Geological Survey Professional Paper 669 The Colorado River Region and John Wesley Powell

A principal conclusion of this study is that Powell was in large part right in his interpretation of the history of the Colorado River—at least, he was less wrong than his later critics implied. Powell recognized the intimate dependence of river history on structural history. He believed that the canyons were antecedent. Their most striking feature—great depth—does indeed seem best explained by his hypothesis. Powell's interpretation needs to be modified, however, because the locations of the canyons, at least of most of them, seem best explained by superposition from an alluvial cover or from erosion surfaces in the shale formations.

In crossing the Colorado Plateau, the Colorado River crosses about 20,000 feet of strata inclined northeastward against the river's course. About half this tilting took place before Cenozoic time and antedates the formation of the Colorado River system. Half the tilting occurred during the Cenozoic while the river system was developing. Throughout their courses, the major streams go from one structural basin to another and must cross the structural barriers between the basins.

The gradients of the streams steepen greatly where they cross the structural barriers. In other drainage basins, such steepened gradients, called knickpoints, have been attributed to lowered base level and retreat of the knickpoint upstream, or to the streambed being held up by resistant rocks. In the Colorado River system, the steepened gradients coincide with the structural barriers, some of which do not involve differences in rock resistance, and these steepened gradients seem best explained by arching of the streambeds. One example is the steep gradient of Cataract Canyon down the east flank of the Henry Mountains structural basin compared with the much flatter gradients of the Green and Colorado Rivers flowing against the dip of the rocks above the junction of the rivers. Other examples are where the Dolores River crosses the Dolores anticline and where the Colorado River crosses the Waterpocket fold. At none of these places is there much difference in rock resistance along the riverbed.

The amount of deformation during Tertiary and Quaternary time in the Colorado River basin seems to have been roughly proportional to the time involved. Geologists profess not to be catastrophists, yet geological literature is replete with "great. upheavals" (technically called orogenies). The geologic history of the Colorado River raises grave doubts whether there was any period of time as long as the Quaternary (2-3 million years) without major earth movement somewhere in the river basin.

My hypotheses for explaining various stretches of the rivers rest heavily on late Cenozoic structural movements, some of them demonstrable, some assumed. The most striking examples of the effect of earth movements on river history are the late Tertiary drainage changes along the main stem of the Colorado River in the Rocky Mountains, the abandonment and uplift of Unaweep Canyon, the abandonment and uplift of the dry canyon at Peach Springs, and the disrupted drainage that coincides with known present-day earth movements west of the lower stretch of the Colorado River. Although many of the structural features began forming in early Tertiary time, in the so-called Laramide orogeny, much of the deformation continued into late Tertiary and even Quaternary time; there does not seem to have been any end to that orogeny.

Evidence of late Cenozoic earth movements is not always present at particular stretches of the river where such movement has been assumed. One of my major assumptions concerns dating the mile or so of uplift of the 1,000-mile-wide structural arch of which the Rocky Mountains and northern part of the Colorado Plateau are part. This regional uplift probably began in Oligocene time after the early Tertiary lakes had been filled with sediment, because middle Tertiary deposits are restricted to structural basins in the Rocky Mountains and Basin and Range province; they are generally lacking on the plateau. The southern part of the plateau has been raised considerably higher than the northern part (fig. 62). In the absence of evidence for a sudden jarring uplift, I assume the Rocky Mountains and Colorado Plateau were raised gradually during the last 35 million years, a rate of uplift of 150 feet per million years, less than 6 inches since the time of Christ. Continued activity on the Hurricane fault and the results of precise leveling in the Lake Mead area can be interpreted to suggest that uplift is still continuing (p. 117). The evidence for late Cenozoic earth movement that is available at some of the local structures has been emphasized because this evidence has been omitted in previous hypotheses, yet these late earth movements can account for most of the apparent conflicts in earlier hypotheses.

Although the streams are inconsequent across the structural basins and barriers, they are very well adjusted to the structural domes at the laccolithic mountains and detour around them. (See Hunt, 1956, p. 82.) The stream courses apparently antedate the laccolithic intrusions. This fact suggests either that the river system was well developed by early Miocene time or that some of the laccolithic mountains are younger than the single age determination (25 million years) made at the La Sal Mountains.

Meanders are deeply incised along the rivers on the Colorado Plateau, and the meander belt changed very little while the canyons were being deepened the last 1,000 feet. Cutoff meanders are scarce. Indeed the greatest incidence of cutoff meanders in the Canyonlands is on the west flank of the Monument upwarp. In addition to the two examples cited along the Colorado River, several in White Canyon produced the natural bridges there. These cutoff meanders could be attributed to renewed uplift at the Monument upwarp which caused the meanders to migrate westward while the canyons were being deepened.

Amounts of erosion seem to have been roughly proportional to the amount of time involved. Some evidence for this is available along the west edge of the Colorado Plateau where cliff retreat has been correlated with faulting (Averitt, 1964a, b). At times, the faulting was accelerated, but between these episodes the maximum amount of cliff retreat was only a few miles. The faulting progressed in repeated small increments without any "great upheaval," and erosion progressed at about the same rate as the deformation. In the river basin as a whole there is no evidence for any long period of crustal stability during the Cenozoic, and there does not seem to have been any "great denudation."

There are, of course, evident changes in kind and rate of weathering and erosion attributable to short-range climatic change, such as occurred during the Quaternary. During periods as brief as this (2-3 million years), episodes of accelerated weathering and erosion have alternated with episodes when rates were slowed, but total erosion during the Quaternary seems to have been roughly equal to the erosion that took place during comparable intervals of time during the Tertiary.

Present erosion above the Grand Canyon, as indicated by the sediment load of the Colorado River, averages about 6.5 inches per 1,000 years (Ritter, 1967). My reconstruction of the history of the river system suggests that rivers in the southern part of the Colorado Plateau began draining to the Basin and Range province before the end of the Oligocene (that is, more than 25 million years ago) and that the whole drainage system has formed since latest Miocene time (about 12 million years ago). At present rates of erosion, the Rocky Mountains and Colorado Plateau would have been lowered, on the average, about a mile in the last 10 million years. Perhaps the erosion averaged no more than that in the preceding 20 million years when most of the drainage was ponded at various places in the north. The amounts are about the right order of magnitude. Even the amount of canyon deepening attributable to the Quaternary, averaging about 500 feet, seems about right, considering the time involved.

This erosion has been differential. The shales, especially the Cretaceous shales, erode readily and must make up by far the greatest proportion of the present sediment in the rivers. Even within the shale formations the erosion is differential, being greatest on bad-land slopes and along arroyos and least on pediments, which are chiefly surfaces of transportation. The occurrence of Cretaceous shale at all the laccolithic mountains implies that when those intrusions formed, presumably in early Miocene time, the Cretaceous formations were essentially continuous across the Colorado Plateau north of the Black Mesa and San Juan basins. Superposition of the river system across the structural barriers may have been from erosion surfaces in the Cretaceous formations. The Colorado Plateau was lower then, but erosion may have been rapid because of the great extent of the easily eroded shales.

Another effect of the differential erosion is to increase the topographic relief at the laccolithic mountains. The mountains are formed of resistant rocks, and the surrounding terrain erodes more rapidly than they do. As a result, these mountains are becoming higher and more rugged as erosion progresses; in effect, as they become older they appear younger.

As a body of water, the Colorado River is small. Its flow is only 5-10 percent of that of other great rivers in the United States, such as the Columbia, Snake, Missouri, Ohio, and St. Lawrence. But the Colorado River is a major geographic force in the American Southwest. The river crosses the arid lands that need its water, some of the arid lands about which Powell concerned himself. Because of its physiographic setting and geologic history, the Colorado River basin is spectacularly scenic. Nearly half our national parks and monuments are within the basin. Even after 100 years, however, the explanation of this landscape still defies us.