To Bretz, the source of the floodwater, although important, was secondary. He had walked over, mapped, and described in great detail the geological products of the flood, and he knew it had happened. But to his critics, more removed from the actual evidence, the source of the floodwaters was crucial. Coupled with their predilection for the slow, stately progress of geological processes, it was a potent reason to resist a catastrophic explanation. One by one they gave their rebuttals. The record of the meeting contains five lengthy statements from prominent geologists, all of them arguing that the features Bretz described did not require the type of flooding he postulated but could actually be the product of “normal” geological processes. One says: “Professor Bretz frankly points out the difficulties in applying his explanation of the origin of the remarkable features of the Columbia Plateau. It is not easy for one, like myself, who has never examined this plateau to supply offhand an alternative explanation of the phenomena . . . but I am left with the feeling that some things essential to the true explanation of the phenomena have not yet been found.” Another critic comments about Bretz’s estimate of the amount of water involved, concluding “criteria used to determine the actual quantities of water involved appear somewhat questionable.” He goes on to say that Bretz’s mechanism for producing this water by a subglacial volcanic eruption is “wholly inadequate.” Yet another, discussing the deep channels in basalt that Bretz had described, comments: “Basalt is a hard rock and very resistant to corraision [a term for erosion by water that is carrying rock particles] . . . I am not convinced that so much work could be done on basalt in so short a time, even by such a flood as is postulated.” Finally, another geologist who had actually seen Grand Coulee, one of the most spectacular of the great dry canyons of the Scablands, concludes: “The dry falls in the Grand Coulee resemble Niagara Falls and are evidently the product of normal stream work. The deep gorge of the coulee below the dry falls was apparently excavated by the same orderly and long-continued process . . . as the gorge below Niagara Falls, and it could hardly have been produced in a short time by a flood of whatever magnitude.”
The published record of the meeting does not indicate whether anyone publicly supported Bretz’s theory. Reading the comments of his opponents, however, it is evident that even in opposition, most of them were troubled by the sheer uniqueness of the Scabland features. Virtually all of them suggested that a lot more work in the region was required before a final pronouncement could be made about the processes that had molded the Scablands.
By the early 1930s, Bretz had effectively completed all of the fieldwork he would do on the Scablands problem. One very interesting and highly important piece of information came to light late in his studies, however. Examining the valleys and streams on the eastern and southern margins of the Scablands, Bretz found that all of them, including the largest, the Snake River, contained evidence for a surge of deep water flowing upstream, to the east, away from the Scablands. This reverse flood had left distinctive sediments along the valleys; in the case of the Snake River, Bretz traced these signs for almost one hundred and fifty kilometers upstream. The only reasonable explanation was that the water level in the Scablands had risen so fast and so high that it simply surged up all of the surrounding low-lying valleys. The force and magnitude of the flood had pushed water uphill, like a wave on a beach. It was further potent evidence for Bretz’s catastrophic flooding theory. But still he was a lonely crusader. Throughout the 1930s, many field trips were organized to the Scablands, and other geologists made their own independent investigations. They floated alternative hypotheses: perhaps most of these features were the direct product of glacial ice flowing over the region (here they were ignoring the well-documented evidence that the ice sheets had terminated to the north of the Columbia Plateau); or perhaps ice had blocked up various key parts of the Snake-Columbia River drainage system, so that some of the region had been submerged under ponded water. If this were the case, rising water might occasionally have spilled over high points of land to create some of the Scabland features. Or perhaps nothing special was needed at all. There were still those who argued that the Scabland topography could be the result of normal stream erosion and sediment deposition. The debate continued quite fiercely during the decade, but Bretz never wavered in his conviction that a catastrophic flood had occurred—even though there was still no adequate explanation for the source of the water.
The turning point for the flood hypothesis came in 1940, in an unexpected way. That year, many of the principal players in the Scablands debate attended a meeting of the American Association for the Advancement of Science (AAAS) in Seattle, where there was a session on the glacial geology of the Pacific Northwest. One of the speakers, Joe Pardee, was a geologist with the U.S. Geological Survey who had worked in the Northwest for years. As early as 1910, Pardee had presented evidence of a huge, glacier-dammed lake in western Montana at the end of the last glacial episode—glacial Lake Missoula, named for the Montana city. The hills above Missoula are marked with multiple rings, easily visible today, that record former shorelines of the glacial lake. Some science historians have concluded that Pardee had realized decades before the Seattle meeting that Lake Missoula could have been the source of water for the Scablands flooding, and that he might even have told Bretz this. Surviving correspondence between the two men suggests that they may have discussed the idea, but if they did, it was never published. At the 1940 AAAS meeting, Pardee’s presentation was about ripple marks in the sediments of the former glacial Lake Missoula. At first glance, that doesn’t seem very catastrophic or exciting. Ripple marks are common enough features that most people have seen along a river’s edge or in the sand of a beach. They are made by flowing water, and usually they are a few centimeters wide and about the same distance apart. If you walk over them in bare feet, it feels as though you are getting a foot massage.
In the title for his talk, Pardee put a question mark after the words “ripple marks.” What he described were definitely not the small ripples on a beach; they were long, rolling ridges up to fifteen meters high and a hundred and fifty meters apart. Pardee commented that calling them ripple marks was really inappropriate, but he couldn’t think of a better term. They occur in valleys near the western margin of the area that had been covered by glacial Lake Missoula, close to the border between Montana and Idaho and directly to the east of the Channeled Scablands. Pardee concluded that they must have been formed by the sudden release of a huge volume of water from the dammed glacial lake. He noted that it would have contained at least 1,700 cubic kilometers of water at its highest levels. Although Pardee didn’t say so explicitly, it was quite apparent that a sudden release of Lake Missoula’s water might have been the source of Bretz’s catastrophic flooding.
Glacial Lake Missoula formed in the mountainous region of Montana now occupied by the Clarke River, which flows across the state from southeast to northwest. Glaciers of the Pleistocene Ice Age had blocked the normal drainage, and all of the valleys and low-lying areas over a vast region simply filled up with water, creating a lake with a complex, crenulated shoreline (figure 12). At its highest levels, the water crept up every little tributary valley in the mountains. It was about 650 meters deep at its deepest points.
The pressure of all that water was tremendous in the narrow valley where the ice blocked the natural drainage to the northwest. As the climate warmed toward the end of the last glacial episode, melting ice added more and more water to the lake and the pressure on the dammed outlet increased relentlessly. One can imagine that one particularly hot summer, a trickle of water forced its way through the rotting ice dam. The trickle quickly became a stream, the stream a torrent. Suddenly, the ice dam gave way altogether, with a roar that must have shaken the countryside for hundreds of kilometers around. Lake Missoula poured out of its valleys and across the Columbia Plateau toward the Columbia River and the Pacific Ocean. Fish swimming in Lake Missoula were floating belly up in the Pacific Ocean a few days later,
if they hadn’t been torn to shreds in transit. Soil, plants, trees, and boulders were caught up in the torrent and carried hundreds of kilomters. Elk, rabbits, and snakes were carried away in the roaring flood, which swept along everything in its path. Pardee calculated that the maximum outflow would have been about 31.6 cubic kilometers of water per hour. He noted that the Mississippi River, at the peak of its February 1937 flood, had discharged 0.16 cubic kilometers every hour. The Lake Missoula flow was nearly two hundred times as great. More recent calculations suggest that Pardee’s estimate was too low, and that the actual peak flow may have been twice as high.
Figure 12.The map shows the location of glacial Lake Missoula as the Pleistocene glaciers retreated into Canada. A lobe of ice blocked the northwestern drainage of the lake; when it gave way a large volume of water flooded westward onto the Columbia Plateau to create the Channeled Scablands. At the Wallula Gap, the water reached another bottleneck and ponded to great depths. Lake Bonneville, which also breached a (rocky) dam to flood the Snake and Columbia Rivers, is also shown.
Pardee’s observations removed the major objection to Bretz’s flood hypothesis: the absence of a source that could supply a very large volume of water in a short time. The giant ripple marks implied just such a source, and the new evidence began gradually to win converts to Bretz’s view. Still, some of his most vocal opponents took a very long time to come around. One, Richard Foster Flint, a widely acknowledged expert in glacial geology who had meticulously developed an alternative noncatastrophic theory for the Scablands landscape, found it particularly difficult to discard his own hypothesis. He had written the comprehensive textbook on glacial geology, but it was not until 1971, in his book’s third edition, that he acknowledged the flood origin of the Channeled Scablands. Even then, he could not bring himself to be effusive about the unique character of the region. In a book of more than eight hundred pages, he devoted one dry paragraph to a discussion of the Grand Coulee and of Scabland features “widely created east of the Grand Coulee by overflow of an ice-margin lake upstream.” He makes cursory reference to two of Bretz’s publications.
Bretz lived to a ripe old age (figure 13). He returned to the Scablands one last time a few months before his seventieth birthday. By that time, there was abundant new information about the area, gathered through surveys by the U.S. Bureau of Reclamation, which had initiated a major irrigation project there. Bretz had access to Bureau of Reclamation excavations, aerial photographs, and new maps. One of the features revealed by the excavations was a complex pattern of layering in many of the Scabland sediment deposits, which suggested that there might have been many floods, not just one. That makes sense. The end of a glacial period is not abrupt; a few decades of cold weather and the glaciers would have advanced enough to again dam the exit from Lake Missoula, only to crumble as the warming continued, releasing yet another flood. Exactly how many occurred is still uncertain—some who have looked carefully at the evidence believe there may have been dozens, of varying size, over the several thousand years of the glaciers’ demise.
At one place along the Columbia River’s course, after it has been joined by the Snake River in southern Washington, there is a particularly narrow section called the Wallula Gap. It is far enough downstream in the Columbia River drainage system for all of the water from each of the Lake Missoula floods to have had to flow through this gorge—it is the only gateway to the west and the Pacific. Bretz found evidence that water had “ponded” behind this gap, backed up like a traffic jam at a narrow bridge. Initially, his critics were incredulous, because even at its highest levels, the present-day Columbia flows easily through the narrow valley. However, Bretz found signs that the flood- waters had risen to heights of at least 300 meters above today’s valley floor and had backed up to flood the valleys of tributary streams such as the Yakima River for tens of kilometers. Wherever it ponded, the water, no longer traveling swiftly enough to hold its heavy load of clay and sand in suspension, began to deposit sediment. Each Lake Missoula flood added its contribution; today, in some low-lying regions of southern Washington, there is layer after layer of such sediment. In an otherwise harsh landscape, fertile, productive soil has developed on these patches of mineral-rich sediment, leaving an unexpected legacy of catastrophic ice age flooding: the burgeoning vineyards of southern Washington.
In 1952, in addition to finding evidence for multiple flooding, Bretz gained some startling new insights into the workings of the floods as he studied the Bureau of Reclamation aerial photos. With the bird’s eye perspective they offered, he found clusters of giant ripple marks in many parts of the Scablands, just like the ones Pardee had described near the exit from Lake Missoula. Although he had crisscrossed most of these areas on foot during his early fieldwork, Bretz had never noticed them. On the ground, pushing his way through sagebrush and a bit like the proverbial flea on a camel’s back, his close-up view had given him no clue to the larger picture. The significance of the regular rise and fall of the land had not sunk in. The scale of the Scabland ripple marks, like Pardee’s Lake Missoula examples, requires deep and fast-flowing currents. Bretz added yet another piece of evidence to his list supporting catastrophic flooding.
Figure 13.J Harlan Bretz at age 95, at home near Chicago. Two years later, in recognition of his work on the Channeled Scablands, he was awarded a prestigious medal by the Geological Society of America. “All my enemies are dead,” he said, “I have no one to gloat over.” Photograph courtesy of the Special Collection Research Center, The University of Chicago Library.
More recent research has shown that most of Bretz’s early conclusions about the processes that formed the Scablands features were essentially correct. Residual doubts about the ability of water to very rapidly generate erosional features on the scale of those seen in the Scablands have been dispelled by engineering studies. Turbulent, high-volume flows have tremendous eroding power, especially when they are heavily loaded with rock particles. Vortices form in the roiling water that can act like a sandblaster, drilling into solid rock, and under some conditions, shock waves produced by breaking bubbles in turbulent flows can generate such large local pressures that they shatter almost any material. Even features as enormous as the Grand Coulee can be formed rapidly given enough water. With such knowledge, Bretz’s work was entirely vindicated. Better late than never, the geological community recognized his contributions by awarding him the Geological Society of America’s highest honor, the Penrose Medal, in 1979. He was 97. His one regret, he is reported to have confided to his son, was: “All my enemies are dead, so I have no one to gloat over.”
There was a precedent for Bretz’s flood hypothesis that was rarely mentioned during the debate. It involved another ice age lake in the western United States, albeit not a glacial lake. In 1890, G.K. Gilbert, a geologist with the U.S. Geological Survey, published a monograph on “Pleistocene Lake Bonneville,” one of many large lakes that formed in low-lying regions of the west during times when the climate in that region was wetter than it is at present. Lake Bonneville occupied a large tract of land in northwestern Utah; the Great Salt Lake is but a small remnant of it. Gilbert’s work describes features to the north of the former lake that appeared to him to be the result of flooding caused by a sudden release of water.
Lake Bonneville was too far south to have been dammed by ice, or to be supplied directly by meltwater from the major Pleistocene glaciers. But it filled and apparently emptied catastrophically toward the end of the most recent glacial episode, in the same time frame as the Lake Missoula floods. The northern exit from Lake Bonneville was blocked by loosely consolidated rocky rubble that filled in a low-lying area between hills. Just as the high water stand of Lake Missoula put enormous pressure on the ice dam that blocked its outflow, so a rising Lake Bonneville pushed mightily against its barrier of rubble. When it broke through, it rapidly cut away the loose material. To the north lay the Snake River plain. The waters of Lake Bonneville, reaching depths more than a hundred meters
above the present valley floors, rushed northward into the Snake River and eventually, like the Scabland floods, emptied into the Columbia River and the Pacific Ocean.
There appears to have been only one major release of water from Lake Bonneville—once the barrier to its northern exit was eroded away, the lake could no longer fill up to such high levels. Estimates of the peak water discharge during the flood vary, but most place it at one-tenth of the maximum Lake Missoula rate, or less—still a very large flow. Potholes, channels, and flood deposits similar to those of the Channeled Scablands can be found for more than a thousand kilometers along the path of the flood. Gilbert’s investigation focused on Lake Bonneville and its vast extent, rather than on the effects of the flood. His fieldwork was excellent and his writing clear, and there was little disagreement about his conclusions. Perhaps surprisingly, his descriptions of the features to the north of the ancient lake, and his interpretation that they were flood-related, did not attract the controversy that Bretz’s work was to generate several decades later. Gilbert, like Bretz, realized that the landscape he was studying had been formed in a catastrophic process. That more traditional geologists did not immediately denounce this interpretation was probably because the scale of the erosional features, while significant, was much smaller than that of the Scablands, and because the source of the flooding was so obvious in the case of Lake Bonneville. G.K. Gilbert died in 1918. Had he been present for the debate about the Channeled Scablands, he would undoubtedly have been on Bretz’s side.
A wonderful thing about science is its interconnectedness, and the fact that one discovery invariably leads to another, often quite unexpectedly. As the concept of catastrophic glacial flooding gradually gained acceptance, it became reasonable to ask other questions. Is there evidence for similar glacial floods elsewhere? Would the rapid addition of all that water to the oceans have had any important consequences? What about the sediment that was swept along by the floods? Could it be traced on the ocean floor? In the past few decades, all of these questions have been answered in the affirmative.
Frozen Earth: The Once and Future Story of Ice Ages Page 13