STOP 12: Mazama ash at Oregon-Washington state line (Fig. 59). Park on the right (southeast) side of U.S. 730. Be alert for traffic and rattlesnakes.

Between the basalt cliffs and the Columbia River are talus slopes and alluvial fans that have formed since the passage of the last Missoula flood about 12,700 years ago. This alluvial fan was partly covered by a debris flow about a decade ago; notice the boulders, particularly near the head of the fan. The channel on the alluvial fan, partly created by the Washington highway department, reveals about 3 m of Mazama ash and fan sediments. Mount Mazama underwent a catastrophic eruption and caldera collapse 6,845 years ago that resulted in Crater Lake (more than 400 km away in Oregon's southern Cascades) (Bacon, 1983). A thin blanket of tephra fell on the Horse Heaven Hills and as far northeast as Alberta (Wilcox, 1965). The ash washed off the Horse Heaven Hills, down gullies, and onto alluvial fans where it was buried and preserved.

Across the river at the base of the cliffs are more alluvial fans (Fig. 60). From here the light-colored lenses of Mazama ash are visible in railroad cuts through the alluvial fans. Gravel bars (Fig. 60) were deposited by the Missoula floods high in canyons tributary to the Columbia River here at Wallula Gap.

After passing through Wallula Gap the Missoula floods poured into the Umatilla Basin and then rushed down the Columbia River Gorge and inundated the Willamette Valley before draining into the Pacific Ocean at Astoria (Fig. 5).

Between U.S. Highway 730 and the railroad tracks is a railroad rockfall warning device designed to alert train engineers if large rocks fall from the cliffs or down the talus slopes. The wire fence here has a mesh size such that small rocks may pass through and land between the tracks. Any rocks large enough to obstruct passage of a train would hit the fence and send an electrical alert signal to trip a semaphore to stop the train.

Continue northeast on U.S. Highway 730, returning to Wallula Junction.

Wallula Junction. End of U.S. Highway 730. Proceed east on U.S. Highway 12 toward Walla Walla. On the right (south) side of U.S. 12 is a row of about a half dozen faceted spurs along the Olympic-Wallowa lineament (Figs. 53 and 61). The faceted spurs are truncated ends of ridges between the gullies that were cut by the intermittent streams that flow north off the Horse Heaven Hills: each truncated spur is a triangle about 100 m high.

Across the road from the Wallula Habitat Management Unit sign, a 1-cm-thick bed of Mount St. Helens set S tephra is exposed in Touchet Beds; this tephra is about 13,000 years old (Mullineaux and others, 1978). Mount St. Helens ash did not fall this far south on May 18, 1980; distribution of ash is controlled by the prevailing winds at the time of the eruption.

The Walla Walla River is meandering across its flood plain in the lower Walla Walla Valley. On the north side of the valley are exposed lava flows of the Frenchman Springs Member (15.5 m.y. old) of the Wanapum Basalt and the Umatilla Member (about 14 m.y. old) and Ice Harbor Member (Martindale flow)(about 8.5 m.y. old) of the Saddle Mountains Basalt (Swanson and Wright, 1981). (See Fig. 62.) The ancestral Clearwater-Salmon river system cut a valley in the Umatilla basalt and partly filled that valley with basaltic, metamorphic, and plutonic clasts. At about 8.5 m.y. ago, the Martindale flow of the Ice Harbor Member capped the sequence (Fecht and others, 1987).

The drainage system of the Walla Walla River was established by the late Miocene to middle Pliocene (Fecht and others, 1987, p. 240). The oldest unconsolidated deposits in the Walla Walla Valley are termed the "old gravel and clay" (Newcomb, 1965, p. 20-22). In the late Pleistocene, the Missoula floods partially filled the Walla Walla Valley with Touchet Beds. In the Holocene, the river has partially excavated these beds, forming a flood plain and locally cutting down to the basalt bedrock.

Bridge over the Walla Walla River at Reese (elevation 389 ft). Roadcuts from here to the top of the hill are in the Frenchman Springs Member of the Wanapum Basalt (Swanson and Wright, 1981).

There is an area of channeled scabland near the hill just southeast of here (elevation 716 ft) and the hill at Divide. There are small buttes, or scabs, of basalt and shallow coulees with southeasterly orientations. The erosion that shaped this area occurred as the Missoula floods overtopped the hills as they poured from the Pasco Basin into the main Walla Walla Valley. With flood levels at 1,200 ft, there was 150 m (500 ft) of water flowing over these hills.

STOP 13: Unconformities at Divide (Figs. 63 and 64). Park at the east end of the roadcut and walk back down the hill. Be alert for traffic.

This is the crest of a north-trending anticline at the west edge of the main Walla Walla Valley (Swanson and Wright, 1981). The Frenchman Springs Member of the Wanapum Basalt is overlain by poorly sorted and crudely stratified caliche-rich strata. Almost all the clasts are basalt, but at least two weathered erratics (one granite and one gneiss) have been discovered near the base of the sediments.

Vrooman and Spencer (1990) interpret these sediments as the remains of two debris flows that were covered by colluvium (talus), loess, and, finally, a soil. According to their scenario, there would have been a hill of basalt just to the east of these sediments. An ancient jökulhlaup left one or more icebergs that were transporting the erratics on the hill. Mass wasting of the western hill slope brought the erratics downslope into the diamictons. The geomorphology, the presence of weathered erratics, and the degree of development of the caliche all suggest great age, perhaps several hundred thousand years, for these Pleistocene diamictons.

Overlying the diamictons and the basalt are about a dozen thin Touchet Beds deposited by the late Pleistocene Missoula floods (Fig. 63). Holocene eolian sediments (mostly loess, but some sand) cap the section.

To the north of the highway are iceberg-transported cobbles and boulders. Walk a short distance (50-100 m) north from the west end of the roadcut to see a variety of iceberg-transported erratics (granite, gneiss, quartzite, sandstone, and siltstone). Bartlett and Carson (1995) reported on a bergmound located 10 km southeast of here. They estimated that the pile of large boulders had been deposited by an iceberg with a volume of 6,000-12,000 m³ (200,000-400,000 ft³).

Here at Divide is evidence for glacial flooding not only in the late Pleistocene but also much earlier in the Pleistocene. The evidence for late Pleistocene jökulhlaups from glacial Lake Missoula is twofold: (1) Touchet Beds near the top of the exposure and (2) erratics at the surface. The evidence for jökulhlaups earlier in the Pleistocene consists of the erratics near the base of the diamictons.

Vrooman and Spencer (1990) interpret the diamictons as follows:

"The lowermost unit consists of a fining-upward diamicton. Large clasts are basaltic except for a few clasts of exotic lithology. Textural parameters suggest deposition by a mudflow, or colluviation concurrent with deposition of loess. A truncated clastic dike in this unit indicates a period of erosion before deposition of the overlying units, which are coarser-grained and caliche-rich. Texturally, these upper units, in which exotic lithologies are lacking, suggest colluvium deposited as a nearby basaltic high weathered and mass-wasted. The upper, caliche-bearing portion of these units is fine-grained, suggesting a period of loess deposition. The presence of caliche indicates that colluviation was followed by long periods of soil development; based on the apparent stage of development, each caliche horizon may represent a minimum of 5-10 Ka, and perhaps considerably more."

The evidence for two distinct ages of flooding is an example of the principle of uniformitarianism, which geologists sometimes state as "the present is the key to the past." On this hill today are basalt outcrops, loess, and erratics from the latest glacial Lake Missoula floods. With time, these materials will creep (a type of slow mass wasting) down the hill, and a soil containing caliche will form. Evidently, this is what happened earlier in the Pleistocene, with the result that the erratics were buried, preserving the evidence for the ancient floods.

Return to your vehicle and continue east on U.S. Highway 12. To the south are the Horse Heaven Hills anticline and fault scarps of the Wallula fault zone. To the east is the much larger Blue Mountains anticline; this northeast-trending anticline is composed mostly of the Columbia River Basalt Group. However, erosion at the crest and in some deep canyons has resulted in limited exposures of Mesozoic granitic rocks and of parts of accreted terranes of mostly Mesozoic metamorphic rocks (Fig. 15).

Cross the Touchet River (near its confluence with the Walla Walla River), then enter Touchet (elevation 443 ft). To the south, on the left bank of the Walla Walla River, is the type locality for the Touchet Beds (Flint, 1938). (There are still better exposures at Stop 14.)

Near the east end of Touchet, turn left (north) on Touchet Road (or Touchet River North Road).

At the south end of the roadcut in Touchet Beds is exposed Mount St. Helens set S tephra (Fig. 65). In places there are two closely spaced thin ash layers, suggesting that the eruption pattern of 13,000 years ago was somewhat like that of 1980—multiple eruptions over a period of time. In 1980, winds distributed ash from Mount St. Helens eruptions as follows: May 18, east-northeast; May 25, northwest; June 12, southwest; July 22, northeast; August 7, north-northeast; October 16-18, southwest (and southeast) (Sarna-Wojcicki and others, 1981). The result was partial geographic overlap of multiple ash falls. Between the thin set S ash layers is a few centimeters of eolian silt, or possibly sheet wash sediment (Waitt, 1980, p. 664-667; 1985, p. 1273).

STOP 14: Touchet Beds and clastic dikes. Park on west side of road. Beware of black widow spiders at this stop.

Although the Touchet Beds were originally believed to represent only two major floods (see review in Carson and others, 1978), most geologists now believe (see, for example, Waitt, 1980, 1985) that each Touchet bed represents one jökulhlaup or glacier outburst flood. Waitt (1980, 1985) has found considerable evidence that the Touchet Beds were deposited by dozens of Missoula floods. Between some flood-deposited rhythmites are slope wash sediments in channels, loess, volcanic ash, rip-up clasts (material pried loose from underlying sediments), and clayey lacustrine beds. Further evidence for a significant amount of time between successive floods includes rodent burrows and reworked shells and vertebrate remains—proof that animals were living in and on the Touchet Beds between these floods (Waitt, 1980, p. 668).

According to Waitt (1987, p. 345), "floodwater backed up dead-end valleys off the main Scabland floodways to form transient ponds in which suspended load settled. Because the side valleys were protected from violent currents, flood-laid strata were not eroded by later floods but became buried and preserved."

Spencer (1989) compared the sedimentology and paleontology at this stop and at the 'Little Grand Canyon' near Lowden (see mile 190.4). The Little Grand Canyon is in the middle of the Walla Walla Valley, but here the Touchet valley is narrower and has a higher elevation and a steeper gradient. Compared with those in the Little Grand Canyon, the graded rhythmites at this stop are poorly developed. More rapid drainage at this site resulted in more vigorous post-flood erosion and removal of the finer upper part of the rhythmites here. At this site Spencer (1989) found rodent fossils (including ground squirrel, vole, and kangaroo rat) and burrows indicating colonization by small animals between flood episodes.

The clastic dikes range in thickness from millimeters to meters and have various strikes and dips—most large ones are nearly vertical. Many consist of alternating vertical laminae of silt and sand (Fig. 66). There are many proposed origins (summarized from Carson and others, 1978): earthquake fissures, cracks that formed as buried ice melted, fissures due to erosion by underground streams, landslide fissures, permafrost-related crevices, desiccation fractures, ground-water injection, and extensional fractures accompanying slumping of unstable rapidly deposited turbidites.

Black (1979) studied the clastic dikes in the Touchet Beds and other geologic material in the Pasco Basin. Most dikes there "are megascopically similar in texture, fabric, and relation to their host." They "display the pronounced vertical layers of sand and silt, and have thick clay and silt coatings separating them and lining the walls of the host. The bulk of the material came down from above aperiodically as openings were enlarged underwater" (Black, 1979, p. 55).

Black (1979) concluded that the clastic dikes do not all have the same origin. However, for most of the dikes, he "hypothesized that hydraulically dammed Pleistocene floodwater that repeatedly covered the Pasco Basin is primarily responsible for the initial fracturing, for the aperiodic widening, and for the primary source of sediments that filled the fractures" (Black, 1979, p. 61).

While most of the dikes were filled from above, some small dikes were produced when liquefied layers of sediment were injected into overlying layers. Sudden loading by a Missoula flood or earthquake vibrations may have caused the unconsolidated sediment below the water table to liquefy. The weight of the overlying sediments forced the liquefied sediments to intrude upward to form the dikes.

Turn around. Drive south down the Touchet valley.

In Touchet, turn left (east) on U.S. Highway 12. All roadcuts between here and Walla Walla expose Touchet Beds.

The Battle of Walla Walla was fought in December 1855; the Oregon Militia and the Walla Walla Indians had a 4-day skirmish from here up the Walla Walla River to Mill Creek (Majors, 1975).

Cross Dry Creek and enter Lowden (elevation 489 ft). To the south is a hill of Touchet Beds, an erosional remnant left as the Walla Walla River cut its flood plain.

Three kilometers to the south is the Little Grand Canyon. The canyon is on private property and entry without permission is forbidden. Because of the potential danger at the Little Grand Canyon, if you plan to visit the site you must have liability insurance specific to this site and a notarized hold-harmless agreement. For information, contact Stuart Durfee, Secretary/Manager, Gardena Farms District No. 13, Box 137, Touchet, WA 99360 (509-394-2331).

The Little Grand Canyon (Figs. 67 and 68) cuts the south side of the ridge visible to the south. The ridge is part of an extensive but dissected surface that slopes gently down the Walla Walla Valley. The surface is at 800 ft just west of College Place, 625 ft here, and 500 ft southwest of Touchet. This surface is the top of the Touchet Beds, with about a meter of loess added. In other words, the surface is the bottom of the intermittent lake that aggraded during each Missoula flood. The surface has been dissected by the Walla Walla River and its tributaries.

Along the crest of the ridge runs the Burlingame irrigation ditch. This ditch is fed by a diversion structure near Mile 36 of the Walla Walla River (southwest of College Place).

The history of the formation of the Little Grand Canyon is summarized from Soil Conservation magazine (December 1935, p. 14-15). In 1904, a diversion canal was built from Burlingame irrigation ditch south to Pine Creek; the purpose was to control discharge in Burlingame ditch, especially when spring winds pile enough tumbleweeds into the ditch to choke the water flow. Until March 1926, the diversion canal was a gully about 10 ft deep and 6 or 8 ft wide. Then for 6 days steady winds caused enough tumbleweeds to accumulate to force the entire 80 ft³/s (2 m³/sec) flow of Burlingame ditch into the diversion canal. The water soon eroded 4,725,000 ft³ (134,000 m³) of Touchet Beds.

In 1927, an overshot flume was constructed to prevent further headward erosion. The flume collapsed into the canyon and another was built. Within a decade the canyon was 100 ft (30 m) wide and as much as 120 ft (36 m) deep. The Soil Conservation Service calculated that 9,588,000 ft³ (271,000 m³) of earth had been removed, and engineers recommended that a wooden flume be constructed around the gully.

The Touchet Beds exposed in the walls of the Little Grand Canyon (Fig. 48) have been studied many times. (See Carson and others, 1978, for review; Bjornstad, 1980; Waitt, 1980, 1985, 1987; Spencer, 1989; O'Connor and Waitt, 1994.) For a comparison of the Touchet Beds in the Little Grand Canyon and those exposed in the lower Touchet valley, see the description for Stop 14.

Leave Lowden. The Blue Mountains anticline rises above the Walla Walla Valley to the east.

The road to the Whitman Mission National Historical Site enters from the right (south). You can see a monument about a kilometer to the southeast on the west end of a ridge forming the drainage divide between Mill Creek and the Walla Walla River. The ridge is composed of Touchet Beds. The Whitmans settled on the flood plain between the creek and the river. A nearby highway sign reads:

"A short distance to the south near the Walla Walla River is Waiilatpu, the place of the people of the rye grass, a mission founded among the Cayuse Indians of the Walla Walla Valley in 1836 by Dr. Marcus Whitman and his wife Narcissa. As increasing numbers of immigrants moved into the Oregon country during the 1840s, Whitman Mission became an important station on the Oregon Trail. Cultural differences climaxed by a measles epidemic that killed many Cayuse ended the missionary effort. A few suspicious Cayuse took the lives of Marcus and Narcissa Whitman and eleven others on November 29, 1847."

The road to College Place enters from the right (southeast). When the highway department improved this intersection in the summer of 1994, new vertical cuts were made in the Touchet Beds and loess on the left (north) side of U.S. Highway 12. Notice all the holes in the roadcuts; within days after the new cuts were made, they became homes for thousands of bank swallows.

Enter Walla Walla; continue east on U.S. Highway 12. Walla Walla lies on the flood plain of Mill Creek. Before the Army Corps of Engineers made a three-pronged attack on Mill Creek there was occasional severe flooding in Walla Walla, particularly in 1906 and 1931.

That attack includes construction of Bennington Lake, channelization of Mill Creek through Walla Walla, and small dams to divert water from Mill Creek into Garrison and Yellowhawk Creeks that run south of the city. From Bennington Lake, potentially one of the largest off-stream flood storage projects in the state (Mills, 1989), water can drain into Russell Creek or back into Mill Creek. The earth-rock dam for Mill Creek Reservoir has never been able to hold much water without leaking. Much of the way through Walla Walla, Mill Creek's channel has concrete sides and bottom. The creek runs beneath many downtown buildings. Garrison, Yellowhawk, and Russell Creeks flow into the Walla Walla River. Floodwaters diverted into these creeks would not re-enter Mill Creek.

Take Rees Avenue exit from U.S. Highway 12 and proceed southeasterly.

Veer southerly on Park Street.

The Whitman College parking lot is to the left (east). This marks the end of field trip.