A presentation on the construction and destruction of a volcanic dam at Mount Meager, BC, 2400 years ago.

1. The life and death of a volcanic dam: the
2360 BP eruption of Mt. Meager, BC
Graham Andrews – Franklin & Marshall College
UBC Collaborators: Kelly Russell, Krista Michol, Martin
Stewart

2. • Why study volcanic dams?
 natural dams 101
 how do dams fail?
 when dams fail…
• Mount Meager and the 2360 BP eruption
Outline
 non-volcanic debris flows
• Volcaniclastic stratigraphy and dam architecture
• Welded vs . non-welded
 porosity
 compaction
• Paleo-Salal Lake
• Lessons for the future?

3. The 2360 BP eruption of Mt. Meager gives us a
unique opportunity to explore:
This study
1. the volcanic damming (and failure) process,
2. the timescales of damming and failure,
3. the influence of changing properties in
volcanic dams, and
4. ways to prepare for and mitigate against
similar events in the future.

4. beaver dam
Natural Dams
Natural dams
form in 3
different ways: Wikipedia
 biological ESA
 geological
 glacial
Lake Sarez,
Tajikistan

5. USGS
Natural Dams - Ice
Hubbard
glacier, AK
In 1986 and 2002 the surging Hubbard Glacier
temporarily created a 5 km3 “Russell Lake” in 24
hours before it burst.
The outburst flood generated was ≤105 m3/s.
Lake Missoula floods ≤107 m3/s

6. Dams typically fail in 1 of 2 ways:
• overtopping,
• undermining by seepage.
How dams fail
1. - Overtopping
gradually but rapidly
erodes the top and the
downstream side of the
dam – like a knick-point
migration – until the
dam fails.
• e.g., “Johnstown Flood”, PA,
1889 – 2,200+ fatalities
Wikipedia

7. Dams typically fail in 1 of 2 ways:
• overtopping,
• undermining by seepage.
How dams fail
2. - Undermining removes a large section of the dam
in one go. It usually follows seepage of water into the
porous dam structure or bedrock.
e.g., Teton Dam, Fremont and Madison Counties, ID, 1976
St. Francis Dam, CA, 1928 – 450+ fatalities

8. How dams fail

9. dam
failure USGS test
excavations
How dams fail
dam
remnant
flood
debris
Teton Dam, ID USGS
Thick dams are rarely completely destroyed – usually most of
the dam is left and the water escapes out the side of the dam.

10. USGS

11. Dam failure - effects Catastrophic dam failures are devastating to the
environment downstream.
Outburst floods form debris flow deposits
commonly contain house-sized clasts of rock / dam
material weighing 10s of metric tons (10,000s of
lbs).
13t piece of St. Francis dam
Volcanic outburst – ½ mile from source
floods and mudflows
are called lahars.
USGS

12. Andrews et al., in prep.
Quaternary Garibaldi
volcanic belt (GVB)
Mount Meager
Northernmost of 3
deeply-eroded felsic
stratovolcanoes:-
Garibaldi, Cayley,
and Meager.

13. Andrews et al., in prep.
Quaternary Garibaldi
volcanic belt (GVB)
Mount Meager
Northernmost of 3
deeply-eroded felsic
stratovolcanoes:-
Garibaldi, Cayley,
and Meager.
Part of the Cascade
arc and Pacific “Ring
of Fire”.

14. Andrews et al., in prep.
Quaternary Garibaldi
volcanic belt (GVB)
Mount Meager
Northernmost of 3
deeply-eroded felsic
stratovolcanoes:-
Garibaldi, Cayley,
and Meager.
Part of the Cascade
arc and Pacific “Ring
of Fire”.
Prospective for
geothermal and
hydroelectric
power.

15. The last eruption formed
the extensive ‘Bridge
River’ tephra (14C - 2360
Mount Meager
BP) – the last explosive
eruption in Canada.
Geological Survey
of Canada

16. The Meager volcanic
complex is a series of
Mount Meager
4 edifices built one on
top of the other.
Each new edifice is
further north.
The volcano has
experienced >5 periods
of glaciation.
The volcano is highly
unstable.
Hickson et al., 1999

17. Meager Peak
With over 1800 Capricorn Plinth Peak (P) –
m of relief Mt Peak 2677 m
Meager is very
Mount Meager
rugged and
steep.
The 2360 BP looking NW
vent (V) is in a
glacier-filled col
~800 m above
the adjacent
Lillooet valley
floor (1 in 3
gradient).
looking SW

18. Mount Meager

19. Andrews et al., in prep.
2360 BP Pebble Creek Formation
Pebble Creek Formation

20. Andrews et al., in prep.
Pebble Creek Formation
misfit stream
and canyon

21. Keyhole Canyon & misfit Lillooet River
slot-canyon =
2300 years of
‘normal’ erosion
looking NW - upstream
looking SE - downstream
90 m
2000 m
300 m

22. The 2360 BP eruption went through 3 major
phases:
2360 BP eruption
1. sub-Plinian explosive eruption deposited
dacite ash across much of British Columbia.
Local pumice fall deposits and thin ignimbrites
(pyroclastic flows).

23. The 2360 BP eruption went through 3 major
phases:
2360 BP eruption
1. sub-Plinian explosive eruption deposited dacite ash across much
of British Columbia. Local pumice fall deposits and thin
ignimbrites (pyroclastic flows).
2. Vulcanian explosions of a hot lava dome
generated welded and non-welded block-and-
ash flow deposits.

24. The 2360 BP eruption went through 3 major
phases:
2360 BP eruption
1. sub-Plinian explosive eruption deposited dacite ash across much
of British Columbia. Local pumice fall deposits and thin
ignimbrites (pyroclastic flows).
2. Vulcanian explosions of a hot lava dome generated welded and
non-welded block-and-ash flow deposits.
3. Collapse of an extrusive dacite lava flow
generated more non-welded block-and-ash
flow deposits
explosive
effusive

25. Montserrat – Feb 5th 2010
Vulcanian eruptions
block &
ash flows
MVO

26. Vulcanian eruptions

27. 2360 BP eruption
Andrews et al., in prep.

28. lake sediments
Pebble Creek Formation
block & ash deposits
lahar deposit
Andrews et al., in prep.

29. 1 Pebble Creek Formation
2
Event stratigraphy
downstream
upstream
Andrews et al., in prep.
3

30. block & ash
deposit dam
Volcanic dam
Michol, Russell, Andrews, JVGR 2008
780 masl
680 masl

31. Volcanic dam
non-welded B&A
welded B&A
Michol, Russell, Andrews, JVGR 2008

32. Volcanic dam welded B&A
• strongly-welded
glassy matrix,
• blocks of dense
dacitic obsidian,
• ~31% compacted,
• deposited ‘hot’
(>600 °C),
• “HARD” like lava
Michol, Russell, Andrews, JVGR 2008

33. Volcanic dam non-welded B&A
• unconsolidated to
weakly-indurated
ashy matrix;
• blocks of dense
dacite;
• deposited ‘cold’
(<600 °C);
• “SOFT” like sand
 easily eroded

34. Lava dome collapse

35. Welded vs. non-welded bubble-wall shard equant, blocky shard
non-welded
Michol, Russell, Andrews, JVGR 2008

36. Welded vs. non-welded bubble-wall shard equant, blocky shard
non-welded
welded
1mm
deformed pumice lapillus flattened shard

37. Welded vs. non-welded
non-welded
loose to moderately lithified
32 – 40% porosity
bulk density 1.4 – 1.5 g/cm3
strongly welded
welded
5 – 16% porosity
bulk density 2.1 – 2.3 g/cm3
1mm
Michol, Russell, Andrews, JVGR 2008

38. block & ash
deposit dam
1 – rapid filling of the
valley by block & ash
flow deposits – dam
axis ~780 masl
2 – 31% compaction and
welding in the B&A
deposits – dam axis
~740 masl
3 – dam breached
eroded by the outburst
flood – canyon floor at
570 masl
Andrews et al., in prep.

39. 1 Pebble Creek Formation
2
Event stratigraphy
downstream
upstream
Andrews et al., in prep.
3

40. Salal Creek
delta
Paleo-Salal Lake
Salal Creek
Andrews et al., in prep.
top of delta & max
elevation of Salal Lake
 740 masl
Lillooet
River
valley floor  dam
680 masl

41. lake
reconstruction
max dam elevation
Paleo-Salal Lake
– 780 masl
Andrews et al., in prep.
delta – 740 masl min dam –
740 masl
max Salal Lake
- 740 masl
Salal Lake  volume = original valley floor
~550 x 106 m3 (~0.55 km3) – 570 masl
• Salal Lake grew no higher than 740 masl (≤160 m
deep)  then breached
• ¼ the volume of American Falls reservoir

42. 1 Pebble Creek Formation
2
Event stratigraphy
downstream
upstream
Andrews et al., in prep.
3

43. downstream –
lahar deposit
Outburst flood
polygonal-jointed margins  huge, rounded
eroded and deposited hot blocks of
welded B&A
deposit
poorly-sorted,
non-welded matrix Andrews et al., in prep.

44. Keyhole Canyon & misfit Lillooet River
cooling joints
perpendicular to
the canyon walls
looking NW - upstream
 excavated
above 600°C
looking SE - downstream

45. • Salal Lake filled for ~90 days at 161 m3/s (at
present flow-rate)
WCS model - instantaneous (catastrophic) 6000 m2 opening
Outburst flood
in the dam [undermining & rapid overtopping]
max. elevation of max. volume of Salal
dam axis and lake at Lake at failure
failure
residual lake residual lake
elevation volume
• Salal Lake drained in ~13 hours; peak deluge flux
of ~2.7 x 105 m3/s Andrews et al., in prep.

46. volcanism and dam-building
Dam evolution
31% viscous
compaction in ~90 days
Andrews et al., in prep.

47. 1. B&A flows rapidly dammed the
Lillooet River and created Salal Lake
(≤160 m deep),
Summary
2. the dam failed after ~90 days,
3. the lake drained catastrophically and
generated an outburst flood (lahar)
that eroded the canyon,
4. welding was interrupted by dam failure
◦ 31% viscous compaction achieved in ~90 days
(consistent with ‘fast’ experimental rates).

48. 1. B&A flows rapidly dammed the
Lillooet River and created Salal Lake
(≤160 m deep),
Summary
2. the dam failed after ~90 days,
3. the lake drained catastrophically and
generated an outburst flood (lahar)
that eroded the canyon,
4. welding was interrupted by dam failure
◦ 31% viscous compaction achieved in ~90 days
(consistent with ‘fast’ experimental rates).

49. 1. B&A flows rapidly dammed the
Lillooet River and created Salal Lake
(≤160 m deep),
Summary
2. the dam failed after ~90 days,
3. the lake drained catastrophically and
generated an outburst flood (lahar)
that eroded the canyon,
4. welding was interrupted by dam failure
◦ 31% viscous compaction achieved in ~90 days
(consistent with ‘fast’ experimental rates).

50. 1. B&A flows rapidly dammed the
Lillooet River and created Salal Lake
(≤160 m deep),
Summary
2. the dam failed after ~90 days,
3. the lake drained catastrophically and
generated an outburst flood (lahar)
that eroded the canyon,
4. welding was interrupted by dam failure
◦ 31% viscous compaction achieved in ~90 days
(consistent with ‘fast’ experimental rates).

51. Sustained high-flux eruptions into drainages will build
volcanic dams.
Natural dams are doomed to fail, usually catastrophically
and without warning.
Proximal and downstream evacuation plans must include
Lessons?
syn- and post-eruption scenarios.
Volcanic dams cannot be geo-engineered to be safe and
stable  must be removed before a large lake builds-up.

52. USGS

53. Welded vs. non-welded