I often get questions about earthquake hazards from people who learn about my background as a federal government seismologist. So I thought a short primer might be valuable given coverage of recent earthquakes.
The most devastating large earthquake to appear in recent news accounts was the Tohoku, Japan, earthquake of March 2011 that caused extensive damage and loss of life from seismic shaking and a massive tsunami, and radioactive leakage from the Fukushima nuclear power plant. At seismic magnitude greater than 9, this was a classic earthquake of the type that occurs at “subduction zones”, where one tectonic plate is sliding under another. Similar sized earthquakes have occurred in recent history on the coast of Chile in 1960, on the coast of Alaska in 1964 and on the coast of Sumatra in 2004.
Staying in Japan and going to a more “local” earthquake, consider the Kobe earthquake of about seismic magnitude 7 that occurred in 1995. This earthquake caused billions of dollars in damage and took about 5000 lives. Japan has state-of-the-art seismic zoning maps in its building codes and uses state-of-the-art earthquake-resistant building design and construction practices. But these could not protect completely against this earthquake because it was unexpected that it would occur directly under the city.
The Japanese seismic zoning maps no doubt considered that such an earthquake was likely within the general region, but it is uneconomical to design and construct buildings to withstand its occurrence under all locations. This is called “acceptable risk”. When that type of earthquake occurs under a third-world city it kills 50 000 people, not 5000, because of the lack of effective design and construction practices. Potential earthquakes of this type, recently discussed in the media, can occur under Los Angeles, away from the San Andreas Fault, and under Vancouver, away from the Cascadia Subduction Zone.
Earthquakes of the Tohoku type occur on the Cascadia Subduction Zone which extends off the coast from northern California to mid Vancouver Island. The last one to occur was on January 26, 1700. This was prior to the time of European settlement, but the earthquake is known in First Nations’ history. Its date and approximate time is known from historical records of the arrival of the resulting tsunami on the coast of Japan.
The geological evidence of large earthquakes on the Cascadia subduction zone is of two main types: tsunami deposits on marine marshes and submarine landslides caused by large earthquakes. When a large earthquake occurs on a subduction zone it sends a large tsunami across the ocean, but it also sends a strong tsunami directly ashore. This local tsunami arrives on shore within 10-15 minutes, so there is little time for a warning to get to higher ground. The tsunami wave picks up sand from the ocean floor and when it washes ashore it deposits the sand on the coastal marine vegetation. Digging down through these layers geologists can date several previous sand layers on top of older marine vegetation, representing previous megaquakes. At a site on the Washington State coast evidence can be seen of five earthquakes dating from 1000 BC to 1700 AD.
A large river like the Columbia River, which reaches the sea about mid way along the Cascadia subduction zone, carries a large amount of sediments and deposits them on the continental shelf. When the sediment load at the edge of the continental slope get too great it becomes unstable and flows down the continental slope as a turbidity current, a dense flow of sediment-laden water that moves rapidly down into the deep ocean.
The sediments deposited by the turbidity currents, called turbidites, can be detected in submarine cores taken from the sea floor. Scientists have been able to correlate the turbidites from one locality to another along the Washington and Oregon coast, suggesting they have a common cause.
Contemporaneous submarine slope failures over such a large area could only be caused by large earthquakes. The current count is 18 events over 10,000 years. The dating of these events is helped by a time horizon in the cores, the volcanic ash from the eruption of Mount Mazama (now Crater Lake, Oregon) 7,400 years ago.
The average interval between these event is about 500 years, but the intervals have varied from about 300 to 800 years. We are just over 300 years from the last one. Seismologists do probability calculations with these numbers. One recent one that I have seen is a 12 per cent probability of occurrence in the next 50 years. (Just remember: The next 50 years includes tomorrow.)
The real story is much more complicated than I can put in this short account; that’s why people write whole books about it. The whole subduction zone does not always break in a great earthquake. Smaller portions of the zone often break in smaller earthquakes. For example, a recent magnitude 8 on the coast of Chile broke the northern end of the zone that broke during the great earthquake of 1960. This makes it difficult to count the large earthquakes and to attribute earthquakes to the geological evidence.
As noted, there can be Kobe-type earthquakes under coastal B.C. cities like Vancouver. The most recent one of this type was magnitude 6.8 centred near Nisqually, Washington, in February 2001. It was felt strongly in Vancouver and lightly in the Okanagan, but did no damage in Canada. A similar earthquake occurred near Courtenay on Vancouver Island in 1946; it did considerable damage but there were no fatalities. These types of earthquakes are caused by the stresses inside the overlying continental tectonic plate caused by the subducting oceanic plate.
All of the earthquakes affecting Canada are accommodated in the Seismic Zoning Maps in the National Building code. But, remember the Kobe experience: It is uneconomical to design for the strong local earthquakes everywhere. A magnitude 7 directly under Vancouver would cause significant damage and loss of life.These zoning maps present degrees of ground shaking throughout the country to which common building must be designed and constructed. The zoning maps that the author was responsible for developing for the 1985 edition of the Canadian building code used a probability of 10 per cent that the ground shaking shown on the maps would be exceeded in 50 years. This level of probability of exceeding the design levels is a societal choice that balances cost of construction against public safety.
Moving inland away from the effects of the Cascadia Subduction Zone into the Okanagan, we need to invoke broader plate tectonic movements to explain our much smaller local earthquakes.
There are numerous faults that geologists can map in the surface rocks of the Okanagan and many more that can be inferred to occur at depth; but these faults have not moved significantly for millions of years. They may have jiggled slightly during the accumulation of thick ice during the glacial period about 100,000 to 10,000 years ago. The small earthquakes are occurring on these faults but a one-to-one correlation between an earthquake and a particular fault has not been made.
The motion on the old faults is caused by the stresses that are applied to them by the slow movement (about five mm/year) of the North American tectonic plate away from the spreading centre along the mid-Atlantic ridge.
The magnitude 6.6 earthquake on the west coast of Vancouver Island on 23 April, 2014 was noticed in highrise buildings in Kelowna. This earthquake occurred just on the northern edge of the Cascadia subduction zone. Whether it was a subduction event or a more common strike-slip event will be determined by further analysis.
The seismic waves generated by an earthquake cover a wide range of frequencies. The high frequency waves attenuate quite quickly away from the epicenter; the low frequency waves travel quite far with little attenuation. When the low frequency waves match the resonant frequency of a highrise building, the building will respond with noticeable effects inside, but at this distance no structural damage. This is what happened on 23 April.
When I was a child sitting in our living room in Westbank on 22 August, 1949, a large picture on the wall began swaying back and forth very quietly; there was no other vibration or movement. We were seeing the effect of the low frequency waves from a magnitude 8.1 earthquake north of Haida Gwaii, about 1500 km away, and the largest earthquake in Canadian history.
If I may make one suggestion for further reading, it is the book At Risk: Earthquakes and Tsunamis on the West Coast (2006) by John Clague et al., published by Tricouni Press, Vancouver, and available in local book stores.
Peter Basham is a seismologist living in Kelowna.