Great Cascadia Earthquake Penrose Conference - The Oregonian Article

The Oregonian: Science - Landslide Sleuths (May 15,2002)

RICHARD L. HILL

CORVALLIS -- The waterlogged stumps sticking out of the Columbia River fascinated Lewis and Clark as they made their way to the Pacific and back again.

The explorers, the first to write about the curious scene in the Columbia River Gorge, accurately described the river as obstructed by large rocks, "which seem to have fallen promiscuously from the mountains into the bed of the river" and with "stumps of pine trees scattered for some distance in the river, which has the appearance of being dammed below."

What they saw is now the Northwest's most famous and studied avalanche -- the Bonneville Landslide -- an event near present-day Cascade Locks that probably led to Native American stories about the "Bridge of the Gods" linking Oregon and Washington.

Two centuries after Lewis and Clark's observations, the ancient landslide and the questions it raises continue to intrigue scientists. Did an earthquake cause the huge slide, possibly the powerful magnitude 9 quake that is thought to have rocked the Northwest in 1700? Could a similar landslide occur in the gorge today?

"We're making some progress in determining when the slide occurred," said Patrick T. Pringle, a Washington Department of Natural Resources geologist who became interested in the Bonneville Landslide about 15 years ago. "But we still have a ways to go. The work is both challenging and can be frustrating at times, but it's important to find out when and why it occurred."

Pringle discussed recent findings about the landslide Tuesday at a regional meeting in Corvallis of the Geological Society of America. He is one of nearly 1,000 scientists at the meeting through today on the Oregon State University campus. The scientists are exploring the landslides, volcanoes, earthquakes and other geologic mayhem that are part of the active Northwest landscape.

Evidence of the date of the slide is hard to come by. Many clues were destroyed during construction in the 1930s of Bonneville Dam. The dam sits on part of the slide and inundated the information-rich tree stumps behind it.

But Pringle and other scientists are making headway with advances in radiocarbon dating, tree-ring dating -- called dendrochronology -- and a technique called lichenometry, which can estimate a landslide's age by measuring the growth rate of certain lichen on the rock surfaces.

A legend it born The massive Bonneville Landslide crashed southward from Washington's Table Mountain about 30 miles east of Portland, covering more than 5.5 square miles and creating a 200-foot-high earthen dam across the Columbia River.

Local tribes might have been able to cross the river on the dam, giving rise to oral histories about a bridge. The slide covered a 3.5-mile stretch of the river, drowning a narrow forest of trees for 35 miles.

Eventually, the rising lake broke through the dam. No one knows how long that took, perhaps two years, but it unleashed a flood that eroded much of the slide and submerged present-day Troutdale under 100 feet of water.

Lewis and Clark estimated that the slide had occurred only a couple of decades before their arrival, but the looks of the slowly decomposing trees were deceiving. Recent investigators have placed the slide as old as 900 years and as young as 250 years.

Radiocarbon dates taken in 1958 from drowned trees indicated that the slide occurred between 1250 and 1280. A quarter-century later, a radiocarbon date of wood samples taken from within and below the landslide deposit put the date at about 1100.

Four years ago, Pringle and Robert L. Schuster of the U.S. Geological Survey had radiocarbon dates taken of a buried Douglas fir that indicated the tree died between 1500 and 1760. That would place the slide close to the earthquake in 1700 that devastated the Northwest coast.

Counting the tree rings, with each ring representing one year of the tree's life, they estimated that the tree died in about 1699.

New twists to mystery But the mystery of the slide's date has taken new twists recently.

Last year, Nathaniel D. Reynolds, then a graduate student at Washington State University in Vancouver, used a technique called lichenometry to estimate the age of the Bonneville Landslide. The dating method uses the growth rate of specific lichen species as an indicator of the age of the surface the lichen is growing on.

Lichens, slow-growing organisms formed from an association between a fungus and an alga, can be used for dating earthquakes and landslides because they quickly colonize fresh rockfalls that occur in the wake of a quake. Once established, they form at a constant rate if left undisturbed.

Reynolds, now with the U.S. Forest Service, said his study indicates the landslide probably happened between 1670 and 1760. "These results demonstrate that the Bonneville Landslide may have occurred more recently than previously believed," Reynolds said in a recent article in Washington Geology.

The dates "provocatively bracket" the powerful offshore Cascadia Subduction Zone earthquake of 1700, he said, but he cautioned that the study doesn't prove that the quake caused the landslide.

The research plot thickened with the recent surprising discovery of tree samples cut from the landslide site in 1934 by the late Donald B. Hamilton of Johns Hopkins University and the University of Minnesota before Bonneville Dam was completed.

Pringle and Reynolds, along with colleagues Jim E. O'Connor, a hydrologist with the Geological Survey in Portland, and Alex C. Bourdeau, an archaeologist with the U.S. Fish and Wildlife Service in Sherwood, had wondered where the samples were. The Western Forestry Center, where some of the samples had been kept, burned to the ground in 1964.

Newfound samples But when Reynolds called the what is now the World Forestry Center in Portland's Washington Park, he learned that a crate of the samples had been stored in the center's basement. They apparently had been in a scientist's lab at OSU when the fire occurred and were returned when the center was rebuilt.

"So the four of us went over there and they drag these slabs out," Pringle said. "We couldn't believe it. They had four slabs -- two were from living old trees and two were from this submerged forest of the Columbia. We were just drooling."

O'Connor and Bourdeau took samples for radiocarbon testing, and Pringle took samples for tree-ring dating.

Pringle found that the sample from a submerged forest tree appears to have died the same year -- 1699 -- as the buried tree that he and Schuster had studied.

"I was amazed when I found that these two trees from different sites had died the same year," Pringle said. "It was a victorious moment. And that 1699 date, almost the same as the 1700 earthquake date, just stopped me cold."

Despite the finding, however, the mystery of the landslide date remains unsolved.

Although the tree-ring and lichen studies point to a slide date around 1700, the radiocarbon dates O'Connor obtained from the tree samples found at the World Forestry Center indicate that the trees died about 1500.

"So now we have conflicting evidence," Pringle said. "We have our work cut out for us in trying to resolve these ambiguities from the different dating techniques."

Pringle said he hopes additional research will help resolve the conflicting dates. The M.J. Murdock Charitable Trust recently gave the Washington Department of Natural Resources a $14,000 grant that will allow Rusty Weaver, a science teacher at Vancouver's Heritage High School, to work with Pringle in taking tree-ring samples from trees that have grown on top of the landslide.

"I'm very excited about getting to do this upcoming work, because it's going to allow us to collect more evidence about the date of this slide," Pringle said. "It's a question that befuddles us. So I think this work will help a lot in helping us find an answer." You can reach Richard L. Hill at 503-221-8238 or by e-mail at richardhill@news.oregonian.com.