Oregon’s side of the Columbia River Gorge is known for its steep cliffsides with cascading waterfalls, including touristy Multnomah falls. Hike up any of the creeks and rivers from the Columbia and you are sure to encounter a waterfall or two or three. Lush forests, gorgeous views of the Columbia and nearby mountains, and heart-pumping climbs, make a visit to the region a popular choice for Portland area hikers.
However, there is more to the region than stunning scenery. The Columbia River Gorge is geologically unique. It cuts through one of only a relatively few flood basalt provinces in the world. In addition, the area has been sculpted by the Columbia, and, most profoundly, by the Missoula Flood events—also relatively rare glacial outburst flood events that help define the region.
To better appreciate the unusual geological history of “the gorge,” I met up with Jim O’Connor from US Geological Survey at the Horsetail Falls Trailhead for a short hike up to Triple Falls.
Quadrangle by Quadrangle
Having not visited Triple Falls in a long while, when I arrived at the trailhead to meet Jim, I was more than ready to explore the trail through his fresh geological eyes. After gathering our gear, we headed across the Columbia Gorge Highway to the Oneonta trailhead to begin our climb up to the falls. As we got started, I asked Jim to tell me about where we were.
Jim explained that the USGS has been mapping the Columbia River Gorge for the past 10 years—quadrangle by quadrangle. A quadrangle is roughly an eight-by-fifteen-mile area.
“In the Gorge, from the Sandy River to the Deschutes River there are probably about twenty-five quadrangles that touch the Columbia River… Right now, we are in the Multnomah Falls quadrangle.”
Most of the mapping has been done, but only a few quadrangle maps have been published so far. The Multnomah Falls quadrangle was mostly mapped about six years ago but publication awaits final review and layout.
I asked Jim why there was such a push to map out the Columbia River Gorge Corridor.
“There are a bunch of reasons for that,” he responded.
One reason, Jim suggested, was to simply tell the history of the place. The Columbia River Gorge, as mentioned early, is a relatively young flood basalt province—built up layer by layer of thin, fluid basaltic lava flows, as you see in a Hawaiian style of eruption.
“There are probably some 50-individual basalt flows as part of a major eruptive episode that occurred mainly around 16 million years ago,” Jim described.
“Originating from fissures in SE Washington, NE Oregon, and Idaho,” he continued, “these flows covered 200,000 square kilometers. Covering much of Eastern Oregon and Washington; flowed through the Cascades, and many flows made it even to the Pacific.”
Other reasons include understanding “the causes and consequences of a large igneous province” and “the history of the Columbia River,” along with the paleogeography and topography.
This Rock is Not Like the Other
One goal of the mapping is to distinguish the different flows apart and track their individual extents.
“They all look the same,” explained Jim. “If we bang on the rocks, we look at fresh faces, sometimes you can see minerals that help tell them apart, but really the only reliable way to tell them apart is geochemically.”
That means heading out into the field and collecting samples of rock.
Jim described hiking up and down the trails and in between with a rock hammer and sampling rocks, being sure to mark each rock’s location. The samples are then sent in for chemical analysis and lines are drawn on the map that traces each flow.
The work is very physically taxing, but the best way to get the job done. Hazards, regulations, and other limitations make it difficult to collect samples during much of the year, so much of the fieldwork is limited to September and October.
With all that in mind, Jim suggested that this sort of mapping would be mostly phased out—probably in the next few decades.
“The mapping we are doing now is probably the last being done ‘boots on the ground’ because it is hard and time-consuming,” said Jim. “Technology is going to develop and remote detection will happen”
In fact, a lot of the technology necessary already exists. Most topographic lines are already drawn using remote sensing, for example. There are even devices (currently used on Martian rovers) that can analyze rocks on the spot. Though Jim said that Columbia River Basalts’ differences are too subtle for the current iteration to work.
“Probably next century [mapping work will be done] with drones,” laughed Jim.
Jointing/Patterns in the Rocks
Once across the road and started hiking up the trail, we were able to get a closer view of the thick layer of rock with a jagged, angular texture.
“What we can see here is one thick massive flow,” Jim explained pointing at the massive rock walk. Jim described the fracture pattern as “Hackly” or “brick bat”—terms that I had not heard but essentially described the sharp and even pattern in the rocks. He also called the area an “entablature” zone—the name for a zone of basaltic lava with this sort of irregular fracture pattern, or ‘jointing’.
“Sometimes the jointing will be in columns,” Jim added—another typical jointing pattern of basalt that reminds me a bit of a bundle of pillars.
He went on to explain how the different jointing patterns develop while the lava cools in place. Jim used the analogy of mud cracks—like mud, it cools (or dries) from the outside in, shrinking and breaking into often beautiful formations.
“Some flows have distinct jointing,” Jim went on, allowing an astute geologist to tell flows apart.
“This is the Downey Gulch flow,” Jim said pointing back up the large rock face. “It commonly has this thick entablature zone.”
Thick entablature zones like this are often where you find waterfalls, Jim explained—“they hang together better than the columnar zones.” They are also “associated with cliff bands,” he added—the resistant tops or layers found in rocks.
Other Lavas
I asked Jim if other lavas had jointing patterns as you see with basalt.
“You will see jointing patterns in all lava flows,” he responded. “Even stickier flows you associate with Hood or Helens have all sorts of cooling fractures.”
“…But in basalt flows,” he went on “the jointing patterns are bolder and more distinctive because they flow farther.”
Columns and College
As we continued along the trail—tracing the contours of the cliffside, we passed by a nice-looking column of basalt laying on its side.
As we walked, Jim told me briefly about how he got started in his career in geology and how long he has been working.
“I got my undergraduate degree 40 years ago,” he shook his head disbelievingly.
Graduate school followed as it was difficult to find a job without it, and eventually, he landed a job with USGS.
“Look,” I pointed to another column laying on its side.
“A perfect hexagonal cross-section,” Jim noted as I snapped a picture.
Making Contact
Jim and I continued up the trail—zig zagging through a rocky cliffside and into the burnt forest. An assortment of shrubs and herbaceous plants lined the path—my favorite, the red thimbleberry, ripe and ready to taste.
However, Jim’s eyes were focused on the rock as we made our way up. He was looking for a particular feature in the rocks we were walking on—a flow contact. An important part of mapping out the area requires knowing when one flow starts and another begins—this is the flow contact.
“It is probably a flow contact up there,” Jim pointed up the trail.
Jim explained that there are many clues to finding a flow contact.
As you reach a flow contact, the rocks look different because the tops and bottoms of flows—as they interact with the surface get gas bubbles trapped in them. This creates a “vesicular” or “bubbly texture,” in the contact zone.
It wasn’t long before we reached the flow contact Jim had suspected from below.
He pointed to the cliff face.
“You can see there is an entablature zone, flow contact, and then columns,” He remarked. “It looks like two flows.”
Jim gestured to the vesicles observed in the rocks where they met in the flow contact zone, as well as a thin layer of weathered rock that lay in between.
This layer is known as a “weathering horizon,” and is another feature of flow contacts.
The tops of many of these basalt flows were weathered during the many thousands of years before the next flow—rain, sun, fire, vegetation growth, and soil formation breakdown the flow top, leaving distinct weathering horizons that were later buried by the next lava flow.
Geological Unit
Looking at the map, Jim hypothesized that we were at the base of the Downey Gulch flow and moving into Grouse Creek—the name given to a younger basalt flow in the overall sequence.
“What makes a unit?” I asked Jim.
Jim explained that each geological unit shares a similar chemistry and are thought to be close together in time.
“Originally, these flows were defined by their magnetic orientation,” Jim continued.
Basically, he explained, the magnetic field has switched at various points during the emplacement of the lava flows. A magnetometer was used to determine the direction of the magnetic field when the flow was emplaced—during a normal orientation or reversal—the sequence of orientations helped to correlate individual flows from place to place as well as to deduce their timing. These correlated flows were then defined as a sequence of geological units.
“We are walking through a magnetic field right now,” Jim proclaimed.
Unstable
As we hiked, it was hard to ignore the blackened trees standing along the trail. The Oneonta Trail is one of many trails in the gorge that was heavily impacted by the Eagle Creek Fire of 2017.
In general, the gorge is an area “susceptible to rock slides,” according to Jim—only the fire has made it “even worse.”
Fires are impactful, especially a couple of years after the fire. The fire kills the trees that normally stabilize the upper part of the soil. As the roots of the trees begin to lose their integrity over the years, that is when rocks and other debris can come loose and fall.
“We have done a bit of work on how the fire has affected the slope processes in the gorge,” Jim remarked. “It really is important work.”
“In January 2021, a woman was killed at the Ainsworth exit by a debris flow,” Jim confided. “That area has been a constant source of debris flow.”
Signs have been erected throughout the burn area for this reason. When entering a burn area, one should prepare for hazards and know the risks.
Columbia River Course
Speaking of water, as Jim and I headed around a bend in the trail, Ponytail Falls came into view. The small falls poured over the basalt cliff into a deep pool below.
The trail took us behind the falls into a cavern where we could get a look at some sediment that was probably present before the lava flow poured through.
“Looks like some sort of floodplain or silty or maybe even a bog,” Jim speculated as he scraped at the baked sediments.
“It is pretty cool to see exposures like this,” he proclaimed. “Paleogeography is what I am interested in… the area between the flows.”
Jim explained how looking between flow deposits can tell you a lot about landscapes of the past. Gravel, for example, indicates a river flowed through—the different sizes of clasts, details about the flow.
One of Jim’s projects is to map out the course of the Columbia River over time by studying these sorts of sediments.
“The Columbia has been roughly in the same place the last 50 million years,” Jim said. But it has wiggled around a bit. Over the last 16 million years the river has been displaced North. In fact, Columbia River Basalt Flows through Silver Falls show that the river came through near Salem during the early phase of the Columbia River basalt flows.
Because the Columbia River system starts at the Continental Divide in the Rocky Mountains, there are a lot more exotic rocks from the east that can be found in the sediments of the Columbia. Mica flakes, mainly from older rocks to the east, are a distinctive component of Columbia River sediment deposits.
The small grain size of the sediment found in the waterfall alcove didn’t show any signs of mica. Whatever stream systems were attached to the body of water that was here in the past must have been more localized.
It’s My Fault
As we walked away from the waterfall, I noticed a large crack in the lava cliff the water flowed over. I asked Jim about it.
“Those could just be from downcutting,” said Jim. As the stream cuts down on the rock, the material is eroded away, and the land is decompressed enough that it may lift and crack.
However, though not likely in this case, cracks can also be formed through faulting. And in the Columbia River Gorge, there are a lot of them.
Jim explained how the state of Oregon is essentially rotating clockwise. The Columbia River is bounded by North trending faults that through rotation are offset laterally, resulting in a collection of strike-slip faults that move land to the west northward.
“Everyone knows about the subduction zones,” said Jim, “but these younger faults are a seismic hazard.”
“One of the main motivations [for mapping the Gorge],” Jim continued, “is to better understand the seismic and other hazards.”
Jim told me about a fault scarp that was discovered just east of Cascade Locks recently. The scarp here is large—five to six feet tall by Jim’s estimation—and would have been the result of a significant earthquake event.
“One idea is that the shaking might have triggered the Bonneville landslide,” said Jim. A landslide that was large enough that it blocked the Columbia River Valley, creating the legendary Bridge of the Gods and the Cascade Rapids.
“We know from tree ring work and carbon dating that the landslide occurred in the 1440s,” said Jim.
With more faults being discovered all the time in the Gorge, the question is when will the next earthquake (and possibly a major slide) occur.
Finding Fault
As we continued away from the falls, I asked Jim how geologists usually identify a fault—are they easy to find?
“There has been a revolution in topographic resolution,” Jim responded. “We can see the landscapes with what we call LiDAR.”
LiDAR works by shooting a laser down from an aircraft that can pick up on the subtle variations in the Earth’s surface.
“More and more of Oregon has been flown for its topography,” said Jim. Now there is imagery of about two-thirds of Oregon, according to Jim.
“It’s like crack for geologists,” he smirked.
Jim also mentioned that it may not always be possible to identify scarps/faults on such steep topography as you find at the Gorge.
“The terrain is so steep,” said Jim, “so scarps don’t last long; things get chopped up.”
Either way, Jim is certain of at least some faults—the nearest being at the Ainsworth exit.
A Room with a View
Jim and I passed by another striking entablature zone, before reaching a junction to the right towards a viewpoint. We decided to stop for a quick detour.
At the end of the short trail are a memorial and a nice view looking to the northeast.
“Well, you can see the toe of the Bonneville landslide,” Jim noted, pointing to a hummocky peninsula jetting out a bit into the Columbia.
“In this area…the uplift has been the greatest,” Jim explained looking out toward the Columbia River. “Everything has arched up 3,000 feet between the Portland Basin and the Dalles.”
Yet, “The river has been at or above sea level the whole time,” continued Jim. “It has been cutting through the whole time… as it cuts through, it undermines the sides.”
Additionally, below the lava flows is the Eagle Creek Formation—a mixture of volcanic sediment from Western Cascade eruptions 20 to 30 million years ago (mya). This underlying material also tilts to the south—giving the Washington side of the Gorge its more gradual slopes compared to Oregon’s steep terrain.
The combination of uplift and a weak, incoherent underlying geology that tilts make the area landslide prone.
“All of that terrain is landslide terrain,” proclaimed Jim, the powerful Columbia rolling past as we spoke.
Beacon to the Past
As we looked out at the scenery, Jim pointed out other features from the viewpoint. One of the most prominent being Beacon Rock.
Long after the flood basalts spread across much of Oregon and Washington, volcanism returned to the region in the form of a large volcanic field about 3.6 mya—reaching from the West Hills of Portland all the way to the Deschutes. Cinder cones, lava flows, and shield volcanoes have been identified in the volcanic field, including Larch Mountain in the Columbia River Gorge Scenic Area and Mount Scott in Portland. Though Jim couldn’t point to an exact source of the volcanism, he was fairly certain it was in part related to subduction.
“Beacon Rock is the youngest in the basin,” said Jim—erupting about 60,000 years ago. Once a small cinder cone, much of the unconsolidated material that surrounded its volcanic neck has been washed away by massive ice age floods.
Now a bare rock sits along the northern shores of the Columbia.
Missoula Floods
This brings us to one of the most recent installments of the Columbia River Gorge’s geological story, the Missoula Floods.
Around 20,000 to 15,000 years ago, Glacial Lake Missoula would form as the Cordilleran ice sheet flowed south from Canada creating an ice dam that blocked the Clark Fork River drainage in Montana. Periodically, this ice dam would fail, and flood waters would pour out of Glacial Lake Missoula through Washington and Oregon as it followed the Columbia River drainage.
The effects were dramatic. Scablands were created through much of eastern Washington, as the water tore up the soil and carried it downstream where much of it settled in the Willamette River Valley or was carried out to the Ocean. The gorge played a significant role in controlling the flood flows.
“The Gorge was a constriction or nozzle the floods were forced through,” Jim explained. “It was mainly erosive through here.
Inverted Topography
After enjoying the views and trying our best to identify the peaks on the horizon—is that Defiance or…?—we turned back to the main trail leading deep into the basalt cliffs that make up the Oneonta Gorge.
As we looked up at the steep cliffs, we could see another entablature, as well as a couple of flows. Jim referred to his geological map to see if we had moved into a different flow unit.
“Nope, still Grouse Creek,” he concluded.
However, looking back and up to the top of the canyon walls, Jim pointed out a lava-capped ridge—Franklin ridge, he called it.
Franklin Ridge is the result of a local volcanic center, Jim explained, erupting somewhere around 1-3 million years ago. At that time, the lava flow was down a valley that led to the Columbia River. But since the lava flow was uplifted and the surrounding rock eroded away, the lava-filled valley bottom “is now a ridge.”
“A topography reversal,” Jim called it, or inverted topography—essentially the low points become the new high points.
“Kind of cool,” Jim remarked.
Water
Jim and I continued to make our way on the dry dusty trail. Soon we passed by a moss-covered basalt rock face dripping with water—not an unusual sight to see.
I asked Jim about what we were seeing.
“It is probably emerging at the top of this flow,” said Jim—at a flow contact.
“Water moves slowly through solid flows,” Jim explained, “seeping through the cooling joints. It moves more rapidly between flows.”
This is visible as wet stripes in the contact zones between basalt flows, especially in the arid east where the moisture contrasts well against the dry rock. You can literally see where the moisture is in the rock, according to Jim.
Transportation of water through the Columbia River Basalts varies a lot but can be rather slow—sometimes taking thousands of years to emerge at a contact zone.
“Some of the water in the layered basalt, especially in the east, certainly dates from the last ice age,” said Jim.
Understanding the basalt stratigraphy for the purposes of managing water resources is another reason mapping the Columbia River Gorge is important work. Knowing the flow contacts and the connections to each local aquifer system, are all part of ensuring water use is sustainable over time.
Lower Falls
Jim and I cruised past another flow top—“See how vesicular it is?” he questioned.
Then, we passed an entablature holding up the cliffs before passing over a deep gorge and reaching the lower falls.
Water ran smoothly down a narrow slide of lava littered with tree trunks—the water cut deeply into the erosion-resistant basalt to form a V-shaped valley.
The Lower Falls reminded me of just how powerful water is as an erosive agent. The canyon walls were high above us. Oneonta Creek had done a fine job cutting down.
Ortley Flow
After climbing a bit deeper into the canyon, we passed by a wilderness sign and an unusual-looking rock face. Jim checked his map and sure enough, we had reached the base of the Ortley flow.
Jim pointed out the differences in morphology that were clues that we were entering a different flow unit.
“One thing that is our first hint is it is oriented sideways,” said Jim. “Cooling fronts are coming from the sides.”
Jim explained that some of the Ortley lava flow must have been locally channelized—flowing through an old river or creek bed.
Vertical columns form slowly as they cool from the bottom and top. These sideways columns were the result of multiple cooling fronts on the sides of the channel.
The bottom of the flow also looked different. Rounded blobs of broken glassy rock were visible, a.k.a. pillow basalt. As Jim explained, when lava encounters water it will either brecciate (break apart into angular fragments) and may form “pillows.”
“Hyaloclastites,” said Jim, referring to the glassy rocks. “The pillows are direct water features.”
Interflow Zone
We walked on up the trail. Only we hadn’t made it far when Jim stopped suddenly.
“I am just noticing some rounded clasts,” he said. “It is making me think there is probably an interflow zone with gravel coming out of it.”
This was Jim’s bread and butter. He started digging through the rocks with intense focus.
“Is there anything other than basalt?” he wondered. “Any hint of the Columbia?”
The rocks were large and rounded, embedded in a “really weathered matrix.”
Jim speculated that the interflow zone was probably a steeper tributary to the Columbia that was back flooded when lava came pouring through the Columbia River drainage.
“This might be related to the pillows we saw earlier,” Jim hypothesized.
He continued poking around in the matrix for loose material until he found a rock that he thought “looked different.”
He picked it up with the thought of cracking it open for further examination.
“Why did you pick that one up?” I asked, curious about what he was seeing.
“It is light colored,” he said with a smile.
It was as simple as that.
Triple Falls
Eventually, we reached Triple Falls—a three-pronged falls flowing over Columbia River Basalts—and our destination for the day. We wondered at the strange flow pattern of Triple Falls and marveled at its beauty.
Triple Falls is classified as a segmented waterfall with flows that drop vertically into a large plunge pool below. Each of the segments flows between bulbous volcanic rock barriers. Were these rocks more resistant than the rocks the water poured over?
Just past the falls, we crossed a bridge and stopped for a short lunch break before heading back to our cars.
Boot Prints in the Sand
On our way back Jim and I talked about all sorts of things—from travel to hiking adventures and pet projects—as we retraced our steps.
Eventually, though, the conversation turned back to the geology of the Columbia River Gorge as something shiny caught Jim’s eye.
“This is interesting,” he said reaching down toward the dusty ground. “This is Columbia River sand.”
The light tan color of the earth faintly shimmered as mica flecks in the sediment caught the sun’s rays, Jim explained that the Columbia River mica flakes are large and platy in structure. When light hits them just right, they shimmer.
“Not sure how it got here,” Jim said quizzically. “It could have blown in…or been brought in by the Missoula floods… or it could be coming out between flows.” There was no way of knowing.
Puzzle
The rest of the hike back went by quickly. Time seems to slip by when you are heading downhill. And before I knew it, we were back at the trailhead saying our goodbyes.
Jim was a lot of fun to hike with. Every time we ran across something new, it was like finding another piece of a geology puzzle.
There is a lot we know about the Columbia River Gorge. Its unique geological history means it has attracted a lot of attention over the years. Yet, there are still a lot of pieces to the puzzle to be found.
Finding those pieces and fitting them together is what geologists, like Jim, do. Offering a glimmer of understanding to folks like me—to help us see the mica through the sand.
Jim O’Connor is a research geologist with USGS. He earned a Bachelor of Science degree in Geology at the University of Washington before going on to earn an M.S. and Ph.D. at the University of Arizona. Jim has been with USGS since 1991 where he has focused his efforts on understanding the processes and events that have shaped the Pacific Northwest.