Powerpoint template for scientific posters (Swarthmore College)

1



Response of the Groundwater Wells in Alaska to
Near and Distant Large Earthquakes Samik Sil
(1) and Jeffrey T. Freymueller (2)University Of
Alaska Fairbanks, Geophysical Institute, 903
Koyukuk Drive, Fairbanks, AK-99775

• Conclusions
• We observed groundwater level changes in Alaska
  following the Nenana Mountain (October, 2002, Mw
  6.7), the Denali fault (November, 2002, Mw 7.9)
  and the Sumtra-Andaman earthquakes (December,
  2004, Mw 9.1-9.2).
• All the wells responded to the Denali earthquake,
  Group-I wells of the Tanana basin area responded
  to all three earthquakes and Group-II responded
  to the Denali and the Nenana earthquakes.
• 3. Response of EDOP wells and KB6 well to the
  Denali earthquake can be explained by poroelastic
  theory.
• Response of Group-I wells were similar (coseismic
  step-like water level increase and gradual
  postseismic decrease) to all the earthquakes. The
  average water level steps is proportional to
  horizontal ground velocities. So dynamic strain
  from ground shaking caused the water level
  changes in this group of wells. We also fit the
  postseismic decay of data using equation 2. A low
  value of decay constant indicates the pore
  pressure sources are very near to the wells.
• Group-II wells showed gradual increases or
  decreases in water level following the Nenana and
  Denali earthquakes. Fitting of postseismic
  response to equation 2 obtained higher value of
  z. So we proposed near by fracture
  formation/cleaning due to ground shaking changed
  the fluid pressure gradient and thus the chages
  occurred.
• 6. Multiple mechanisms are responsible for the
  water level changes in Alaska due to near and
  distant large earthquakes.

Introduction We observed water level changes in
22 groundwater wells in Alaska following the
large 2002 Nenana Mountain, the Denali fault and
the Sumatra-Andaman earthquakes. Multiple
mechanisms are responsible for the variable
temporal pattern and the magnitude of the
observed water level changes. For the wells in a
consolidated confined aquifer system, diffusion
of earthquake induced pore pressure is the main
cause of water level changes Roeloffs, 1998.
Whereas for the wells in partially confined
unconsolidated aquifer, fracture formation due to
an earthquake Brodsky et al., 2003 and
consolidation of unconfined aquifer by earthquake
induced dynamic strain (liquefaction) Manga et
al., 2003 are both important mechanisms. For
each well the dominant mechanisms are the same
for all earthquakes. Though more than 10000 km
away, the Sumatra earthquake induced water level
changes are very much consistent with the changes
due to the local large earthquakes. Along with
the determination of the tectonics mechanics of
the well water level changes in Alaska, we also
determined general hydrological parameters of the
wells which can be helpful for future studies.
The Sumatra-Andaman Earthquake

• Basic Theory
• 1)Poroelastic Theory of water level changes
• (1)
• For a poroelastic material water level change (h)
  is directly proportional to the volumetric
  strain( ).B is the Skemptons coefficient and
  ?u is the undrained Poissons ratio. B is
  estimated from the response of atmospheric
  pressure changes.
• 2)Shaking induced changes
• Dynamic strain Water level changes are directly
  proportional to the horizontal ground velocities
• Fracture formation Ground-shaking induced
  fracture formation can develop a new pore
  pressure gradient, which diffuses according to
  the following equation
• 
  where (2)
• Here p is the pore pressure and c is the
  hydraulic diffusivity. Higher value of decay
  constant (t) indicates a large distance of pore
  pressure source (z).

Sil and Freymueller 2006 detected step-like
water level changes in Group-I wells following
the 2004 Sumatra-Andaman earthquake. The changes
occurred during the arrival of largest surface
waves from Sumatra to Fairbanks. No other wells
showed any detectable changes in water level.
Fig. 6. Step-like changes are observed in Group-I
wells following the Sumatra-Andaman earthquake
Discussions
Using equation 1 and the calculated volumetric
strain (Fig. 4), expected water level changes are
calculated for the EDOP and KB6 wells. The
expected water level changes from poroelastic
theory matches the observed water level changes
excellently. This indicates poroelasticity is the
cause of water level changes in EDOP and KB6
wells.
The Denali Fault Earthquake
Fig. 2. Step-like water level drops were observed
and estimated at the EDOP wells. Step-like rises
were observed and estimated in Group-I and KB6
wells. For the estimation of steps following
equation is used to fit the time series
Well locations and aquifers The EDOP
wells are drilled into an upland aquifer system
contained within highly fractured bedrock
consisting of metamorphic and igneous rocks. The
well MCGR is situated almost 15 km east of the
EDOP wells, and is drilled into a confined
aquifer system that consists of Quartz-mica
schist of pre-Jurassic age. A cluster of 18 wells
are situated southwest of the city of Fairbanks,
in the Tanana Valley area of Alaska, which is
covered by thick deposits of alluvium and loess.
They are all drilled into Quaternary Chena
alluvial deposits. The aquifer system is
unconsolidated and is considered to be confined
during the winter because of the presence of a
permafrost layer. During summer seasons the
aquifer is not confined and the hydrographs of
the wells vary systematically with the variation
of the water levels of the near by Chena (Group I
wells of the figure) and Tanana rivers (Group II
wells). Another well, KB6, is situated 29 km
north of Anchorage, drilled into Quaternary sand
and gravel. Water level data were collected at an
interval of one hour in all of the wells using a
submersible pressure transducer at the time of
the Denali earthquake. At present, the water
level are monitored at 15 minute intervals for
the Tanana basin wells. Resolutions of water
level measurement are 0.3 mm and 3 mm, for the
Tanana basin and other wells respectively.

Fig. 7. Comparison between predicted and observed
water level changes from the EDOP and KB6 wells
using eqn.1.
Literature cited Brodsky, E., E. Roeloffs, D.
Woodcock, I. Gall, and M Manga, A mechanism for
sustained groundwater pressure changes induced
by distant earthquake. J. Geophys. Res.,108,
2390, 2003. Manga, M., E. Brodsky, and M, Boone,
Response of streamflow to multiple earthquakes,
GRL,30,2003. Roeloffs, E., Poroelastic
techniques in the study of earthquake related
hydrologic phenomena, Advances in Geophysics, 37,
135-189,1996. Sil, S., and J. T. Freymueller.
Well water level changes in Fairbanks, Alaska,
due to the great Sumatra-Andaman earthquake, EPS,
58, 181-184, 2006.
Group-I wells always showed a water level rise
after all 3 earthquakes. The average water level
rise from all the wells of this group is
proportional to the horizontal ground velocities
due to 3 studied earthquakes. This indicates
dynamic strain is the cause of water level
changes in Group-I wells. Fitting of postseismic
water level changes with an error function and
decay constant (equation. 2) returned a small
value of t, which indicates a highly localized
pore pressure source.
Fig. 3. Transient changes were observed in
Group-II wells after the Nenana Mountain and
Denali Fault earthquakes .
Fig. 1. Studied well locations with names. RHS is
the zoomed view of the red box. The Tanana basin
wells are divided into two groups according to
their seasonal behaviors.
Fig. 7. Comparison between average water level
rises and ground velocities for the Group-I wells
for all 3 earthquakes.
Fig. 4. We determine the volumetric strain due to
the Denali fault earthquake. Only EDOP wells and
the KB6 well showed the direction of water level
changes consistent with poroelastic theory. Water
level rise in Group-I wells and transient changes
in Group-II wells are not being explained by this
theory.
Fig. 8. Postseismic water levels of the Group-II
wells are modeled with an error function with a
decay constant for the Nenana (left) and the
Denali (right) earthquakes.
Acknowledgments We thank Ms. Heather R. Best of
USGS and Dr. David Barnes of UAF for providing us
water level data. We are also thankful to Dr.
Michael Manga of University of California,
Berkeley and Dr. Emily Brodsky of University of
California, Los Angeles, for several technical
discussion which help to improve this work. This
project is supported by National Science
Foundation EAR-03/04/0 For further information,
contact (1)ftss_at_uaf.edu (2) jeff_at_giseis.alaska.edu
.
The Nenana Mountain Earthquake
Fig. 5. After the Nenana Mountain earthquake
step-like water level rises and transient changes
(Fig.3) were observed in Group-I and Group-II
wells respectively. The similar pattern of water
level changes from both the earthquakes indicates
mechanisms of water level changes are same as for
the Nenana Mountain earthquake.
Postseismic gradual changes of water level
following the Nenana mountain and the Denali
earthquake for Group-II wells are fitted with a
error function and decay constant. The very high
value of the decay constant indicates a larger
value of z (equation. 2). We proposed fracture
cleaning/formation near well sites and thus
changing the hydraulic gradient is the cause of
water level changes in this group of wells.