THE CARVING OF THE SCABLANDS
As noted earlier the lava field is shaped somewhat like a giant saucer, tilted to the southwest. Slopes (gradients) down the tilted field are as much as 25 to 35 feet per mile (for comparison, the Spokane River between Coeur d'Alene and Spokane has an average gradient of only 9 feet per mile). When the flood reached the lava field and started down this sloping surface, the enormous volume, velocity, and turbulence of the water provided the erosional energy required to sweep away the loess and expose the jointed basalt underneath. The currents were so turbulent and so powerful that they were able to pluck out and transport blocks of basalt, some measuring more than thirty feet across. Deep canyons were eroded into the basalt and, where cascades developed, plunge pools and cataracts formed. Some of these depressions are more than 200 feet deep. Many of the lake basins in the northeastern part of the area were formed in this manner. In other places, great amounts of stream-borne rock debris were dumped to form immense gravel bars and deltas.
Three giant rivers raced across the lava field (Fig. 13, page 12). The easternmost stream, 20 miles wide in places and locally 600 feet deep, carved the widest channel, the Cheney-Palouse Tract. A middle river carved the Crab Creek Channel and its tributaries, a tract about 14 miles wide crossed near its head by U. S. Highway 2 between Davenport and Creston. The third and western most river, which may have carried the greatest volume of water, carved Grand Coulee.
These three major torrents, together with dozens of smaller streams, flowed simultaneously, and at times criss-crossed their channels until the 500 cubic miles of water stored in Glacial Lake Missoula had drained away.
At Sprague, in part of the Cheney-Palouse Tract, flood waters swirled along a channel eight miles wide and more than 200 feet deep? Here it stripped away 100 to 150 feet of Palouse soil and parts of jointed lava flows, leaving a wild rough wasteland in its wake. This jumble of mesas and depressions is so irregular that an area of more than 70 square miles contains no throughgoing streams.
The sides of the scabland channels in the eastern part of the lava field are marked in many places by steep slopes cut in the Palouse soil revealing in spectacular fashion the depth of the soil. South of Ewan, Washington, for example, eroded scarps are nearly 200 feet high.
The largest of the scabland channels is Grand Coulee, a two-stage canyon 50 miles long and as much as 900 feet deep. Its ancestral channel was cut by the Columbia River when it was forced to flow across the lava field by the Okanogan ice lobe. When flood waters cascaded down the steep southeast-facing slope of the monocline, this fold was eroded upstream across the zone of broken rocks and into the higher, flat-lying basalt, forming a typical recessional gorge now called Upper Grand Coulee (Fig. 18). Erosion and cataract retreat continued in Upper Grand Coulee until the basalt flows at the head of the Coulee were completely removed. This exposed the granitic rocks that form the Coulee floor and the foundations for Coulee Dam today. The granite is tough and hard, but most importantly, it is not jointed like the basalt, and therefore could not be plucked and removed by the flood. But erosion did lower the divide between the Columbia and the Coulee to a point where the falls essentially destroyed itself.
Lower Grand Coulee, cut along the foot of the monocline, also developed a series of falls and plunge pools as erosion progressed upstream. Dry Falls, a 350-foot high, 3-mile-wide group of scalloped cliffs shows today the position of the falls at the end of the flood.
South of Grand Coulee the water from the Coulee channel was joined by the great river that boiled down Crab Creek. Together they surged into Quincy Basin, spread, slowed, and dumped a great blanket of boulders, sand, and gravelmaterial mostly derived from flood erosion upstream. This flood debris now covers an area of about 500 square miles to a depth of at least 125 feet. As Quincy Basin filled, some of the water overflowed through two large cataracts into the Columbia River Valley at the Potholes and Frenchman Springs dischargeways (Fig. 13, page 12). But most of the water escaped to the south around the east end of Frenchman Hills.
At the east end of Frenchman Hills the flood crossed a 50 square mile area covered by a layer of weak sedimentary rocks called the Ringold Formation which was deposited during early Pleistocene time (2 to 3 million years ago). This part of the Scablands, called the Drumheller Channels, is perhaps the most spectacularly eroded area of its size in the region (Fig. 13, page 12). The erosional features can be seen especially well from the road built on the crest of O'Sullivan Dam, along the north edge of the Channels (Wash. State Highway 11G). South of the dam is a wild jumble of cliffs, depressions, ponds, and remnants of lava flows. This channel area has a local relief of 150-200 feet.
South of Frenchman Hills another branch of the flood flowed eastward in Lower Crab Creek Valley along Saddle Mountain, and near here this branch was joined by more flood waters from Washtucna and Lind Coulees. These surging torrents cut the Othello Channels at the east end of Saddle Mountain (Fig. 13, page 12).
Like roads to Rome, all of the scabland rivers led to Pasco Basin, a down-warped area of more than 500 square miles at the southwest corner of the lava field. The southwestern edge of Pasco Basin is marked by the Horse Heaven Hills, which was an effective barrier to the escape of water except for a single outletWallula Gap. Because all flood waters from the lava field had to pass through this narrow squeezeway, a huge lake formed in Pasco Basin and extended more than 100 miles upstream in the valley of the Snake River. At the junction of the Snake and Clearwater Rivers at Lewiston, Idaho, this impounded water was almost 600 feet deep (Fig. 13, page 12).
Leaving the Gap the flood waters flowed westward through the Columbia River Gorge and on past the Cascade Mountains to the Portland, Oregon area. There the great stream created a large delta and continued on up the Willamette Valley to form a short-lived lake nearly 400 feet deep. As this lake gradually receded, huge icebergs loaded with rock debris were stranded along its shoreline. When the ice melted, the debris, containing boulders as much as seven feet across, was left behind as mute evidence of the immense size and power of the flood.
Did early man witness the fearful destruction of the land by the raging torrent? No one knows for sure, because the earliest known evidence of man in the region has been dated at about 10,130 years B.P. (before the present), or a few thousand years after the flood as determined in the Carbon-14 laboratory of the U. S. Geological Survey from charcoal collected at the "Marmes Man" site in southeastern Washington (Fig. 13, page 12). This prehistoric rock shelter, carved along the base of a basalt cliff by the Palouse River, is believed to be among the oldest known inhabited sites in the United States. Thus, at the present time, definite proof is lacking that man was living in the Pacific Northwest at the time of the flood.
The duration of the flood is not known, but a reasonable guess is that at any one point on the lava field it consisted of a series of surges, and that the crest was short-livedperhaps lasting only a day or two. At Wallula Gap, where the maximum flow was calculated at nearly 40 cubic miles per day, the main part of the flood probably lasted 2 to 3 weeks.
The maximum rate of flow at the site of the ice dam has been estimated at 9.5 cubic miles of water per hour. If this rate were sustained, Glacial Lake Missoula would have drained in little more than two days. The maximum rate of flow at Wallula Gap has been estimated at 39.5 cubic miles per day, which would drain the lava field in about two weeks. But these maximum rates of flow probably could not have been maintained for that long. A realistic estimate of the duration of the flood, from the time when water first spilled over the ice dam to the time when scabland streams returned to normal flow, was probably about a month.
The unique combination of geologic events, beginning with a vast series of lava flows, then regional tilting of the land, followed by deposition of a one hundred-to-two-hundred-foot layer of wind blown silt, and ending with a glacial lake dammed and suddenly released to form earth's greatest known flood, involved such a large area that only parts of the scabland picture can be seen at one time. Now, as a result of the detailed studies of many geologists working in the area for more than fifty years, the many fragments of evidence have been pieced together to support Bretz' concept of the Great Spokane Flood.
(from material supplied by Paul L. Weis and William L. Newman)
Last Updated: 28-Mar-2006