Professional Documents
Culture Documents
10 1130@G22681 1 PDF
10 1130@G22681 1 PDF
䉷 2006 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or editing@geosociety.org.
Geology; October 2006; v. 34; no. 10; p. 841–844; doi: 10.1130/G22681.1; 3 figures; Data Repository item 2006182. 841
ern U.S. and highlight the importance of local notion that scatter among boulder-exposure
moisture sources on glacier mass balance near ages is caused by geologic phenomena, name-
the end of the last glaciation. ly partial shielding of boulder surfaces during
a period of moraine stabilization, along with
METHODS analytical errors. The uncertainties associated
Cosmogenic Surface-Exposure Dating with the weighted-mean boulder-exposure age
To set limits on the timing of the LGM in estimates incorporate analytical uncertainties
the southwestern Uinta Mountains, we target- of each individual age.
ed the most prominent lateral and terminal
moraines from the last glaciation, located at LGM Glacier ELA Reconstructions
the mouths of the Lake Fork and Yellowstone ELAs were estimated for all reconstructed
Canyons (Fig. 1). We sampled 7 boulders on LGM valley glaciers in the Uintas during the
the outermost lateral moraine from the last LGM (n ⫽ 44) using a weighted average of
glaciation in Yellowstone Canyon and 14 individual estimates obtained from three sep-
boulders on a terminal moraine in Lake Fork arate ELA estimation techniques: accumula-
Canyon. The latter is a compound feature; tion area ratio (AAR ⫽ 0.65), toe-headwall
seven samples are from the distal ridge (la- altitude ratio (THAR ⫽ 0.40), and the upper-
beled LF1 in Fig. 1) and seven samples are most elevation of continuous lateral moraines
from the proximal ridge (labeled LF2 in Fig. (LM) after Meierding (1982). Weighting fac-
1). The Lake Fork terminal moraine complex tors were AAR ⫽ 3, THAR ⫽ 2, and LM ⫽
was targeted to determine the time interval be- 1, following Locke (1990) and Munroe and
tween the formation of the two ridges. Suit- Mickelson (2002). The LGM extent of valley
able boulders of metaquartzite or weakly glaciers was reconstructed for the northern
Figure 1. A: Map of glacial localities in metamorphosed sandstone were sampled with Uintas by Munroe (2001, 2005) and for the
central and northern Rocky Mountains a sledgehammer and chisel. Laboratory pro- southern Uintas by Shakun (2003). Paleogla-
south of the Cordilleran and Laurentide Ice cedures used to isolate beryllium in boulder
Sheets (margins shown schematically along cier extents in the western Uintas were taken
samples were adopted from Bierman et al. from Oviatt (1994).
northern part of map) recently dated by cos-
mogenic surface-exposure or radiocarbon (2003) and were completed at the University
methods. Dark gray polygons are locations of Wisconsin Cosmogenic Nuclide Prepara-
of Last Glacial Maximum glacier systems in
RESULTS
tion Lab (laboratory methods are summarized
Wallowa Mountains (WM; Licciardi et al., Cosmogenic Surface-Exposure Dating
in the GSA Data Repository1). Concentrations
2004), Sawtooth Mountains (SM; Thackray et Seven boulders on the Yellowstone Canyon
al., 2004), Yellowstone Plateau (YP; Licciardi
of 10Be were measured at the Purdue Rare Iso-
moraine (Table DR1; see footnote 1) and 16
et al., 2001), Wind River Mountains (WR; tope Measurement Lab by accelerator mass
boulders on the Lake Fork moraines (Table
Benson et al., 2004), north-central Colorado spectrometry (AMS).
Rockies (NCCR; Benson et al., 2005), south- DR2) yield in situ cosmogenic 10Be exposure
Boulder-exposure ages were calculated
western Colorado Rockies (SWCR; Benson ages that range from 11.5 ⫾ 1.2 to 19.9 ⫾ 2.0
from measured 10Be concentrations using the
et al., 2005), and Uinta Mountains (U; this ka (Fig. 2). The exposure ages from the Yel-
study). Mean 10Be boulder-exposure ages age equation from Lal (1991) and a produc-
lowstone Canyon moraine span this relatively
from Benson et al. (2005) are recalculated tion rate of 5.1 ⫾ 0.3 (2) atoms g SiO2⫺1
broad range, but all of the ages overlap at 2
using production rates employed in this yr⫺1 scaled for site elevation and geographic
study and in Licciardi et al. (2001, 2004). Box (Fig. 2; Table DR1). A long interval of
latitude following Stone (2000). We consider
indicates location of Lake Fork and Yellow- moraine-crest stabilization and/or subsequent
stone River Canyons. B: Shaded-relief map
the effects of topographic shielding, boulder
boulder exhumation are possible explanations
of Lake Fork (LF) Canyon and Yellowstone erosion, and snow cover to be minimal at all
sample locations (age calculations and mo- for the younger ages on this moraine, although
(YS) Canyon terminal moraines. Long-
dashed lines indicate Pinedale-equivalent raine boulder properties are described in Ta- our observations of moraine morphology
moraines and short-dashed lines indicate bles DR1 and DR2). Although previous stud- failed to identify evidence of variable
pre-Pinedale moraines. Circles mark loca- moraine-crest erosion rates among the sample
tions of sampled boulders. ies suggest that uncertainties of scaling
production rates are as much as 20% (e.g., sites. Nonetheless, the scatter of boulder-
Bierman et al., 1999), we report only the an- exposure ages on the Yellowstone moraine
the largest pluvial lake in the Great Basin alytical uncertainty of the AMS measurements precludes a precise determination of the time
(Gilbert, 1890; Oviatt, 1997), allowing the in our age estimates. The mean of acceptable of moraine abandonment. Thus, we consider
possible significance of a local moisture boulder-exposure ages (⫾2) from each mo- the weighted-mean age of the seven boulders,
source to be evaluated. raine, weighted according to the inverse- 15.2 ⫾ 2.6 ka, to be a working estimate for
Here we report the first cosmogenic variance of measurement precision, was com- the age of the Yellowstone moraine until more
surface-exposure ages that delimit the timing puted for individual moraines in the data are available, and base our following in-
of the local LGM in northeastern Utah. We Yellowstone and Lake Fork canyons. This terpretations of the deglacial chronology for
also present reconstructed equilibrium line al- method of age interpretation is based on the the southwestern Uintas on the cosmogenic
titudes (ELAs) for Uinta Mountain glaciers surface-exposure ages yielded by boulders on
during the LGM, and then discuss the climatic 1GSA Data Repository item 2006182, Tables the Lake Fork moraines (Tables DR1 and
and temporal relationship between the mass DR1 and DR2, cosmogenic methods, data analysis, DR2).
balance of these glaciers and the hydrologic and expanded results, is available online at www. In the Lake Fork Canyon, a single sample
geosociety.org/pubs/ft2006.htm, or on request from
balance of Lake Bonneville. Our results add editing@geosociety.org or Documents Secretary,
that returned an exposure age of 11.5 ⫾ 1.2
to the current understanding of the latest Pleis- GSA, P.O. Box 9140, Boulder, CO 80301-9140, ka on ridge LF1 is a statistical outlier at 2
tocene behavior of alpine glaciers in the west- USA. compared to six other samples (Fig. 2; Table
DR2 [see footnote 1]), and is therefore omit- 18.7–20.8 ka for the Wind River range (as- at the highest ELAs, and western Uinta gla-
ted from further calculations. The six oldest suming no snow cover, and with 10Be ages for ciers must have received more precipitation
boulder-exposure ages from LF1 are indistin- the Wind River range adjusted to the produc- than those farther east in order to reach their
guishable from one another and have a tion rate used here). However, the exposure maximum positions. Given the sensitivity of
weighted mean of 18.0 ⫾ 1.1 ka. The mor- ages determined for the Uinta moraines are ELAs to precipitation and temperature (Selt-
phology of the Lake Fork terminal moraine consistent with those from the western Wa- zer, 1994), and assuming the modern summer
system requires that the glacier retreated after satch Mountains (Lips et al., 2005) that indi- lapse rate, the ELA depression requires that
ca. 18 ka to form ridge LF2. The seven cate a glacial maximum as late as 17–15 ka. the western Uinta glaciers received at least
boulder-exposure ages from this inner ridge The Wasatch dates are preliminary, but their 1000 mm more winter accumulation than gla-
are also indistinguishable from one another overlap with the Uinta results reported here ciers at the eastern end of the range, where
with a weighted mean of 16.8 ⫾ 0.7 ka. Be- suggests that the LGM and subsequent retreat ELAs were higher and ablation season tem-
cause LF2 is the innermost ridge in the Lake in these two ranges roughly coincided with the peratures were presumably cooler (Munroe
Fork terminal moraine system (Fig. 1) and hydrologic peak of Lake Bonneville and its and Mickelson, 2002). This difference exceeds
yields a robust set of moraine-boulder expo- subsequent climate-driven fall below the Pro- the modern orographic effect in the range,
sure ages (Fig. 2; Table DR2), we consider vo shoreline after ca. 14.0 14C ka (16.2–17.1 which is on the order of a few hundred mil-
this age to best represent the onset of degla- ka) (Oviatt, 1997). Previous studies suggest limeters of water-equivalent precipitation.
ciation in the southwestern Uinta Mountains. that northward migration of the polar jet Thus, the lower ELAs in the western Uintas
stream during the latest Pleistocene in re- reflect substantial enhancement of the modern
ELA Reconstructions sponse to the orographic collapse of the Lau- orographic precipitation pattern.
The weighted means of the reconstructed rentide Ice Sheet altered the climate of the When considered together, the ELA pattern
ELAs rise ⬃600 m in elevation from west to northern Great Basin in a way that allowed and the synchroneity between the hydrologic
east along the 120 km length of the Uinta transgression of Lake Bonneville several mil- highstand of Lake Bonneville and the peak of
Mountains; most of that rise occurs in the lennia after the global LGM (e.g., Benson and glaciation in the southwestern Uintas suggest
western 80 km of the range (Fig. 3). Assuming Thompson, 1987; Oviatt, 1997). This phe- that Lake Bonneville was responsible for the
a uniform temperature depression at 40⬚N lat- nomenon may also have supported the growth additional precipitation (Munroe and Mickel-
itude in northern Utah, this ELA pattern in- of alpine glaciers in the Wasatch and south- son, 2002). Modeling studies have concluded
dicates the presence of a profound east-west western Uinta Mountains during the latest that much of the lake remained ice free in Jan-
precipitation gradient during the LGM. Fur- Pleistocene through increased precipitation uary (Hostetler et al., 1994), and this large
thermore, the lowest ELAs in the western Uin- and/or reduced summer temperatures. Thus, area of relatively warm water only ⬃100 km
tas and Wasatch Mountains are as much as the broad synchroneity of the local LGM in upwind from the western Uintas would have
⬃400 m lower than those predicted by the re- northeastern Utah and the lake highstand like- enhanced winter snowfall considerably be-
constructed regional LGM ELA pattern in ly reflects a regional climate forcing. yond modern levels. This assertion is sup-
Porter et al. (1983), suggesting that a local Evidence for climate forcing at a subre- ported by glacier modeling by Laabs et al.
source may have enhanced moisture delivery gional scale is provided by the westward slop- (2006), which concluded that the Little Cot-
to these glaciers. ing ELA gradient, which indicates that the tonwood glacier in the western Wasatch
magnitude of the glacial advance in the west- Mountains immediately downwind of Lake
DISCUSSION AND CONCLUSIONS ern Uintas was influenced by a local moisture Bonneville was sustained at its LGM extent
Our interpretation that deglaciation began source. Given the near constancy of modern under a temperature depression of ⬃7 ⬚C and
in the southwestern Uintas ca. 16.8 ⫾ 0.7 ka summer adiabatic lapse rates (6.1⬚C/km) a precipitation increase of ⬃3.5 times modern.
suggests that glaciers there reached, or per- across the Uintas (Laabs et al., 2006), it is Similar results were obtained by earlier efforts
sisted at, their maximum extents considerably unlikely that there was appreciable spatial var- directed at modeling Lake Bonneville and the
later than glaciers in neighboring ranges of the iability in altitude-temperature relationships in Little Cottonwood glacier (McCoy and Wil-
Rocky Mountains. For example, Benson et al. the range during the LGM. Thus, mean abla- liams, 1985). Additional modeling is needed
(2005) documented mean ages for the onset of tion season temperatures at the western ELAs, to determine how far downwind the effects of
deglaciation of 18.4 ka in north-central Col- which are ⬃600 m lower that those in the east, Bonneville moisture would have propagated,
orado, 18.9 ka in southwestern Colorado, and would have been ⬃3.7 ⬚C warmer than those but the ELA pattern is strong evidence of en-