THE RELATION OF GREAT BASIN LATE QUATERNARY HYDROLOGIC
AND CRYOLOGIC VARIABILITY TO NORTH ATLANTIC CLIMATE
OSCILLATIONS
L. Bensona, R. Spencerb, D. Rhodec, L. Louderbackd, R. Ryee
aU.
bDepartment
cDesert
S. Geological Survey, 3215 Marine Street, Boulder, CO 80303, USA
of Geology and Geophysics, University of Calgary, Alberta T2N 1N4, Canada
Research Institute, Division of Earth and Ecosystem Sciences, 2215 Raggio Parkway,
Reno, NV 89512, USA
dAnthropology
Department, University of Nevada,1664 North Virginia Street, MS 096,
Reno, NV 89557, USA
eU.S.
Geological Survey, MS 963, Denver Federal Center, Lakewood, CO 80225, USA
HEINRICH AND DO EVENTS
Heinrich events were first recognized in marine sediments as layers
containing high percentages of carbonate fragments in the 180-µm
to 3-mm sediment-size fraction (Heinrich, 1988). Broecker et al.
(1992) interpreted the Heinrich layers as debris released during
massive influxes of icebergs into the North Atlantic.
The carbonate fragments in Heinrich events 1, 2, 4, and 5 indicate that much of the ice that transported the
carbonate came from the LIS (Bond et al., 1992). Heinrich events occur during cold DO stadials.
There are two principal theories regarding the relation of Heinrich events to variability in the strength of the
AMOC. In one theory (SURGE FIRST), the Heinrich event surges ice from the LIS producing a freshwater
cap on the North Atlantic which shuts down the production of NADW and shuts down the AMOC
(MacAyeal, 1993). In the other theory (SURGE LAST), freshwater is released from ablation of the LIS or
melting of pack ice in the North Atlantic, shutting down the AMOC, which through a variety of possible
mechanisms leads to the ice surging associated with a Heinrich event (Clark et al., 1999; Calov et al.,
2002) .
Both theories are consistent with the observation that abrupt climate change in almost all climate models is
triggered by changes in the surface freshwater balance of the North Atlantic.
During a non-Heinrich DO stadial, air and water temperatures in the North Atlantic region also decline in
concert with reductions in the strength of the AMOC (Boyle, 2000). Wunsch (2006) has suggested that
DO oscillations observed in Greenland ice were not generated by shifts in North Atlantic ocean circulation,
given the small contribution of heat from the high-latitude ocean to the overall meridional heat flux.
Instead, Wunsch suggests that DO oscillations are a consequence of the interactions of the windfield with
Northern Hemisphere ice sheets and that changes in ocean circulation also are a consequence of wind
shifts.
Surging Ice:
Heinrich 1 (15.8 ka)
Outburst floods:
Heinrich 2 (23.9 ka)
8.2 ka event (8.4-8.0 cal ka)
Preboreal Oscillation (11.3-10.7 cal ka) Heinrich 3 (30.0 ka)
Heinrich 4 (38.5 ka)
Younger Dryas (12.9-11.6 cal ka)
Melting Ice:
Dansgaard-Oeschger interstadial 1-12
(A) Blue Lake/Bonneville core, (B) Wilson
Creek Fm, (C, D) Owens Lake cores, (E)
Pyramid Lake core.
The age model for the Wilson Creek
Formation (Fig. B) was constructed using
the GISP2 ages of H1, H2, H3, H4 and
their paleomagnetic secular variationbased locations in the sediment column
together with the GISP2 age of the MLE
(33.9 cal ka) and the calibrated 14C age of
the top of the stratigraphic section
(14.14 cal ka) (Benson et al., 1998).
All 14C ages were assigned a zero
reservoir age except those for Pyramid
Lake which were assigned a 600-yr
value before calibration (Fig. E).
However, sediments in Bonneville
associated with the Mazama tephra are
1800 14C yr too old.
Age-model calibration done using CALIB
5.0.1 (Stuiver et al., 2005) for 14C ages
<19,250 yr. Older ages calibrated by
interpolation of Cariaco-based chronology
discussed in Hughen et al. (2005).
8-m sediment core
1-cm sampling interval
1-cm ~ 60 yr
(1298 m)
The Bonneville climatic highstand/Provo
was terminated by the warming that
followed H1.
Periods of aridity occur during H3 and
slightly before H2 and H4. If we invoke
a 1000-yr reservoir correction then H2
and H4 are associated with aridity and
H3 is not.
The YD and PBO are associated with
abrupt increases in lake size.
The overall shape of the Bonneville δ18O
record resembles the GISP2 δ18O
record.
BONNEVILLE CARBONATE
MINERALOGY
Aragonite, high-Mg calcite, lowMg calcite form in sequence as
salinity decreases.
In general the data support the
isotope and TIC records.
Pe
Bodia
try str
oc um
oc
cu
s
Po
ta
m
og
et
Sp
on
or
e
A
ac
ea
e
Cy
pe
r
Ty
Typha
ph la
a tif
an oli
gu a
st
ifo
lia
Cattails
Sedges
1000
6
2000
3000
5
4000
5000
4
6000
7000
8000
3b
9000
10000
3a
11000
YD
12000
13000
2
14000
15000
1
16000
17000
Age
40
(cal BP)
1000
2000
2000
4000
400
1000
2000 40 15
Pollen grains per cubic centimeter/years per sample
Appendix A. Aquatic pollen accumulation rates. Accumulation rates were calculated by dividing the pollen
concentration by the number of years represented by the sample (Data of Lisbeth Louderback)
Zones
Summer Lake and GISP2 records
tied together by the Mono Lake
excursion (Zic et al., 2002).
IRM record stretched to match the
GISP2 δ18O record. Cold N. Atlantic
temps. occur during low lake levels.
Low IRM occurs during low-lake
levels when magnetite is lost due
to oxidative decomposition of TOC.
Summer Lake δ18O record indicates
low lake levels during Heinrich events
but GISP2 interstadials do not usually
line up with Summer Lake δ18O
maxima (Benson et al. 2003).
M
L
E
Mono Lake record indicates
aridity during H2 and H1 and
possibly during H4.
Two highstands (δ18O minima):
One during glacial maximum
(20 cal ka) and another at 14.9
cal ka (terminated by BØA warm
Interval).
TIC values between 30 and 15.5
cal ka indicates that Tioga
glaciation was also terminated
by the BØA.
Pyramid Lake δ18O record
indicates aridity during H1 and a
highstand at 15.5 cal ka that
was terminated by BØA.
The Pyramid Lake TOC record
indicates that the TIOGA
glaciation began ~28 cal ka and
probably ended during or after
the BØA.
Note that shape of Pyramid
Lake δ18O record is similar to
GISP2 δ18O record but shifted in
time and that the frequency of
TOC oscillations is similar to the
frequency of GISP2 δ18O
oscillations.
Owens Lake desiccated during H1.
The TIC record indicates Tioga
glaciation occurred between 29.5
and 15.5 cal ka and was terminated
by BØA.
The TOC record indicates that six
pre-Tioga Sierran glacier advances
occurred during GISP2 stadials.
The YD and PBO oscillations were
associated with peaks in Owens Lake
overflow (δ18O and TOC minima).
GLACIER RECESSION IN THE UINTA, WASATCH
AND SIERRA NEVADA MOUNTAINS
Well-constrained cosmogenic surface-exposure 10Be ages for two LGM
moraines in the southwestern Uinta Mountains have ranges,
respectively, of 16.7 ± 1.5 to 19.9 ± 2.0 ka and 16.1 ± 1.5 to 18.0 ± 1.9
ka (Munroe et al., 2006). Bear Lake IRM data indicate glacial activity
between 32 and 17.5 cal ka (Joe Rosenbaum, pers. comm.).
10Be
ages of moraines deposited after the Bonneville flood, downvalley
of the mouths of Little Cottonwood and Bells canyons on the western
flank of the Wasatch, have younger ages, ranging from 16.9 ± 0.4 to
15.2 ± 0.4 ka (Elliot Lips, pers. comm.).
Sierran Tioga 4 glaciers retreated from Chiatovich Creek between ~15
and 14 36Cl ka (Phillips et al.,1996).
SUMMARY
•
H1 was a dry period in the Great Basin.
•
H2 and H4 may also have been associated with dry periods.
•
The YD and PBO were associated with abrupt increases in lake size.
•
The overall shapes of the Lake Bonneville and Lake Lahontan δ18O records resemble
the GISP2 δ18O record shifted in time.
•
Alpine glaciers and highstand lakes were terminated by the warming that followed H1
(~15.5 cal ka) and which caused changes in the topography of the LIS.
•
Between 39 and 29.5 cal ka, Sierran alpine glacial advances were associated with
four DO stadials and two Heinrich events (which also occur during DO stadials).
•
The lack of consistent coherence between the GISP2 and Great Basin hydrologicbalance records suggests that the Heinrich/DO signal may have been shifted in time
and/or space before it reached the Great Basin, or that the precipitation field over the
Great Basin was spatially inhomogeneous, or that the age models used to construct
the hydrologic-balance records are flawed, or that older carbonate-containing
sediment may have been reworked and transported to core sites during lowstand
lakes that accompanied Heinrich events and may have confounded isotopic records
of low-lake levels.
Chappell (2002) calculated that 10-15 m increases in sea level occurred
during H4 and H5; Clark et al. (2007) recently modeled the response of sea
level to changes in the mass balance of Northern Hemisphere ice sheets.
These simulations, indicate that H4 through H6 were accompanied by 1017 m of sea-level rise). An isotope-based calculation by Roche et al.
(2004) suggested that H4 released only 2 ± 1 m sea-level equivalent of ice;
Bintanja et al. (2005) has calculated the amount of water stored in the LIS
during the past 1.2 Ma. We used the amount of water stored in the LIS just prior
to H1, H2, and H4 and estimates of sea level change, ranging from 2-15 m, to
calculate the percentage of LIS ice transported to the North Atlantic during a
Heinrich event, assuming all the ice came from the LIS.
Percent Reduction in Size of
Laurentide Ice Sheet
Increase in sea level
Heinrich event
2m
5m
10 m
15 m
H1
3.1
7.7
15.4
23.1
H2
3.3
8.3
16.7
25.0
H3
4.0
10.0
20.0
30.0
H4
5.0
12.5
25.0
37.5
HEINRICH EVENT FORCING OF THE WINDFIELD
We offer the hypothesis that the change in size and shape of the LIS
associated with warming after H1 and with Heinrich events, in general,
caused a shift in the mean position of the PJS away from the catchment
areas of one or more Great Basin lakes. For example, Dyke et al.
(2002) have suggested that H1 “probably drew down the entire central
ice surface” positioned over Hudson Bay”. Such a change in topography
should have affected the windfield, including the trajectory of the PJS,
which is in agreement with the hypothesis of Wunsch (2006).
Changes In the Trajectory Of The
Polar Jet Stream
Many of the lakes discussed in this paper indicate
increases in wetness at 20 ± 1 and 15.5 ± 0.5 cal
ka. Mono Lake experienced highstands at 20.3
and 15.0 cal ka, and Lake Mojave to the south
experienced prominent highstands centered at
21.7 and 14.8 cal ka.
16-15 ka
These records indicate that the mean position of
PJS reached its most southerly extent during the
LGM (21-20 cal ka).
21-19 ka
At ~16 cal ka, Pyramid Lake/Lake Lahontan
(~40ºN) experienced its maximum highstand;
however, Mono Lake and Lake Mojave also
experienced highstands, suggesting that the PJS
still affected latitudes south of Pyramid Lake.