The Earth’s crust seems pretty quiet most of the time. Although we now know that the Puget Sound region is seismically active, you and I can drive from Portland, Oregon, to Vancouver, B.C., along Interstate 5 and never feel an earthquake.
I was a graduate student at the University of Washington in the 1950s, and I never thought about earthquakes. If I had arrived in Seattle a half-dozen years earlier, I would have experienced a magnitude 7.1 earthquake in 1949 that did a lot of damage and caused loss of life. And if I had stuck around a few years longer, I would have been shaken by the Seattle Earthquake of 1965, which produced more damage and fatalities. Even though I didn’t feel anything during the short time I lived in Seattle, the Seattle area was experiencing normal seismic activity during that time. Modern seismicity maps of the Puget Sound show lots of black dots, although most identify earthquakes that are too small to be felt by anything other than sensitive seismographs.
How long is a long time for a geologist? Look at Table 1, which shows a series of time scales, each encompassing a longer period of time than the last. The first scale is historical, the time of recordkeeping, starting with the arrival of Western explorers two centuries ago. The next two scales are in thousands rather than hundreds of years; the written history of the Pacific Northwest spans only a brief part of the Late Quaternary time scale. The Late Cenozoic scale is in millions of years, and the Older Earth History scale covers four and a half billion years.
OK, I’m a geologist, and I am supposed to think in these great lengths of time. But I still consider it a long time when I’m stuck in traffic on Interstate 5. When I was growing up, I thought it was an unacceptably long time until Christmas or my birthday. You may agree that it is a very long time before you graduate from college, or get your kids raised, or retire, and so it is probably tough to envision even the two hundred years people have been keeping records in the Northwest.
Now that I am older, I have learned to take a somewhat longer view of time (except when I’m stuck on the freeway). I knew both my grandfathers, who told me stories about the horse-and-buggy days. I enjoy reading about the early settlers in the Willamette Valley and Puget Sound 150 years ago, and that, to me, seems an unbelievably long time ago.
But in fact, our recorded history in the Northwest (Historical Time Scale, Table 1) is short. The stretch of the coast from Alaska to California was the last region of the Pacific Rim to receive settlers willing to record their history, a fact that will become significant when we consider the great Cascadia Earthquake of A.D. 1700.
Spanish explorers reached the southern Oregon coast around A.D. 1600, and a Greek adventurer, Ioánnis Phokás, known by his name in Spanish, Juan de Fuca, may or may not have discovered the strait that bears his name. Captain George Vancouver and Spanish sea captains visited Puget Sound in the late 1700s, followed by Meriwether Lewis and William Clark, who arrived for a winter layover in 1806, complained about the rain, and went home. But they did blaze the trail, and fur traders set up posts at Fort Vancouver and Astoria. Soon after, many settlers from the eastern United States came to Oregon (which, as the Oregon Territory, included at that time most of the Pacific Northwest south of Canada). New towns were established west of the Cascade Mountains, and along with towns and farms, people built roads, established land claims, and started newspapers. By the 1840s, less than two centuries ago, people were keeping written records more or less continuously throughout the area west of the Cascades. This means that we know only that the Pacific Northwest has been free of great earthquakes since that time. To a geologist, that is not a very long time, not at all.
Native Americans were here long before that, of course, but they did not keep written records. Their rich oral traditions are another matter, though, and some of their stories document great earthquakes and earthquake-induced waves from the sea.
To a geologist, two centuries is like the blinking of an eye. The Earth is more than four and a half billion years old. The evidence from the rocks shows that the Pacific Northwest is much younger than that, and only in northeastern Washington and adjacent Idaho and British Columbia do we find rocks that are more than a billion years old. Most of the rocks in western Washington and Oregon are less than sixty million years old. But that is still an incredibly long time. A geologist can easily talk about sixty million years but it is just as hard for a geologist to imagine such a long period of time as it is for anybody else.
If the length of time that geologic processes have operated in the Northwest is unimaginably long, the rates of these processes are incredibly slow, about as fast as your fingernails grow.
|Table 1: Time Scales|
|2000||Age of computers, logging cutbacks, the decline in state services, increased population, Nisqually Earthquake in 2001|
|1980||Mt. St. Helens erupted. Space exploration and men on the Moon; Vietnam War|
|1960||U.S. interstate highway network. End of World War II atmospheric testing of nuclear weapons|
|1940||Roaring Twenties followed by the Great Depression and World War II|
|1914-18||World War I|
|1900||Extensive logging and development of farmland; autos replaced horses|
|1880||Development of rail network|
|1860||U.S. Civil War; present U.S.-Canada border established after the Pig War in the San Juan Islands|
|1840||Pioneers headed west to Oregon; settlement of Willamette Valley, Puget Lowland, Fraser Delta, southern Vancouver Island, newspapers established.|
|1820||Astoria and Fort Vancouver fur trade centers established.|
|1800||Native Americans were in charge but left no written records. Lewis and Clark's expedition began great westward migration.|
|1780||Explorers reached coasts of British Columbia, Washington, and Oregon.|
|1700||Cascadia Subduction Zone Earthquake recorded by a tsunami in Japan.|
|1600||Spanish explorers reached the Southern Oregon coast.|
|2000||Today. Last great subduction-zone earthquake Jan. 26, 1700|
|1500||Columbus discovered America but not the Pacific Northwest.|
|1000||Large earthquake(s) on Seattle Fault around A.D. 900|
|500||Three subduction-zone earthquakes between A.D. 500 and 1000. Long interval with no earthquakes between B.C. 500 and A.D.500.|
|B.P., which used to mean “Before the Present” before nuclear fallout messed up our dating scales, now means “Before A.D. 1950”|
|5,000 B.P.||Same as B.C. 3000; 5,000 years before A.D. 1950. Sea level approached present position; Mt. Mazama erupted to form Crater Lake|
|11,700||End of Pleistocene and beginning of Holocene. Sea level rising. Eighteen subduction-zone earthquakes during the Holocene. Great Missoula floods 15,000 to 12,000 years ago|
|15,000||Ice caps retreating and sea level rising rapidly|
|20,000||Glacial ice as far south as Olympia and Spokane, Washington; shorelines nearly 400 feet lower than today.|
|Late Cenozoic (Age, in thousands of years)|
|0||Today. Sea level is 20 feet lower today than 124,000 years ago|
|500||500,000 years. Several ice advances and retreats. Earth’s magnetic field reversed at 780,000 years; previously, compass needle pointed south.|
|1,000||More glacial advances and retreats.|
|2,600||Beginning of Pleistocene 2,600,000 years ago|
|2,500||First major ice age started about 2,400,000 years ago. Still in the Pliocene, which started about 5,300,000 years ago.|
|Older Earth History (Age, in millions of years)|
|2.4||Beginning of Ice Ages|
|15-17||Great eruptions of Columbia River Basalt|
|66||Asteroid slammed into southern Mexico, dinosaurs became extinct.|
|245||Greatest mass extinction in the history of the Earth|
|570||Beginning of trilobites and shelly fossils|
|4,570||Age of the Earth, 4,570,000,000 years|
When I talk about the motion of the oceanic plate northeastward toward Oregon, Washington, and Vancouver Island, and I say that the motion is a little more than an inch and a half per year, I sometimes lose my audience. Here we’re talking about increasing the speed limit on Oregon freeways, and this guy is worried about speeds of an inch and a half a year? Give us a break! But this rate is faster than the rate of a little more than an inch per year at which coastal California is grinding past the rest of North America on the San Andreas Fault. Even with that slow rate of travel, the San Andreas Fault has had great earthquakes in 1812, 1857, and 1906. If you continue this slip rate for five million years, coastal California will move northwest more than eighty miles. Keep that up long enough, and—hold your breath—Los Angeles will become part of the Pacific Northwest!
Let’s suppose that one gigantic earthquake ruptured the entire Cascadia Subduction Zone in 1700 AD, prior to the start of recordkeeping in the region, and caused displacement of 65 feet, which many scientists believe is possible. And let’s suppose also that this earthquake relieved all the strain that had been slowly building up at a rate of 1.6 inches per year. Dividing 1.6 inches per year into 65 feet, you find that it would take almost five hundred years for the crust to recover that strain so that the subduction zone could rupture again in the next earthquake. Now that’s a long time, about two and a half times our recorded history in the Pacific Northwest since the expedition of Lewis and Clark.
But we’ve already used up more than two hundred years of recorded history with no monster earthquake, and, as will be shown below, there is geologic evidence from Brian Atwater’s subsided marshes and historical evidence from a tsunami in Japan in 1700 AD that we have already used up more than three hundred years. Should we forget about it, inasmuch as we still might have two hundred years to go?
Unfortunately not, because the repeat time of earthquakes can be highly variable. In southern California, a section of the San Andreas Fault ruptured in 1812 and again in 1857, just forty-five years later. Yet more than one hundred and fifty years have gone by without another major earthquake along that same section of the fault. The sourthernmost part of the San Andreas fault has not had a major rupture in more than 300 years. The Cascadia Subduction Zone could have much longer than two hundred years to go, or we could have the next great Cascadia earthquake much sooner, maybe in our lifetime, maybe tomorrow.
Another reason that we can’t laugh at 1.6 inches per year is the massive amount of rock that is building up strain. The oceanic slab that is forcing its way under the edge of the North American continent is about 40 miles thick and 740 miles long, extending from Vancouver Island to northern California. So, even though the movement rate is slow, the bodies of rock that are being strained are titanic in size.
Because the times for geologic processes to work are so ponderously long, geologists have devised time scales (see Table 1), analogous, perhaps, to historians referring to the Middle Ages or the Renaissance. At first, this was done using fossils, because organisms have changed through time by evolution, and distinctive shells or bones of species that had become extinct were used to characterize specific time intervals called periods and epochs. In the past few decades, it has become possible to date rocks directly, based on the extremely regular rate of decay of certain radioactive isotopes of elements such as uranium. These atomic clocks enable us to date the age of the Earth at about four and a half billion years and, in addition, to date the age of trilobites, of dinosaurs, and of other dominant groups of organisms that are now extinct.
In our study of earthquakes, we do not need to be concerned about most of the geologic periods and epochs, including the ages of trilobites and dinosaurs. We do need to know about those times when the geologic processes that produce today’s earthquakes have been operating: the Tertiary and Quaternary periods, together known as the Cenozoic Era. We need to know something about the geologic history of the later part of the Tertiary Period, but we are most concerned about the Quaternary, which started 2.6 million years ago (Table 1). We divide the Quaternary into the Pleistocene and the Holocene epochs, with the boundary between the two dated at about eleven thousand years ago. The Pleistocene Epoch, covering most of the Ice Ages, saw much of the evolution of human beings, as well as saber-tooth tigers, mastodons, and great cave bears.
But it is the Holocene, the last 11,700 years, that concerns us most. During the latest Pleistocene and early Holocene, the great ice caps of North America and Europe melted away, and the addition of all that meltwater to the world’s oceans caused sea level to rise hundreds of feet. During the last half of the Holocene, civilizations arose in Mesopotamia, Egypt, and China, and written records began to be kept.
If geologists can show that a fault sustained an earthquake during the Holocene, it is placed in a special category of hazard. If it ruptured that recently, it is likely to rupture again, and it is called an active fault. This classification based on the time of most recent activity is written into law in some states and into regulations by federal agencies such as the U.S. Nuclear Regulatory Commission and the U.S. Army Corps of Engineers.
To learn the age of an earthquake, we have written records only for the last part of the Holocene, and for the Pacific Northwest, the historical record only slightly longer than two hundred years. But we can use one of the nuclear clocks to date formerly living organisms for the last twenty to thirty thousand years. This is radiocarbon dating, based on the natural decay of a radioactive isotope of carbon (carbon 14) into stable carbon (carbon 12). Carbon 14 starts off as ordinary nitrogen, which makes up the greater part of the atmosphere. The stable isotope of nitrogen, nitrogen 14, is bombarded by cosmic rays from outer space, changing it to carbon 14, which is radioactive and unstable. Organisms, including you and I, take up both the radioactive and stable isotopes of carbon in the same proportions as in the atmosphere. After the organism dies, carbon 14 decays to carbon 12 at a precise rate, so that half of the carbon 14 is gone in 5,730 years. In another 5,730 years, half of what’s left decays to carbon 12, and half of that decay in another 5,730 years, until finally there is too little radioactive carbon 14 to measure. We say that 5,730 years is the half-life of the radioactive decay of carbon 14 to carbon 12.
Unfortunately, the radiocarbon clock is not as precise as we would like. Radiocarbon dating cannot get us to the exact year, but only to within a few decades of the actual age. An example of a radiocarbon age is 5,300 ± 60 radiocarbon years, an expression of the laboratory precision in counting the atoms of carbon 14 relative to carbon 12. Radiocarbon years are not the same as “calendar” years because the cosmic radiation that creates carbon 14 is not constant, but has changed over the years. Minze Stuiver and his colleagues at the University of Washington designed a conversion scale that changes radiocarbon years to calendar years, and in most reports today, this conversion has already been made, using a computer program. A radiocarbon age or a calendar age of, say, 5,300 years is stated as 5,300 years B.P., meaning Before Present. But “Present” is not really today, because the atmospheric fallout from nuclear weapons testing after World War II completely messed up our dating. To get around that, we refer to “present” as A.D. 1950.
In addition, the geologist or archaeologist must ensure that the carbon sample being dated (charcoal, shell fragment, bone fragment) is the same age as the deposit in which it is found. The charcoal in a deposit may have been washed in from a dead tree that is hundreds of years older. Or the charcoal may be part of a root from a much younger tree that grew and died long after the deposit was buried by other sediments.
Finally, the ratio of carbon 14 to carbon 12 in lakes and in parts of the ocean may not be the same as it is in the atmosphere. To accurately date the remains of organisms that died in these environments, it is necessary to figure out what the carbon isotope ratios are under these conditions and make a reservoir correction.
To conclude our discussion of time, we need to think of earthquakes in two ways. On the one hand, an earthquake takes place in a matter of seconds, almost (but not quite) instantaneously. But on the other hand, an earthquake marks the release of strain that has built up over periods of hundreds, thousands, even tens of thousands of years. We use radiocarbon dating to learn how long it has taken the strain to build up enough to break a large mass of rock in an earthquake over the last thirty thousand years. We can also use tree rings to determine within one year when a particular tree growing in a coastal forest was suddenly buried below sea level.
To understand the earthquake hazard, it is not enough to figure out what will happen in a future earthquake. To make progress in forecasting earthquakes, we need to know how long it takes a fault to build up enough strain to rupture in an earthquake, and how large that earthquake is likely to be. When? Where? How big? The answers to those questions rest on our ability to respond to the earthquake danger and to survive it.
Suggestions for Further Reading
Levin, H. L., 1999. The Earth Through Time, Sixth Edition. New York: Saunders College Publishing, 568 p., 7 appendices.
Pellegrino, C. R. 1985. Time Gate: Hurtling Backward through History. Blue Ridge Summit, PA: TAB Books, Inc., 275 p. An explanation of the vastness of time by looking backward through ever-increasing time spans to the very beginning; written for the layperson. Table 1 is based on this idea.
Wicander, R., and J.C. Monroe, 2012, Historical Geology, 7th Edition. CENGAGE Learning, print or ebook.
Yeats, R. S., K. E. Sieh, and C. R. Allen. 1997. The Geology of Earthquakes. New York: Oxford University Press, Chapter 6, p. 116-38.