Have you ever stood on a Pacific Northwest beach, your feet in the sand with the wind whisking by, watching the waves, and wondered, “What if the earthquake happened while I stood here?”
A tsunami would reach the shore in as little as ten minutes after the shaking stops, but… when the shaking starts, the land along the coast will drop in elevation by a few feet, a phenomenon known as co-seismic subsidence. When it does, will the ocean water rush in to cover the beach instantly at that new, lower elevation? Will the land drop slowly over the duration (≈5 minutes) of shaking with the water coming in later—hopefully after the tsunami, leaving time to get to higher ground? Some other scenario all together?
In essence, what might this sequence of events, shown here to explain “ghost forests,” look like for someone standing on the beach when the Cascadia fault ruptures??
For complicated questions like this, I reach out to the experts. They never disappoint! Rather than trying to paraphrase (I couldn’t do this topic justice), I’ve pasted their full answers below, as well as the questions I asked in its entirety.
I’m hoping this group can help me better understand the timing relationship for the shaking, subsidence, and 1st tsunami wave after a CSZ event. For an area expected to drop 5 feet in elevation:
- Does the drop in elevation (subsidence) happen more or less evenly (≈a foot/minute) over the course of the shaking?
- Does the ocean water begin to flood the beach at the same rate of subsidence DURING the shaking so that what sat at 5 feet in elevation before the earthquake sits underwater when the shaking stops?
- Or does the receding ocean start pulling water from the coastline right away so that the beach doesn’t flood until the tsunami wave comes?
- Which brings me to the next question… How soon does the ocean water start to recede after the shaking? Is it slowly over the 10/15 minutes before the first tsunami wave comes, or does it happen in the last ≈2 minutes before the first tsunami wave comes?
- How far out does the water recede, in general? A football field? Is there a visual for how far out the ocean is expected to recede along the coastline, before the water rushes back in?
If there are good visuals to help explain this sequence of events, please let me know! Thank you!!
“I’ll start: land subsidence would happen in less time by far than the shaking, at a given location. Subsidence happens due to the specific local area’s slip on the fault, relaxing the flexed crust above. Someone will have a more specific answer from studies, but my guess would be about 10-30 seconds. Next, yes the nearby water would flood into that subsided area as fast as it can rush in. This is not really part of the tsunami. Third question: ok, first of all, the leading front of the tsunami isn’t necessarily a recession – it could lead with a crest rather than a trough. That depends on specific cases of the offshore seafloor displacement by the quake slip. If it does recede, then that IS the first tsunami wave, in a sense – just happens to be a trough. It will arrive at the coast some minutes after the quake, seemingly pulling the water out. Yes, it can go out hundreds of meters or more, it depends on the nearshore slope.”
—Harold Tobin, Director of the Pacific Northwest Seismic Network and University of Washington Professor
“The issue of flooding due to subsidence is nuanced. The water above subsided crust subsides as well, so that flooding of the subsided area is generated by the gradient of the water surface and so is initiated offshore rather than locally. That implies some delay. If the initial wave is a trough, which depends on the distance from shore as well as the initial condition, then flooding due to subsidence may not happen at all. In any case, it is small (on the order of tidal fluctuations).”
—Tim Walsh, Assistant State Geologist for the Washington State Geological Survey
“I had a discussion like this about the water response with Anne Trehu once. She pointed out that the water would not rush in immediately because the water all subsided with the land at the same time. This in fact forms the trough of the leading tsunami wave, which is when the water rushes in, 15-20 min later, but not before. After a bit of head scratching, I think she is right.”
—Chris Goldfinger, Oregon State University Professor Emeritus
“We have an animation of data from the 2011 earthquake off of Japan that is a great illustration of when the motion occurs. The data show that displacements do not occur until the S waves arrive — it took a while after the earthquake started before anything was detected on land because of the distance. Once the S waves arrived, the displacements happen very quickly in any given location, so they would be early in the period of strong shaking. After that, you just see a mixture of data noise and later seismic phases like surface waves passing through.”
—Jeff Freymueller, Director of the EarthScope National Office and Michigan State University Professor, Endowed Chair for Geology of the Solid Earth
“Jeff Freymueller, this is the animation I was mentally picturing when I wrote my answer – thanks for posting. Using my best playback and pause skills, I see that the vertical subsidence of the nearest (and ultimately largest subsidence) station appears to get going by 90 seconds after origin time (05:47:30 frame) and to be essentially complete and permanent by about 90 seconds after that (05:49:00 frame). Meanwhile strong shaking at that location went on for many minutes.”
—Harold Tobin
“Hmm. I agree with Chris (more or less), but disagree with Harold and John.
Subsidence will begin immediately at the epicenter (which I’m assuming is offshore), but not at the shoreline, because rupture speed is finite. If rupture were instantaneous, then the entire overriding block would subside coherently (meaning subsidence everywhere, including at the shoreline, starting at earthquake origin time). But since rupture speed is finite, and since the overriding block must deform, it will be a while before subsidence begins at the shoreline.
When will shoreline subsidence begin? My guess is when rupture has extended downdip to a point beneath the shoreline. If the epicenter is about 50km offshore, that would be 15-20 seconds after OT.
What about the tsunami? The tsunami will begin radiating away from the epicenter at OT, with the instantaneous source of the radiation expanding as rupture expands. But the tsunami will travel much slower than the rupture so it will be left behind. What you’ll see at the shoreline is a withdrawal starting about 15 seconds after OT, with that withdrawal becoming more and more rapid as components of the tsunami, generated from points farther and farther offshore, arrive. And yes, those components will all arrive as a withdrawal if the only slip is on the décollement (if other faults are letting go at the same time, you might get a positive wave first, but I’m having difficulty imagining how unless you have a landward-verging splay).
Or I could be totally out to lunch.”
—Gerard Fryer, Affiliate Researcher at the University of Hawaii at Manoa and the former Senior Geophysicist for the NOAA Pacific Tsunami Warning Center
“Gerard Fryer, of course we don’t know where a CSZ event will begin, nor where the large slip patches will be, but I was implicitly imagining nucleation near the downdip end, meaning directly beneath the shoreline. Then slip has to propagate up-dip yes, but even more so along strike, like e.g. Sumatra 2004. So the timing of *local* fault displacement within the overall rupture depends where along-strike you are.”
—Harold Tobin
“Tsunami waves are much slower than the seismic waves, so one can think of the rupture as nearly simultaneous, taking just a couple of minutes at the very most, and even less in the trench-perpendicular direction, across only tens of seconds. As several have mentioned, the initial wave mostly depends on the slip distribution underneath the coastline.
A slip gradient such that there is greater amplitude of uplift in an inland direction will tilt the seafloor to make water flow outward away the beach. If, in contrast, the vertical motion is greater toward the offshore, the land will tilt the other way so water initially flows in.
Vertical deformation tends to be proportional to local fault slip, but also in inverse proportion to the depth to the fault plane.
So I think if the big slip asperity is offshore, the uplift offshore tilts so that the initial wave comes inward, if the big slip asperity is onshore, resulting in the opposite tilt, the water first wave moves out to sea.
As Harold says, effects start Immediately as the fault breaks underneath the coast, but the biggest waves come from the biggest motions in deeper water, delayed by the time it takes for the waves to travel in to the coastline (their speed is less than 1 km/s, and way slower just offshore). This leads to 5-20 minute delays after the shaking.Even less intuitive is that motions are generated initially by the gradient in the wave level, not the amount of uplift or drop – water seeks to form a flat surface – only in the longer term does the flat level need to correspond with global sea level.
Still more difficult to imagine is that the waves carry kinetic energy and inertia – these are traveling waves that reinforce and cancel. This is why guys like Tim, Chris, and Randy Leveque resort to suites of scenario calculations and the resulting inundation maps.”
—John Vidale, former Director of the Pacific Northwest Seismic Network and the Southern California Earthquake Center and current Dean’s Professor of Earth Sciences at USC Dornsife
Surviving Cascadia 