Glacial Lake Missoula formed during the last ice age when an ice dam blocked the Clark Fork River in what is now northern Idaho. This immense lake held as much water as Lakes Erie and Ontario combined. There is debate around how many times the ice dam catastrophically failed, releasing massive floods across the region. Studying the bottom sediments of the lake can provide clues about its history. A large exposure of well-preserved Lake Missoula bottom sediments near Missoula, Montana shows characteristics like varves that can help unravel the number and timing of floods from the lake.

1. 1
Bottom
sediments
provide
clues
to
glacial
Lake
Missoula’s
history
by
Richard
L.
Chambers
and
David
Alt
INTRODUCTION
During
the
last
ice
age
an
immense
lake
existed
in
the
western
Montana
basins
when
the
Clark
Fork
River
was
dammed
by
the
Purcell
Lobe
of
the
Cordilleran
ice
sheet
at
the
site
occupied
today
by
Lake
Pend
Oreille,
in
northern
Idaho.
This
lake,
known
as
glacial
Lake
Missoula,
held
as
much
water
as
the
combined
volumes
of
Lakes
Erie
and
Ontario.
Once
the
water
level
reached
a
critical
depth,
the
ice
dam
began
to
float
allowing
water
to
find
its
way
under
the
ice,
when
suddenly
the
ice
dam
collapsed
releasing
an
enormous
discharge
of
water
that
had
only
one
way
to
go.
In
only
a
few
days
nearly
2100
cubic
kilometers
of
floodwater
swept
across
northern
Idaho,
eastern
Washington,
and
Oregon
and
down
the
Columbia
River
gorge
finding
its
way
to
the
Pacific
Ocean
at
Portland,
Oregon.
In
its
wake
it
left
giant
flood
bars
and
trains
of
huge
current
ripples
and
a
ravaged
landscape
known
as
the
“Channeled
Scablands”
in
eastern
Washington.
The
Lake
Missoula
floods
probably
involved
the
largest
freshwater
discharges
in
the
geologic
record;
about
20
times
greater
than
the
average
worldwide
runoff
(Bretz,
1930,
1969;
Pardee,
1942;
Baker,
1973;
Baker
and
Bunker,
1985).
A
topic
still
debated
is
the
number
of
times
these
floods
occurred,
and
whether
each
of
the
lake
drainages
was
the
result
of
a
catastrophic
collapse
of
the
ice
dam
impounding
the
lake
and
the
torrential
release
of
its
enormous
water
volume.
Some
studies
suggest
dozens
of
floods
and
even
as
many
as
40
or
more
(Alt
and
Chambers,
1970;
Chambers,
1971,
1984;
Waitt,
1980,
1984,
1985;
Alt,
2001;
Hanson,
et
al.,
2012).
Because
each
successive
flooding
event
would
rework
and
even
remove
evidence
of
earlier
floods,
a
study
of
the
bottom
sediments
that
accumulated
in
the
still
waters
in
glacial
Lake
Missoula
basins
should
provide
some
of
the
answers
to
these
questions.
Despite
all
of
the
studies
on
the
Lake
Missoula
floods,
surprisingly
little
was
known
about
the
nature
of
the
bottom
sediments
and
their
contribution
to
unraveling
the
lake’s
history.
D.
M.
Sieja,
a
graduate
student
in
geology
at
Montana
State
University,
conducted
the
earliest
known
quantitative
study
of
the
Lake
Missoula
bottom
sediments
in
1959.
The
main
purpose
of
his
study
was
to
document
the
clay
mineralogy
of
the
glacial
Lake
Missoula
varves
and
to
see
if
the
clay
minerals
changed
in
quantity
vertically
in
an
exposure
and
laterally
from
one
location
to
another
across
the
Missoula
Valley
basin.
In
his
study,
Sieja
found
three
varve
types:
simple,
composite,
and
drainage;
however
he
made
no
detailed
interpretation
for
their
origin,
other
than
to
note
that
the
clays
probably
settled
to
the
lake
floor
through
calm
water,
whose
source
were
the
Blackfoot
and
Rattlesnake
valley
glaciers.

2. 2
In
his
comprehensive
review
of
the
evidence
for
repeated
catastrophic
outbursts
from
glacial
Lake
Missoula,
Bretz
(1969)
suggested
that
each
lake
sequence
should
be
separated
by
an
unconformity
and
that
bogs
and
forests
would
occupy
the
drained
lake
floor
only
to
become
buried
when
a
new
lake
formed.
At
the
time
of
Bretz’s
1969
review,
he
suggested
that
the
lake
sediments
should
contain
varves
and
randomly
distributed
ice-­‐rafted
fragments.
The
term
varve
is
a
derivative
of
the
Swedish
word
“varv”
whose
meaning
includes
“a
periodical
iteration
of
layers.”
Varves
are
typically
associated
with
glacial
lake
deposits
consisting
of
two
layers
–
a
lower
light-­‐colored
layer
composed
primarily
of
silt
deposited
during
the
summer
by
melting
glaciers,
and
an
upper
darker-­‐
colored
layer
composed
mainly
of
clay
and
organic
matter
deposited
during
the
winter
by
slowly
settling
sediment
through
calm
water.
Varves
were
recognized
as
yearly
deposits
as
early
as
1832
and
have
been
successfully
used
to
correlate
different
localities
and
by
counting
varves,
like
tree
rings,
it
is
possible
to
establish
a
time
scale
for
glacial
lakes
and
glacial
retreat.
CHARACTERISTICS
OF
THE
LAKE
MISSOULA
BOTTOM
SEDIMENTS
A
large,
well-­‐preserved
exposure
of
glacial
Lake
Missoula
bottom
sediments
(Fig.
1)
is
located
about
40
km
west
of
Missoula,
Montana
along
Highway
90,
near
the
juncture
of
the
Clark
Fork
River
and
Ninemile
Creek.
The
exposure
is
a
road
cut
about
250-­‐m
long
and
10-­‐25-­‐m
in
height.
Rather
than
finding
a
single
thick
sequence
of
glacial
lake
varves
that
could
Figure
1.
Ninemile
Creek
exposure
of
Lake
Missoula
bottom
sediment.
be
easily
counted
to
date
the
exposure,
Alt
and
Chambers
(1970)
and
Chambers
(1971,
1984),
discovered
a
much
more
complex
and
far
more
interesting
situation.
What
they
found
was
about
40
short
sequences
with
glacial
lake
varves
sandwiched
between
layers
of
fine-­‐grained
sand
and
silt.
The
varves
record
the
times
when
a
deep,
calm
lake
existed.
After
a
detailed
study
of
the
sand-­‐silt
layers,
they
became
convinced
that
streams
traversing
the
drained
lake
floor
deposited
these
sediments
into
fairly
shallow
water
as
the
lake
began
to
fill
once
the
ice
dam
reformed,
blocking
the
outflow
of
the
Clark
Fork
River.
So,
Bretz’s
1969
hypothesis
of
multiple
lake
sequences
proved
correct,
however
Alt
and
Chambers
(1970)
and
Chambers
(1971,
1984)
did
not
find
any
evidence
of
tree
stumps
or
bogs
between
the
sequences,
therefore,
suggesting
only
short
periods
of
time
between
successive
lake
fillings,
thus
preventing
the
formation
of
deep
soil
profiles.

3. 3
Lake
Missoula
Rhythmites
Well-­‐developed
small-­‐scale
cycles,
up
to
several
meters
thick,
characterize
the
glacial
Lake
Missoula
bottom
sediments.
The
rhythmically
bedded
deposits
show
a
very
distinctive
pattern
of
alternating
light
and
dark-­‐colored
layering
at
the
outcrop
scale
(Fig.
1).
A
typical
rhythmite
has
a
light-­‐colored
base
of
fine-­‐grained
sand
and
silt
passing
upward
into
a
darker-­‐toned
sequence
of
glacial
lake
varves
composed
of
silt
and
clay-­‐size
sediment.
Stream
Deposits
The
basal
sand-­‐silt
layer
(Fig.
2)
present
in
each
rhythmite
is
characterized
by
planar
bedding,
with
about
half
of
them
containing
ripple
cross
bedding
and
other
sedimentary
features
suggestive
of
deposition
by
streams.
The
lower
contact
between
successive
rhythmites
is
unconformable
to
the
underlying
sediment,
usually
an
earlier
sequence
of
varves.
The
feature
outlined
in
Figure
2
is
interpreted
to
be
gravel
deposited
by
a
local
stream
that
flowed
across
the
exposed
lake
floor
eroding
the
varves
deposited
in
a
former
lake.
Figure
2.
An
example
of
the
basal
silt
unit
in
a
Lake
Missoula
rhythmite
and
a
small
channel
deposit
of
gravel
(outlined).
Varves
The
basal
sand-­‐silt
layer
passes
upward
into
a
sequence
of
varves
(Fig.
3).
The
varved
couplets
tend
to
thin
upward
in
any
sequence
indicating
a
deepening
lake
that
continued
to
deepen
until
the
ice
dam
failed
(Chambers,
1971,
1984).
The
varves
have
a
sharp,
unconformable
contact
with
the
next
rhythmite
deposited
in
a
new
lake.
Although
no
deep
soil
profiles
were
found,
thin
zones
of
desiccated
and
weathered
varves
were
found
between
22
rhythmites.
Two
varve
sequences
were
so
deeply
weathered
that
no
varve
count
was
possible.
The
top
most
7
cycles
became
so
deeply
weathered
since
the
final
retreat
of
the
glaciers
about
12,000
years
ago,
that
detailing
those
cycles
proved
impossible.
Figure
3.
Rhythmite
20
showing
22
varves.

4. 4
In
1971,
Chambers
counted
766
varves
at
the
Ninemile
Creek
road
cut,
however
this
number
does
not
take
into
account
the
22
zones
of
weather
varves,
or
the
varves
in
cycles
33
to
40.
In
that
study,
Chambers
estimated
that
more
than
100
varves
needed
to
be
included
raising
the
count
to
about
900
varves.
Hanson,
et
al.
(2012)
visited
the
Ninemile
Creek
exposure
and
counted
only
583
varves,
however
they
apparently
did
not
include
an
estimate
of
varves
in
the
weather
zones.
The
discrepancy
between
766
and
583
could
be
explained
by
1)
weathering
of
the
upper
part
of
the
exposure
over
the
past
several
decades
preventing
Hanson,
et
al.
to
count
varves
in
the
upper
section,
2)
Chambers
(1971)
counted
composite
varves
as
simple
varves,
thereby
increasing
the
number,
or
3)
that
Hanson
et
al.
(2012)
grouped
thinner
varves
into
composite
varves
resulting
in
a
fewer
number
of
varves.
Subaerial
exposure
of
the
lake
floor
indicates
that
glacial
Lake
Missoula
drained
or
partially
drained
about
40
times.
The
22
weathered
zones
range
in
thickness
from
2.3
to
18-­‐cm
and
average
about
7.2-­‐cm
(Fig.
4).
Although
the
top
of
about
12
varve
sequences
do
not
appear
weathered,
it
is
possible
that
some
material
is
missing
because
of
stream
erosion
and
subsequent
deposition
of
the
next
sand-­‐silt
layer.
Frost
cracks
(Chambers
1971)
or
ice-­‐wedge
casts
(Hanson,
et
al.,
2012)
suggest
that
the
lake
floor
was
exposed
to
subaerial
conditions
for
several
years;
a
period
of
time
much
longer
than
simple
lake-­‐level
fluctuations.
The
ice-­‐wedges
were
filled
with
weathered
varves
or
sand.
Figure
4.
A
varve
sequence
illustrating
a
weathered
material
infilling
a
frost
crack
created
in
the
exposed
lake
flood
sediment.
Discussion
About
40
last
glacial
Lakes
Missoula
are
documented
in
the
Ninemile
road
cut,
however
there
is
no
conclusive
evidence
that
each
drainage
was
complete
or
catastrophic
(Chambers,
1971,
1984;
Hanson,
et
al.,
2012),
but
the
lake
did
drain
below
an
altitude
of
985-­‐m.
Because
the
number
of
varves
in
any
given
sequence
ranges
from
9
to
58,
Alt
and
Chambers
(1970)
inferred
periods
of
several
decades
between
lake
drainages;
Alt
(2001)
thought
that
the
average
interval
between
lake
fillings
to
be
about
50
years.
The
deduction
that
each
lake
draining
resulted
in
a

5. 5
catastrophic
flood
was
made
by
Alt
(in
Alt
and
Chambers,
1970),
then
by
Waitt
(1980,
1984,
1985),
and
again
by
Alt
(2001).
Looking
at
the
Ninemile
road
cut,
it
is
easy
to
observe
that
each
younger
cycle
is
thinner
than
the
preceding
cycle,
suggesting
that
each
filling
of
glacial
Lake
Missoula
contained
less
water
than
the
one
before.
This
seems
consistent
with
the
observation
that
the
greater
number
of
varves
occur
lower
in
the
section,
becoming
fewer
in
number
at
the
top
of
the
road
cut.
Thus,
in
the
waning
stages
of
the
last
ice
age,
the
ice
dam
became
thinner
and
less
able
to
impound
great
quantities
of
water,
thus
reducing
the
amount
of
time
between
lake
drainages.
Chambers
(1971,
1984)
also
noted
that
multiple
cycles
of
lake
deposits
are
preserved
within
the
troughs
of
the
giant
current
ripples
in
Camus
Prairie.
This
led
him
to
conclude
that
the
later
drainages
must
have
been
much
slower
than
earlier
discharges
otherwise
the
fragile
lake
sediment
would
have
been
flushed
from
the
basin.
It
is
very
probable
that
the
unconsolidated
lake
bottom
sediment
could
be
preserved
even
when
the
lake
drained
catastrophically.
A
good
analogy
of
glacial
Lake
Missoula
draining
is
to
observe
a
bathtub
empty
when
filled
with
water.
The
drain
plug
would
be
the
ice
dam,
whereas
the
back
of
the
bathtub
represents
the
Missoula
and
Bitterroot
valleys.
When
the
plug
is
pulled,
all
the
action
happens
at
the
front
of
the
tub,
at
the
drain
hole,
however
the
water
level
at
the
back
of
the
tub
gently
lowers
with
no
turbulence.
Once
the
ice
dam
burst,
Camas
Prairie,
the
Little
Bitterroot
and
Mission
valleys
would
empty
first,
followed
by
the
slower
emptying
of
the
Missoula
and
Bitterroot
valleys.
The
Ninemile
Creek
road
cut
is
almost
245
km
from
the
site
of
the
ice
dam
and
at
that
distance
the
water
level
most
likely
gently
lowered
preserving
the
lake
bottom
sediments.
References
Alt,
D.,
2001.
Glacial
Lake
Missoula
and
its
Humongous
Floods,
Mountain
Press,
Missoula,
197
p.
Alt,
D.
and
R.
L.
Chambers,
1970.
Repetition
of
the
Spokane
flood:
American
Quaternary
Association
Meeting
1,
Yellowstone
Park
and
Bozeman,
Montana,
Abstracts.
Montana
State
University,
Bozeman,
p.
1.
Baker,
V.
R.,
1973,
Paleohydrology
and
Sedimentology
of
Lake
Missoula
Flooding
in
Eastern
Washington:
Geological
Society
of
America
Special
Paper
144,
73
p.
Barker,
V.
R.
and
R.
C.
Bunker,
1985.
Cataclysmic
late
Pleistocene
flooding
from
glacial
Lake
Missoula:
a
review:
Quaternary
Research,
27:
182-­‐201.
Bretz,
J
H.,
1969.
The
Lake
Missoula
floods
and
the
Channeled
Scabland:
The
Journal
of
Geology,
77:
505–543.
Bretz,
J
H.,
1930.
Lake
Missoula
and
the
Spokane
Flood:
Geological
Society
of
America
Bulletin,
38:
385-­‐422.
Chambers,
R.L.,
1971.
Sedimentation
in
glacial
Lake
Missoula,
MSc.
Thesis,
University
of
Montana,
Missoula.

6. 6
Chambers,
R.L.,
1984.
Sedimentary
evidence
for
multiple
glacial
Lakes
Missoula,
in:
McBane,
J.
D.
and
P.
B.
Garrison
(Eds.),
Northwest
Montana
and
Adjacent
Canada.
Montana
Geological
Society,
Billings,
pp.
189–199.
Hanson,
M.
A.,
Olav,
B.
L.,
and
J.
J.
Claque,
2012.
The
sequence
and
timing
of
large
late
Pleistocene
floods
from
glacial
lake
Missoula,
Quaternary
Science
Reviews,
31:
67-­‐81.
Pardee,
J.
T.,
1942.
Unusual
currents
in
glacial
Lake
Missoula,
Montana:
Geological
Society
of
America
Bulletin,
53:
1569–1599.
Sieja,
D.
M.,
1959.
Clay
Mineralogy
of
Glacial
Lake
Missoula
Varves,
Missoula
County,
Montana,
MSc.
Thesis,
Montana
State
University,
Montana.
Waitt,
R.
B.,
1980.
About
forty
last-­‐glacial
Lake
Missoula
jökulhlaups
through
southern
Washington,
Journal
of
Geology,
88:
653–679.
Waitt,
R.
B.,
1984.
Periodic
jökulhlaups
from
Pleistocene
Glacial
Lake
Missoula
—
new
evidence
from
varved
sediment
in
northern
Idaho
and
Washington,
Quaternary
Research,
22:
46-­‐48.
Waitt,
R.
B.,
1985.
Case
for
periodic,
colossal
jökulhlaups
from
Pleistocene
glacial
Lake
Missoula,
Geological
Society
of
America
Bulletin,
96:
1271–1286.