reachable on foot in 10–15 minutes.

Ocosta Elementary School in Westport, Washington — completed in 2016, the first school in the United States specifically built to serve as a tsunami evacuation refuge — can shelter 1,000 people on its elevated roof and upper floors, designed to withstand 3-meter overtopping wave loads. Similar structures have been built or are under construction in Cannon Beach, Seaside, and other Oregon coastal communities. But the current inventory of completed TVESs falls far short of the total capacity needed to shelter the full at-risk population in all vulnerable communities — particularly in the lower-income unincorporated coastal areas where capital investment for public safety infrastructure is most difficult to secure.

Seaside, Oregon: The Hardest Case

Seaside, Oregon — a coastal resort city of approximately 7,000 permanent residents at the mouth of the Necanicum River — is widely considered the most challenging community for Cascadia tsunami preparedness in the United States. The city sits almost entirely within the projected tsunami inundation zone: most of the developed area is less than 5 meters above sea level, the nearest high ground is 2–3 kilometers away, and the primary evacuation routes (US-26 east and US-101 north and south) both lead through inundation zones before reaching high ground. The projected tsunami runup from a full Cascadia rupture reaches 8–20 meters at Seaside — enough to overtop most multi-story buildings and to sweep the city floor completely.

The Seaside School District's decision to relocate all of its schools to high ground — a project completed in 2019 at a cost of $100 million funded by a combination of local bonds and FEMA Hazard Mitigation Grant funds — is the most consequential single tsunami preparedness investment in Oregon's history. On a school day, the previous school sites would have held more than 1,000 students and staff in the middle of the inundation zone. The relocation to high ground near the Necanicum Ridge eliminates the most concentrated life-safety exposure in the most vulnerable community on the Oregon coast.

Slow Slip and the ETS Connection: Is the Clock Ticking Faster?

As discussed in the context of slow earthquakes, the Cascadia subduction zone produces regular episodes of episodic tremor and slip (ETS) — slow slip events at the base of the locked zone that release M6.5–7.0 equivalent energy silently every 12–16 months. Each ETS event transfers a small additional stress increment to the locked megathrust above it. The question of whether any specific ETS episode could trigger megathrust rupture is not answerable with current science — the stress increments from individual ETS events are small compared to the total stress deficit accumulated over 325 years of interseismic loading. But the cumulative effect of 20–25 ETS cycles since 1700, each incrementally loading the locked zone, contributes to the earthquake's overall probability.

No current scientific evidence suggests an elevated near-term Cascadia rupture probability relative to the long-run average. The PNSN (Pacific Northwest Seismic Network) monitors the ETS cycle and publishes updates with each new event — none of the observed ETS characteristics (recurrence interval, duration, spatial extent, associated tremor intensity) show the kind of anomalous behavior that would indicate a transition toward megathrust failure. The probability remains at its long-term average: not imminent by any scientific measure, but not remote by any honest accounting.

What Has Been Done: Progress Since the Scientific Wake-Up

The Cascadia hazard was effectively unknown to the general public before Kathryn Schulz's 2015 New Yorker article "The Really Big One" — which won the Pulitzer Prize and introduced millions of Americans to a risk they had never heard discussed. In the decade since, the pace of preparedness investment in Oregon and Washington has accelerated substantially, even if the gap between current state and what the scenario demands remains large.

Oregon has completed its first comprehensive statewide resilience plan for a Cascadia earthquake, establishing prioritized investment targets for transportation, water, energy, and healthcare infrastructure. Washington has funded several rounds of school seismic safety assessments and has begun mandatory retrofits of the highest-risk K-12 buildings. ODOT and WSDOT have prioritized seismic retrofit of the most critical bridge crossings on the I-5 and I-205 corridors, with the Columbia River bridges under active engineering study. FEMA has funded multiple rounds of hazard mitigation grants for coastal TVES construction, school relocations, and community preparedness programs.

The Oregon Resilience Plan, published by the Oregon Seismic Safety Policy Advisory Commission in 2013, established explicit resilience targets: critical transportation routes restored within 1 month, hospitals operational within 3 days, critical utility corridors operational within 1 week. Achieving those targets requires capital investment of approximately $30–$40 billion over 50 years — an enormous but not unprecedented infrastructure commitment. Current investment rates fall far short of what the timeline requires.

The Inland Preparedness Calculus: Portland and Seattle

For the 5 million people in the Portland and Seattle metro areas — who will experience severe shaking but no tsunami — the Cascadia scenario demands a different preparedness framework than coastal survival planning. The challenge for inland communities is not immediate physical survival (most people in modern buildings in the I-5 corridor will survive the shaking) but functional recovery: how does a metropolitan area of millions remain functional for weeks to months when its water system is damaged, its fuel supply is interrupted, many of its bridges are closed, and it is simultaneously responding to the catastrophic destruction of its coastal neighbors?

The honest answer is that Portland and Seattle are not currently prepared for this. The building stock includes significant quantities of unreinforced masonry (Oregon has no mandatory URM retrofit law; Washington has some provisions but incomplete coverage), concrete tilt-up buildings, and soft-story wood frames that will perform poorly in 3–5 minutes of sustained shaking. The water systems have known seismic vulnerabilities that are being addressed on decade-long timelines. The bridge networks have been assessed but not yet fully retrofitted. The cultural and institutional knowledge of earthquake response is substantially lower than in California, reflecting the absence of damaging earthquakes in the modern era of these cities.

Conclusion: The Scale of the Problem and the Adequacy of the Response

The Cascadia subduction zone will produce another great earthquake. The timing is unknown — it could be tomorrow, or in 200 years — but the 30-year probability of a full-margin M9+ event is roughly 10–15%, and the probability of any damaging Cascadia megathrust rupture in 30 years is approximately 37%. These are not alarming numbers in the sense of imminent threat; they are sobering numbers in the sense of planning horizons. A 37% probability over 30 years means that an infrastructure investment made today — a seismically isolated bridge, a relocated school, a TVES built on solid ground — has a better-than-one-in-three chance of performing its life-saving function within the lifetime of the people who built it.

The FEMA estimate of 13,000 deaths is not a fixed number — it is a function of preparedness. The 2004 Indian Ocean tsunami killed 230,000 people across 14 countries, in part because coastal communities had no warning system, no evacuation culture, and no vertical evacuation structures. Japan's Tohoku coast — which had all three, despite the 2011 tsunami's waves exceeding many communities' designed protection levels — had casualty rates an order of magnitude lower per unit of inundation. The difference between 2,000 deaths and 20,000 deaths in the Cascadia scenario is not geology — it is the behavioral and infrastructure investments made in the decades before the rupture.

The Pacific Northwest has 326 years of interseismic quiet behind it and an unknown number of years ahead. The geological record says those years will end. The question the region faces every year it defers investment — in bridge retrofits, school relocations, TVES construction, URM policies, community training — is not whether the money spent will be needed. It is whether the earthquake will wait long enough for the spending to be completed.