 |
| From The Seattle Times, January 21, 2018 |
Update January 31, 2018
Sandi Doughton, science reporter for the Seattle Times, has another article in the paper this morning with some breaking news. Because the landslide is no longer accelerating, no longer moving faster and faster, two things have happened to the experts' assessments:
- the landslide's movement has fallen off the track of predictable landslide failures, so the geologists and engineers have canceled any predictions of when it is likely to fail
- the currently favored prediction is that it will keep moving more or less steadily, dropping boulders and chunks of earth into the quarry site at the bottom end of the landslide
 |
| Table copied from page 17 of "Rattlesnake Hills Landslide Evaluation" by Norman Norrish. |
Norrish's provisional conclusion, that the landslide will continue moving slowly and is unlikely to fail catastrophically, largely agrees with the earlier prognosis reached by Cornforth Consultants, the company hired by the quarry to monitor and assess the landside.
The caveat remains that this is what is currently considered likely, based on the record of how the hillside has cracked and moved so far.
Update January 23, 2018
On his blog, Reading the Washington Landscape, Dan McShane posted some comments on the landslide based on newspaper articles published the weekend of January 21-22 in the Seattle Times and the New York Times.
Dan pointed out an error that the New York Times journalist made about how Union Gap was formed, in an article that Dan otherwise recommends. The author stated that Union Gap was eroded through Rattlesnake Ridge by the glacial Lake Missoula floods, which is incorrect. In my earlier post about the landside, I explained how Union Gap formed gradually, as Rattlesnake Ridge slowly rose up, over several million years, due to tectonic compression of the crust across the area. The Yakima River maintained its pre-existing course by keeping a gap eroded though the ridge.
 |
| Google Maps image of Union Gap before the landslide. Labels added. |
The January 21 Seattle Times article delves into how people are trying to solve the problem of when the landslide is likely to collapse. By the way, the author of that article is Sandi Doughton, who recently published a fascinating and somewhat alarming book, "Full Rip 9.0: The Next Big Earthquake in the Pacific Northwest."
We expect a very-high-magnitude earthquake and concomitant tsunami to hit the Pacific Northwest coast's subduction zone one of these days, but we cannot predict when.
Although we cannot predict earthquakes, for landslides that are already starting to move there is a track record of being able to predict when the main slope failure will occur. The track record is a bit spotty, but it has led to some standard methods for measuring slope offset and fitting the data to some formulas to make the "time of failure" prediction.
The main things the formula depends on are (1) how fast the slope is moving, and (2) how much its rate of motion is increasing each day. For you math nerds, these are the first and second derivatives of slope displacement with respect to time, also known as the velocity and the acceleration.
Once monitoring equipment is deployed on the moving mass to track its movement, data can be gathered to apply the formula and see if the landslide is signaling when it will fail.
 |
| This graph, from The Seattle Times, depicts how the landslide engineers monitoring the Rattlesnake Ridge landslide use the data to project forward in time to when the landslide is likely to fail. The lowering of the acceleration rate (the easing up of the slope in the data trend), starting in late December 2017, has put the estimated time of failure window further into the future, probably in March according to the most recent trend. |
If the unstable, moving slope is steadily accelerating its rate of movement, moving faster and faster each day, the formula for estimating the time of failure can be applied with some confidence. The Seattle Times article makes this clear, so I will quote from it:
"The method works because of the way many landslides unfold, said University of Utah engineering geologist Jeffrey Moore. As a slope cracks and begins to slide, the forces holding the mass of soil and rocks in place weaken and the underground surface the mass is sliding on — usually a weak layer of soil — gets slicker. As a result, the mass gains momentum and starts to slide even faster. Gravity eventually overcomes friction, and the slope collapses.
'There’s enough case histories behind the method that when the data look right and you are able to identify trends, you can accurately predict the time of failure,' Moore said."
But if the slope is not steadily accelerating - not moving faster and faster the same amount each day - it becomes a lot more iffy as to what is going to happen, and when. And that is the problem lately with the Rattlesnake Ridge landslide at Union Gap.
To quote again from the Seattle Times article:
“The geologic foot is coming off the gas,” [UW geologist David] Montgomery said. “Even though it’s moving faster than it was, the rate of acceleration is declining.”
A set of images have been used to create a time-lapse video of the landslide moving over the last few weeks, as shown below. The recent snow highlights, by contrast, the growing cracks.
 |
| From The Seattle Times, January 18, 2018. |
If the landslide is not sliding faster and faster at a steadily accelerating rate, it does not fit the model a landslide heading for a predictable failure, which reduces confidence in predictions of when it will tumble.
Or, as has happened lately with the Rattlesnake Ridge landslide, if it changes its rate of acceleration, that changes the time of failure estimate.
It will take awhile to see if the Rattlesnake Ridge landslide will remain at its new rate of acceleration. If it does, if the landslide's new trend of accelerating movement continues, that will raise our confidence in being able to estimate the time of failure.
By the way, although landslides on the surface are different from the ruptures that occur within the Earth on deep faults, it is possible that if we ever succeed in being able to tell when a fault is going to fail and an earthquake occur, we may be helped by what we learn about predicting landslides.
Ultimately, to be most effective at keeping people safe, earthquake early warnings will have to be made farther in advance than when the first seismic waves hit the Earth's surface at the epicenter.
The first major seismic waves arriving at the Earth's surface near the epicenter allows early warnings to go out to places farther away, much faster than the seismic waves can move, giving up to an hour of warning in the best cases, for people hundreds of miles from the epicenter. Mexico has developed the most advanced earthquake early warning system of this type. Such early warnings of up to an hour ahead of the earthquake waves arriving help save lives.
 |
| From the US Geological Survey earthquake early warning systems information page. |
But if the build-up to a fault rupturing and an earthquake occurring is a process that can be detected many hours or even days or weeks in advance, before the fault ruptures, before the earthquake occurs, we could save even more lives.
There is a chance that the ability to predict earthquakes will follow a technique similar to how well we can pin down when landslides will occur, based on the rate at which the Earth deforms, and acceleration of the rate of deformation, in the vicinity of the fault.
Scientists have been trying to predict earthquakes that way, and by other methods, for years, but the progress is limited and success has not been achieved. It's harder to measure underground fault offset than surface slope movement, and the methods and theories are not yet there.
Going off a bit more on my earthquake tangent, the following image is not about landslides. It is a shake map showing how much the different parts of Washington state would be shaken - and subject to earthquake damage as a result of the ground shaking.
Although landslides and earthquakes should not be confused, another one of their connections is that a common trigger of landslides, probably the second most common landslide trigger after heavy rainfall, is earthquakes.
Let's not tangle earthquakes and landslides together too much - the map only shows shaking from an earthquake, and the possible levels of damage that may result.
 |
| This interactive map of the amount of ground shaking and damage across Washington state from a theorized magnitude 9.0 Cascadia subduction zone earthquake was created by the Washington State Department of Natural Resources, the US Geological Survey, and other agencies. |
We still hope to be able to predict the "time of failure" of major faults someday, so we can tell when a major earthquake is coming, but that goal will probably take decades to achieve.
In the meantime, we can see how well "time of failure" predictions for landslides, which have a record of partial success, work in the case of the Rattlesnake Ridge landslide.
Update January 7, 2018
For a balancing of viewpoints, see the last part of this Seattle Times article about the landslide and how residents from the evacuation zone are being given temporary lodging in motel units with kitchens.
The relevant part of the newspaper article is subtitled "Experts disagree." It features a response to Bruce Bjornstad from Stephen Reidel, a geologist who is a research professor at Washington State University - Tri-Cities.
Steve has worked with the Columbia River Basalt and its sedimentary interbeds, and their geological contexts, structures, and origins, as much as anybody I know. As with Bruce Bjornstad, I have seen Steve give talks to the public and to professional geology groups, and have been on field trips he helped lead. I will finish with a diagram provided by Dr. Reidel to the Seattle Times, showing a schematic, simplified cross-section view of the landslide situation on Rattlesnake Ridge.
 |
| From the Seattle Times. |
Here is a four-page fact sheet about the landslide from the Washington Geological Survey, which is in the Washington State Department of Natural Resources-Division of Geology and Earth Resources.
Co-agencies on the fact sheet are the Washington State Department of Transportation and the Yakima County Office of Emergency Management.
Below is an image that the the Yakima County Office of Emergency Management tweeted on January 7. What is interesting about this image is that it maps, as red lines, cracks west of the ridge crest, making an arc when joined with the more obvious cracks everybody can easily see above the quarry on the south-facing slope of the ridge. If those cracks west of the ridgeline are there and are interpreted as marking the limit of the landslide, that may be a big reason why the geotechnical company hired by the quarry considers it most likely that the landslide is a slab which will continue moving south toward the quarry.
 |
| From Yakima County Office of Emergency Management (Yakima OEM). The original caption is, "Map of DNR's geological monitoring equipment network deployed at #RattlesnakeRidge to track changes and provide instant data to incident managers." The map is excerpted from the four-page fact sheet posted by the Washington State Department of Natural Resources. |
On this still from a drone video, you can see what may be the continuation of the crack on on the hillside west of (to the left of) the ridge crest.
 |
| From a video on YouTube made by KING5 News, Seattle. (Arrows added.) |
Update January 6, 2018
The Seattle Times ran an article this morning, by Hal Bernton and Sandi Doughton: "Geologist warns Yakima-area landslide could be worse than officials expect." The geologist is Bruce Bjornstad. Bruce is a colleague of mine with whom I have interacted on a few field excursions, and heard during some talks he has given to the public and to groups of geologists. Bruce has written two in-depth, well-illustrated guidebooks to the geology of the Channeled Scablands in eastern Washington.
In the comments after the article online, and in Facebook where I follow Bruce Bjornstad, there are some differing opinions offered about the Rattlesnake Hills landslide and how it is being reported in the news.
Bjornstad is emphasizing the worst possible scenario of landsliding, in his view. Based on his experience studying the geology of the area, and his knowledge of a prehistoric landslide a few miles down the Yakima Valley from the currently active site, he thinks it likely, or at least quite possible, that the slide will travel west, onto I-82 and possibly into the Yakima River. Here is a diagram he drew on a Google Earth image:
 |
| Bruce Bjornstad’s depiction of the potential landslide in what may be a worst-case, and according to other geological analyses unlikely, scenario. (Copied from nwnewsnetwork.org) |
The quarry operators hired a geotechnical consulting company several months ago, Cornforth Consultants from Portland, and their preliminary report is available online. The report points out that (1) the layers of rock in the hillside are tilted south, towards the quarry, at 10-15 degrees of dip, and (2) their measurements of motion of the hillside found that, above the quarry, it is moving south toward the quarry. They report that the landslide is likely to continue sliding toward the quarry, even if it moves bigger and faster, and in the end it may not do more than fill the quarry.
The preliminary report discusses and shows (page 3 and Figure 8) previously mapped landslides along the north edge of Rattlesnake Ridge, on the other side of the hill from the quarry.
In addition, the report suggests that there may be some older landslides in the area that were not previously recognized. The largest of these newly proposed landslides overlaps and contains the quarry and the currently moving zone of concern above the quarry. There are, it seems to me, some questionable aspects to this proposed ancient landslide, including how small a distance it would have moved, given that the top layer of basalt within it still extends to the top of the ridge.
In addition, the report suggests that there may be some older landslides in the area that were not previously recognized. The largest of these newly proposed landslides overlaps and contains the quarry and the currently moving zone of concern above the quarry. There are, it seems to me, some questionable aspects to this proposed ancient landslide, including how small a distance it would have moved, given that the top layer of basalt within it still extends to the top of the ridge.
Comments
Post a Comment