- UW Department of Geology and Geophysics
The Laramide synorogenic strata of the Denver Basin record the uplift and denudation of the central and southern Front Range of the Rocky Mountains. Synorogenic sedimentation took place in two distinct pulses, the first spanning the Cretaceous–Tertiary boundary and extending into the early Paleocene. The second occurred during the latest Paleocene and early Eocene. Facies patterns reflect proximal to distal fluvial environments. Progressive unconformities mark the western side of the basin, and lignite beds and ponded deposits characterize the distal part of the first pulse. The second pulse preserves a more uniform fluvial succession characterized by alternating arkosic channel sandstones and olive-brown overbank deposits. The lateral and vertical facies changes found in these strata have engendered a complex history of nomenclature. By considering the units as a pair of unconformity-bounded sequences called the D1 sequence and the D2 sequence, a simpler pattern emerges that allows regional facies variability to be understood on a basin-wide scale. Two complete cores are used to calibrate geophysical logs from oil and water wells, providing a subsurface database that is correlated to outcrops. This allows the creation of cross sections that permit isolated outcrops and fossil occurrences to be correlated into an integrated basin-wide framework. The paleontologic record is used together with radiometric dating and magnetostratigraphy to define the time span during which sediment accumulated. The episodic accumulation of sediment is interpreted to reflect episodic displacement along the thrust faults bounding the Front Range. The first episode of sedimentation, defined as the D1 sequence, is interpreted to represent uplift of that part of the Front Range bounded by the Golden and Rampart Range faults. During this period of sedimentation, andesitic volcanic rock that covered much of the Front Range was stripped from the uplift and deposited in the D1 sequence. The second episode of sedimentation, the D2 sequence, is interpreted to involve sediment eroded from the mountains west of the Colorado Springs area, perhaps as a result of uplift of the Pikes Peak area by the Ute Pass fault. Depositional patterns in the basin are a logical response to differential uplift and variations in fluvial geomorphology. Understanding these systems can help quantify depositional patterns and subsurface distribution of bedrock aquifers.
The Late Cretaceous through early Tertiary Laramide orogeny disrupted and partitioned the regional near sea-level coastal plain that characterized the west-central United States following the retreat of the Cretaceous Western Interior Seaway. A series of compressional Precambrian basement-cored mountain uplifts bound intervening and peripheral downwarped basins. Structural relief often exceeds six kilometers (Dickinson et al., 1988; Gries et al., 1992; Snoke, 1993). The history of the Laramide orogeny is recorded in the texture and character of the synorogenic sediments contained within these basins. Late Tertiary epeirogenic uplift has rejuvenated drainage networks across and among these basins, and the Laramide basin-filling strata now being exhumed are available for our study.
The Denver Basin lies immediately east of the central and southern Front Range uplift in central Colorado (Figs. 1 and 2). The basin has no well-defined eastern structural boundary (Hemborg, 1996; Weimer, 1976, 1996) and is considered a perimeter basin (Dickinson et al., 1988). It is bounded on the west by the uplifted Front Range, on the north by the Greeley Arch which separates it from the Cheyenne Basin (Kirkham and Ladwig, 1979), on the east by a gentle onlapping relationship onto the stable mid-continent platform, and on the south by a saddle in the Pueblo area (the Apishapa Arch) separating it from the coeval Raton Basin. In the Denver Basin, a subset of what was once a more extensive Laramide basin-filling succession is preserved in the area south and east of Denver (Fig. 3). These strata are dissected by tributaries to the South Platte and Arkansas Rivers.
Defining mappable units in the synorogenic strata is difficult because of abrupt facies and compositional changes. To add to the challenge, rock outcroppings are relatively poor and discontinuous in this low-relief and vegetated basin. As a consequence, surface mappers have used a variety of strategies to define and depict field-mappable units (see reviews by Soister, 1978b; Crifasi, 1990; note variable mapping strategies in three adjacent quadrangles by Scott, 1962 and Mayberry and Lindval, 1972, 1977).
This study considers depositional environments and subsurface data to arrive at a simplified regional interpretation of the stratigraphic architecture of the synorogenic Laramide strata preserved in the Denver Basin. Many of the data described in this paper are derived from wells and cores, from which depth measurements are recorded in feet. As a consequence, this paper uses both English and metric units.
EARLIER WORK ON SYNOROGENIC STRATA
Patterns and complexities of facies variability in the Denver Basin were recognized early and have been the subject of considerable discussion (see, for example, Eldridge in Emmons et al., 1896, p. 153 and Crifasi, 1990). The history of stratigraphic studies in the basin is daunting. Rock units have been named and renamed. In some cases, boundaries have been based on subtle differences. In many other cases, boundaries are quite clear where defined only to vanish laterally amidst facies changes or covered ground. One of the principal conclusions of the work presented in this paper is that a simple, two-part subdivision of the Laramide synorogenic strata is consistent with the data.
The following section recapitulates some of the historically most important observations in chronological order.
Emmons et al. (1896) compiled a major monograph on the Denver Basin summarizing many years of work by the U.S. Geological Survey and the Territorial Surveys. They reviewed early literature and described the Laramie Formation (p. 72–77). They defined the Arapahoe Formation as unconformably overlying the Laramie (p. 151–155) and the Denver Formation as characterized by andesitic clasts (p. 160). They also defined the Monument Creek Formation (p. 253–254), which includes strata later known as the Dawson Arkose and the Castle Rock Conglomerate.
Richardson (1915, p. 7) named the Dawson Arkose after exposures on Dawson Butte.
Johnson (1930) summarized the paleontologic record with emphasis on “splendidly preserved” leaves of the Denver Formation. He suggested abandoning the term Arapahoe Formation as it and the Denver Formation grade laterally into the Dawson Arkose.
Dane and Pierce (1936) presented detailed mapping on the east side of the basin. They speculated on orographic effects of uplift (p. 1326) and divided the Dawson Arkose into a coarse upper and a heterolithic lower member.
Brown (1943) determined the location of the Cretaceous–Tertiary boundary based on fauna, most notably at South Table Mountain. He projected the K–T boundary around the basin with great prescience.
LeRoy (1946) presented a synthesis of stratigraphy of the Denver area. He used the phrase “Arapahoe–Denver Formation” to encompass the post-Laramie strata of the Denver area.
Reichert (1956), having completed a Ph.D. dissertation on Green Mountain, presented a synthesis of the Denver Basin in which he proposed that the term Denver Formation be used for all rocks dominated by andesitic clasts (p. 110–111). He proposed that the Denver Formation is underlain by the Arapahoe Formation and overlain by the Dawson Formation.
Epis and Chapin (1975) presented an analysis of the Rocky Mountain erosion surface in the Front Range area, concluding that it represents a regional pre-Oligocene erosion surface.
Having published a series of quadrangle maps and open-file reports on the Denver Basin, Soister (1978a, 1978b; Soister and Tschudy, 1978) presented a series of three papers recapitulating the stratigraphy and age control of the synorogenic strata. Working with palynologist R. H. Tschudy, he recognized the Denver Basin paleosol, and noted that Dawson Arkose strata above it are Eocene. He also noted that strata below the paleosol belonging to the Dawson Arkose and Denver Formation are Paleocene and Cretaceous, and he reported rare upper Paleocene pollen.
Morse (1979) completed a Ph.D. dissertation reviewing the facies and depositional environments of the Dawson Arkose and the Castle Rock Conglomerate. He included the entire synorogenic package in the Dawson Arkose and mapped an upper unit on the basis of a coarsening phase in the arkosic package developed below Soister's paleosol.
Kluth and Nelson (1988) described progressive unconformities in the Dawson Arkose in the southwest corner of the Air Force Academy and used pollen to date the deformation as latest Cretaceous.
Crifasi (1992), with a background in groundwater, examined the challenging stratigraphic nomenclature and emphasized the difficulty of employing compositionally based terms in subsurface work where only geophysical log control is available. He used sand/shale ratios determined from the logs to map sand-rich lobes in the aquifer units and suggested that vertical variation in sand/shale implies episodic deposition (p. 26).
Raynolds (1997) suggested that Soister's paleosol can be used to divide the synorogenic strata into two unconformity-bounded sequences termed D1 and D2, thus simplifying the nomenclature.
As the Cretaceous Western Interior Seaway retreated from Colorado to the northeast during the Baculites clinolobatus and the Hoploscaphites birkelundae ammonite zones (Scott and Cobban, 1965, 1986; Landman and Cobban, in press), marine strata of the Pierre Shale gave way to the nearshore and shoreface environments represented by the regressive Fox Hills Sandstone.
The Fox Hills Sandstone is a series of aggradational sandstone bodies that accumulated during the episodic regression of the interior seaway. These sandstone units are superimposed, and commonly three or four beds can be identified at a single locality. The Fox Hills style of regressive coastal sediment accumulation seen in the Denver Basin is similar in motif to that documented in other Cretaceous regressive sandstone beds of the Western Interior (e.g., Devine, 1991). Stratigraphic studies of the Fox Hills Sandstone have been conducted for uranium prospecting (Reade, 1978) and for regional correlation (Kiteley, 1978). A detailed study examining internal facies distribution patterns using outcrops, geophysical logs, cores, and modern analogues has been carried out by Nibbelink (1983) north of the Greeley Arch. South of the Denver Basin, the equivalent sandstone is termed the Trinidad Sandstone and has been the subject of two master's theses (Billingsley, 1977; Manzolillo, 1976). Internal facies patterns in the Trinidad Sandstone have been well documented by Flores and Tur (1982).
The base of the Fox Hills Sandstone is transitional into the Pierre Shale (Lovering et al., 1932; LeRoy, 1946). Typically, thin distal storm beds displaying hummocky cross stratification merge gradationally downward into uniform, well-bedded open-marine shale. Because of this transitional contact relationship, workers commonly differ in picking the base of the Fox Hills Sandstone or top of the Pierre Shale. The body of the Fox Hills Sandstone is comprised of from one to four sandstone beds, each typically 10–15 meters thick. These are generally comprised of well-sorted medium- to fine-grained quartz-rich sandstone with planar cross bedding and common Ophiomorpha trace fossils. Immediately north of Colorado Springs, an outcrop of Fox Hills Sandstone contains phosphate pebbles up to 4 cm in diameter and well-rounded quartzite and volcanic pebbles up to 6 cm in diameter.
The top of the Fox Hills Sandstone is more distinctive. It is generally defined by a sharp transition from clean beach or near-shore sandstone into the carbonaceous shales and coal beds of the lowest Laramie Formation (Emmons et al., 1896, p. 72–77). While over the years there has been confusion over this boundary (e.g., see discussion in Knowlton, 1922, p. 18–25), I follow Lovering et al. (1932, text of committee report quoted in LeRoy, 1946, p. 87) and LeRoy (1946) in considering the Fox Hills Sandstone to consist of the near-shore and beach facies sandstone beds, while assigning the generally conformable overlying nonmarine strata to the Laramie Formation. In the Denver Basin, this transition is often well defined in outcrops and on geophysical logs; however, the boundary can be difficult to define in isolated outcrops or on individual geophysical logs where fluvial sandstone beds in the Laramie Formation immediately overlie the Fox Hills Sandstone. Because of the shingled nature of the Fox Hills Sandstone it is possible to find situations in which marine shale arguably correlative to the Pierre Shale occurs in tongues above a layer of Fox Hills Sandstone (e.g., see Fig. 18 of Weimer, 1976). On cross sections that include numerous geophysical logs, the upper Fox Hills exhibits abrupt seaward step-ups where Fox Hills Sandstone shingle boundaries are encountered. Outside the Denver Basin, this boundary can be difficult to define where widespread estuarine conditions characterize the transition from marine to continental environments.
Regional thickness patterns suggest the Laramie Formation is about 400–500 feet thick on the western side of the basin and 200–300 feet thick on the eastern side (Fig. 4). Laramie Formation strata have been studied with an emphasis on coal resources (Emmons et al., 1896; Kirkham and Ladwig, 1979; Brand and Eakins, 1980; Spencer, 1986; Eakins and Ellis, 1987). These coal beds were primarily exploited from the late 1800s through the mid 1900s. Abandoned coal mines have been used for natural gas storage and present a subsidence hazard along the rapidly developing Front Range urban corridor. An appraisal of remaining Laramie Formation coal resources in the central part of the Denver Basin (Eakins and Ellis, 1987) concluded that although over 20 billion tons of coal are present, most of it is too deep or too thinly bedded to be mined economically. On the east side of the basin, individual Laramie Formation coal beds are typically five to 15 feet thick. The coal beds are concentrated near the base of the Laramie Formation; the overlying strata, dominated by overbank mudstone beds, are interbedded with relatively uncommon fluvial channel sandstone units. Fluvial sandstone beds in the Laramie Formation are thicker and more prevalent on the western side of the basin.
The principal stratigraphic record of the Laramide orogeny is contained in rocks overlying the Laramie Formation. In the Denver Basin, these strata have been named the Arapahoe and Denver Formations and the Dawson Arkose. The historical context is summarized in the cross sections of Figure 5.
The term Arapahoe Formation (originally defined by Emmons et al., 1896, p. 153 at Willow Creek southwest of Denver) has been variously used to encompass the basal conglomerates and the lower sandstones of the basin-fill. I concur with Brown (1943, p. 78), who suggested that the term Arapahoe conglomerate be restricted to the conglomeratic facies (Lower Division of the Arapahoe Formation of Emmons et al., 1896), which generally is present at the base of the synorogenic sediments and which lies unconformably on top of the Laramie Formation. Whereas Reichert (1956) extended the thickness of the unit upward to create a mappable formation, I suggest the term be used solely for the basal conglomerate and not be given formational status.
The Denver Formation, named after exposures in the vicinity of the City of Denver, was first defined by Cross (in Emmons et al., 1896, p. 155–156) and is characterized by fluvial strata containing significant amounts of andesitic volcanic debris. The Denver Formation is conglomeratic in the vicinity of Green Mountain and is finer-grained where exposed in the south and east.
The Dawson Arkose is named from exposures on Dawson Butte, located south of Denver (Richardson, 1915, p. 7–14). The unit is characterized by alternating arkosic sandstone and mudstone beds. While the Dawson Arkose generally overlies the Denver Formation, the two units are observed to interfinger in the Littleton area (Scott, 1962). In other areas, significant thicknesses of arkosic strata overlie the basal conglomerates and little if any andesitic Denver Formation is exposed. Considerable confusion has arisen because these two units are defined on the basis of compositional characteristics which vary laterally and vertically.
Local member names have been proposed (e.g., near Colorado Springs by Kittleman, 1956 and Thorson et al., 2001; near Green Mountain by Hares, 1926; in the Table Mountain area by LeRoy, 1946). Many of these and other efforts at stratigraphic subdivision have been reviewed by Crifasi, 1990. The subdivided units cannot be correlated for long distances in the basin and are therefore of only local use.
Regional compilers such as Reichert (1956), Trimble and Machette, (1979a, 1979b) Tweto (1979), and Bryant et al. (1981) have grouped units using a variety of criteria that have required the lumping and splitting of previously defined formations and units (e.g., Tweto, 1979). This is especially the case where parts of the Dawson Arkose are separated by a mapped boundary between “Upper part of Dawson Arkose” and “Denver Formation or Lower part of Dawson Arkose”). Below, I take a similar approach in lumping and splitting and subdivide the Dawson Arkose and Denver Formation into two packages representing two distinct phases of basin fill.
Hydrogeologists working in the Denver Basin have developed a nomenclature to characterize the bedrock aquifers for administrative purposes (e.g., Robson and Romero, 1981a, 1981b; Robson et al., 1981a, 1981b; Robson, 1983, 1987, 1989; Romero, 1975; and Crifasi, 1990, 1992). These aquifer units (Fig. 5B) are considered to be regionally continuous and are administered by the Office of the State Engineer (Graham, 1999). Whereas the internal stratigraphic variability of these aquifers is significant, present State models may represent an incomplete evaluation of water resource potential (Lapey, 2001; Woodard et al., this issue).
METHODS AND DATA SETS
The new stratigraphic framework for the Laramide synorogenic strata of the Denver Basin presented here integrates information derived from surface geological mapping and subsurface stratigraphic studies.
Geologic mapping has been carried out by the U.S. Geological Survey starting with monographs published in 1896 by Emmons et al. and continuing through a series of scattered 1:24,000-scale maps and the 1:100,000-scale Front Range Urban Corridor series (Trimble and Machette, 1979a, 1979b), the 1:250,000-scale one by two degree map (Bryant et al., 1981), and the updated geologic map of Colorado (Tweto, 1979). In 1999, the Colorado Geological Survey started a 1:24,000-scale geologic mapping program in the southwestern part of the basin (Carroll and Crawford, 2000; Thorson et al., 2001). A wide variety of approaches has been taken to handle the rapidly varying facies changes in the basin. Some mappers have defined formations (e.g., Scott, 1962). Some have mapped facies (e.g., Mayberry and Lindvall, 1972, 1977; Thorson et al., 2001) while others have broadly combined units to obtain regional maps (Reichert, 1956; Tweto, 1979).
Electric Logs and Cored Wells
Hundreds of geophysical logs are available from water wells and oil and gas test holes. Stratigraphic relationships mapped at the surface can be recognized in the logs of wells that are adjacent to those outcrops; this facilitates interpretation of subsurface stratigraphy. Data from two cored wells located at Castle Pines (Robson and Banta, 1993) and Kiowa (Raynolds et al., 2001; Wilson, this issue; Fig. 2) provide additional valuable subsurface data. To complement the core data, cuttings from selected water wells south of Denver have also been examined. The geophysical log signature of the units in the Denver Basin can be recognized based on these correlations. Figure 6 illustrates the typical log responses that have been used to develop cross sections (Figs. 5 and 7) and maps. Similar geophysical log cross sections have been constructed and correlated across the basin using the cored wells as tie points. These log cross sections permit a series of regionally useful boundaries to be recognized. They form the basis for the stratigraphic nomenclature proposed herein.
Recognition of Key Horizons in Surface and Subsurface Settings
A critical aspect of stratigraphic research in the Denver Basin is selection of mappable units. The ability to recognize rock units in outcrop is often hampered by discontinuous and small-scale exposures. By establishing a three-dimensional framework of correlatable rock units that integrate surface and subsurface observations, a regionally applicable and simplified set of mappable units is defined (Figs. 5 and 8).
Projection of Subsurface Horizons to Surface
The stratigraphic units defined through the integration of surface and subsurface observations can be evaluated in three dimensions and their surface projections identified. The results of this are shown in the map pattern on Figure 2.
NEW STRATIGRAPHIC FRAMEWORK FOR LARAMIDE SYNOROGENIC STRATA
Previously described mapping units and the aquifer nomenclature are not readily suited to study the regional evolution of synorogenic strata of the Denver Basin. Both existing systems are designed to rationalize a specific suite of observations (surface facies patterns and water-well performances), and neither allows for regional interpretation of basin-filling patterns. I subdivide the synorogenic sediments into two regionally extensive, unconformity-bounded sequences termed the D1 and D2 sequences. The D stands for Denver and the numerical convention is that the older sequence is the first. I suggest that future use of the names Arapahoe Formation, Denver Formation, and Dawson Arkose be done only with careful definition of the sense intended. Regional discussions of the Denver Basin designed to place it in a context with other Laramide basins, together with discussions dealing with time-calibrated analyses of faunal and floral evolution (e.g., Nichols and Fleming, this issue) will be better served by using the proposed new sequence nomenclature.
Definition of Two Unconformity-bounded Units
The unconformity at the base of the D1 sequence is defined by an abrupt facies change at the top of the Laramie Formation. The basal Arapahoe conglomerate unconformably overlies the fine-grained Laramie strata. These high-energy fluvial deposits lie on an erosional surface from which an unknown amount of rock has been removed by erosion. Geophysical logs show an abrupt coarsening of the strata that overlie the unconformity. This abrupt coarsening is present across all but the northwest quadrant of the basin. Here the basal synorogenic material is generally fine-grained, and the top of the Laramie Formation is difficult to pick with confidence on geophysical logs. Projections from outcrop contacts are used in this area to define the boundary. There is a change in the pollen zonation at this boundary (Nichols and Fleming, this issue). The Laramie Formation is in the Aquilapollenites striatus Interval Zone whereas the overlying D1 sequence strata are in the Wodehouseia spinata Assemblage Zone. In the Castle Pines well, there appears to be a magnetic polarity transition at this boundary from reversed to normal polarity (Hicks et al., in press). The time span marked by this unconformable boundary probably represents less than a million years; the underlying Laramie Formation was deposited above marine strata with ammonites dated at 69 to 70 Ma, and the base of the overlying D1 sequence can be estimated to be approximately 67 to 68 Ma based on linear rates of sediment accumulation (Hicks et al., in press).
The top of the D1 sequence is defined at the base of the prominent regional paleosol series first identified by Soister and Tschudy (1978) and recently reviewed by Farnham (2001) and Farnham and Kraus (this issue). It was originally thought that the paleosol series marked the long hiatus separating the two synorogenic sequences. The absence of lateritic minerals reported by Farnham, 2001, (at variance to an earlier report by Soister, 1978b), however, and the presence of aggradational features such as channel sandstones within the paleosol series, suggests that the paleosol series represents a phase of intense weathering associated with the earliest stages of accumulation of the overlying D2 sequence (rather than a prolonged weathered surface worn onto the surface of the D1 sequence; Farnham and Kraus, this issue). As discussed below, the geochronologic data suggest that this unconformity represents approximately five to eight million years not represented by rock. This unconformity defines the boundary between the D1 and D2 sequences. In outcrop and cores, the precise location of the unconformity can be difficult to define as two fluvial systems with attendant internal erosion surfaces are superimposed.
The paleosol series generally caps a shale-rich zone at the top of the D1 sequence. This shaly zone can be identified on gamma ray and resistivity logs from water wells and oil and gas test wells. Well penetrations of the paleosol in the center of the study area are interpreted based on correlations with adjacent well logs and ultimately tied to surface outcrops of the paleosol interval. A structure contour map of the top of the paleosol series is shown in Figure 9.
The top of the D2 sequence is defined by the unconformity beneath the Wall Mountain ignimbrite dated at 36.73 ± 0.07 Ma (McIntosh and Chapin, 1994). The ignimbrite erupted across the Rocky Mountain erosion surface (Epis and Chapin, 1975). This surface has been widely mapped in the Front Range (Scott and Taylor, 1986) and may have correlative counterparts elsewhere in the Rocky Mountains (Mears, 1993; Bradley, 1987). Based on the co-planar nature of the two surfaces, I correlate it with the beveled surface below the Miocene Ogallala strata in eastern Colorado and surrounding areas (Fig. 10). Where the surface is present above the paleosol, the entire thickness of the D2 sequence can be observed. Where the top of the D2 sequence has been eroded, one can restore an original thickness by interpolating the elevation of the Rocky Mountain surface and determining the difference in elevation to the base of the paleosol. Figures 11 and 12 illustrate isopach maps of the D1 and D2 sequences. Table 1 tabulates the subsurface control points used in making the structure and isopach maps.
Stratigraphy of Synorogenic Sequences
Both the D1 and the D2 sequences were deposited by fluvial systems draining from the Front Range. Ponding is more common in the D1 sequence than in the D2 sequence. The mineral compositions of the two sequences are distinctive. The D1 sequence is frequently characterized by abundant volcanic rock fragments and the D2 sequence is almost completely arkosic (Wilson, this issue). Locally, studies have differentiated heavy mineral suites that mirror the observed bulk composition of the strata (Curtis, 1942) with volcanic-derived pyroxenes and amphiboles characterizing the andesitic facies of the D1 sequence. In general, an unroofing character can be seen in the entire succession with recycled sedimentary grains being more common near the base of the D1 sequence. The D2 sequence is characterized by grains derived from deep erosion of the Front Range (Kelley, this issue; Wilson, this issue).
STRATIGRAPHY OF D1 SEQUENCE
The onset of sedimentation in the D1 sequence is marked by the Arapahoe conglomerate. In my terminology, the Arapahoe conglomerate refers only to the basal conglomerates that, while ranging up to 10 meters in thickness, can also be absent along paleo-interfluves (e.g., Fig. 10 of Weimer, 1976). In this case the contact may be difficult to discern.
Exposures of this basal facies are found in Golden, on the west side of the campus of the Colorado School of Mines, on Route 6 west of the Jefferson County building, on Alameda Parkway just east of C-470 (Fig. 13), and at the mouth of Jarre Canyon, 40 km south of Denver. The conglomerate is comprised of a variety of resistant clasts derived from the erosion of Mesozoic and Paleozoic strata together with granitic and gneissic clasts derived from the Precambrian bedrock of the Front Range. At some localities, Precambrian crystalline clasts exceed 75 percent of the clast population. As noted by Cross (in Emmons et al., 1896, p. 211), in the Denver area there is generally no vertical increase in the proportion of granitic clasts, thus there exists no unroofing succession in the Arapahoe conglomerate. The composition of the Arapahoe conglomerate indicates that in places the onset of synorogenic sedimentation was contemporaneous with incision through the veneer of Paleozoic and Mesozoic strata on the Front Range. At Colorado Springs, the basal D1 sequence Arapahoe conglomerate is dominated by chert, reworked phosphate nodules, and Cretaceous fossils. In this area, incision into the Precambrian granite is not reflected until higher in the sedimentary succession. The Arapahoe conglomerate facies is present across the Denver Basin; gravels have been described from east of Kiowa (Dane and Pierce, 1936, p. 1320), and granitic pebbles up to five cm in diameter were recovered from the Arapahoe conglomerate in the Kiowa core.
Heterolithic Facies of D1 Sequence
Overlying the Arapahoe conglomerate is a series of alluvial, fluvial, and paludal strata (mapped variously as the Arapahoe and Denver Formations and Dawson Arkose) that accumulated at the foot of the growing Front Range mountains (Kirkham and Ladwig, 1979). Progressive unconformities in gravelly facies characterize the proximal reaches (Kluth and Nelson, 1988) and the coarse clastic fluvial facies rapidly grade distally into fine-grained, paludal, and lignitic deposits (Soister, 1978a; Kirkham and Ladwig, 1979, 1980; Brand and Eakins, 1980; Eakins and Ellis, 1987). Figures 14 and 15 show the character of these strata. The isopach map (Fig. 11) demonstrates the thinning of the D1 wedge away from the active thrust faults (Boos and Boos, 1957; Harms, 1964; Jacob and Albertus, 1985) that bounded the mountain front during D1 sequence deposition.
The uppermost part of the D1 sequence is comprised of fine-grained mudstones which help make the boundary with the D2 sequence detectable on geophysical logs. These mudstone beds, which crop out on the western side of the basin, are laminated olive-green silty mudstone often with millimeter-scale bedding that suggests accumulation in low-energy, perhaps lacustrine conditions. Occasional coarse beds of sandstone containing highly weathered andesitic clasts intertongue with these mudstone beds suggesting that high energy streams drained into the ponded landscape.
Genesis of D1 Sequence
The D1 sequence preserves a diverse series of facies and lithologies that reflect the denudation of the uplifting Front Range. Unroofing successions are preserved wherein the composition of the clasts reflects the progressive erosion of sedimentary, volcanic, and basement lithologies. The succession of rock composition preserved in the sequence reflects the eruption and subsequent complete erosion of an andesitic volcanic terrain that once mantled much of the Front Range. This pattern is well seen in the composition of clasts at Green Mountain west of Denver, where early granitic clasts are replaced by andesitic clasts and ultimately in turn replaced by granite clasts as one ascends the mountain. The andesitic debris has been used in the northern part of the basin to define the Denver Formation, here considered a component of the D1 sequence. Near Colorado Springs, andesitic debris near the base of the D1 sequence was recently mapped as the “Lower part of the Dawson Formation” by Thorson et al. (2001). These workers have identified and mapped a total of four facies packages in the D1 sequence, thereby documenting in map view the compositional complexity of the succession noted earlier by Kittleman (1956).
These four heterolithic facies accumulated under settings with relatively steep topographic gradients. A combination of high sediment supply and rapid changes in accommodation space yielded a varied paleogeography that included proximal alluvial aprons spreading through distributary fluvial systems into paludal environments similar to those found today in the region east of the Andes in northern Bolivia and in the Minto Flats area north of the Alaskan Range (field reconnaissance by author). These factors translate into a rock assemblage that is made up of packages of characteristic facies and composition reflecting an evolution in both the source terrain that provided the sediments to the basin and in the patterns of the distributary systems that deposited them in the evolving foredeep (Horton and DeCelles, 2001).
The isopach patterns (Fig. 11) suggest that the sediments accumulated at the foot of the Front Range concurrent with fault motion on the Golden and Rampart Range faults.
STRATIGRAPHY OF D2 SEQUENCE
Denver Basin Paleosol
The onset of sedimentation in the D2 sequence is heralded by the Denver Basin paleosol series, first recognized as a widely correlatable unit by Soister and Tschudy (1978). The paleosol horizon is exposed in numerous outcrops around the Denver Basin south and east of Denver. The paleosol is compound and is properly considered a paleosol series. Mature oxisol horizons with brilliant red and purple iron oxide colors are common. Intense mottling and development of limonitic nodules in zones and beds are common features; burrows and root-casts occur as well. Coarse gravelly channel fills occur within the paleosol interval. These channel facies exhibit a striking diagenetic pattern where the granitic clasts have had all feldspar altered to clay minerals and the clasts are preserved as rounded boxworks of quartz and other less labile minerals. The geochemistry and internal character of the paleosol series is reported by Farnham and Kraus (this issue).
The paleosol series is distinctive in outcrop (e.g., Richardson, 1911) and is used as a feed-stock for the manufacture of bricks. Several clay-pits have been operated by brick companies in this horizon for over 50 years. In the course of 1:24,000-scale quadrangle mapping, the paleosol was termed the “variegated beds” by Mayberry and Lindvall (1972, 1977), but its significance as a regional mappable horizon was not recognized until the work of Soister. In tracing the outcrop pattern of the paleosol, it is evident that the paleosol series was variably developed across the ancient landscape. In places the paleosols aggregate more than 10 meters of red to purple clays, while in other places the paleosol is only an orange, clay-rich zone. In some places, strata of the D2 sequence rest on strata of the D1 sequence with no discernible weathering zone at the boundary. This variability is particularly evident on the eastern side of the basin where soil development can be examined along a well-exposed, north–south-trending escarpment immediately west of Bijou Creek (Farnham, 2001).
On the west side of the basin, mapping by the Colorado Geological Survey in the Monument area (Thorson, personal communication, 2002) and field observations in the Sedalia area indicate that there are multiple reddish paleosol horizons in the lower part of the synorogenic strata. This may result from more severe weathering on the uptilting margin of the subsiding foreland basin.
Heterolithic Facies of D2 Sequence
Overlying the paleosol are heterolithic arkosic strata characterized by coarse, multi-storied channel sandstone beds separated by overbank mudstone. Weakly developed paleosols occur within the overbank mudstone beds. The coarse, trough cross-bedded fluvial strata extend across the entire outcrop belt.
Good exposures of the D2 sequence occur in the Castlewood Canyon area, 50 km south-southeast of Denver. To the east, the resistant D2 strata form a carapace over the less resistant D1 shales and lignites. The resistant ledges and mesa rims also preserve good outcrops of D2 strata (Fig. 16).
The D2 sequence strata exhibit a facies and thickness distribution pattern quite different from that of the D1 sequence. The facies are more uniform and have less lateral compositional and textural variability. The observed thickness relations, together with northerly paleocurrent vectors observed south of Denver (Morse, 1979), suggest a source in the Pikes Peak area west of Colorado Springs. Subsidence of the Denver Basin area generated the accommodation space needed to preserve the strata. However, because the accommodation was not focused in a narrow zone parallel to the Golden and Rampart Range faults along the edge of the Front Range, coarse fluvial D2 sequence sediments were carried (and preserved as strata) well beyond the mountain front.
Genesis of D2 Sequence
The composition of material comprising the D2 sequence strata suggests it is the product of the redistribution of a grus mantle derived from the Front Range Precambrian granite. The Pikes Peak batholith weathers into grus that rarely includes clasts much over 2–4 cm, and generally it breaks down to a coarse arkosic sandy gravel. The sky-blue feldspar, amazonite, is a distinctive, though rare, constituent of the D2 sequence. Amazonite is well known from pegmatite veins in the area near Crystal Peak, north of Pikes Peak. It is probable that an uplift event in the central and southern Front Range caused the dispersal of the D2 sequence strata. The accumulation of D2 sequence arkose may be coincident with uplift in the Pikes Peak area and eastward thrusting of the Ute Pass fault. An alternative explanation for the deposition of D2 sequence strata could involve eastward tilting of the Front Range in response to a period of westward thrusting on the Elkhorn thrust (Kluth and Nelson, 1988).
CHRONOLOGY OF SYNOROGENIC STRATA
Age Constraints on D1 Sequence
Five lines of evidence are used to date the time spanned by the D1 sequence — palynology, paleobotany, vertebrate paleontology, isotopic dating, and magnetostratigraphy.
Maastrichtian pollen has been reported from D1 sequence strata at the Air Force Academy (Kluth and Nelson, 1988). Soister and Tschudy (1978) noted that strata below their Dawson Arkose (mostly D1 sequence) are of Late Cretaceous and Paleocene age. Recent work by Nichols and Fleming (this issue) has identified uppermost Cretaceous and earliest Paleocene pollen sampled from D1 strata in the Castle Pines and Kiowa cores, confirming earlier USGS observations. The Paleocene pollen includes species of Momipites characteristic of palynostratigraphic Zones P1 to P3 of Nichols and Ott (1978).
The plant remains include fossil leaves that can be correlated to fossil floras collected and dated elsewhere in the Western Interior. D1 sequence strata contain Cretaceous and early Paleocene fossil leaf localities (Johnson et al., 2000; Johnson, 2001; Johnson et al., in press). These include the Cretaceous floras excavated near Colorado Springs in Jimmy Camp Creek (Benson, 1998) and the early Paleocene floras excavated at the Denver International Airport, the Plains Conservation Center (Barclay et al., 2001, in press), and at Castle Rock (Ellis and Johnson, 1999; Johnson et al., 2000; Johnson, 2001; Johnson et al., in press).
Dinosaurian remains are abundant in the lower part of D1; they include some of the earliest described dinosaurs (Marsh, 1877) and a partial skeleton of Tyrannosaurus discovered in Littleton in 1992 by Charles Fickle. A variety of other latest Cretaceous dinosaurs has also been recognized from these beds (Carpenter, this issue). Brown (1943) reported the faunal transition from dinosaur-bearing beds into early Tertiary mammal-bearing beds that occurs within the D1 sequence on South Table Mountain. Numerous earliest Paleocene (Puercan) mammal localities have been identified in D1 sequence strata, principally at Corral Bluffs, South Littleton, West Bijou Creek, and South Table Mountain (Middleton, 1983; Eberle, in press).
Several radiometric dates have been obtained from the D1 sequence. During the construction of Denver International Airport, a volcanic ash interbedded within lignite beds was collected and dated at 65.7 ± 0.4 Ma using Ar40/Ar39 (Obradovich, this issue). Lava flows capping North and South Table Mountains have been dated at between 63 and 64 Ma (Mutschler et al., 1987; Obradovich, this issue). Three volcanic ashes found east of Kiowa near the top of the D1 sequence have been dated at 63.62 ± 0.3, 64.13 ± 0.21, and 65.03 ± 0.25 Ma (Obradovich, this issue). A volcanic ash from the top of Green Mountain has been dated at 63.94 ± 0.28 Ma, interpreted to be near the top of the D1 sequence (Obradovich, this issue).
Magnetostratigraphic analyses of the continuous cores from Castle Pines and Kiowa have yielded reversal polarity stratigraphies in D1 sequence strata that can be correlated to the geomagnetic polarity time scale. Hicks et al. (in press) have determined that the uppermost Cretaceous and earliest Tertiary paleomagnetic records are preserved in these cores with chrons 30N through 28N interpreted in the cored material.
Age Constraints on D2 Sequence
The age of the D2 sequence is constrained by palynology, paleobotany, vertebrate paleontology, and isotopic dating.
Fossil pollen in D2 sequence strata indicate an Eocene age (Soister and Tschudy, 1978). This has been corroborated by Nichols and Fleming (this issue), who confirm that basal D2 strata contain rare fossil pollen, Platycarya platycaryoides, interpreted to be earliest Eocene based on correlations to other basins in the Rocky Mountain area. These authors also report finding latest Paleocene pollen at the base of this unit.
Two Eocene leaf localities have been identified in D2 strata by Johnson et al. (in press).
A fragmentary tooth from an Eocene Coryphodon was found in D2 sediments in Elbert County by volunteers from the Denver Museum of Nature and Science (DMNH locality #1999).
A thin, biotite-rich airfall volcanic ash found in the D2 sequence strata by Soister and Tschudy (1978) yields a K/Ar date of 54.3 ± 0.7 Ma. The ash occurs in a low-relief setting, and correlations to the basal D2 unconformity indicate that the ash is near the base of the D2 sequence.
The chronologic controls on the D1 and D2 sequences are summarized in Figure 17. The upper age limit of the D2 sequence is not well defined, so the time span and rate of accumulation of this sequence remain interpretive.
After accumulation of the D2 sequence, the basin witnessed a prolonged period of renewed stability. There is no evidence for subsidence or significant basin formation younger than the late Eocene. During this time interval, the Front Range area was beveled, forming the Rocky Mountain erosion surface.
In the latest Eocene (36.7 ± 0.7 Ma, McIntosh and Chapin, 1994), the Wall Mountain ignimbrite (Epis and Chapin, 1974) was spread catastrophically across the low-relief landscape. This isochronous (possibly single-day) event forms an elegant bookmark in the evolution of the Denver Basin. The present distribution of the welded tuff from this eruption confirms that there has been little or no post-Laramide differential motion on the Front Range fault system on the west margin of the Denver Basin (Leonard and Langford, 1994). Shortly after the eruption of the Wall Mountain ignimbrite sheet, the rock record reveals a series of ponding episodes (e.g., at Florissant) and the catastrophic dispersal of the Castle Rock Conglomerate by a series of breakout flood events (Chapin and Wyckoff, 1969; Morse, 1985). These drainage disruptions and dramatic floods presumably result from the reestablishment of drainage networks on the surface mantled by the ignimbrite (Epis and Chapin, 1975; Scott and Taylor, 1986).
Widespread ash-rich loess deposits mantled the area in the late Eocene–Oligocene (White River Group) and may well have veneered the study area. Today, the best exposures of these strata are in northeastern Colorado along the Pawnee Buttes escarpment.
During the interval that separates the D1 and D2 sequences and again during the late Tertiary, the Front Range piedmont area was a low-relief region with sediment lapping up against basement massifs (Tweto, 1975; Epis and Chapin, 1975; Trimble, 1980; Chapin and Kelley, 1997). An analogous environment is seen today in the Granite Mountains of Wyoming where inselbergs rise above a sediment-filled plain drained by mature meandering streams.
Regional Tilting and Incision
Poorly understood factors have caused regional uplift and eastward tilting of the Rocky Mountain region during the late Miocene, and the process continues today. This epeirogenic uplift is manifest in the stratigraphic record by the eastward dispersal of the Miocene Ogallala Formation. This unit forms a veneer across the entire tilted surface (e.g., Frye et al., 1956), extending from the mountain front toward the Mississippi River. It is characterized by a polymict gravel the composition of which reflects a large and lithologically diverse provenance. Headward erosion and downcutting by the Platte, Arkansas, Canadian, and Pecos Rivers have incised and beheaded most of the primary gravel sources for the Ogallala Formation. Headward recycling thus contributes to the gradual eastward migration of the gravel lag surface. The Ogallala Formation is characterized by low accommodation potential, low accumulation rates, reworking, and long-lived surfaces such as the Llano Estacado, the relict Gangplank of southeastern Wyoming, and the high surfaces in the Black Forest area of the Denver Basin. Due to its generation by epeirogenic tilting rather than asymmetric orogeny, it has low potential for preservation.
Regional uplift of the Rocky Mountain region and the headward erosion and incision of the drainage network have resulted in the dissection of the Laramide basins and exhumation of the Laramide uplifts, perhaps to a grandeur they never exhibited during the Laramide orogeny. The dispersed sediments have accumulated on the margin of the Gulf of Mexico where the recycled remains of the Laramide orogeny comprise a significant part of the Late Tertiary sedimentary record (Xue, 1997).
The age of the onset of the Laramide orogeny is difficult to define. Early faulting involving the upper part of the Pierre Shale has been documented east of the Front Range near Boulder (Davis and Weimer, 1976). Rare, outsized volcanic clasts in the Fox Hills Sandstone together with observations that this unit contains Cretaceous volcanic grains (Kelley, this issue) indicate that an early phase of volcanic activity in an undetermined location presaged uplift of the Front Range. Thickness patterns in the Laramie Formation and the concentration of fluvial sandstone units on the west side of the Denver Basin suggest that an early sag or trough formed prior to advent of the D1 sequence.
The widespread distribution pattern of the basal D1 sequence Arapahoe conglomerate indicates that subsidence at the mountain front was not pronounced enough to restrict the conglomerate facies. The dispersal of this facies over 60 km into the basin suggests it was deposited during a time of low regional accommodation, interpreted to be the onset of significant subsidence in the Denver Basin. The D1 and D2 sequences of the Denver Basin are interpreted to portray the major periods of Laramide deformation of the central and southern Front Range in the Denver area. The two pulses of sedimentation are linked to two pulses of orogenic activity. The first pulse involved sudden onset of relief due to movement on the Golden fault and its allies with a coeval accumulation of synorogenic strata up to 2,000 feet (610 m) thick. Dramatic facies partitioning is reflective of the heterogeneous alluvial, fluvial, and paludal environments. During this uplift episode, a volcanic province developed on the Front Range and mantled the terrain so completely with volcanic rocks that the sedimentary debris which accumulated near Colorado Springs and west of Denver contains almost 100 percent volcanic clasts. During late D1 sequence sedimentation, these volcanic edifices had been stripped away such that the eroded debris was derived entirely from the granitic bedrock. A second pulse of orogenic activity resulted in deposition of over 1,200 feet (366 m) of strata with a more uniform fluvial facies architecture. Other areas in the Rocky Mountains have similar records of multiple Laramide pulses (e.g., Kraus, 1985; Gries, 1990; Bergh and Snoke, 1992).
The interplay of sediment supply and accommodation through time can be used to explain the vertical and lateral facies distribution patterns seen in the Denver Basin. Environments of sediment accumulation such as those found at the eastern foot of the Andes or northern flank of the Alaska Range are proposed as modern analogs for the processes and depositional patterns inferred from the synorogenic rock record in the Denver Basin. Logical and predictable structural patterns and resultant sediment accumulation patterns permit a better understanding of the heterogeneous and variable synorogenic strata preserved in the Denver Basin. Recognition of these patterns permits the prediction of facies distribution in areas with less control and also allows for a more coherent model to be created for the distribution of facies. Recognition of the patterns may have application, as well, to the study of bedrock aquifers in the Denver Basin.
This work has benefited from discussions in the field with Doug Burbank, Chuck Chapin, Bob Crifasi, Bill Dickinson, Kirk Johnson, Glenn Scott, Ned Sterne, and Jon Thorson. Conversations with Ken Carpenter, Tim Cross, Emmett Evanoff, Glenn Graham, Doug Nichols, Peter Robinson, and Robert Weimer have been very valuable. Colleagues involved in the Denver Museum of Nature & Science's Denver Basin Project have been helpful and supportive in many ways. I am particularly grateful to Kirk Johnson, Michele Reynolds, Richard Barclay, and Tim Farnham. Glenn Scott and Robert Weimer kindly reviewed an earlier version of the manuscript and provided many useful suggestions. Insightful reviews by Bob Kirkham and Mark Kirshbaum were extremely helpful. Major funding for this research comes from NSF Grant EAR-9805474 and from the Colorado Water Conservation Board.
- Received August 17, 2001.
- Revision received September 4, 2002.
- Accepted October 17, 2002.