Timing and patterns of basin infilling as documented in
Lake Powell during a drought
Lincoln Pratson1, John Hughes-Clarke2, Mark Anderson3, Thomas Gerber1*, David Twichell4, Ronald Ferrari5,
Charles Nittrouer6, Jonathan Beaudoin2, Jesse Granet3, John Crockett6†
1

Nicholas School of the Environment, Duke University, P.O. Box 90227, Durham, North Carolina 27708, USA
Ocean Mapping Group, University of New Brunswick, P.O. Box 4400, Fredericton, New Brunswick E3B 5A3, Canada
3
National Park Service, Glen Canyon NRA, P.O. Box 1507, Page, Arizona 86040, USA
4
U.S. Geological Survey, 384 Woods Hole Road, Quissett Campus, Woods Hole, Massachusetts 02543-1598, USA
5
Bureau of Reclamation, P.O. Box 25007, Denver, Colorado 80225, USA
6
School of Oceanography, University of Washington, P.O. Box 357940, Seattle, Washington 98195-7940, USA
2

ABSTRACT
Between 1999 and 2005, drought in the western United States led to a >44 m fall in the level
of Lake Powell (Arizona-Utah), the nation’s second-largest reservoir. River discharges to the
reservoir were halved, yet the rivers still incised the tops of deltas left exposed along the rim of
the reservoir by the lake-level fall. Erosion of the deltas enriched the rivers in sediment such
that upon entering the reservoir they discharged plunging subaqueous gravity flows, one of
which was imaged acoustically. Repeat bathymetric surveys of the reservoir show that the
gravity flows overtopped rockfalls and formed small subaqueous fans, locally raising sediment accumulation rates 10–100-fold. The timing of deep-basin deposition differed regionally
across the reservoir with respect to lake-level change. Total mass of sediment transferred from
the lake perimeter to its bottom equates to ~22 yr of river input.
INTRODUCTION
The timing, location, and nature of deposition in a basin are largely dictated by base
(i.e., water) level, basin-floor subsidence, and
sediment supply (Posamentier et al., 1988). The
impacts of base-level change and subsidence on
basin infilling have been explored extensively
through modeling (Paola, 2000), experiments
(e.g., Heller et al., 2001), and analyses of basin
stratigraphy (e.g., Van Wagoner et al., 1990).
These same methods have also been used to
evaluate the effects of sediment supply (e.g.,
Perlmutter and Matthews, 1989), but to a much
lesser extent and with fewer conclusive and
generally applicable results. This is because of
uncertainties over (1) how sediment flux to a
basin varies with respect to base-level change
(e.g., are changes in the two in or out of phase?),
and (2) the source of the sediments, from inside
or outside the basin (Pratson et al., 2007).
Here we present new insights on these issues
gained through repeat bathymetric surveys of
Lake Powell (Arizona-Utah) (Fig. 1A) during
a fall and rise in reservoir level between 1999
and 2005 (Fig. 1B). Lake Powell proved exceptional for this purpose because its physiography,
time of operation, and water and sediment input
are so well constrained. Water outflow occurs
through penstocks (i.e., pipes) embedded in the
Glen Canyon Dam >80 m above the adjacent
*Current address: ChevronTexaco Energy Technology Company, 1500 Louisiana St., Bellaire,
Texas 77002, USA.
†
Current address: San Diego State University Research Foundation, 5250 Campanile Dr., San Diego,
California 92182, USA.

reservoir floor (Fig. 2A), so essentially all sediments that enter the reservoir are trapped within
it. This trapping began when the dam was completed 11 April 1963. Pre-reservoir topography
was accurately surveyed beforehand (United
States, 1963) and lake level has been recorded
since (Fig. 1B). Input to the reservoir is gauged
and dominated by the Colorado and San Juan
Rivers, which supply ~95% of the water and
~90% of the sediment (Potter and Drake, 1989).
Deltas have been rapidly growing beyond the
mouths of both rivers (Fig. 2A), with maximum
sediment accumulation rates through 1986
being measured in meters per year (Ferrari,
1988). The narrowness of the lake (generally
≤1.5 km wide) (Fig. 1A) prevents significant
lateral sediment movement, so the deltas are
essentially two-dimensional.
BACKGROUND AND METHODS
Between fall 1999 and fall 2004, average
river inflow to Lake Powell was approximately
halved by a drought in the southwestern U.S.
(Fig. 1B). As withdrawals from the reservoir
exceeded the inflow, lake level fell >44 m
(Fig. 1B). Inflow then returned to its historical
norm in spring 2005, raising lake level >9 m by
early summer (Fig. 1B).
Within this period, three bathymetric surveys of Lake Powell were carried out. The
surveys were conducted early (April 2001) and
late (May 2004) during the fall in lake level,
and (June 2005) when lake level had risen >9 m
from its lowest point two months prior (Figs.
1B and 2A). Note that all the surveys were done
as snowmelt from the Colorado watershed was

arriving at Lake Powell and inflow was at its
yearly maximum (Fig. 1B).
The surveys were of varying extent and used
different high-resolution mapping systems
capable of detecting lake bottom to within
<3 cm. In 2001, the main axis of the Colorado
arm was surveyed using a Garmin 240B singlebeam 50/200 kHz fathometer integrated with
a P-CODE global positioning system (GPS)
receiver. In 2004, the main axes of both the
Colorado and San Juan arms were surveyed
with an EdgeTech 512i 0.5–12 kHz Chirp (compressed high-intensity radar pulse) sonar and a
GPS receiver. In 2005, a Kongsberg EM3002
300 kHz multibeam sonar system, coupled with
short-baseline differential GPS, was used to
completely map the Colorado, San Juan, and
Escalante arms of the reservoir. A 500 kHz
SonTek acoustic Doppler profiler, with GPS,
was used to profile the topsets of the Colorado
and San Juan deltas, where water depths were
too shallow to take the EM3002 (i.e., <10 m).
The EM3002 automatically applied sound
speed corrections to the soundings and translated them into georeferenced depths. For
the other surveys, depths were computed by
multiplying the bottom echo traveltimes with
an average water sound speed for each survey. The average sound speeds were derived
from control points with known elevations
located throughout the reservoir. Finally, depth
measurements from all the surveys were converted into elevations using the appropriate
daily reservoir levels posted by the U.S. Bureau
of Reclamation (USBR).
The surveys were placed within the longerterm context of reservoir infilling by coregistering them in the geographic information system
(GIS) with the few previous surveys of Lake
Powell: (1) a detailed topographic survey made
by the USBR before the Glen Canyon Dam was
built (United States, 1963); (2) a bathymetric
profile up the axes of the Colorado and San Juan
arms compiled in 1976 by Potter and Drake
(1989); and (3) a series of cross-lake bathymetric profiles measured by the USBR every
few kilometers up each waterway of the reservoir in 1986 (Ferrari, 1988).

© 2008 The Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or editing@geosociety.org.
GEOLOGY,
November
2008
Geology,
November
2008;
v. 36; no. 11; p. 843–846; doi: 10.1130/G24733A.1; 3 figures.

843

A
M

A

R

L

O

R

A

D

O

2

C

O

37.5° N

ES

Full Pool

CA

LA

June 2005

NT

Axis of 2005 Survey

E

1

S

J

A

R

M

UTAH

37.0° N
111.0° W

Glen
Canyon
Dam

ARIZONA

0

110.5° W

30
km

108

B

1125

Inflow (m3/s)

1105
2 × 108
1085
1065

1 × 108

Lake level (msl)

3×

A N

N
U A

1045

2004

2000

1996

1992

1988

1984

1980

1976

1972

1968

1025
1964

0

Year
Figure 1. A: Map of Lake Powell. Full pool is maximum reservoir level (1141 m). Regions
1 and 2 correspond to Figures 3A and 3B, respectively. B: Inflow to Powell and lake level
(msl—meters above sea level) since closure of Glen Canyon Dam (http://www.usbr.gov/uc/
crsp/GetSiteInfo). Gray bars indicate when reservoir was surveyed.

RESULTS
The earlier bathymetric surveys of Lake
Powell (Fig. 2A) show that prior to the 1999–
2004 drought, sedimentation was greatest on the
foresets of deltas building out from the Colorado
and San Juan Rivers, with accumulation rates at
these sites reaching 2–3 m/yr (Ferrari, 1988).
As lake level fell during the drought, deposition shifted to the deeper part of the reservoir
(Figs. 2A–2C). However, the location and extent
of this deposition varied basinward of the two
rivers, the differences persisting through the
subsequent lake-level rise.
In the Colorado arm of the reservoir, the delta
foreset continued to prograde and steepen even
as lake level fell and the delta top was incised by

844

the Colorado River (Fig. 2A). During the lakelevel fall between 2001 and 2004, mean accumulation rates remained highest on the delta
foreset, but sediments were deposited >70 km
basinward at an average rate of 0.75 m/yr
(Fig. 2B). A significant fraction of this deposition became ponded behind three large rockfalls
that cross the reservoir floor (RF1–3, Fig. 2A).
These were created decades before by failures of
the sandstone cliffs that border much of the lake
(Potter and Drake, 1989). The sediment ponding was greatest behind the rockfalls closest
to the delta (RF1 and RF2; Fig. 2A, and 1 in
Fig. 3A), but sediments were also carried over
these obstacles to at least as far as a third rockfall near where the Colorado and San Juan arms

meet (RF3, Fig. 2A). The bathymetric changes
show that sediment accumulation then waned in
the deep part of the lake during lake-level rise,
and deposition shifted back up onto the delta
foreset, where the maximum accumulation rate
jumped to >13 m/yr (Fig. 2B).
In the San Juan arm of the lake, a different
pattern of deposition ensued. The delta here
prograded no more than 1 km during the fall
and subsequent rise in lake level with little if
any steepening (Fig. 2A). Sediments tended to
bypass the entire delta and deposit beyond its
foreset all the way to the Glen Canyon Dam,
~165 km beyond the inflow (Figs. 2A and 2C).
Sedimentation in this deep, lowermost stretch of
Lake Powell did not increase significantly until
well into the drought. The 210Pb measurements
from a core collected in the lower lake in 2001
(Figs. 2A and 2C) indicate that for more than
10 yr sediment accumulation rates at the site were
0.01 m/yr. This rate agrees with an average rate
of 0.04 m/yr determined ~1 km away by differencing soundings from the 1986 and 2001 bathymetric surveys. However, by the 2004 survey,
mean sediment accumulation rates in the lower
part of the lake increased almost two orders of
magnitude, to meters per year (Fig. 2C). Furthermore, small subaqueous fans had formed at the
mouths of at least two side canyons, Antelope
(SF3) and Forbidden (SF1) (Fig. 2A). The Forbidden fan was imaged in the 2004 and 2005 surveys (1 in Fig. 3B). It was not present during the
earlier 1986 survey and was not mapped in 2001,
but the 2004 Chirp subbottom profile reveals a
reflection at an equivalent depth to the lake floor
in 2001 that can be traced under the fan (3 in
Fig. 3B). This suggests that the fan formed during the 3 yr period when lake level fell. Fan formation may have then ceased; 1.5 m of sediment
ponded behind the up-lake levee of the fan as
sediment continued to come down the main axis
of the lake (3 in Fig. 3B). Unlike in the Colorado
arm, deep-water deposition below the San Juan
arm did not decline during the subsequent lakelevel rise. Instead, accumulation rates through
the 2005 survey remained as high as during the
period of lake-level fall (Fig. 2C).
The deep-water deposits were laid down by
subaqueous gravity flows. Such events are not
restricted to periods of low lake level. Turbid,
bottom-hugging density currents commonly
emanate from the Colorado and San Juan
Rivers (http://www.gcmrc.gov/products/water_
quality/; Johnson and Merritt, 1979). These hyperpycnal flows were probably frequent between
1995 and 2000 when lake-level was high, for
suspended-sediment concentrations averaged
≥0.5 g/L at the Green River, Cisco, and Bluff,
Utah, gauging stations 100–150 km upriver of
the reservoir (http://co.water.usgs.gov/sediment/
seddatabase.cfm). However, the jumps in deepreservoir sediment accumulation rates (Figs. 2B

GEOLOGY, November 2008

1150

San Juan
Delta

Full Pool

1125

Colorado
Delta

Elevation above sea level (m)

Lake level 2001

1100

2005

2004

1075

Penstocks

1050

Dam
bypass
tubes

RF4
RF2

RF1

A

RF5

1963
1976

1025

1986

Dead Pool

1000
975

2001

SF1
SF3

RF3

SF2

2004
2005

core
site

Distance uplake from Glen Canyon Dam (km)

950
50

2001–1986
2004–2001
2005–2004

100

C
Accumulation
rate (cm/yr)

1000

Colorado
Arm

300

B
1000

San Juan
Arm

100

100
Lake Below
Confluence
10

10
2001–1986
2004–2001
2005–2004

surveys
core

0

50

Accumulation
rate (cm/yr)

0

1
100

150

200

250

Figure 2. A: Successive bathymetric profiles along main axes of Powell. RF—rock fall; SF—
subaqueous fan. Profiles portrayed at same irregular several-kilometer sample spacing as
1986 survey (Ferrari, 1988). Note that RF3 separates floor of Colorado arm from that of San
Juan arm to Glen Canyon Dam. B, C: Sediment accumulation rates from differencing surveys
indicated and dividing by elapsed time. Except for soundings adjacent to dam, 2001 survey
stops ~80 km up reservoir of dam. Dead pool is unusable water, defined by elevation of tubes
for bypassing water around dam.

and 2C) clearly point to increased deposition by
subaqueous gravity flows during the drought.
At least one of these subaqueous gravity flows
was imaged in the 2005 survey (Fig. 3A). Adjacent multibeam bathymetry swaths collected
over five days captured the progress of the flow
as it moved down the Colorado Delta foreset (2 in
Fig. 3A). The type of flow is unclear. It could have
been a viscous debris flow or just the bedload
component of a slow-moving turbidity current.
In either case, its sediment concentration was
great enough to produce a distinct acoustic reflection delineating the front of the flow as an abrupt
20–40 cm rise above the reservoir floor (3 in
Fig. 3A). Differences in the position of the front
over time (2 in Fig. 3A) indicate that the flow was
progressing down the 0.1° incline at 6 m/h.
DISCUSSION
As in natural basins, water and sediment
supply to Lake Powell are coupled through the
tie between regional climate and hydrology

GEOLOGY, November 2008

(Scholz, 2001). Because of this coupling, our
observations of change in Lake Powell offer
new insights on sediment supply to basins during climate-driven base-level change.
The first of these insights is that subaqueous
gravity-flow dynamics may affect the timing
and distribution of basin sedimentation during
base-level change. In Lake Powell, a far greater
fraction of incoming sediment bypassed the San
Juan Delta foreset than the Colorado, and the
bypassing was not limited to falling base level,
but continued into the lake-level rise, although
why this occurred is unclear. Until recently, the
San Juan Delta foreset had always been steeper
than the Colorado, but by spring 2005 the latter
had achieved the same 0.1° dip (Fig. 2A, and
1 in Fig. 3A). While the mean water discharge
from the San Juan River is little more than 15%
that of the Colorado, sediment discharge from
the two rivers is roughly equivalent (Potter and
Drake, 1989). The higher concentration of suspended sediment in the San Juan River could

imbue gravity flows with greater excess density
to move beyond the delta foreset. Whatever
the reason, the timing and extent of deep-basin
sedimentation in Lake Powell were not simply
a function of lake level or its rate of change,
as sequence stratigraphic concepts postulate
(Posamentier et al., 1988). The character and
dynamics of the subaqueous gravity flows
involved were also factors.
The second insight is that base-level falls
that expose unconsolidated strata can lead to
increased hyperpycnal flows in a basin even
when river inflow is reduced. Whereas hyperpycnal flows are common in lakes (Fan and
Morris, 1992), long-term recordings of coastal
river discharges have suggested that such flows
require floods to initiate where rivers empty into
the ocean (e.g., Dadson et al., 2005; Johnson
et al., 2001; Mulder and Syvitski, 1995). Like
all rivers, sediment discharge in the Colorado
and San Juan Rivers is directly proportional to
water discharge, which declines during droughts
(Committee to Review the Glen Canyon Environmental Studies, 1991). Consequently, the
halving of water discharge to Lake Powell during the 1999–2004 drought should have also
reduced sediment discharge to the lake and
presumably hyperpycnal flows. However, our
observations indicate that such flows increased
over this period. The primary sediment sources
for these flows were the local unconsolidated
delta topsets exposed by the lake-level fall.
Although diminished, river-mouth water discharges incised these strata with relative ease,
becoming rich in suspended sediment and thus
negatively buoyant as they crossed into the reservoir. It seems feasible that marine hyperpycnal
flows could be similarly generated at river
mouths during sea-level lowstands if exposed
basin strata are easily eroded, even when river
water discharges are lower than those needed for
the phenomena to occur today.
Finally, the vast majority of sediments moved
deep into a basin over the course of low base
level may actually come from erosion of strata
within the basin exposed by the base-level
fall, and not from an up-river source. In Lake
Powell, the drought triggered the redistribution of ~1 km3 of sediments from along the lake
perimeter to below the elevation of tubes that
can bypass water around the Glen Canyon Dam,
i.e., the “dead pool” level (Fig. 2A). Assuming
that the average porosity and grain density of the
strata in Lake Powell are 60% and 2500 kg/m3,
respectively, the volume of sediments moved
below the dead pool level equates to a mass of
1000 × 106 Mt. Combined, the Colorado and
San Juan rivers annually deliver ~41 × 106 Mt
of sediment into Lake Powell, which represents
90% of the total sediment load to the lake (Potter
and Drake, 1989). This suggests that possibly as
much as 22 yr of river sediment input to Lake

845

Distance (km)

A
210

200

220

230

B

240

250

80

RF2

225

230

RF1

1040

1
≤ 35

Swath
width

15-5-05

18-5-05

19-5-05

2

100 m

Canyon wall

36
14-5-05

Moving
flow

990
200 m

U

985

1

1020

Flat lake
bed

95

SF1

995

Ship
track

Water depth (m)

Elevation (m)

1060

1040
220

1963
1986
2001
2004
2005

3

1045

1080

90

Flow

1050

Elevation (m)

1055

85

1000

1060

1100

Distance (km)

260

n

Fa

980

Leveed
channels

D

2m

2

D
Shaded
relief

U

Onlapping
wedge

3
2 km

Chirp sonar (2004)

37

Figure 3. A: 1—Colorado Delta profile (100 m horizontal sample spacing); note filling behind RF1 (RF—rockfall) and 2 between 2001 and
2004. 2—Moving subaqueous gravity flow evident in shaded relief imagery. Arrows point to locations of flow front on successive dates (day,
month, year). 3—Flow is evident in 2005 bathymetric profile of Colorado Delta, rising 3 m above surrounding reservoir floor. B: 1–2—Subaqueous fan (dashed outline) emanating from Forbidden Canyon into Lake Powell. SF—subaqueous fan. Fan formed by 2004. 3—Chirp
(compressed high intensity radar pulse) sonar profile, collected in 2004.

Powell was remobilized to the deepest parts of
the lake in ~20% of the time. While the speed
and scale of this redistribution are undoubtedly
tied to the rate and magnitude of the base-level
fall in Lake Powell as well as to the erodability
of the exposed strata, our estimates suggest that
deep-water sediments deposited in many basins
during low base level could be overwhelmingly
derived from erosion of shallow-water strata in
the basin exposed by the fall.
ACKNOWLEDGMENTS
Grants to Pratson from the U.S. Office of
Naval Research (N00014-03-1-0141) and the U.S.
National Science Foundation (EAR-98-96392 and
OCE-04-05515) helped support data collection,
analysis, and manuscript preparation. The 2005 multibeam survey was made possible by a Natural Sciences and Engineering Research Council Discovery
grant to Hughes-Clarke. We thank Carlos Pirmez and
an anonymous reviewer, whose comments greatly
improved the manuscript.
REFERENCES CITED
United States, 1963, Colorado River Storage Project,
Glen Canyon Unit, Lake Powell, area and
capacity tables: Salt Lake City, Utah, U.S.
Bureau of Reclamation.
Committee to Review the Glen Canyon Environmental Studies, 1991, Colorado River ecology and
dam management: Washington, D.C., National
Academy Press, 288 p.
Dadson, S., Hovius, N., Pegg, S., Dade, W.B., Horng,
M.J., and Chen, H., 2005, Hyperpycnal river
flows from an active mountain belt: Journal

846

of Geophysical Research, v. 110, F04016, doi:
10.1029/2004JF000244.
Fan, J., and Morris, G.L., 1992, Reservoir sedimentation I: Delta and density current deposits: Journal of Hydraulic Engineering, v. 118,
p. 354–369, doi: 10.1061/(ASCE)0733-9429
(1992)118:3(354).
Ferrari, R.L., 1988, 1986 Lake Powell Survey: Denver, Colorado, U.S. Bureau of Reclamation
Technical Report REC-ERC-88-6, 67 p.
Heller, P.L., Paola, C., Hwang, I.-G., John, B., and
Steel, R., 2001, Geomorphology and sequence
stratigraphy due to slow and rapid base-level
changes in an experimental subsiding basin
(XES 96–1): American Association of Petroleum Geologists Bulletin, v. 85, p. 817–838.
Johnson, K.S., Paull, C.K., Barry, J.P., and Chavez,
F.P., 2001, A decadal record of underflows
from a coastal river into the deep sea: Geology,
v. 29, p. 1019–1022, doi: 10.1130/0091-7613
(2001)029<1019:ADROUF>2.0.CO;2.
Johnson, N.M., and Merritt, D.H., 1979, Convective
and advective circulation of Lake Powell, UtahArizona, during 1972–1975: Water Resources
Research, v. 15, p. 873–884, doi: 10.1029/
WR015i004p00873.
Mulder, T., and Syvitski, J.P.M., 1995, Turbidity currents generated at river mouths during exceptional discharges to the world oceans: Journal
of Geology, v. 103, p. 285–299.
Paola, C., 2000, Quantitative models of sedimentary
basin filling: Sedimentology, v. 47, p. 121–178,
doi: 10.1046/j.1365-3091.2000.00006.x.
Perlmutter, M.A., and Matthews, M.D., 1989,
Global cyclostratigraphy; a model, in Cross,
T.A., ed., Quantitative dynamic stratigraphy:
Englewood Cliffs, New Jersey, Prentice Hall,
p. 233–260.

Posamentier, H.W., Jervey, M.T., and Vail, P.R.,
1988, Eustatic controls on clastic deposition
I—Conceptual framework, in Wilgus, C.K.,
et al., eds., Sea-level changes: An integrated
approach: Society of Economic Paleontologists and Mineralogists Special Publication 42,
p. 109–124.
Potter, L.D., and Drake, C.L., 1989, Lake Powell:
Virgin flow to dynamo: Albuquerque, University of New Mexico Press, 311 p.
Pratson, L.F., and 21 others, 2007, Seascape evolution on clastic continental shelves and slopes,
in Nittrouer, C.A., et al., eds., Continentalmargin sedimentation: From sediment transport to sequence stratigraphy: International
Association of Sedimentologists Volume 37:
Malden, Massachusetts, Blackwell Publishing,
p. 339–380.
Scholz, C.A., 2001, Applications of seismic sequence
stratigraphy in lacustrine basins, in Last, W.M.,
and Smol, J.P., eds., Tracking environmental
change using lake sediments. Volume I: Basin
analysis: Dordrecht, Kluwer Academic Publishers, p. 7–22.
Van Wagoner, J.C., Mitchum, R.M., Campion,
K.M., and Rahmanian, V.D., 1990, Siliciclastic
sequence stratigraphy in well logs, cores, and
outcrops: Concepts for high-resolution correlation of time and facies: American Association
of Petroleum Geologists Methods in Exploration 7, 55 p.
Manuscript received 11 January 2008
Revised manuscript received 8 July 2008
Manuscript accepted 19 July 2008
Printed in USA

GEOLOGY, November 2008