Journal: News & Comment

Thursday, December 30, 2004
# 10:33:00 PM:

Answers to tsunami questions

Permalinks to this entry: individual page or in monthly context. For more material from my journal, visit my home page or the archive.

NOTE: I've combined all my various writings about tsunamis from December 2004 and January 2005, including this entry, into a single tsunami article, which might make a good introductory reference.

Following my earlier posts about the Indian Ocean tsunami, a couple of people have asked questions about it. Here are my responses. Keep in mind that I am not a tsunami expert, just someone with a marine biology degree who did some research on them 15 years ago, and who has maintained an interest in this fascinating but awful phenomenon.

Tsunami "drawdown": why did the water retreat first?

The first was from Petula Brown, who was actually vacationing in Phuket when the tsunami hit and witnessed it first-hand. She wrote:

I am still confused about the water retreating as where we were the water disappeared for five or nearly ten minutes allowing people to wander out to investigate more. Why would the water remain out for so long. The first wave was more like a fast rising tide but the waves to follow were crashing monsters destroying everything in its way. Fortunately the sleeping rooms of our resort were 105 steps up from the beach so people were injured but no one from our resort died.

Dr. Stephen Nelson's very technical page from Tulane University provides an explanation of why the water retreated before the tsunami waves arrived:

If the trough of the tsunami wave reaches the coast first, this causes a phenomenon called drawdown, where it appears that sea level has dropped considerably.  Drawdown is followed immediately by the crest of the wave which can catch people observing the drawdown off guard. When the crest of the wave hits, sea level rises (called run-up). [...]

Because the wavelengths and velocities of tsunami are so large, the period of such waves is also large, and larger than normal ocean waves.  Thus it may take several hours for successive crests to reach the shore.  (For a tsunami with a wavelength of 200 km traveling at 750 km/hr, the wave period is about 16 minutes).  Thus people are not safe after the passage of the first large wave, but must wait several hours for all waves to pass. The first wave may not be the largest in the series of waves. For example, in several different recent tsunami the first, third, and fifth waves were the largest."

Essentially, the low water level is just as much a part of the tsunami as the big wave, because tsunamis are waves with very long wavelengths, and waves have both crests (the peak of the tsunami wave) and troughs (the bottom of the same wave).

If the trough reaches shore first, the water level will lower dramatically, far below the normal lowest tide, before the crest comes rushing in. Some areas will encounter the peak first, some the trough, which is why low water levels are not always present before tsunami waves arrive. Following the crest that caused the initial destruction, the following trough is what dragged people and debris back out to sea, followed by the subsequent waves.

How big does an earthquake have to be to create a tsunami?

It's not so much how big, as what kind and where. Obviously, a huge earthquake far inland can't create a tsunami in the ocean, but Walt Davis had an interesting question:

I read your article on tsunami wave and found the information useful. I read a U.S. Goverment paper that stated that only a 7.5 quake or larger can produce a tsunami. Do you know if this is corrrect?

As far as I understand it, that's not true, because it doesn't have to be an earthquake that causes a tsunami, and it has more to do with the type and location of the earthquake as well as the topography of the sea bottom and shoreline than earthquake magnitude.

A tsunami can arise (at least locally) from something like a landslide in a lake or fjord. The key is that a significant amount of water has to be displaced suddenly—if it is confined to a narrow channel, "significant" may not be all that much, compared to big tsunamis like that in the Indian Ocean. Some examples from my part of the world:

  • A landslide with no significant associated earthquake generated waves that ran up 150 metres (450+ feet!) in (uninhabited) Lituya Bay, Alaska, in 1936.
  • A similar event, also in Alaska, as recently as 1994 created waves 7.6 m (25 feet) high, and there are lots more on this page, plenty of which are below 7.5 magnitude.
  • At the extreme end, an 8.2 magnitude quake in 1958 in southeastern Alaska didn't create a Pacific-wide tsunami, but created local tsunamis that ran up more than 500 metres (1500 feet) above sea level (the largest ever recorded)—again at Lituya Bay.

You could, however, argue that many of these events are simply monstrous waves, and not really tsunamis—if your definition of a tsunami is that it must be not only big, but energetic enough to cross significant stretches of ocean. It may be that a large, ocean-wide tsunami created by a seafloor quake alone does require a magnitude of 7.5+ to generate enough energy, but there are also many other circumstances (volcanic eruptions, meteor impacts) that could do so otherwise.

In addition, a big quake that moves laterally (sideways, not up or down) can be way above 7.5 magnitude, but not cause a tsunami—which is why the 1906 San Francisco quake and later smaller ones in southern California, which occur on the lateral-moving San Andreas fault, haven't created tsunamis.

Risky geology off the Pacific Northwest coast?

Finally, Walt also asked:

Another questain is about a underwater fault line that runs off the Northwest coast. Do you happen to know the name of it.

There are five separate tectonic plates in this region: in order of size, the Pacific Plate, the North America Plate, the Juan de Fuca Plate, the Explorer Plate, and the South Gorda Plate.

The two major north-south lateral faults are the San Andreas, running through California and reaching the sea at San Francisco Bay, and the Queen Charlotte, running just west of the Queen Charlotte Islands. There are ocean spreading ridges between some of the plates, and the Cascadia Subduction Zone forms a trench just offshore from Vancouver Island and the Washington, Oregon, and northern California coasts.

I guess the best name for what you're looking for is the Cascadia Trench or Cascadia Subduction Zone—that's the region most prone to a strong vertical earthquake, and to potentially generating a tsunami.


Journal Archive »

Template BBEdited on 29-Apr-2010

Site problems? Gripes? Angst? - e-mail
Site contents © 1997–2007 by Derek K. Miller

You may use content from this site non-commercially if you give me credit, under the terms of my Creative Commons license.

eXTReMe Tracker