Geological Survey Professional Paper 1356 Neogene Tectonics and Geomorphology of the Eastern Uinta Mountains in Utah, Colorado, and Wyoming

GENERAL FEATURES AND DISTRIBUTION

The Bishop Conglomerate was named by Powell (1876, p. 44, 169) for its occurrence on Bishop Mountain (now called Pine Mountain), just north of the Utah-Colorado State line in southern Wyoming (fig. 2), where it is well exposed. Exposures are even better a few miles to the northwest, along the southeast side of Little Mountain and on Black Mountain, west of Flaming Gorge Reservoir in southern Wyoming. Powell's original name, "Bishop Mountain Conglomerate," was shortened by later geologists. The same formation was called "Wyoming Conglomerate" by the geologists of the 40th Parallel Survey (King, 1878, atlas, maps 1 and 2). Its distribution along the north flank of the Uinta Mountains is about as shown by Bradley (1936, pl. 34). Its distribution in the Eastern Uinta Mountains shown on figure 13 is generalized from the detailed map of Rowley and others (1979). Its distribution and character in the Western Uinta Mountains have been under study recently by Bruce Bryant of the U.S. Geological Survey. Remnants of the Bishop Conglomerate are more extensive on the south flank of the mountains than on the north but are widespread on both (figs. 5, 13). In the higher, western part of the range, the distribution is somewhat obscured locally by glacial deposits.

Although the Bishop Conglomerate is very widespread on the flanks of the Uinta Mountains, its present distal limit is erosional, and it must have once extended considerably farther from the mountains than now. Presumably it reached nearly to the centers of the adjacent basins, where it may have merged with deposits from other rimming mountain ranges, though it must have been much thinner there than closer to the mountains. Both the Duchesne River and White River flow to the Green in courses that are roughly parallel to the Uinta Basin axis (fig. 1)—the Duchesne all the way from its confluence with the Strawberry, which continues the parallelism westward, and the White downstream from its crossing of the Douglas Creek arch. These rivers likely were established in consequent courses on the old basin floor following the disappearance of Eocene Lake Uinta, but their courses now are well south of the basin axis. Perhaps they were displaced southward by alluvial fans building out from the mountains as the Bishop Conglomerate spread southward, or by a basinward equivalent of the Starr Flat Member (of Andersen and Picard, 1972) of the Duchesne River Formation, or by both formations. At this writing (1982) the possibility seems good that the Bishop and the Starr Flat are equivalent, a view shared with me by Bruce Bryant (written commun., 1982).

In its most characteristic form the Bishop Conglomerate consists of rather loosely cemented bouldery, cobbly conglomerate and coarse, poorly sorted, pebbly, friable sandstone. Locally, in deposits that probably originated as debris flows, the clasts are largely matrix supported. Clasts tend to be subangular to subrounded, but some are very well rounded. The firmness of cementation is varied, and surface exposures are mostly loose and gravelly. The conglomeratic character of the rock is most evident on the sides of steep hills and bluffs where the calcareous cement has not been dissolved away and where the rock may form low cliffs and overhanging ledges. On Black Mountain, west of Flaming Gorge Reservoir (fig. 2), resistant conglomerate forms nearly continuous cliffs more than 2.2 km long.

The coarseness of the rock varies laterally and vertically. In most places the upper meter or so is finer grained than the rock at depth—commonly as fine as pebble gravel but with scattered cobbles and boulders. Figure 14 depicts this characteristic very well. Except in slopeside exposures, the formation may appear to be deceptively fine grained. Away from its source, which was the crestal part of the Uinta Mountains, the rock is progressively less coarse, just as one would expect, owing to attrition of the clasts and to reduced competence of the ancient streams. Along a given drainage line the size (volume) of the largest boulders diminishes exponentially with distance. Regionwide it tends to do the same (table 1). The median grain size of all clasts probably does also, but the formidable sampling required for proof has never been attempted. On any exposures, the maximum clast size greatly exceeds the median. Boulders 2 m or more in longest diameter are occasionally found several kilometers from the nearest possible source. On the slopes of Diamond Peak in Colorado (not be confused with Diamond Mountain in Utah), where a possible source was only 3-5 km away, many boulders exceed 1 m across, and 2-m boulders are not uncommon. Weber (1971, p. 168) noted an angular block of red quartzite "over 8 feet long" (>2.4 m) near the top of Diamond Peak. Such a block must have been transported by mass wastage, perhaps by a debris flow.

TABLE 1.—Maximum observed sizes of clasts in the Bishop Conglomerate and distances from nearest possible sources in the Uinta Mountains

¹Probably transported by mudflow.

²The Elk Springs deposit may have had a slightly nearer source at Cross Mountain, possibly 18 km.

³Flat Top Mountain supports no conglomerate—only a scattering of resistant pebbles, mostly cherts and quartzites, including a distinctive red chert common in the Round Valley and Morgan Formations of the Eastern Uintas.

FIGURE 14.—Bishop Conglomerate capping south end of Little Mountain, a high mesa south of Rock Springs, Wyo., elevation 2,788 m. Surface rises toward the south (left) at a rate of 24 m/km. Conglomerate at surface is deceptively fine grained; for comparison with more typical rock, see figure 15, which was taken to left of automobile and just below rim of mesa. Uinta Mountains in background have been lowered by post-Bishop subsidence.

Rich (1910, p. 610) observed that boulders <1.5 m in diameter were common at the southern end of Miller Mountain 24 km from the nearest possible source, although cobbles and pebbles predominate. I measured one boulder there that was 2 m across. At Lookout Mountain in Sand Wash basin, 19-24 km from the nearest possible source, pebbles and cobbles predominate, although occasional boulders exceed 0.76 m. Thirteen to fourteen kilometers farther distant, just south of Powder Wash gas field, a gravelly bench that is geomorphically continuous with Lookout Mountain contains very few boulders larger than 46 cm, and most of the clasts are less than 20 cm across. This deposit is about 32 km from the nearest possible source. At Elk Springs Ridge, near the east end of the Uintas, boulders 1.2 m in diameter must have travelled 18-19 km. Here, also, pebbles and cobbles predominate. At Wapiti Peak, which is near the east end of Elk Springs Ridge, Dyni (1980) noted boulders as much as 1.5 m across, 14 km from a nearest possible source. At the shoulder of Aspen Mountain, south of Rock Springs and 48 km from the nearest possible source, pebbles predominate, and the coarsest material consists of small cobbles (about 8-20 cm). Size gradations are similar on Cedar Mountain, west of the Green River in Wyoming (Bradley, 1936, p. 172).

On the south flank of the range, from Ashley Creek eastward to Little Brush Creek, occasional groups of large angular quartzite blocks, "some individuals weighing several tons," were noted by Kinney (1955, p. 115), who attributed their transport to mudflows. Just west of Utah Highway 44, I measured one tabular boulder that was 7.6 m long and saw others that may have been larger. The very poorly sorted, matrix-supported texture of the conglomerate in other places also suggests mass transport. In the Yampa Plateau—Ruple Ridge area of Dinosaur National Monument I have seen 2-m boulders 19 km from their nearest possible source, but boulders this size in some places are clast supported and appear to be fluvial rather than colluvial. Figure 15, for instance, shows three very coarse, inversely graded, clast-supported bedding units passing upward from pebble conglomerate at the base to boulder conglomerate at the top. Bedding thicknesses are about 1—2-1/2 m. Upward coarsening of flood deposits has been noted by previous workers. In a tractive load under high-energy transport, the coarsest particles tend to rise toward the top (Bagnold, 1954; Scott and Gravlee, 1968). This bedding habit, however, may also result from torrential flooding, rising to a gradual peak of intensity, then declining rapidly. If normally graded sedimentation accompanied the decline, as seems likely, its deposits may have been stripped by the next succeeding flood, inasmuch as the tops of the bedding units show evidence of scour.

FIGURE 15.—Bishop Conglomerate at the southwest rim of Little Mountain, Sweetwater County, Wyo. Three coarse, inversely graded bedding units grade from pebble conglomerate at the base to boulder conglomerate at the top. Clasts are subrounded to well rounded. The hammer is about 33 cm long.

Coarsening upward on a sequential scale, rather than a bedding scale, is well exposed at Sand Wash (fig. 13), where the Bishop rests unconformably on the Bridger Formation (fig. 16). Several bedding units are involved in a gradual upward coarsening over a stratigraphic distance of 20 m or more, from pebbly silty sandstone at the base to boulder conglomerate at the top. Similar, but cyclothemic, gradations elsewhere (Steel and others, 1977, for example) have been attributed to tectonic activity. Increased aridity, accompanied by more intense but less frequent flooding, might have achieved the same end at Sand Wash.

FIGURE 16.—A coarsening-upward sequence of beds at the base of the Bishop Conglomerate at Sand Wash, Moffat County, Colo. Dipping to the right, the Bishop rests unconformably on the Bridger Formation. About 30 m of section is visible in this view, but the upper part is greatly foreshortened by perspective. Boulders near the top of the picture are as much as 1.2 m across.

LITHOLOGY AND THICKNESS

Local sources of the Bishop Conglomerate are apparent in the varied lithology of the formation. Red quartzite or sandstone derived from the Uinta Mountain Group is the most abundant clast in general, but in some places gray Paleozoic limestone clasts outnumber those of all other lithologies, including red quartzite. At Cedar Mountain in Wyoming the abundance of limestone cobbles and the paucity of red quartzite suggest that limestone was more widely exposed in the mountains south of there when the Bishop was deposited than it is now, much of it having since been stripped away. Limestone predominates on Black Mountain also. The chief sources were the Madison and Round Valley Limestones, although some limestone may have been reworked from Tertiary conglomerate at the flank of the mountains. Much of the Uinta Mountain Group in the source area at that time must have been concealed. In the Eastern Uinta Mountains also, the Round Valley and the Morgan (Pennsylvanian) were important sources for the Bishop Conglomerate along the south slope of Blue Mountain and in local remnants along Douglas Mountain; these formations yielded distinctive clasts of red chert as well as limestone. Along Wolf Creek, Bear Valley, and Klauson Pasture, limestone cobbles greatly outnumber red quartzite because of the wide extent of the outcropping limestone and the distance to a source of the Uinta Mountain Group. South of Rock Springs on Little, Miller, Pine, and Aspen Mountains, but not on adjacent Diamond Peak, distinctive clasts of white to pale green metaquartzite, amphibolite, and other metamorphic rocks were derived from sources to the south in the Red Creek Quartzite (Early Proterozoic or Archean). The transporting streams simply picked up and redeposited whatever rocks were in their drainage paths, and the more abundant, more resistant rock types selectively became predominant.

The thickest sections of Bishop Conglomerate are on the south flank of the Uinta Mountains. Near the Uinta River at Jefferson Park, Kinney (1955, p. 115) noted a thickness of 244 m consisting mostly of light-tan to red boulders in a matrix of coarse sandstone. Bruce Bryant (oral commun., 1981) suspected that at least part of this section should be assigned to the Starr Flat Member (Oligocene?) (of Andersen and Picard, 1972) of the Duchesne River Formation. Kinney (written commun., 1959) expressed a similar suspicion years before. The Bishop may thus merge downward and basinward in the Uinta Basin with the Duchesne River Formation, although the contact relationship remains unclear, and a sizable hiatus could exist. Bryant (written commun., 1982) recently obtained fission-track dates very close to the age of the Bishop from tuff beds that he considers to be Starr Flat: 30.0, 30.6, and 34.0 m.y.

East from Jefferson Park the Bishop Conglomerate thins rather uniformly, according to Kinney (1955), to about 90 m near the west end of Diamond Mountain Plateau. Near the southeast end of the plateau, however, a magnificently exposed section about 200 m thick is nearly all conglomerate. In other places on Diamond Mountain Plateau, coarse, poorly sorted, light-gray sandstone predominates over conglomerate. The sandstone is very calcareous, and some of it is tuffaceous. Kinney (written commun., 1959) suspected that all these rocks correlate in time with the White River Formation (Oligocene) of the High Plains region and the Wyoming Basin, a correlation since confirmed by radiometric dating. (See "Tuff" section, below.) Tuff beds on Diamond Mountain, the Yampa Plateau, and Blue Mountain include air laid(?) tuff that contains phenocrysts of fresh black biotite and hornblende.

Near Jones Hole, the thickness of the Bishop Conglomerate, including tuff and sandstone, may reach 150 m (Hansen, 1977b). No such thickness exists anywhere farther east on either flank of the Uinta Mountains. At the type locality on Pine Mountain the thickness is 60-75 m (Roehler, 1972a, b). On the Yampa Plateau and on Blue Mountain, thicknesses range from as much as 60 m to as little as a thin skin, depending largely on topographic irregularities on the local substrate and on random post-Bishop erosion (Hansen and Rowley, 1980a, b). At Klauson Pasture at the east end of the Uinta Mountains, the conglomerate is about 60 m thick (Rowley and others, 1979). Bradley (1936, p. 172) noted that, on the north flank of the range, the formation attains its maximum thickness near the Utah-Wyoming State line (generally less than 60 m) and thins irregularly to the north and south. It rises in altitude southward and thins against the bare Gilbert Peak erosion surface. Northward the conglomerate "thins very gradually and before dissection *** presumably extended far out into the basin" (Bradley, 1936, p. 172). Bruce Bryant (written commun., 1983) has observed stratigraphic sections as thick as 150 m just north of the North Flank fault zone in the Blacks Fork drainage, west of where Bradley did most of his work.

SANDSTONE

Much of the upper part of the Bishop Conglomerate as mapped on the south flank of the Uinta Mountains, and locally the lower part, is friable calcareous sandstone, light gray to pale pink, coarse grained, poorly sorted, and poorly bedded. This sandstone resembles somewhat the sandstone in the Browns Park Formation and might in part be stratigraphically equivalent, but my impression is that it is much less well sorted and that its individual grains, which commonly are pink, are generally subangular, whereas most of the sandstone in the Browns Park is fine grained, is moderately well sorted and well bedded, and has well-rounded grains (Luft and Thoen, 1981). The sandstone in the Bishop also generally contains abundant small rock fragments, mica, and scattered dark grains. Individual beds in both formations, however, depart widely from the perceived norm, and they range through a wide spectrum of color, grain size, and sorting. The Bishop contains nothing comparable to the striking eolian sand that is so abundant in the Browns Park Formation in the Maybell—Elk Springs area or to the fresh vitric, crystal-poor, mafic-poor tuffs of the Browns Park Formation in the type area. These tuffs are much younger than any known in the Bishop.

I suspect that deposition continued without much interruption on the south flank of the Uinta Mountains, generally fining upward, while the sands were accumulating above the conglomerate. Locally steep dips in the Pot Creek area suggest eolian deposition. Deposition ended when tilting and faulting changed the hydrologic regime and initiated dissection. At the east end of the range, however, large-scale subsidence produced a structural sag that provided catchment for sediments throughout most of succeeding Browns Park time. In Browns Park itself, several million years elapsed after deposition of the Bishop Conglomerate before the Browns Park Formation began to accumulate, and a great deal of geomorphic history intervened.

TUFF

Tuff was introduced at times into the Bishop Conglomerate, explosive volcanism being widespread during that period in the Western Interior (Axelrod, 1981, fig. 2). Datable tuffs on Diamond Mountain and the Yampa Plateau provide clues to the minimum age of the conglomerate. Southern Utah and central Nevada are possible sources (Rowley and others, 1975, p. B9; Burke and McKee, 1979, p. 183). Biotite and hornblende from a sample collected above the main body of conglomerate by Glen A. Izett and me in the SW1/4 sec. 13, T. 2 S., R. 23 E., have radiometric potassium-argon ages of about 29 m.y. (biotite, 29.50+1.08 m.y.; hornblende, 28.58+0.86 m.y.) as determined by Harald H. Mehnert of the U.S. Geological Survey, using the following constants:

⁴⁰K λ_ε=0.581 X 10^-10/yr λ_β=4.962x 10^-10/yr ⁴⁰K/K=l.167 X 10^-4

The height of this tuff above the base of the conglomerate can be estimated at about 90 m if the dip (3° NNE.) is projected northward to the outcrop from the basal contact at the rim of Diamond Mountain Plateau and if corrections are added for about 30 m of displacement on intervening faults.

This tuff bed and others overlie most of the conglomerate, but they are interbedded with and overlain by loose, coarse-grained pebbly sandstone. Some of the sandstone might be appreciably younger than the dated tuff, therefore, and might be temporally equivalent to some part of the Browns Park Formation. The lower, coarsely conglomeratic part of the formation, however, could be appreciably older than the tuff; its maximum possible age is limited by the subjacent Duchesne River Formation. Emry (1981) has suggested that the vertebrates of the Duchesnean land mammal age are partly Eocene Uintan (Brennen Basin Member of Andersen and Picard, 1972) and partly Oligocene Chadronian (Dry Gulch Creek and Lapoint Members of Andersen and Picard, 1972). The Lapoint Member underlies the Starr Flat, which in turn may be equivalent to the Bishop Conglomerate.

The dated tuff is light gray, compact, firm, and flecked with abundant euhedral biotite, much of which is coarser than 0.5 mm across (Hansen, 1965, table 3). The tuff is further distinguished by abundant crystal fragments of feldspar and quartz aggregated with clay, by abundant euhedral hornblende, and by euhedral pyroxene and magnetite commonly adhering to glass shards. The euhedra are deeply etched, and the shards are partly altered to clay. Zircon forms inclusions in the feldspar. What might be the same tuff bed was sampled by Winkler (1970) and dated by Damon (1970, p. 52) at 26.2+0.7 m.y. (biotite). A tuff near the Harpers Corner road in Dinosaur National Monument is finer grained and rather more altered but is similar in most other respects.

These tuffs resemble the tuffs in the lower part of the Browns Park Formation in the Little Snake River—Maybell area (Izett and others, 1970, p. C151). This resemblance is the chief reason why the Bishop on the south flank of the Uinta Mountains had been correlated with the Browns Park Formation in recent years (Kinney and others, 1959; Hansen and others, 1960; Untermann and Untermann, 1965).

A tuff collected by Winkler on the Yampa Plateau, about 37 m above the base of the Bishop, was dated by Damon (1970, p. 52) at 41.3±0.8 m.y. from a biotite separate. The biotite was brownish black and slightly resinous (Damon, p. 53), which suggests incipient alteration (Mauger, 1977, p. 23). I am inclined to discount this date because of its wide disparity with the ages from the nearby Diamond Mountain localities and because of contradictory geomorphic evidence for the age of the Bishop. This date lies within the age of the saline facies of the Uinta Formation at Duchesne, Utah (Damon, 1970, p. 51; Mauger, 1977, p. 19, 32), and is close to the age of the Green River—Uinta Formation boundary in the western Uinta Basin. A correlation of the Bishop with these rocks is improbable. The Eocene rocks on both flanks of the Uinta Mountains are deeply truncated by the Gilbert Peak erosion surface, including the Wasatch, Green River, and the Bridger Formations and probably the Duchesne River Formation, and an enormous amount of post-Green River erosion took place before the overlying Bishop was deposited.

PHYSICAL GEOGRAPHY AT THE ONSET OF DEPOSITION OF THE BISHOP CONGLOMERATE

When the Eocene lakes disappeared from the basins bordering the Uinta Mountains 45-40 m.y. ago (Mauger, 1977, p. 37), they were supplanted by alluvial plains. A detailed discussion of this complex transition is outside the scope of this report, but a brief summary might help place into perspective the middle Tertiary events that followed. Ryder and others (1976, p. 511) aptly described the extinction of Lake Uinta south of the Uinta Mountains in just two sentences:

Following the deposition of strata containing the richest oil shale of Lake Uinta, the water became hypersaline, and the remaining oil-shale and mud-supported carbonate units are therefore associated with sodium-rich evaporite and Magadi-type chert beds. By late Eocene and early Oligocene time, the last vestiges of Lake Uinta were buried by coarse alluvial sediments (Uinta and Duchesne River Formations) derived mainly from the Uinta uplift.

West of the present Uinta Basin, Lake Uinta persisted locally, as a freshwater lake, into Oligocene time (Weiss, 1982), but as the basin gradually filled with sediment, its rim was overtopped to the south, and drainage was southward out of the basin from that time on (Hunt, 1969, p. 94). The disappearance of the lakes was accompanied by climatic cooling that Roehler (1974, p. 58) has suggested was triggered by volcanism, a concept recently emphasized in a broad sense by Axelrod (1981) for the whole of the Western Interior during Cretaceous and Tertiary time. Roehler (written commun., 1983) cites evidence of worldwide cooling 38 m.y. ago. Tectonism probably was a factor also. A continued rise of the Uintas and other Rocky Mountain uplifts surely would have altered the local climate.

The extinction of Lake Gosiute north of the Uinta Mountains in middle Eocene time was caused by sedimentation in a lake that was a playa most of the time, in the view of most investigators (Eugster and Surdam, 1973; Bradley, 1973; Eugster and Hardie, 1975; Surdam and Wolfbauer, 1975; Surdam and Stanley, 1979), although Desborough (1978) and Boyer (1982) have recently raised doubts about the playa-lake model. At any rate, the last major stage of Lake Gosiute (the Laney stage) was predominantly fresh until its disappearance and must, therefore, have had an outlet, presumably south into Lake Uinta (Bradley, 1929, p. 89, 1964, p. A2; Roehler, 1965, p. 147; Surdam and Stanley, 1979, p. 101, 105) by way of the Sand Wash Basin. A physical connection at times between the two lakes, moreover, is indicated by their common fish faunas.

One or the other or both of the lakes probably drained east at times to the Mississippi Valley. If Lake Uinta received the overflow of Lake Gosiute, it in turn may have drained east across Colorado during its freshwater stages via some unknown outlet to the Mississippi Valley, where the affinities of the well-known fish faunas of the Green River Formation chiefly lay. The numerous gars, catfish, perches, bowfins, paddlefish, mooneyes, herrings, and perhaps sunfish (Priscacaridae) (Grande, 1980) were all members of eastern American families (Miller, 1958; Uyeno and Miller, 1963), none of which is indigeneous to the Colorado River system, and most of which are unknown in the Eocene faunas of the Pacific Northwest (Grande, 1980). If these fishes entered Lake Uinta from Lake Gosiute, rather than vice versa, the chief high-water outlet of Lake Gosiute might have been east across Wyoming to the Mississippi. Such easterly overflow might have followed a rise of both lakes to a common high water level. In either event, the fish faunas suggest that the Continental Divide in Eocene time was west of the Green River lakes and the Uinta Mountains, a likelihood that is compatable with Eocene paleogeography. The Colorado River system at that time did not yet exist. None of the 14 species of fishes native or endemic to the present Upper Colorado River system (Behnke and Benson, 1983), moreover, had progenitors in the Green River lakes except, perhaps, the wide ranging suckers (Catostomidae), which are Nearctic in their distribution. The indigenous Salmonidae (trouts) and Cottidae (sculpins) of the modern Upper Green River and the Columbia are known in North America from the Eocene (Miller, 1965; Patterson, 1981, p. 276) but are not present in the fish fauna of the Green River Formation.

A contributing cause of the disappearance of both lakes was a decrease in tectonic downwarping, which enabled depositional filling to overtake subsidence. The Uinta Mountains, however, probably continued to rise as shown by the accumulations of coarse conglomerate above the Green River Formation close to the south flank of the mountains, in the Duchesne River Formation (Andersen and Picard, 1972). In most places this conglomerate is gently tilted, locally to as much as 30 degrees. On the north flank at Phil Pico, a thick pile of conglomerate contains rocks of Wasatch, Green River, and Bridger age (Bradley, 1964, p. A53; Rowley and others, 1979). Downwarping continued in the Uinta Basin while most of the Duchesne River Formation was being deposited, followed by regional uplift, probably starting in early Oligocene time. This uplift terminated the long depositional cycle that had begun in the basin with the Laramide orogeny. To the north in the Green River Basin, reelevation of the Rock Springs uplift steepened dips locally to as much as 13 degrees after the Bridger Formation had been deposited but before the Bishop Conglomerate had (Bradley, 1964, p. A14). The stage was being set for the cutting of the Gilbert Peak erosion surface.

Before the onset of pedimentation the gradient between the foot of the Uinta Mountains and the central parts of the adjoining basins had to be increased enough to change the drainage regimen in the basins from a depositional to an erosional mode. This gradient was achieved by deformation and concomitant erosion, the extent of which has been described by Bradley (1964, p. A16) as follows:

Without departing from its original depositional dip, the Bishop Conglomerate bevels the tilted Bridger and Green River Formations and the more steeply inclined Upper Cretaceous rocks of the Rock Springs uplift.

The extensive erosion surface on which the Bishop Conglomerate was deposited indicates qualitatively that the post-Eocene deformation must have occurred appreciably earlier than the cutting of this erosion surface. In the Green River Basin this extensive surface bevels beds high in the Bridger Formation, and farther east it cuts progressively lower in the Tertiary; at Pine Mountain and Diamond Peak (near the junction of Wyoming, Colorado, and Utah) it bevels rocks that belong to a tongue of the Wasatch Formation which overlies the basal unit of the Green River Formation. These formations, therefore, must have been deformed long enough prior to the deposition of the Bishop Conglomerate to permit the removal of several thousand feet of beds over a large area and to allow for the reduction of the terrain to a remarkably smooth surface.

To erode all that rock, the drainage needed a new base level, and strong uplift was needed to provide the gradient. At the close of the Green River epoch the floors of the basins were only 150-300 m above sea level (Bradley, 1929, p. 89; MacGinitie, 1969, p. 52, 73), and the gradient of the basin floors was only about 0.2-0.4 m/km (Bradley, 1964, p. A16). But the graded relief on the Gilbert Peak erosion surface is hundreds of meters. The surface truncates rocks of Precambrian to Eocene age, and it slopes as much as 7-19 m/km across the Tertiary basin sediments alone. The relief is more than 600 m across the southeastern part of the Green River Basin between Pine Mountain on the south and Pilot Butte on the north. This profile, which cuts deeply into the Cretaceous of the Rock Springs uplift, may have been warped, according to Bradley (1936, p. 186), but the relief is just as great on undeformed profiles 100-115 km to the west, cut entirely on the Bridger Formation.

The Gilbert Peak erosion surface thus could not have been cut without strong prior uplift and dissection of the floors of the basins, without a drastic change in base level and local physiography. There also seems to be no way for the Gilbert Peak surface to have been graded to a master stream flowing out of the Green River, Washakie, and Sand Wash Basins into the Uinta or the Piceance Creek Basin, as will be pointed out, although a hydraulic connection across the Uinta axis probably had existed during the deposition of the Laney and the Parachute Creek Members of the Green River Formation (Bradley, 1929a, p. 89; Roehler, 1965, p. 147; Surdam and Stanley, 1979, p. 101, 105). This connection would have formed a narrow strait across the axis at Axial Basin and perhaps would have had an appreciable current. Surdam and Stanley (p. 105) have described a dispersal of volcanic lithic detritus into the Piceance Creek Basin at that time from the Lake Gosiute Basin, an event that would have signaled the filling and demise of Lake Gosiute. Other sources of volcanic detritus are plausible also, as noted further on, and the likelihood that north-to-south drainage persisted across Axial Basin into Gilbert Peak time, at any rate, is remote. Axial Basin was relatively high structurally and topographically by the end of Eocene time and in Gilbert Peak time (before post-Bishop collapse of the eastern Uinta Mountains), and the threshold or strait across Axial Basin would have been correspondingly high.

If early southward drainage ever crossed Axial Basin, it probably was halted before or by the time of the sharp anticlinal rise of Axial Basin at the end of Eocene time. Sears (1924b, p. 300) was first to point out that the early Tertiary rocks of that area (the Fort Union, Wasatch, and Green River Formations) rest on the Cretaceous with no discordance of dip, and, therefore, that Axial Basin was uplifted after their deposition. This time of uplift probably marked the beginning of the disturbance that deformed the Eocene rocks in the adjacent basins and terminated their deposition.

The Washakie Formation of Wyoming and the Uinta Formation of Utah and Colorado are partly equal in age, but "there is no evidence to suggest that they ever were physically connected across the Uinta uplift" (Roehler, 1973, p. 9). Roehler noted lithologic and color differences between the Washakie and the Uinta that indicate independent sources for their clastic fractions. Much of the Washakie, according to Roehler, consists of arkose derived from the Sierra Madre to the east, whereas the Uinta Formation had local sources in the Uinta Mountains. Volcanic ash constituents of the two formations could have had a common source but need not have. Volcanism was widespread at that time (Lipman and others, 1972), and many centers could have contributed. The source of both ash and volcaniclastics in the Washakie appears to have been the Absaroka volcanic field of northwestern Wyoming (Ebens, 1963; Roehler, 1973), but sources in Idaho are possible also (the Challis Volcanics, for example; Armstrong, 1978; Axelrod, 1981, p. 5). The Rattlesnake Hills of central Wyoming (Carey, 1959; Love, 1970) could have contributed, given favorable winds aloft during an eruption. Absaroka ash could have settled in the Uinta Basin, and probably did, but ash may also have traveled from igneous centers in western Nevada (Armstrong, 1963) if intrusive bodies there vented to the surface. Tuffs younger than 45 m.y. may have entered the Uinta Basin from sources in the Sevier orogenic belt (Dyni, 1981, p. 14).

Detrital volcanic material in the Uinta Formation may have had multiple sources also, possibly including nearer sources in Middle Park, Colo., or even the Front Range, where lower Tertiary volcanic and intrusive rocks are abundant (Izett, 1968, p. 22-25; Tweto, 1976, 1979) and where the volcanic cover then must have been wider than now. This detritus might even have reached the Uinta Basin by way of the ancestral Colorado River, which, according to Hunt (1969, p. 67, 70), once flowed west into the Uinta Basin via the White River.

By late Eocene time, beginning about 41 m.y. ago, and lasting generally into the Miocene, airborne and perhaps stream-deposited volcanic material might have entered the Uinta Basin from extensive new igneous centers in the Great Basin and adjacent areas, including the Wasatch, Oquirrh, Tintic, and Marysvale areas of Utah, where intrusive and volcanic activity overlapped spatially and temporally (Bassett and others, 1963; Moore and others, 1968; Laughlin and others, 1969; Moore and Lanphere, 1971; Crittenden and others, 1973; Bromfield and others, 1977). Volcanism started in the Marysvale district somewhat later than elsewhere, about 31 m.y. ago, and persisted intermittently into the Quaternary (Rowley and others, 1975). Once basin-and-range faulting was underway, however, stream-carried detritus no longer could have reached the Uinta Basin from the down-faulted sources in or beyond the Wasatch Range. (For example, see Stokes, 1976; Hunt, 1982.) By various estimates, regional extension, which led to the faulting, could have started as early as 30 m.y. ago (Ingersoll, 1982) or locally as late as 17 m.y. ago (Christiansen and McKee, 1978, Steven and others, 1978).

The strongest evidence that drainage did not cross Axial Basin in Gilbert Peak or Bishop time is in the physiographic form of the Bishop Conglomerate itself and its transport direction. Stanley and Surdam (1978) and Surdam and Stanley (1979, fig. 19) reported evidence of south to southeast flow of water from Lake Gosiute toward the Piceance Creek Basin in middle Eocene time. This direction, however, is 100°-180° from the later transport direction and slope of the Bishop Conglomerate on the Gilbert Peak surface, which was north to northeast from the Uinta Mountains across the Rock Springs uplift and the Sand Wash Basin (fig. 17). Any southerly drainage across Axial Basin into the Piceance Creek Basin, therefore, should have ended long before the Gilbert Peak erosion surface was formed. Modern south-flowing drainages (such as Sand Wash, the Little Snake River, and Vermillion Creek) postdate the deformation of the Gilbert Peak erosion surface. For example, the Bishop Conglomerate east of Sand Wash had to come from sources to the west or south, opposite or across the present drainage direction; Vermillion Creek flows southwest, but Bishop remnants there slope northeast. Similarly, the course of the Green River farther west is directly opposite the old pediment slope of Hickey, Cedar, Black, Little, Miller, Flattop, and Pine Mountains.

FIGURE 17.—Directions of foreset bedding and cross stratification in the Laney Member of the Green River Formation (small arrows; from Surdam and Stanley, 1979) and transport directions of Bishop Conglomerate (large arrows) greater Green River Basin area.

An alternative and more likely direction for a base leveling stream in the Green River Basin in Gilbert Peak time, therefore, was eastward toward the North Platte River, as first suggested by Bradley (1936, p. 177) and as elaborated by Hansen (1965, 1969b). Ritzma had endorsed the idea in 1959; Sears had hinted at it in 1924 (1924a, p. 304), but it had never received serious attention until my reconnaissance studies of the 1960's. Such a stream would have provided a geomorphically credible base level for the Bishop Conglomerate. It may have crossed the present Continental Divide at the Great Divide Basin near Tipton, Wyo., skirting the north side of the Rawlins uplift, and perhaps merging with the Sweetwater River. (See section on the capture of the master drainage of the Green River Basin, p. 67.) According to Denson and Chisholm (1971, p. C125) the main drainage framework in central and eastern Wyoming was in existence at the close of the Eocene, including the courses of the Sweetwater and the North Platte, and persists today with only minor modifications. The Tertiary history of central Wyoming is complex (Love, 1970, 1971), and the exact pattern of drainage is problematical, but it is not incompatible with easterly escape of drainage from the Green River Basin. Even now the Continental Divide at Tipton is 245 m lower (altitude about 2,070 m) than the base level of the Gilbert Peak erosion surface near the center of the Green River Basin (altitude about 2,315 m).

ALTITUDE AND RELIEF

To get some idea of regional altitude and relief at the onset of deposition of the Bishop Conglomerate, one may use the top of the Tipton Member of the Green River Formation as a datum and a basis for speculation, since this horizon has wide areal extent. Along with the overlying Bridger and Washakie Formations, the Green River Formation was deformed at the end of the Eocene, and deposition was supplanted by erosion. From the center of the Green River Basin west of Rock Springs to the flank of Pine Mountain about 80 km to the south, the difference in altitude of the top of the Tipton is about 760 m, judging from maps by Bradley (1964) and Roehler (1972, a, b). This figure essentially represents pre-Bishop differential uplift following deposition of the Bridger and Washakie Formations. Perhaps 150 m should be subtracted from the 760 m total to account for an initial dip of 0.2-0.4 m/km on the Tipton, for stratigraphic transgression of the top, and for an unknown but possibly appreciable later Tertiary and Pleistocene differential uplift of the mountains. Thus, between the end of the Eocene and the deposition of the Bishop Conglomerate, the northeast flank of the Uinta Mountains may have risen some 600 m relative to the centers of the adjacent basins.

Differential uplift, however, was accompanied and reinforced by regional (epeirogenic) uplift, which raised the centers of the basins as well as the mountains high enough—perhaps 900 m—to change the climate from the subtropical one of the middle Eocene to the more temperate climate of the post-early Oligocene, as indicated by floristic and faunal studies (Dorf, 1959; MacGinitie, 1969; Love, 1971; Leopold and MacGinitie, 1972). Nine hundred meters of regional uplift would also provide a reasonable gradient for drainage out of the Green River Basin to an ultimate base level far to the east, a gradient somewhat less than but not too different from that of the present Sweetwater and North Platte Rivers across central Wyoming. The center of the Green River Basin, therefore, may have been about 1,200 m above sea level when the Gilbert Peak erosion surface began to form. The mountains were more than 1,600 m higher, because the surface itself has 1,600 m or more of relief, and the peaks stand 300-600 m above it. Perhaps they stood 3,000-3,700 m above sea level. The crestline was much degraded during Gilbert Peak and Bishop time, and the coarseness of the Bishop suggests that local relief was greater then than now.

RATES OF UPLIFT

Figures in the previous paragraph provide a means of estimating minimum uplift rates in the Lake Gosiute area between the end of the Eocene and the deposition of the Bishop Conglomerate. The maximum elapsed time was about 9 m.y., inasmuch as the Eocene ended about 38 m.y. ago, and a tuff well up in the Bishop is about 29 m.y. old, as dated by H. H. Mehnert (Hansen and others, 1981). Based on these dates and on the altitude differences of the Tipton datum before and after uplift, regional uplift in the basin exceeded 10 cm per thousand years (900 m/9 m.y.), and the rate of rise along the basin rim exceeded 16.6 cm per thousand years (1,525 m/9 m.y.). The actual uplift, however, probably took much less than 9 m.y., perhaps half as much, because a long period of crustal and climatic stability had to follow uplift and accompany the cutting of the Gilbert Peak erosion surface to allow so broad and featureless a plain to form. This period itself probably lasted millions of years, maybe most of Oligocene time. The Rock Springs uplift, for example, had to be truncated by the Gilbert Peak erosion surface before the Bishop was deposited while, concomitantly, hundreds of square kilometers of hard rock was being bevelled along the Uinta Mountains. Reasonable rates of uplift, therefore, perhaps would have been closer to 20 cm per thousand years near the center of the basin and 30 cm per thousand years along the mountain flank. By way of contrast, the earlier rate of Laramide uplift of the Uinta Mountains was a half to a whole order of magnitude higher, in the range of 1-2 m per thousand years (about 9,150 m in 5-10 m.y.; Hansen, 1984, p. 15). This figure compares favorably with a calculated deformation rate for the Wind River Mountains in Wyoming—1.3 m/1,000 yrs for the displacement of the Wind River fault (Hunch and Smithson, 1982, p. 1559).

In the time since the Bishop Conglomerate was deposited the average rate of uplift has been very low. The center of the basin has risen about 1,100 m in the last 29 m.y.—about 3.8 cm per thousand years (or, in round numbers, 4 cm). During that 29-m.y. interval the rate probably was uneven and may have varied greatly from time to time.