Ice Age Floods National Geologic Trail
Ice Age Floods National Geologic Trail
W line sec 25, T2N R43E, at intersection of side road with Palouse Highway; 3 ft of superficial loess on 2 feet of laminated pale silt on 8-11 ft of till with "Palouse soil"-like matrix, on "Palouse soil" Flint 1937 Drift Paper
W line sec 25, T2N R43E, at intersection of side road with Palouse Highway; 3 ft of superficial loess on 2 feet of laminated pale silt on 8-11 ft of till with "Palouse soil"-like matrix, on "Palouse soil" Flint 1937 Drift Paper— Flint 1937
This is one of more than 1,800 field sites cataloged in the early scabland surveys — the bedrock of the Ice Age Floods scientific record. The Bretz-era researchers walked the ground first; modern cosmogenic dating, LiDAR, and remote sensing have since extended and refined what they mapped.
The Idaho Panhandle reach of the Ice Age Floods story covers the corridor where Lake Missoula water actually left the ice dam and entered the broader Columbia River system. The region includes Lake Pend Oreille, Lake Coeur d'Alene, the Rathdrum Prairie (between Athol, Rathdrum, and Post Falls), Spirit Lake, the Pack River delta near Sandpoint, the Purcell Trench between the Selkirk and Cabinet Mountains, Cabinet Gorge at the Idaho-Montana border, Farragut State Park at the south end of Lake Pend Oreille, and the Spokane River outflow at Post Falls. Hydrologically this is the immediate downstream end of every Missoula flood: water released from glacial Lake Missoula passed through this funnel before spreading west onto the Channeled Scabland.
The Purcell Trench is a north-south, glacially modified valley between the Selkirk Mountains to the west and the Cabinet and Purcell Mountains to the east. Its origin is tectonic (Eocene normal faulting), but during the late Pleistocene the Cordilleran Ice Sheet pushed a tongue of ice south down the trench from British Columbia. This tongue, the Purcell Trench lobe, repeatedly filled the Lake Pend Oreille basin to thicknesses estimated at 4,000+ feet and widths of 30+ miles, scouring the bedrock floor and damming the Clark Fork River where it now enters the lake near the town of Clark Fork, Idaho. The ice dam stood roughly 2,000 feet high, and water from the upper Clark Fork drainage ponded behind it as glacial Lake Missoula — at maximum, about 3,000 square miles in area and about 500 cubic miles in volume.
The dam failed catastrophically, repeatedly. Estimates put the number of flood cycles at 40 or more across roughly 2,000 years of the late Pleistocene, with recurrence intervals on the order of 10-60 years. Failure mechanisms remain debated (overtopping, flotation, subglacial tunneling, thermal erosion); the 2020 USGS review concluded that overtopping or tunneling, possibly with thermal erosion, are the most defensible options. The largest documented flood is dated at 18.2 ± 1.5 ka via cosmogenic 10Be on flood-transported boulders (Balbas et al., 2017); that early flood routed northwest down the Columbia valley before the Okanogan lobe advanced across and diverted later floods south through the Channeled Scabland.
Balbas et al. dated the near-maximum lateral moraine of the Purcell Trench lobe at 15.7 ± 1.3 ka (averaged 10Be ages on moraine boulders). A boulder in till and outwash at the southwest margin of Lake Pend Oreille returned 14.3 ± 1.2 ka, dating final retreat of the lobe out of the lake basin. The last great Missoula flood is dated at 14.7 ± 1.2 ka, marking the moment the Purcell Trench lobe withdrew far enough north to stop blocking the Clark Fork. After that, the ice dam was gone and no more Missoula floods occurred.
The Rathdrum Prairie is what those floods left behind. It is a flat, broad outwash plain stretching roughly 40 miles southwest from the south end of Lake Pend Oreille at Bayview through Athol and Rathdrum to Post Falls, where the Spokane River cuts out of it. Beneath the prairie surface sits 150 to more than 800 feet of unconsolidated, coarse-grained sand, gravel, cobbles, and boulders — the Spokane Valley-Rathdrum Prairie aquifer, deposited almost entirely by the Missoula floods. Material coarsens toward the center of the valley and is mantled by mega-bedforms along its margins. Balbas et al. dated three boulders on broad expansion bars at the head of the Rathdrum valley to a range of 14.3 ± 1.2 to 17.2 ± 1.4 ka, with the 14.7 ka cluster marking the last flood across the prairie.
The most accessible flood bedforms in the region are the giant current ripples east of Spirit Lake along Idaho Highway 54. Built of gravel and pebble foresets and oriented transverse to flow, individual ripples reach 15 to 50 feet high, 100 to 250 feet wide, and 300 feet to a half-mile long. The cusps are convex upstream, arms downstream — the same morphology as small bedload ripples but scaled for floodwaters moving on the order of tens of meters per second under hundreds of feet of head. The Hoodoo Channel, which crosses U.S. 95 about three miles south of Careywood, was the major flood outlet for the later (and possibly final) outbursts, draining south and west out of the Pend Oreille basin past Spirit Lake. Spirit Lake itself is a debris-dammed marginal lake left behind when the flood retreated.
Lake Pend Oreille is the deepest survival of the system. The Purcell Trench lobe gouged the basin to depths now reaching 1,152 feet — fifth deepest in the United States — and floods later scoured an additional 1,500 feet of glacial-fluvial sediment out of the deepest holes. The lake is dammed at its southern end by the same flood gravels that built the Rathdrum Prairie.
J Harlen Bretz worked the Idaho Panhandle in his second and third field seasons. His 1923 paper "The Channeled Scabland of the Columbia Plateau" (Journal of Geology v. 31) established the flood hypothesis; the follow-on two-part 1925 paper "The Spokane Flood beyond the Channeled Scablands" (Journal of Geology v. 33, parts I and II) extended that argument to the upstream reach in northern Idaho and the Spokane Valley. The 1925 paper is the foundational description of the Rathdrum Prairie as the immediate exit of the floods and of the Spokane Valley as the conduit carrying flood water from Lake Pend Oreille west onto the scabland.
Bretz's localities from these seasons turn up in the P1-23 / P2-23 / P3-23 series of field-note designators (his shorthand for Pleistocene/Plateau localities catalogued in his University of Chicago notebooks) and continue into the 1925 numbering. Many of the northern Idaho station numbers — Rathdrum Prairie cuts, Spirit Lake area exposures, the Spokane River channel through Post Falls, gravel quarries near Coeur d'Alene, and the Pend Oreille shoreline — come from these two seasons. Bretz's original journals from 1919-1929 were digitized in 2024 by Nick Zentner and the University of Chicago Library and given public usage rights by the Bretz family. Glenn Cruickshank has mapped the journal localities into per-year Google Earth files, which now make it possible to put any Bretz station on the modern landscape and re-walk what he saw.
Bretz did not have a chronology. He had landforms — the bedforms, the gravel bars, the scoured channels, the immense aquifer — and he correctly inferred catastrophe from morphology alone. The cosmogenic dating that anchors the modern story came decades later.
The chronology now rests on cosmogenic 10Be exposure dating of flood-transported boulders and moraine boulders, primarily from Balbas et al. (2017, Geology v. 45). Key results for this region:
The natural starting point is the Museum of North Idaho in Coeur d'Alene (115 Northwest Blvd, reopened April 2024 at the foot of Tubbs Hill), where the permanent Missoula Floods exhibit interprets the geography and geology of the panhandle alongside the Schitsu'umsh, fur-trade, and mining stories of the same landscape.
Farragut State Park, on the south end of Lake Pend Oreille near Bayview, sits at the breakout point where the ice dam failed and water and ice burst from the lake basin. The visitor center carries flood interpretation, and the park itself contains erratics, gravel bars, and the upper end of the Hoodoo Channel. From Farragut, Idaho Highway 54 runs west through the Spirit Lake mega-ripples, where the road cuts the ripple field on a clean transverse section.
Cabinet Gorge has an AVISTA-maintained viewpoint with Ice Age Floods signage at the dam, marking the likely position of the ice plug. North of Lake Pend Oreille, Highway 200 west of Hope offers a pullout looking down the lake into the trench, and the Clark Fork Drift Yard sits on the delta at the river's mouth — the same delta the floodwater poured off of. Erratics scattered along Highway 200 east of Clark Fork, in Sandpoint's City Beach, and at Farragut record ice-rafted transport from upstream. The view from City Beach in Sandpoint or from Schweitzer Mountain Road is the cleanest way to see the Purcell Trench itself.
The IAFI's Idaho activity is anchored by the Coeur d'Alene chapter and by periodic field meetings — including the 2005 biannual IAFI meeting in Sandpoint, which produced field-trip guides covering the Clark Fork ice dam and adjacent flood features that are still in circulation.
Active: 1925-1976 (Spokane-region glacial mapping 1935-1938) Affiliation: Yale University (faculty 1925-1976, full professor 1945) Key paper: Flint, R.F. (1938). "Origin of the Cheney-Palouse Scabland Tract." Geological Society of America Bulletin 49(3): 461-524.
Flint conducted three field seasons of glacial mapping in the Spokane region beginning in 1935 and was, for the first two decades after Bretz's 1923 paper, the most prominent skeptic of catastrophic flooding. His 1938 paper argued the scablands were cut by ordinary glacial meltwater operating over long durations, an interpretation that delayed acceptance of Bretz's hypothesis until Pardee's 1942 paper made the catastrophic source unavoidable. Flint's larger contribution to flood-corridor geology was methodological: his 1947 textbook Glacial Geology and the Pleistocene Epoch (and its 1957 and 1971 revisions) became the standard reference that defined how the Cordilleran Ice Sheet's southern margin, ice-dam configurations, and lobate retreat were mapped across the Pacific Northwest. He helped establish the Yale Geochronometric Laboratory and brought radiocarbon dating into glacial stratigraphy. He died of a stroke in 1976.
Source: Richard Foster Flint - Wikipedia; GSA Memorial, Stephen Porter
Every site along the trail will receive the full Terrain360 capture treatment: ground-level 360° panoramas, drone aerial imagery, and photogrammetry-based 3D models that visitors can spin in their browser. This page reserves the slots; the imagery flows in as field capture completes.
Ground-level 360° panorama, every step along the feature, captured by Terrain360 field crews.
Drone flyovers reveal the geometry of catastrophe — ripple marks, gravel bars, and scour patterns invisible from the ground.
Photogrammetry and Gaussian-splat models let visitors rotate, measure, and inspect features in detail-page WebGL viewers.
These till deposits help define the maximum extent of Cordilleran Ice Sheet lobes into the Columbia Plateau. Modern dating techniques have refined the timing of ice advances, showing they correlate with periods when flooding would have been blocked by ice, adding complexity to the flood chronology.