The Cascadia subduction zone in the Pacific Northwest is known for two things – earthquakes and tsunamis. This is particularly true since The New Yorker article proclaiming that “everything west of I-5 will be toast,” brought considerable attention to the seismic hazards spanning from northern California to southern British Columbia. Most people, however, are unaware that most of the damage from a magnitude 9 earthquake will be from landslides. Despite this looming threat, there is a general lack of understanding of how earthquakes affect the landscape in the Pacific Northwest. The last major fault rupture occurred on January 20, 1700; the earthquake lowered the coastline and unleashed a tsunami that reached Japan. However, no single landslide from that earthquake has been definitively pinpointed. In the Oregon Coast Range, there are greater than 30,000 large bedrock landslides; yet, none of these landslides have been historically observed, though shallow landslides and debris flows abound in the rainy season each year.
Klickitat Lake in the Oregon Coast Range has many still-standing Douglas Fir trees that were drowned after a landslide formed the lake. The age of the lake-forming landslide can be ascertained when dendrochronological techniques are applied to wedges extracted from these trees.
So, how can we quantify hazard from landslides during major earthquakes if we do not even know how common landslides were during the last earthquake?
Many researchers have aimed to date as many landslides as possible in the Pacific Northwest to ascertain if any are from the 1700 earthquake. Radiocarbon dating, where carbon from landslide organic material is compared to the modern atmosphere, is a common landslide-dating technique. Other researchers have measured the surface roughness of landslide surfaces to estimate the time of failure. While both techniques are useful and can provide rapid dating of landslides for cataloging, there is error associated with each that makes pinpointing a landslide age to a single year and earthquake exceptionally difficult. Fortunately, some sites in Cascadia have an additional resource that can help us date young landslides. Trees!
Some landslides in Oregon have dammed rivers, resulting in lakes or marshes upstream. Where these lakes form, they drown Douglas-fir trees. Long after these trees have died, many still stand, forming ‘ghost forests. Given these trees died synchronously with the landslide, the tool of dendrochronology, or tree ring science, provides the means necessary to date the landslide.
Collected slabs, or wedges, from Douglas-fir snags at landslide-dammed lakes. Ring width variations record climatic forcings such as drought, as visible on the right. For example, several small rings beginning in 1739 correspond to regional drought that is visible in tree rings throughout western Oregon. The end of the ring record here corresponds to tree death in 1751.
In order to date a landslide-dammed lake, we collect slabs from the standing trees for analysis. We sand and polish the slabs to a brilliant shine, taking off material as small as 12 micrometers, and using high-resolution scans, we measure the rings on all the slabs. Since we are not interested in how old the trees are, but the year in which they died, we cannot simply count the rings. We must compare the ring widths to a known record. Fortunately, dendrochronologists devote much of their time to building tree ring chronologies from live trees: records of ring growth that are high-precision records of climate, particularly drought. Using one such chronology from the Oregon Coast Range, we use statistical tools to match the measurements from our ‘ghost forests’ to the tree ring chronology. If we have prepared and measured our slabs well, and there is overlap between the measurements and the chronology, we find a date!
Our work to date landslides in the Oregon Coast Range using Douglas-fir tree rings has so far resulted in two landslide ages. The landslides that formed Wasson and Klickitat Lakes occurred in the years 1819 and 1751, respectively. Based on the ring growth, we even know that the landslides occurred in the winter of 1819/20 and 1751/52. While neither of these landslides formed in the year 1700 during the last earthquake, our dates are noteworthy due to their unprecedented accuracy. Dating a landslide with radiocarbon methods often results in an error spanning a 60-year period. Some have claimed that they have dated landslides that correspond to the 1700 earthquake, though none have circumvented the error that distorts their landslide ages. We have found at Wasson and Klickitat Lakes that while radiocarbon dating may suggest that 1700 is a possible date for the landslides, tree ring data negate such a possibility. So, while radiocarbon dating is useful and can help narrow the range of possible ages, it is currently not possible to date a landslide with the sub-annual accuracy common to dendrochronology.
As the people of the Pacific Northwest confront the hazards posed by the Cascadia subduction zone, they must consider the threat of landslides in steep terrain, in addition to the hazards of ground motion and tsunamis. As landslides in the Oregon Coast Range continue to be dated, and potentially connected to previous earthquakes, communities must utilize these improved hazard assessments.
See our recent article in Eos, Hunting for Landslides from Cascadia’s Great Earthquakes, for more information on dating landslides in the Pacific Northwest.
0
0
0
0
0
0
Will Struble
Latest posts by Will Struble (see all)
- Dating Oregon landslides with ‘ghost forests’ - November 8, 2018