Earth Surface Processes and Landforms
Earth Surf. Process. Landforms 27, 773–787 (2002)
Published online 24 June 2002 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/esp.352
26
Al AND 10 Be DATING OF LATE PLEISTOCENE AND HOLOCENE
FILL TERRACES: A RECORD OF FLUVIAL DEPOSITION AND
INCISION, COLORADO FRONT RANGE
TAYLOR SCHILDGEN,1 DAVID P. DETHIER,1 * PAUL BIERMAN2 AND MARC CAFFEE3†
1 Department of Geosciences, Williams College, Williamstown, MA 01267, USA
Department of Geology and School of Natural Resources, University of Vermont, Burlington, VT, USA
Center for Accelerator Mass Spectrometry, Lawrence Livermore National Laboratory, Livermore, California, USA
2
3
Received 26 June 2000; Revised 9 January 2002; Accepted 17 January 2002
ABSTRACT
26
10
14
Cosmogenic Al, Be, and C dating of fluvial fill terraces in steep canyons of the Colorado Front Range provides
a temporal framework for analysing episodic aggradation and incision. Results from Boulder Canyon show that terrace
heights above the modern channel (grade) can be divided into: (1) Bull Lake (100 ka; 20–15 m above grade); (2)
Pinedale (32–10 ka; 15–4 m above grade); and (3) Holocene age (<4 m above grade). No pre-Bull Lake deposits are
preserved along Boulder Canyon, and only three small remnants >15 m above grade record Bull Lake deposition. Wellpreserved terraces of Pinedale age suggest that the range of terrace height above grade reflects short-term fluctuations
in the river profile during periods of rapidly changing stream load and power. Net river incision apparently occurred
during transitions to interglacial periods. Soil development and stratigraphic position, along with limited cosmogenic
and 14 C dating, suggest that ¾130 ka terraces in Boulder Canyon correlate with the Louviers Alluvium, and that 32 to
10 ka fills in the canyon correlate with the Broadway Alluvium on the adjacent High Plains. Late Pleistocene incision
rates (¾0Ð15 m ka1 ) along Boulder Canyon exceed pre-late Pleistocene incision rates, and are higher than middle to
late Pleistocene incision rates (¾0Ð04 m ka1 ) on the High Plains. This study provides an example of how modern
geochronologic techniques allow us to understand better rivers that drain glaciated catchments. Copyright  2002 John
Wiley & Sons, Ltd.
KEY WORDS: cosmogenic nuclides; fill terraces; incision rates; Boulder Canyon
INTRODUCTION
Fluvial fill terraces in Boulder Canyon, Colorado Front Range, record late Quaternary downcutting, punctuated
by periods of aggradation that likely reflect glaciation of the upper reaches of Boulder Creek (Madole, 1991).
Terraces thus register perturbations of the catchment sediment budget in a system driven by incision through
bedrock since at least Pliocene time (Epis et al., 1980; Scott and Taylor, 1986; Gregory and Chase, 1994).
Heights of fill terraces above grade reflect temporal changes in base level and the catchment sediment budget,
as well as evolution of the longitudinal profile of Middle Boulder Creek. In the Boulder Creek catchment and
in other large Front Range catchments, millennial-scale changes in sediment budgets likely record the effects
of advancing and retreating glaciers and other major shifts in climate (Madole, 1991) whereas decadal-scale
changes may reflect forest fires and large storms (Schildgen and Dethier, 2000; Kirchner et al., 2001).
Previous studies of fluvial terraces in Boulder Canyon analysed stratigraphy, soil development, and clast
weathering to derive qualitative age estimates (Netoff, 1977; Barber, 1983). Sparse organic matter for 14 C
dating has limited the use of terraces as a tool for studying river dynamics over time. However, recent
advances in the use of in situ produced cosmogenic nuclides, such as 26 Al and 10 Be, provide a basis for
estimating exposure ages of terraces and other fluvial landforms (e.g. Burbank et al., 1996; Chadwick et al.,
* Correspondence to: D. P. Dethier, Department of Geosciences, Williams College, Williamstown, MA 01267, USA.
E-mail: [email protected]
† Present address: PRIME Laboratory, Physics Department, Purdue University West Lafayette, IN, USA.
Copyright  2002 John Wiley & Sons, Ltd.
774
T. SCHILDGEN ET AL.
1997; Phillips et al., 1997). These nuclides allow us to: (1) correlate downstream terraces with upstream
moraines; (2) calculate incision rates; and (3) provide calibration for other age estimates based primarily on
rock weathering and soil parameters.
The nature of the change in climate or rock–uplift rates that initiated incision of the Colorado Front Range
remains controversial after a century of geologic discussion (e.g. Davis, 1911; Epis et al., 1980; Jacob and
Albertus, 1985; Leonard and Langford, 1994; Gregory and Chase 1994; Chapin and Kelly, 1997). According
to Chapin and Kelly (1997), fission track dating of the Pike’s Peak batholith in the southern Front Range
argues against significant post-Laramide (75–60 Ma) uplift. Rather, they argue, the Front Range was protected
from incision prior to Pliocene time by extensive alluvial aprons that extended over the High Plains.
We investigated the age of fill terraces and incision in a small Front Range catchment to add evidence to the
discussion about processes that drove canyon cutting in the Front Range. Our field observations indicate that,
in Boulder Canyon, downcutting incised bedrock. In some places, fluvial deposits may have been eroded,
particularly between about 15 ka and the present. Dated terraces and correlation with the broad alluvial
surfaces of the western High Plains (Madole, 1991) allow us to study the evolution of Middle Boulder
Creek’s longitudinal profile, as well as the detailed history of Front Range incision. Incision rate estimates
derived from terrace ages allow us to explore the correlation between climatic events and the initiation
of incision and to comment on the origin (Merritts et al., 1994; Pazzaglia et al., 1998) and persistence of
convex-upward longitudinal profiles or ‘knickpoints’.
Formation of fluvial fill terraces
Fill terraces form by valley aggradation of alluvial material above a channel, followed by channel incision
through the alluvium. Such terraces often record aggradation in response to an increased sediment load
followed by incision when the sediment load is reduced (Merritts et al., 1994; Pazzaglia et al., 1998). In
glaciated regions of the Rockies, widespread exposure of unconsolidated sediment during and immediately
after glacial retreat led to high sediment loads and aggradation for some distance downstream (Madole et al.,
1998). When the sediment supply diminished, streams cut down through the alluvium, stranding terraces, a
process originally discussed by Huntington (1907) and described in detail by Church and Slaymaker (1989).
However, subsequent erosion generally prevented terraces from providing a complete, unambiguous record of
geomorphic response to a glacial episode. Coarse glacial outwash that lies beneath terraces is unstable and is
preserved best where it is boulder rich or covered with coarse material such as colluvium. In a downcutting
river system, stranded fluvial deposits can be preserved near eroding valley walls, where overlying colluvial
deposits provide protection. In such settings, fill terraces may record a more complete history of repeated
upstream events than does source area evidence (moraines), which may be removed by subsequent glacial
advances.
Late Pleistocene chronology
The pre-Bull Lake, Bull Lake, and Pinedale glaciations are not well dated in the Front Range, but regional
correlations suggest probable ages. The age of pre-Bull Lake deposits is least-well constrained, with a possible
range between 750 and 230 ka supported by (1) 36 Cl and 10 Be ages of moraines in the Wind River Range,
Wyoming (Chadwick et al., 1997), and (2) 36 Cl and 10 Be in boulders on moraines in the same region (Phillips
et al., 1997). Bull Lake advance(s) and retreat(s) probably occurred between 150 and 100 ka at sites in
Wyoming, southwest Montana, and Colorado. Age control is provided by: (1) 36 Cl and 10 Be dating of boulders
in Wind River Range in Wyoming (Phillips et al., 1997); (2) obsidian rind thicknesses near West Yellowstone
(Pierce et al., 1976); (3) a uranium-trend age in the North St Vrain drainage basin in the Colorado Front
Range (Shroba et al., 1983); and (4) 26 Al and 10 Be ages of moraines in Boulder Canyon, Colorado Front
Range (Dethier et al., 2000).
In contrast, the age of Pinedale events is relatively well known. Data from Colorado, southwest Montana,
and Wyoming suggest glaciation between ¾35 and 15 ka with most glacial maxima between 23 and 16 ka,
based on cosmogenic nuclide exposure ages and calibrated 14 C ages. We have changed uncalibrated 14 C
ages in the literature to calibrated years before present (cal BP) using the CALIB v4Ð2 program (Stuiver and
Reimer, 1993) to compare more easily 14 C with cosmogenic exposure ages. We report cosmogenic exposure
Copyright  2002 John Wiley & Sons, Ltd.
Earth Surf. Process. Landforms 27, 773–787 (2002)
775
DATING PLEISTOCENE AND HOLOCENE FILL TERRACES
Table I. Height, age, and correlation of High Plains alluvial surfaces
Alluvium
Nussbaum
Rocky Flats
Verdos
Slocum
Height
above grade
(m)1
Local
height2
(m)
Correlative
glaciation3
Age
estimate
Basis for
age
estimate
>1Ð35 Ma
1Ð35 Ma
640 ka
190 š 50 ka
stratigraphy
soil development
Lava Creek B ash
U-series bone
129 š 10 ka
86 š 6 ka
<38 ka
<10 ka
U-series bone
U-series bone
14
C
14
C
¾140
100 š 10
70 š 5
6–30
20–25
7
pre-Bull Lake
pre-Bull Lake
pre-Bull Lake
Bull Lake
Louviers
<10
4–7
Bull Lake
Broadway
Piney Creek
s<9
<5
4
<4
Pinedale
Holocene
Reference
Birkeland et al. (1999)
Hunt (1954), Scott (1962)
Scott and Lindvall (1970),
Szabo (1980)
Szabo (1980)
Machette (1977)
1: Height above grade from Madole (1991) and Scott (1960, 1962).
2: Local height above grade from Boulder Quadrangle geology, Wruck et al. (1967).
3: From Birkeland et al. (1999); correlation derived from glacial chronology of the Front Range.
and other radiometric ages in ka. Regional limiting dates are derived from: (1) 14 C dating of sediments
from glacial Lake Devlin, Colorado Front Range (Madole, 1986); (2) 14 C dating of sediment from the upper
valley of the Colorado River on the west side of the Colorado Front Range (Madole, 1976, 1980); (3)
obsidian hydration ages in West Yellowstone (Pierce et al., 1976); (4) 10 Be dating of boulders in the Wind
River Range, Wyoming (Gosse et al., 1995); and (5) 26 Al and 10 Be dating of boulders near Boulder Canyon,
Colorado Front Range (Schildgen and Dethier, 2000). We can limit Pinedale ice retreat in the Front Range to
before 14 940–13 810 cal BP, based on late glacial ‘Santanta Peak’ readvances high in Front Range cirques
(Benedict, 1973, 1981; Menounos and Reasoner, 1997).
Several alluvial surfaces on the High Plains east of the Front Range have been correlated with Pleistocene
glaciations or changed climate (Scott, 1960, 1962; Madole, 1991; Birkeland et al., 1999) and may correlate
with terrace surfaces within Boulder Canyon. The High Plains surfaces include late Pliocene Nussbaum Alluvium, early Pleistocene Rocky Flats Alluvium, middle Pleistocene Verdos Alluvium, latest middle Pleistocene
Slocum and Louviers Alluviums, late Pleistocene Broadway Alluvium, and Holocene age pre-Piney Creek,
Piney Creek, and post-Piney Creek deposits (Table I). Madole (1991) noted that the height of geomorphic
surfaces above grade based on rivers in the High Plains is considerably higher than height ranges in the
Boulder, CO, quadrangle (Wruck et al., 1967). This discrepancy may reflect different incision histories of
drainages that head in glaciated catchments, or may result from miscorrelation of local units.
SETTING
Boulder Canyon is typical of the steep, deeply incised canyons that cut through the gently undulating upland
surface of the Colorado Front Range east of the Continental Divide (Figure 1). The canyon drops 785 m in
25 km, from 2425 m at Barker Reservoir in Nederland to 1640 m west of the city of Boulder (Figure 2).
The average gradient of 0Ð032 steepens in places to 0Ð085 and the canyon is locally as deep as 300 m and
as narrow as 50 m at the valley floor. North Boulder Creek joins Middle Boulder Creek 12Ð5 km west of
Boulder. The total drainage area of the combined branches where Middle Boulder Creek enters the city
of Boulder is 347 km2 , whereas the drainage area of Middle Boulder Creek at Nederland is 93 km2 (US
Geological Survey, unpublished data).
Fluvial terraces, marked by upper strath surfaces 1 to 12 m above the present level of Middle Boulder
Creek, are common in the inner canyon, as are extensive areas of colluvial debris and localized alluvial fans.
Terraces are best preserved upstream of the confluence with North Boulder Creek. The paucity of terraces and
fans downstream may reflect a change in river dynamics (such as stream power), or may result from human
manipulation of deposits during road, railroad, and pedestrian path construction. Exposed bedrock slopes and
Copyright  2002 John Wiley & Sons, Ltd.
Earth Surf. Process. Landforms 27, 773–787 (2002)
776
T. SCHILDGEN ET AL.
EXPLANATION
R.
COLORADO
Bull Lake ice limit
Middle Boulder
Creek catchment
0
1
2 mi
PLAINS
S.
R.
Pl
at
te
do
ora
Col
Pinedale ice limit
DI
VI
BOULDER
DE
0 1 2 km
>3500 m
B
North
Boulder
Creek
Nederland Barker
Middle
Reservoir
Sampling
area
HIGH
CONTINENTAL
Sunnyside
ch
ran
Figure 1. Map showing location of Boulder Canyon and Middle Boulder Creek catchment. Dashed line separates High Plains from
Front Range
cliffs are common on steep valley walls of Boulder Canyon, but a thin cover of colluvial debris with local
bedrock outcrops predominate where slopes are <30° . In the gently rolling upland, granitic bedrock is deeply
weathered, grussified, and readily eroded where exposed.
METHODS
Field
We mapped and sampled fluvial terraces along Middle Boulder Creek in the Tungsten, CO, 7Ð50 quadrangle
and adjoining areas, using hand levels to measure the height of terrace remnants above Middle Boulder Creek.
Errors in terrace height resulting from local variations in river gradient and changes in stage are probably
<1 m. We measured the longitudinal profile using 1 : 24 000 topographic maps (contour interval 12 m) and
surveyed an ¾2Ð2 km profile along the steepest reach using compass, tape, and hand level (Figure 2). Where
terraces were buried by colluvium, we used the highest occurrence of stream-rounded cobbles to approximate
terrace height. To help provide qualitative age estimates, we noted soil development, bedrock weathering, and
stratigraphic relationships with other terraces.
We collected samples for cosmogenic dating from six terraces along Middle Boulder Creek (Figure 2),
including several sites where age had been estimated from the degree of soil development (Table II). In
selecting samples (Figure 3), we chose the largest boulders from a surface to maximize chances that the rock
had been stable since deposition. In the absence of boulders, or where terraces were buried by colluvium,
we sampled cobbles from the B-horizon beneath the terrace, noting the depth and density of cover material.
In addition to terrace samples, we collected one sample from a boulder, the lithology and shape (Figure 3)
of which suggested that it fell into Middle Boulder Creek from a nearby cliff and was sculpted in place
(99-T60-1). We collected another sample from a polished edge of a pothole (99-T26-1). With the exception
Copyright  2002 John Wiley & Sons, Ltd.
Earth Surf. Process. Landforms 27, 773–787 (2002)
777
2.3
43
60
4140
2.1
2.0
1.9
1.8
1.7
1.6
1.5
0
5
10
80
60
40
20
0
0
0.5
1.0
1.5
2.0
2.5
Distance downstream (km)
10
18
72
71
26
Sunnyside
N. Boulder Cr.
Elevation (km)
2.2
4
Exposed bedrock
and thin alluvium
100
Boulder
2.4
64
120
Waterfalls
Nederland
2.5
Waterfalls
Fresh rockfall
Local channel height (m)
DATING PLEISTOCENE AND HOLOCENE FILL TERRACES
15
20
25
30
Distance downstream from Nederland (km)
Figure 2. Profile of Middle Boulder Creek showing sample locations (sites are designated as 99-T and the number plotted). Inset shows
surveyed reach
of 99-T60-1, all samples had smoothed or polished surfaces, indicating minimal erosion since deposition. We
measured the angle to the horizon at each sample site to make geometric corrections for horizon shielding,
sample thickness, and the dip angle of each sample surface for azimuth corrections.
Table II. Description, soil characteristics, and age estimates for sample sites
Sample
Distance east of
Nederland
gauging
station
(km)
Terrace
height
(m)
Profile
thickness
(cm)
99-T11
99-T18
99-T10
13Ð9
14Ð4
13Ð2
21
12
10
nd
203
nd
99-T64
99-T43
99-T45
2Ð6
4Ð1
4Ð5
9
7
7
180+
127
80
99-T60
5Ð9
4Ð3
5Ð4
7Ð7
7Ð6
3Ð9
5
6Ð2
5
4Ð5
2
2Ð5
nd
89
216
87
70
69
99-T41
99-T40
99-T4
Copyright  2002 John Wiley & Sons, Ltd.
Soil characteristics
Maximum
Soil
colour
horizons
nd
10YR 4/4
Oxidized around
boulders
10YR 5/8
10YR 4/6
Oxidized near
surface
10YR 4/3
nd
nd
nd
nd
Age
estimate
Reference
nd
Bt/BC
nd
Bull Lake
Pine/B. Lake
late Pinedale
this study
BC-12 (Netoff, 1977)
this study
A/BC
A/B/BC
nd
Pinedale
late Pinedale
late Pinedale
this study
BC-8 (Netoff, 1977)
this study
A/Bw/BC
A/B/BC
A/B/BC
A/BC
A/C
A/C
Pinedale
Pinedale
Pinedale
early Holocene
Holocene
Holocene
this study
BC-9 (Netoff, 1977)
BC-10 (Netoff, 1977)
this study
this study
this study
Earth Surf. Process. Landforms 27, 773–787 (2002)
778
T. SCHILDGEN ET AL.
a)
d)
e)
b)
f)
c)
Figure 3. Cosmogenic sample sites noted with arrows: (a) 99-T26, collected from polished bedrock above pothole, 6Ð7 m above Middle
Boulder Creek; (b) 99-T43, boulder collected from terrace surface, 7Ð5 m above Middle Boulder Creek; (c) 99-T10-1, cobbles collected
from 65 cm below the surface of 10 m terrace; (d) 99-T71-1, polished top surface of boulder (2 m ð 1Ð5 m ð 0Ð8 m) on 12 m terrace;
(e) 99-T71-2, cobbles collected from 30 cm below surface of 12 m terrace; (f) 99-T60-1, sample collected from beneath boots of author
on boulder (4Ð2 m ð 2Ð8 m ð 2Ð2 m); sample 5Ð0 m above Middle Boulder Creek
Sample preparation and analysis
Samples were processed for 10 Be and 26 Al at the University of Vermont using standard techniques. Rock
samples were wire brushed, crushed, ground, and sieved isolating the 250–850 µm fraction. Sediment samples
were sieved and the sand fraction retained (250–850 µm). For samples 99-T64-40, -60, -80, -100, and -120,
pebbles with diameters <4 cm were crushed and added to the sand fraction so that sufficient quartz was
available for analysis.
Copyright  2002 John Wiley & Sons, Ltd.
Earth Surf. Process. Landforms 27, 773–787 (2002)
DATING PLEISTOCENE AND HOLOCENE FILL TERRACES
779
Heated ultrasonic etching with 6 M HCl, followed by repeated 1% HF and HNO3 etches, was used to purify
29 to 40 g of quartz so that the purified material contained 11 to 79 µg g1 of 27 Al (Kohl and Nishiizumi,
1992). The quartz was dissolved in HF along with 250 µg 9 Be carrier and sufficient 27 Al carrier so that at
least 2000 µg of Al was available for isotopic analysis. Total Al was measured using inductively coupled
argon plasma (ICP) spectrometry optical emission on two aliquots and normalized to four standards run as
unknowns. Average Be yield for all samples was 98Ð1%; average yield for the three Al blanks was 101Ð5%.
Anion and cation chromatography, along with pH-specific precipitation, were used to purify Al and Be, which
were precipitated as hydroxides and burned to produce oxides.
Oxides were mixed with Ag (Al) or Nb (Be), packed into targets, and transported to Livermore National
Laboratory for accelerator mass spectrometric (AMS) analysis. All AMS analyses are referenced to standards
produced by K. Nishiizumi and corrected for ratios measured in blanks processed along with the samples.
Measured isotopic ratios ranged from 1Ð9 ð 1013 to 4Ð8 ð 1012 for 10 Be/9 Be and 3Ð2 ð 1013 to 1Ð6 ð 1011
for 26 Al/27 Al. Blank subtractions were substantial (7 to 13%) only for 10 Be in the least dosed samples (99-T64
series). For the other samples, blank corrections represented at most several percent of the measured ratios.
In this sample set, 26 Al and 10 Be are extremely well correlated (r 2 D 0Ð997) and we use the average of
26
Al and 10 Be ages in our discussion. The 26 Al/10 Be ratio resulting from linear regression is 6Ð02. This ratio
is indistinguishable from the production ratio of 6Ð04 measured by Nishiizumi et al. (1989) and similar to the
average value for our data set of 6Ð16 š 0Ð36. All but one sample (99-T60-1) have nuclide ratios consistent (at
2) with simple exposure history and no extended (×100 ka) burial. Both isotopic data (26 Al/10 Be D 4Ð97)
and geologic data (the sample appears to have fallen from a cliff) indicate that sample 99-T60-1 has a
complex exposure history unrelated to the sample’s current position and exposure geometry; therefore, we
did not consider this sample in our data analysis, nor did we use data from this sample to estimate a terrace age.
Cosmogenic data reduction and assumptions
We used interpretive models to make exposure age estimates from measured nuclide abundances (Lal,
1988). These models assume continuous cosmic ray dosing of sampled boulders, outcrops, and sediment
without burial by snow, without erosion of sampled surfaces since deposition, and without inheritance of
nuclides from prior periods of exposure, such as on cirque headwalls. Violation of these assumptions can
result in model ages that are inaccurate. For example, burial or erosion of boulder surfaces will lower 26 Al
and 10 Be model ages. Nuclide inheritance from periods of prior exposure will result in age overestimates
(Bierman et al., 1998; Colgan et al., submitted).
Nuclide production rates change over time, and with altitude and latitude (Lal, 1991; Clapp and Bierman,
1995). Both integrated production rates and corrections for latitude and altitude remain uncertain to at least
10%, and perhaps as much as 30% (Clark et al., 1995; Dunai, 2000). In this paper, we calculate model ages
using the established 10 Be and 26 Al production rates of Nishiizumi et al. (1989); recent work suggests that
these production rate estimates may be 10 to 20% too high (Clark et al., 1995; Stone et al., 1998); if so, then
our age estimates are too low by a similar percentage. Because production rates and the site-specific corrections
we made are uncertain, we propagate a 20% uncertainty in exposure ages (Bierman, 1994). Although the
model exposure ages of samples may change as production rate estimates are further refined, the relative ages
of samples will not change. We correct for exposure altitude and latitude using Lal (1991), considering only
neutrons. We correct for exposure geometry using the scaling factors in Dunne et al. (1999), accounting for
both horizon geometry and surface dip.
To calculate production rates for buried samples, we assumed that colluvium buried terraces soon after
deposition of the fill and used a neutron attenuation factor of 165 g cm2 to calculate subsurface production
rates, assuming that shielding depth and density ( D 1Ð5 g cm3 ) have not changed over time. This assumption is generally supported by the degree of soil development and its continuity from the colluvium into fluvial
deposits, and by field observations of active processes (i.e. creep and landslides) along the steep, modern valley margins. Only observations at site 99-T10 suggest that the colluvial cover is substantially younger than
the fill. Given this ambiguity, we calculated production rates at this site given the two limiting cases: colluvial
deposition immediately after terrace abandonment, which gives a maximum age estimate (19Ð1 š 4Ð0), and
Copyright  2002 John Wiley & Sons, Ltd.
Earth Surf. Process. Landforms 27, 773–787 (2002)
780
T. SCHILDGEN ET AL.
colluvial deposition just prior to sampling, which gives a minimum age estimate (12Ð6 š 2Ð6). This calculation illustrates that, if the assumption of instantaneous colluvial burial is incorrect, the terrace model ages we
present are overestimates.
BOULDER CANYON TERRACES
Morphology, correlation and weathering
As height above grade increases, preservation of fluvial deposits decreases and soil development increases.
Low terraces (0Ð5 to 3 m above grade) form surfaces up to 50 m wide and hundreds of metres long in the
broadest portions of the valley and are continuous for up to several kilometres in the upper canyon. In the
lower canyon, low terraces are more difficult to identify, since many have been extensively disturbed by
development. Holocene terraces and adjacent fans show minimal soil development (Table II; Netoff, 1977).
Terrace remnants 4 to ¾9 m above grade are also common and are best preserved in the upper canyon.
They have weakly developed B-horizons that suggest latest Pinedale age (Netoff, 1977). Terraces ¾9 to 14 m
above grade are exposed within ¾2 km of Sunnyside (Figure 2) and display better-developed Bt-horizons,
suggesting an early Pinedale or Bull Lake age. Near Sunnyside, we mapped three exposures of terraces 14
to 21 m above grade, each consisting of a layer of stream-transported cobbles in an oxidized, coarse-grained
matrix. Topographic position of these exposures above inset early (?) Pinedale terraces suggests a pre-Pinedale
age. In this region, where high terraces are preserved, low (0Ð5 to 3 m) terraces are uncommon.
We found no terrace remnants >21 m above Middle Boulder Creek, despite a diligent search along the
upper reaches of the valley walls and noted only a few small patches of rounded gravel and cobbles higher
than 15 m above grade. The best-preserved terrace remnants are typically covered by colluvium derived from
adjacent valley walls. Exposed bedrock straths are also rare due to rapid weathering of bedrock surfaces;
accordant remnant toeslopes on valley walls may indicate past river positions >30 m above grade.
Cosmogenic ages
Fluvial fill terraces in Boulder Canyon give cosmogenic exposure age estimates (Table III) that are broadly
consistent but younger than previous age estimates for terraces based on soil development, clast weathering,
and stratigraphic position (Netoff, 1977; Barber, 1983). The single sample from a terrace deposit 16 m above
grade has an age estimate of 130 š 27Ð5 ka, consistent with the 101 š 21 and 122 š 26 ka age estimates of
two Bull Lake moraines near Nederland (Schildgen and Dethier, 2000). Terrace remnants are dominated by
Pinedale cosmogenic exposure ages ranging from 29 to ¾11 ka. Several terraces noted as ‘early Wisconsinan/Bull Lake’ in age, based on soil development and stratigraphic position (Table II; Netoff, 1977), have
cosmogenic age estimates consistent with the Pinedale glaciation. Terraces ¾9 to 14 m above grade have early
Pinedale exposure ages (29 to 16 ka), whereas terraces 4 to ¾9 m above grade have late Pinedale exposure
ages of 14 to 10Ð5 ka. We assume that all samples give reliable exposure ages except for sample 99-T60-1
and the samples at site T64, which are discussed below.
Depth profile
Cosmogenic nuclide activity in the depth profile at site T64 and in a rounded clast on the terrace indicate
that the material we sampled has not been continually exposed for much, if any, of Holocene time. This is
surprising, because the surface of the deposit is 9 m above grade, typical of Pinedale-age terraces. Model
ages show no clear depth relationship (Figure 4); rather, there appears to be slight variation about a mean of
¾3 ka. We consider three potential explanations for the lower than expected nuclide abundance and unusual
depth profile. First, the deposit could have accumulated rapidly during the late Holocene, and variation
in the profile reflects low and somewhat variable nuclide inheritance from prior upstream exposure. This
explanation is unlikely, as construction of an ¾9 m terrace at this wide valley site would imply extensive
sediment transport caused by a Holocene event for which we observed no evidence downstream. Second,
until recently a few metres of fine-grained sediment could have buried the terrace, restricting cosmogenic
nuclide accumulation. This also seems unlikely, as we did not map caps of fine-grained fluvial or aeolian
sediment locally or elsewhere in the field area. Third, the deposit may have resulted from or been heavily
Copyright  2002 John Wiley & Sons, Ltd.
Earth Surf. Process. Landforms 27, 773–787 (2002)
Copyright  2002 John Wiley & Sons, Ltd.
2Ð451
2Ð451
2Ð451
2Ð451
2Ð451
2Ð451
2Ð451
2Ð451
2Ð024
2Ð024
1Ð951
1Ð878
2Ð415
2Ð354
1Ð939
1Ð939
1Ð951
9Ð5
9Ð5
9Ð5
9Ð5
10Ð0
10Ð0
12Ð5
6Ð7
7Ð5
4Ð0
12Ð0
12Ð0
16Ð0
Elevation
(km)
9Ð5
9Ð5
9Ð5
9Ð5
Terrace
height (m)
4424200
4424200
4424200
4424200
4427800
4427800
4428100
4428000
4424600
4425600
4428100
4428100
4428100
4424200
4424200
4424200
4424200
N,
N,
N,
N,
N,
N,
N,
N,
N,
N,
N,
N,
N,
N,
N,
N,
N,
459200
459200
459200
459200
466700
466700
467600
469000
460400
461500
467800
467800
467650
459200
459200
459200
459200
Location
(UTM)
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
0.137 š 0.007
0.103 š 0.004
0.128 š 0.004
ND
0.069 š 0.028
0.108 š 0.004
0.086 š 0.004
0.100 š 0.003
0.166 š 0.005
0.270 š 0.008
0.270 š 0.008
0.287 š 0.008
0.261 š 0.007
0.425 š 0.011
1.343 š 0.040
0.561 š 0.014
0.561 š 0.014
2.34 š 0.08
0.910 š 0.066
0.621 š 0.038
0.811 š 0.097
0.384 š 0.027
0.407 š 0.026
0.641 š 0.039
0.564 š 0.033
0.687 š 0.045
1.032 š 0.053
1.512 š 0.085
1.512 š 0.085
1.774 š 0.092
ND
2.558 š 0.167
6.678 š 0.320
3.477 š 0.167
3.195 š 0.170
14.13 š 0.67
Be measured2 26 Al measured3
(106 atom g1 (106 atom g1 10
Be/26 Al
6.64 š 0.57
6.01 š 0.42
6.32 š 0.79
ND
5.87 š 0.44
5.92 š 0.41
6.53 š 0.48
6.86 š 0.51
6.21 š 0.37
5.59 š 0.35
5.59 š 0.35
6.19 š 0.36
ND
6.01 š 0.42
4.97 š 0.28
6.20 š 0.34
5.69 š 0.33
6.03 š 0.35
10
3.8 š 0.8
2.8 š 0.6
3.5 š 0.7
ND
1.9 š 0.4
2.9 š 0.6
2.3 š 0.5
2.7 š 0.6
4.5 š 0.9
19.9 š 4.0
13.1 š 2.7
15.9 š 3.2
13.8 š 2.8
11.9 š 2.4
39.8 š 8.1
22.4 š 4.5
31.1 š 6.3
133 š 28
Be model
age4 (ka)
10
4.2 š 0.9
2.8 š 0.6
3.6 š 0.8
1.7 š 0.4
1.8 š 0.4
2.8 š 0.6
2.5 š 0.5
3.1 š 0.6
4.6 š 1.0
18.3 š 3.8
12.0 š 2.5
16.2 š 3.4
ND
11.8 š 2.5
32.7 š 6.8
22.9 š 4.8
29.7 š 6.2
139 š 31
Al model
age4 (ka)
26
4.0 š 0.8
2.8 š 0.6
3.5 š 0.7
ND
1.9 š 0.4
2.9 š 0.6
2.4 š 0.5
2.9 š 0.5
4.5 š 0.9
19.1 š 4.0
12.6 š 2.6
16.0 š 3.2
ND
11.8 š 2.4
35.6 š 8.1
22.6 š 4.5
30.4 š 6.3
136 š 28
Average
age5 (ka)
Ł For 99-T64 profile samples, last number is depth below surface in cm; sample 99-T64-r is a clast on the surface. 2: Uncertainty is for counting statistics and carrier (2%) only.
3: Uncertainty is for counting statistics and stable Al (4%) only.
4: Calculated using Nishiizumi et al. (1989) production rate estimates of 10 Be D 6Ð03 atoms g1 year1 ; 26 Al D 36Ð8 atoms g1 year1 ; for T64 series, age is surface exposure
equivalent; 20.
5: Average (weighted) calculated considering uncertainty of each measurement.
99-T64-r
99-T64-0
99-T64-20
99-T64-40
99-T64-40 (rep)
99-T64-60
99-T64-80
99-T64-100
99-T64-120
99-T10-1
99-T10-2
99-T18
99-T26-1
99-T43
99-T60-1
99-T71-1
99-T71-2
99-T72
SampleŁ
Table III. Location and model cosmogenic nuclide ages for terraces along Boulder Canyon, Colorado
DATING PLEISTOCENE AND HOLOCENE FILL TERRACES
781
Earth Surf. Process. Landforms 27, 773–787 (2002)
782
T. SCHILDGEN ET AL.
disturbed by 19th century anthropogenic activities that reworked Holocene-age sediment. The extensive history
of placer and hardrock mining in the region (Lovering and Tweto, 1953) makes this final explanation the
most tenable, despite a lack of field observations to support it. Despite the ambiguous origin of this deposit,
nuclide measurements for this profile and surface clast provide a noteworthy result: low nuclide abundance
suggests minimal nuclide inheritance within this fill deposit, consistent with our results from other fill deposits
throughout Boulder Canyon.
DISCUSSION
Terrace heights and changing base levels
Cosmogenic age estimates for terraces in Boulder Canyon and correlation with surfaces in the adjacent
High Plains allow us to reconstruct late Pleistocene base-level changes along Middle Boulder Creek. There is
no record of alluvium older than ¾130 ka preserved in Boulder Canyon (Figure 5). Based on terrace heights
and cosmogenic age estimates, base level defined by Middle Boulder Creek was ¾20 m higher than today at
130 ka. By ¾20 ka, incision of bedrock dropped base level to between ¾9 and 14 m above present grade.
Before ¾12 ka, terraces formed 4 to ¾9 m above Middle Boulder Creek, and during the past 10 ka the base
level has been as much as 3 m above grade. The range of heights represented by terraces within Boulder
Canyon contrasts markedly with those of High Plains surfaces thought to correlate with Bull Lake and Pinedale
terraces. In the Boulder Quadrangle, where Boulder Canyon emerges on the High Plains, piedmont surfaces
correlated with Bull Lake and Pinedale events are only 7 m and 4 m respectively above grade (Table I),
suggesting downstream convergence of profiles, and slower net incision in the High Plains (Figure 6).
Because terraces of similar age occur at different heights above grade, it is impractical to use only height
to subdivide deposits of a single glacial episode along Middle Boulder Creek. Different terraces at essentially
the same height above grade (¾12 m) and <50 m apart differ in model age by nearly 10 ka, as illustrated by
samples at T71 (¾23–29 ka) and at T18 (¾16 ka). We interpret such data as a record of base-level fluctuation
along Middle Boulder Creek; little net downcutting occurred during most of the Pinedale and probably during
the previous glacial episode. It seems likely that the major stream load and power changes during glacial
periods could have resulted in repeated formation and degradation of fill terraces within ¾10 m of a relatively
constant base level. Late Holocene changes in discharge and local sediment budgets have produced a series
of inset fills along Boulder Canyon that range in height by at least 3 m.
As an approximate check on this hypothesis, we consider the feasibility of constructing multiple fill deposits
within a single glacial episode. A fill 10 m thick, 75 m in average width, and 12Ð5 km in length (distance
between Boulder Dam and the confluence with North Boulder Creek) requires approximately 4Ð7 ð 106 m3 ,
or ¾7 ð 106 Mg of sediment assuming a density of 1Ð5 g cm3 . If the Pinedale sediment yield from the
0
Depth below surface, in cm
Surface clast
20
weighted average
(26Al and 10Be)
40
60
80
100
120
140
0
2
4
6
8
10
Cosmogenic exposure age, in ky
Figure 4. Cosmogenic exposure ages, assuming surface exposure in soil profile at site T64; points record samples from soil profile and
clast on surface. Bars are 1 counting errors
Copyright  2002 John Wiley & Sons, Ltd.
Earth Surf. Process. Landforms 27, 773–787 (2002)
783
DATING PLEISTOCENE AND HOLOCENE FILL TERRACES
1
Bull Lake
late (?)
Pinedale
10
0
early (?)
Pinedale
100
Holocene
Sample age, in ka
1000
5
10
15
20
Terrace height above grade, in m
Figure 5. Terrace age versus height above grade; horizontal bars show height range within glacial periods. Vertical bars show calculated
age (Table III) š20 %
20
?
15
Boulder Canyon
0.1 mkyr -1
?
10
High Plains
-1
5
0.6 mkyr
Height above grade (m)
25
?
0.1
0.02 mkyr-1
mk
-1
yr
0
200
150
100
50
0
Age, in kyr
Figure 6. Rates of river incision for Boulder Canyon and the High Plains based on age and terrace height above grade. Curves show
assumed rapid incision after 190 and 130 kyr, respectively
93 km2 drainage basin of Middle Boulder Creek at Nederland was 0Ð5 Mg ha1 yr1 (¾3 ð background
sedimentation rate; Dethier et al., 2000), a 10 m thick fill could be constructed in 1500 years. Considerable
material would be transported through the canyon during such episodes of aggradation, but it seems likely
that fills 10 m thick could accumulate and erode several times within a 20 000 year glacial period such
as the Pinedale. The age and height of the terraces (Figure 5) suggest that net incision occurs during the
transition to interglacial periods, when available sediment decreases concurrent with continued periods of
peak discharge. The geomorphic result is a series of inset terrace remnants within approximately 10 m of a
base level characteristic for each glacial period (Figure 7).
Copyright  2002 John Wiley & Sons, Ltd.
Earth Surf. Process. Landforms 27, 773–787 (2002)
784
T. SCHILDGEN ET AL.
calibrated 14C age
cosmogenic exposure age
130 ka
Bull Lake fill
terrace
30 ka
colluvial
cover
16 ka
Pinedale fill terraces
11 ka
12 ka
Holocene fill
terrace
3 ka
Bedrock
Middle Boulder
Creek
V.E. ~ 8:1
~20 m
Figure 7. Schematic cross-section showing fill terraces, colluvium and bedrock along Middle Boulder Creek
Long-profile evolution of Middle Boulder Creek
The pattern of preserved terraces reflects late Pleistocene evolution of the long profile of Middle Boulder
Creek and may record long-term persistence or slow migration of a bedrock knickpoint. Differences in
terrace records preserved above and below the extremely steep reach (knickpoint) on Middle Boulder Creek
near Sunnyside (Figure 2) complicate generalizations about terrace height. Bull Lake and early Pinedale
deposits are preserved only near Sunnyside at the base of the knickpoint, and low (Holocene) terraces are
conspicuously absent in this area. Late Pinedale terraces are preserved throughout the canyon (Figure 8).
Major perturbations of the sediment budget may have produced thick fills in the vicinity of the knickpoint.
Downcutting through such fills likely proceeded rapidly, stranding high terraces near Sunnyside as the steep,
bedrock-controlled channel became re-established. Upstream of the knickpoint, shallower bedrock may have
slowed incision and promoted channel meandering, eroding the fill and preserving high terraces downstream.
Alternatively, the sharp meander bend may have served to protect deposits as Middle Boulder Creek cut down.
The Preservation of both Bull Lake and Pinedale terraces, and the absence of Holocene deposits suggest that
only during glacial periods did river dynamics permit deposition on this steep meander of Middle Boulder
Creek.
Disparate late Pleistocene incision rates measured in lower Boulder Canyon and on the adjacent High
Plains (Figure 6) can be extrapolated, with caution, to examine long-term evolution of the canyon. If we
extrapolate late Pleistocene incision rates from lower Boulder Canyon (¾0Ð15 m ka1 ) to 2500 ka (initial
continental glaciation in North America), we calculate 375 m of net incision in Boulder Canyon. This
value exceeds the depth of the modern canyon. Early Pleistocene incision rates thus must be less than
late Pleistocene rates measured in Boulder Canyon near Sunnyside, but greater than rates extrapolated
from High Plains surfaces near Boulder (0Ð04 m ka1 ). Presumably, incision at Sunnyside would have been
rapid as the knickpoint migrated from east to west, and slower prior to knickpoint migration. However,
Copyright  2002 John Wiley & Sons, Ltd.
Earth Surf. Process. Landforms 27, 773–787 (2002)
785
DATING PLEISTOCENE AND HOLOCENE FILL TERRACES
2800
Holocene
l. Pinedale
e. Pinedale
Bull Lake
2800
2600
2400
2400
Moraines
Slocum
Alluvium
2200
Broadway
Alluvium
2000
2200
2000
Middle Boulder
Creek profile
1800
Terrace elevation + offset (m)
River elevation (m)
2600
+ 100 m
+ 200 m
+ 300 m
+ 400 m
1800
High
Plains
1600
0
5
10
15
20
25
1600
30
Distance downstream from Nederland gaging station (km)
Figure 8. Profile of Middle Boulder Creek with Pleistocene and Holocene terraces. For clarity, terraces are plotted at fixed heights (see
legend) above profile. Position of the Slocum Alluvium and the Broadway Alluvium from Wruck et al. (1967)
away from Sunnyside, we have not mapped terraces older than 20 ka, preventing us from testing this
hypothesis.
Relationship between fill terrace deposition and glaciation
Our terrace age estimates are similar to those for the late Pleistocene glaciation of the Colorado Front
Range. The most extensive terraces within Boulder Canyon are <8 m above grade. We can correlate these
terraces with a number of terraces from 7 to 12 m above grade that have exposure ages <14 ka. If we
consider the 14 940–13 810 cal BP estimate of Pinedale ice retreat from the Colorado Front Range based on
late glacial ‘Santanta Peak’ readvances in Front Range high cirques (Benedict, 1973, 1981; Davis, 1982,
1987; Menounos and Reasoner, 1997) and on deglaciation at Lake Emma in the San Juan Mountains (Elias
et al., 1991) to be a limiting age for ice retreat, then most of the terrace deposits preserved within Boulder
Canyon would have formed during or soon after the retreat of Pinedale ice. This suggests that a peak in
terrace formation followed Pinedale ice retreat from the Boulder Creek drainage. The two samples at site
T71 at 12 m both have 22 ka model exposure ages, and thus may represent a minor retreat or a significant
period of melting within the Pinedale glaciation. The single sample dated from a Bull Lake age terrace
(130 š 27Ð5 ka), though imprecise, can correspond with a similar scenario, as the age is within the estimated range of Bull Lake ice retreat (¾100–150 ka). Fill terraces in Boulder Canyon probably formed at
times of high water and sediment availability, i.e. when ice retreat exposed widespread unconsolidated sediment and provided considerable meltwater for sediment transport. Uncertainties in cosmogenic production
rates prevent us from making isotopic revisions to the late glacial chronology of the local area. However,
our results show that fluvial fill terraces dated with cosmogenic nuclides may be used to limit the timing of major glacial retreats, provide information concerning minor retreats within a glacial episode, and
provide a quantitative context within which to explore long-term changes in fluvial dynamics and canyon
evolution.
Copyright  2002 John Wiley & Sons, Ltd.
Earth Surf. Process. Landforms 27, 773–787 (2002)
786
T. SCHILDGEN ET AL.
ACKNOWLEDGEMENTS
We acknowledge J. Larson and B. Copans for help in sample processing, we thank P. Birkeland for many
helpful suggestions in the field, as well as help with charcoal collection, and we thank A. Matmon, K. Nichols,
and an anonymous reviewer for reviews.
REFERENCES
Barber LB. 1983. Correlation of river terrace deposits along Boulder Creek east slope, Colorado Front Range. Unpublished manuscript,
University of Colorado.
Benedict JB. 1973. Chronology of cirque glaciation, Colorado Front Range. Quaternary Research 3: 584–599.
Benedict JB. 1981. The Fourth of July Valley – glacial geology and archeology of the timberline ecotone. Ward, Colorado, Center for
Mountain Archeology Research Report 2.
Bierman PR. 1994. Using in situ produced cosmogenic isotopes to estimate rates of landscape evolution: a review from the geomorphic
perspective. Journal of Geophysical Research 99(B7): 13 885–13 896.
Bierman PR, Davis PT, Marsella K, Colgan P, Mickelson DM, Larsen P, Caffee M. 1998. What do glaciers take away? What do they
leave behind? Geological Society of America Abstracts with Programs 30(7): A–299.
Birkeland PW, Miller DC, Patterson PE, Price AB, Shroba RR. 1999. Soil–geomorphic relationships near Rocky Flats, Boulder and
Golden, Colorado area, with a stop at the Pre-Fountain Formation paleosol of Wahlstrom (1948). Geological Society of America
Field Trip No. 18.
Burbank DW, Leland J, Fielding E, Anderson RS, Brozovic N, Reid MR, Duncan C. 1996. Bedrock incision, rock uplift and threshold
hillslopes in the northwestern Himalayas. Nature 379: 505–510.
Chadwick OA, Hall RD, Phillips FM. 1997. Chronology of Pleistocene glacial advances in the central Rocky Mountains. Geological
Society of America Bulletin 109(11): 1443–1452.
Chapin CE, Kelley SA. 1997. The Rocky Mountain erosion surface in the Front Range of Colorado. In Geologic History of the Front
Range, Bolyard DW, Sonnenberg SA (eds). 1997 Rocky Mountain Section – AAPG Field Trip No. 7; 101–113.
Church M, Slaymaker O. 1989. Disequilibrium of Holocene sediment yield in glaciated British Columbia. Nature 337: 452–454.
Clapp E, Bierman P. 1995. First geomagnetic-based, in situ produced cosmogenic isotope calibration program. Geological Society of
America Abstracts with Programs 27: A–59.
Clark D, Bierman PR, Larsen P. 1995. Improving in situ cosmogenic chronometers. Quaternary Research 44: 366–376.
Colgan PM, Mickelson DM, Bierman PR. Submitted. Deglacial timing, cosmogenic 10 Be and 26 Al evidence, south central Wisconsin,
USA. Geological Society of America Bulletin.
Davis PT. 1982. Chronology of Holocene glaciation, Arapaho cirque, Colorado Front Range. In Abstracts for the XI Congress of the
International Union for Quaternary Research, vol. II, Moscow; 54.
Davis PT. 1987. Late Pleistocene age for type Triple Lakes moraines, Arapaho cirque, Colorado Front Range. Geological Society of
America Abstracts with Programs 19(5): 270.
Davis WM. 1911. The Colorado Front Range, a study on physiographic presentation. Association of American Geographers Annals 11:
21–83.
Dethier DP, Schildgen TF, Bierman P, Caffee M. 2000. The cosmogenic isotope record of late Pleistocene incision, Boulder Canyon,
Colorado. Geological Society of America Abstracts with Programs 32(7): 473.
Dunai TJ. 2000. Scaling factors for production rates of in situ produced cosmogenic nuclides: a critical reevaluation. Earth and Planetary
Science Letters 176: 157–169.
Dunne J, Elmore D, Muzikar P. 1999. Scaling factors for the rates of production of cosmogenic nuclides for geometric shielding and
attenuation at depth on sloped surfaces. Geomorphology 27: 3–11.
Elias SA, Carrara PE, Toolin LJ, Jull JT. 1991. Revised age of deglaciation of Lake Emma based on new radiocarbon and macrofossil
analyses. Quaternary Research 36: 307–321.
Epis RC, Scott GR, Taylor RB, Chapin CE. 1980. Summary of Cenozoic geomorphic volcanic and tectonic features of central Colorado
and adjoining areas. In Colorado Geology, Kent HC, Porter KW (eds). Rocky Mountain Association of Geologists: Denver, CO;
135–156.
Gosse JC, Evenson EB, Klein J, Lawn B, Middleton R. 1995. Precise cosmogenic 10 Be measurements in western North
America – support for a global Younger Dryas cooling event. Geology 23: 877–880.
Gregory KM, Chase CG. 1994. Tectonic and climatic significance of a late Eocene low relief, high-level geomorphic surface, Colorado.
Journal of Geophysical Research 99(B10): 20 141–20 160.
Hunt CB. 1954. Pleistocene and recent deposits in the Denver area, Colorado. US Geological Survey Bulletin 996-C; 91–140.
Huntington E. 1907. Some characteristics of the glacial period in non-glaciated regions. Geological Society of America Bulletin 18:
351–388.
Jacob AE, Albertus RG. 1985. Thrusting, petroleum seeps, and seismic exploration, Front Range south of Denver, Colorado. In Rocky
Mountain Field Trip Guide – 1985 , Macke DL, Maughan EK (eds). Rocky Mountain Section – DEPM: Denver; 77–96.
Kirchner JW, Finkel RC, Riebe CS, Granger DE, Clayton JL, King JG, Megahan W. 2001. Mountain erosion over 10 yr, 10 k.y., and
10 m.y. time scales. Geology 29: 591–594.
Kohl CP, Nishiizumi K. 1992. Chemical isolation of quartz for measurement of in-situ-produced cosmogenic nuclides. Geochimica et
Cosmochimica Acta 56: 3583–3587.
Lal D. 1988. In situ-produced cosmogenic isotopes in terrestrial rocks. Annual Reviews of Earth and Planetary Science 16: 355–388.
Lal D. 1991. Cosmic ray labeling of erosion surfaces, in situ production rates and erosion models. Earth and Planetary Science Letters
104: 424–439.
Leonard EM, Langford RP. 1994. Post-Laramide deformation along the eastern margin of the Colorado Front Range – a case against
significant deformation. The Mountain Geologist 31: 45–52.
Copyright  2002 John Wiley & Sons, Ltd.
Earth Surf. Process. Landforms 27, 773–787 (2002)
DATING PLEISTOCENE AND HOLOCENE FILL TERRACES
787
Lovering TS, Tweto O. 1953. Geology and ore deposits of the Boulder County Tungsten District Colorado. US Geological Survey
Professional Paper 245.
Machette MN. 1977. Geologic Map of the Lafayette Quadrangle, Adams, Boulder, and Jefferson Counties, Colorado. United States
Geological Survey Geologic Quadrangle Map GQ-1392, scale 1 : 24 000.
Madole RF. 1976. Bog stratigraphy, radiocarbon dates, and Pinedale to Holocene glacial history in the Front Range, Colorado. US
Geological Survey Journal of Research 4(2): 163–169.
Madole RF. 1980. Time of Pinedale deglaciation in north-central Colorado: further considerations. Geology 8: 118–122.
Madole RF. 1986. Lake Devlin and Pinedale glacial history, Colorado Front Range, Colorado. Quaternary Research 25: 43–54.
Madole RF. 1991. Colorado Piedmont, Section in: Quaternary geology of the Northern Great Plains. In Quaternary Nonglacial Geology:
Conterminous US , Morrison RB (ed.). The Geology of North America, vol. K-2. Geological Society of America: Boulder, CO;
441–476.
Madole RF, VanSistine DP, Michael JA. 1998. Pleistocene glaciation in the upper Platte River drainage basin, Colorado. US Geological
Survey Geologic Investigations Series I-2644.
Menounos B, Reasoner MA. 1997. Evidence for cirque glaciation in the Colorado Front Range during the Younger Dryas Chronozone.
Quaternary Research 48: 38–47.
Merritts DJ, Vincent KR, Wohl EE. 1994. Long-river profiles, tectonism, and eustasy: a guide to interpreting fluvial terraces. Journal
of Geophysical Research 99(B7): 14 031–14 050.
Netoff DI. 1977. Soil clay mineralogy of Quaternary deposits in two Front Range–Piedmont transects, Colorado. PhD Thesis, University
of Colorado.
Nishiizumi K, Winterer EL, Kohl CP, Klein J, Middleton R, Lal D, Arnold JR. 1989. Cosmic ray production rates of 10 Be and 26 Al in
quartz from glacially polished rocks. Journal of Geophysical Research 94: 17 907–17 915.
Pazzaglia FJ, Gardner TW, Merritts DJ. 1998. Bedrock fluvial incision and longitudinal profile development over geologic time scales
determined by fluvial terraces. In Rivers Over Rock: Fluvial Processes in Bedrock Channels. Geophysical Monograph 107. American
Geophysical Union: 207–235.
Phillips FM, Zreda MG, Gosse JC, Klein J, Evenson EB, Hall RD, Chadwick OA, Pankaj S. 1997. Cosmogenic 36-Cl and 10-Be ages
of Quaternary glacial and fluvial deposits of the Wind River Range, Wyoming. Geological Society of America Bulletin 109(11):
1453–1463.
Pierce KL, Obradovich JD, Friedman I. 1976. Obsidian hydration dating and correlation of Bull Lake and Pinedale glaciations near
West Yellowstone, Montana. Geological Society of America Bulletin 87: 703–710.
Schildgen TF, Dethier DP. 2000. Fire and ice: using isotopic dating techniques to interpret the geomorphic history of Middle Boulder
Creek, Colorado. Geological Society of America Abstracts with Programs 32(7): 18.
Scott GR. 1960. Subdivision of the Quaternary alluvium east of the Front Range near Denver, Colorado. Geological Society of America
Bulletin 71: 1541–1543.
Scott GR. 1962. Geology of the Littleton Quadrangle, Jefferson, Douglas, and Arapahoe Counties, Colorado. United States Geological
Survey Bulletin, 1121-L.
Scott GR, Lindvall RM. 1970. Geology of new occurrences of Pleistocene bisons and peccaries in Colorado. US Geological Survey
Professional Paper 700B; B141–B149.
Scott GR, Taylor RB. 1986. Map showing late Eocene erosion surface, Oligocene–Miocene paleovalleys and Tertiary deposits in the
Pueblo, Denver, and Greeley 1 ° ð 2 ° quadrangles, Colorado. United States Geological Survey Miscellaneous Investigations Series
Map I-1626, scale 1 : 250 000.
Shroba RR, Rosholt JN, Madole RF. 1983. Uranium-trend dating and soil B horizon properties of till of Bull Lake age, North St. Vrain
drainage basin, Front Range, Colorado. Geological Society of America Abstracts with Programs 15(5): 431.
Stone JOH, Evans JM, Fifield LK, Allan GL, Cresswell RG. 1998. Cosmogenic chlorine-36 production in calcite by muons. Geochimica
et Cosmochimica Acta 62(3): 433–454.
Stuiver M, Reimer PJ. 1993. Extended 14 C database and revised CALIB radiocarbon calibration program. Radiocarbon 35: 215–230.
Szabo BJ. 1980. Results and assessment of uranium-series dating of vertebrate fossils from Quaternary alluviums in Colorado. Arctic
and Alpine Research 12(1): 95–100.
Wruck CT, Wilson RF, Bucknam RC, Jones LR, Mendes P, Gardner ME. 1967. Geology of the Boulder Quadrangle, Colorado. US
Geological Survey, unpublished map, scale 1 : 24 000.
Copyright  2002 John Wiley & Sons, Ltd.
Earth Surf. Process. Landforms 27, 773–787 (2002)