2nd Volcano-Ice Interaction
on Earth and Mars
Conference
University of British Columbia, Vancouver, BC, Canada
June 19th-22nd 2007
Abstracts received
Injection of CO2 From Sills Into Sub-cryospheric Aquifers as a Mechanism of Geyser
Eruption on Mars (alternative title: women are from Venus, geysers are from Mars)
A.S. Bargery (Department of Environmental Science, Lancaster University, Lancaster, UK,
LA1 4YQ; ph: +44-1524-593-975; fax: +44-1524-593-985; email: a.bargery@lancaster.ac.uk
); [*L. Wilson*] (Department of Environmental Science, Lancaster University, Lancaster,
UK, LA1 4YQ; ph: +44-1524-593-889; fax: +44-1524-593-985; email:
l.wilson@lancaster.ac.uk)
Since the water that formed the outflow channels on Mars originated from depth within the
crust, the likelihood is that it interacted, at least during part of its evolution and migration to
the surface, with CO2, and most likely had a high solute proportion of CO2 (Sears et al.,
2004). The water forms aquifers within porous rock, beneath a water-ice-rich cryosphere
(Clifford, 1993). At these depths, the pressure is a few tens of MPa, and the temperature is
~350 K. Therefore, H2O and CO2 will both be in the liquid state (Miller, 1974; Wagner et al.,
1994; Span and Wagner, 1996; Stewart and Nimmo, 2001) and will be completely miscible
with each other; if they co-exist, then they can be treated as a mixture.
One of the questions that should addressed is how could the CO2 have come to exist in the
groundwater? Several pathways have been suggested for the CO2 cycle. CO2 is extant at the
Martian poles and could have been re-deposited at lower latitudes in a similar fashion to
water ice during periods of higher obliquities via sub-basal melting of the polar caps
(Clifford, 1987, 1993; Clifford and Parker, 2001). However, Gaidos and Marion (2003) have
argued that this is unlikely without a cold early Mars to sequester most of the CO2 in the
early Noachian atmosphere as carbonates in the upper regolith.
Here, an alternative mechanism is proposed, whereby emplacement of a thick sill beneath an
aquifer can readily supply an aquifer with a significant amount of magmatic CO2 as the sill
solidifies and degasses. There is an exchange of CO2 between the aquifer, the crust, and the
atmosphere, due to the pressure of the volatile. This scenario would be compatible with a
probable crustal geology dominated by extrusive and intrusive volcanics (Nimmo, 2005). The
extensive layering observed in the walls and floors of Valles Marineris, for example, may
include intrusive igneous sills (Malin and Edgett, 2000). This process would also be
independent of the climatic conditions, and would require only that an aquifer was present in
the vicinity of the intrusion.
Fountain and geyser activity on Mars may not require a large geothermal heat flux. A cold
geyser appears to be a contradiction in terms, but a combination of CO2 and effervescing
groundwater can create a spectacular water fountain on Earth. For example, Crystal Geyser in
northern Utah (e.g., Waltham, 2001) forms with no geothermal heat and no steam flashing.
Instead, it is driven by CO2, and there are many similarities with hot geysers in the cyclic
production of gas to power the fountains. The Little Grand Wash Fault constitutes a
groundwater conduit from deep artesian aquifers. Analogues to faulting in the Martian crust
include graben features such as the source regions of the Amazonian outflow channels, for
example Cerberus Fossae, Medussae Fossae. Calcite travertine is sometimes found in the
outcrop vicinity of these cold geysers on Earth; other minerals may be found in the vicinity of
the Martian geysers or fissures. If such evaporite minerals were identified at these source
regions, it would provide evidence of evaporite deposition from CO2 – H2O fountains.
POSTER
CORRESPONDING AUTHOR: A.S. Bargery
FIRST AUTHOR IS A STUDENT
1
Distinctive Landforms Produced by Permafrost-Volcano Interactions, Arctic Alaska
[*James E. Beget*] (Alaska Volcano Observatory and Geophysical Institute, University of
Alaska, Fairbanks, AK. 99775-5780; ph: 01 907 474-5301; fax: 01 907 474-5163; email:
ffjeb1@uaf.edu); Jeffrey Kargel (University of Arizona, Tucson, Arizona, 85721); Rick
Wessels (Alaska Volcano Observatory, U. S. Geological Survey, Anchorage, AK. 99508),
and Paul Layer (Geophysical Institute, University of Alaska, Fairbanks, AK. 99775-5780)
Volcano-Permafrost interactions are responsible for several different kinds of distinctive
landforms at different sites in Arctic Alaska, where basaltic eruptions of various ages and
sizes have occurred in areas containing disseminated and massive ground ice. On the Seward
Peninsula at ca. 66° N, a series of giant composite explosion craters as much as 10 km in
diameter were produced by numerous explosions where basaltic magma encountered water
released by cryo-magmatic interactions. The resultant Espenberg Maars are the largest known
maar craters on earth (Beget et al., 1996). New Ar/Ar dates indicate they formed at ca. 18 ka,
70 ka and 150 ka, and so are correlative with times of cold climate and ground ice more than
ca. 100-m-thick ground ice during past full glacial periods (i.e. during marine isotope stages
2, 4 and 6). The sedimentology of frozen ground and the thermodynamics of magma-ground
ice interactions can facilitate the generation of repeated fuel-coolant explosions.
On the Seward Peninsula at Imuruk Lake at ca. 65° N the Lost Jim lava flow advanced for
more than 5 km over ice-rich permafrost and into the lake. The eruption occurred only a few
thousand years ago. The basaltic lava flow is bounded by steep flow fronts and marginal
levees as much as 20 m high. These flow margins terminate in zones of complex thermokarst
collapse features that record melting of ground ice under the lava. These features indicate
significant permafrost-magma interactions occurred at the thin lava flow margins, and these
interactions strongly influenced the ability of the lava flows to travel across ice-rich areas,.
The thermokarst features also demonstrate that the lava flows were able to melt significant
amounts of ground ice even at thin lava flow margins several kilometers from the vent.
On the Yukon-Koyukuk Delta the Ingakslugwat Hills area at ca. 61.5° N. contains several
unusual composite volcanoes as much as 10 km long and 400 m high composed largely of
unconsolidated pyroclastic ejecta. Ingakslugwat, the Inupiaq name for this site, can be
translated as "mud hills", but coarser volcaniclastic material is common around the vents. The
volcanoes preserve numerous intersecting arms of explosion craters of various ages that
record multiple cycles of cryo-magmatic explosive volcanism. Our currently preferred model
for the origin of the "Ingaksluwat" volcanoes involves cycles of eruptive activity occurring
along linear fissures interspersed with intervals of quiescence and ground ice formation.
ORAL
CORRESPONDING AUTHOR: James Beget
Ice Thickness Estimates Based on Volatile Content of Pillow Glasses from the
Mount Edziza Volcanic Complex, northern British Columbia
[*B.I. Cameron*] (Department of Geosciences, University of Wisconsin-Milwaukee,
P.O. Box 413, Milwaukee, WI, 53201, USA, ph: 414-229-3136; fax: 414-229-5452;
email: bcameron@uwm.edu); M.D. Wright (Department of Geosciences, University of
Wisconsin-Milwaukee, P.O. Box 413, Milwaukee, WI, 53201, USA); I.P. Skilling
(Department of Geology and Planetary Science, 200 SRCC, University of
Pittsburgh, Pittsburgh, PA, 15260, USA); B. Edwards (Department of Geology,
Dickinson College, Carlisle, PA, 17013, USA)
2
The volatile content of glaciovolcanic pillow glasses offers a promising
means of estimating ice thickness. The pressure dependence of water solubility
in magma allows the estimate of ice thickness based on the measured volatile
concentrations in pillow glasses formed by volcano-ice interactions. Simply,
the higher the (lithostatic, hydrostatic, cryostatic) pressure, the more water
that can be dissolved in the magma. Previous studies have measured changes in
water content in glasses from vertical sections though pillows and
hyaloclastites deposits to provide information on fluctuations in ice thickness
through time as the edifice grew under ice cover. In this study we take the
approach of sampling glasses from a single spatially extensive pillow unit to
assess ice surface topography at a particular eruptive moment in the edifice's
history. Basaltic pillow lavas erupted from Tennena Cone at the Mount Edziza
Volcanic Complex (MEVC) in northern British Columbia occur as sinuous ridges
with near vertical margins. Basaltic lavas with steep margins suggest
confinement by glacier ice. Thus, these R-channel fills provide convincing
evidence that the pillow lavas formed in an entirely sub-ice environment.
Voluminous sheet flows of pillow lava proximal to the vent fed the narrow
channels from the east through a notch in a sub-ice cliff. A reduction in
meltwater volume with distance from the vent and/or ponding of meltwater and
lava against a thicker ice wall downstream of the sub-ice cliff might explain
the transition from proximal pillow sheets to more distal small channel.
Initial water concentrations and hydrogen isotopic composition of the
pillow glasses were measured by manometry and mass spectrometry in the Stable
Isotope Laboratory at Southern Methodist University. Water content of the
glasses was also measured by Fourier Transform Infrared (FTIR) spectrometry at
the University of Wisconsin-Milwaukee. Initial water contents and hydrogen
isotopic compositions from the Tennena Cone pillow rinds range from 0.66 to
0.83 wt% and –78.5 to -127.1 per mil, respectively. The eruption pressures were
estimated using the measured water contents and the VolatileCalc program.
Assuming a basaltic composition with 49 wt% silica, an eruption temperature of
1000 degrees Celsius, and 0 ppm carbon dioxide, eruption pressures range from
39.8 to 65.0 bars. Assuming an ice density of 0.917 g cm-3, these pressures
correspond to ice thicknesses between 442 to 722 m. These ice thickness
estimates are strongly dependent on the carbon dioxide content of the glasses.
The absence of carbonate peaks during the FTIR analysis suggests that the
glasses contain less than 30 ppm, the accepted detection limit. Assuming a
carbon dioxide concentration of 20 ppm, the ice thickness would range between
934 and 1215 m. Ice thickness estimates over the west-dipping surface
topography generally increase westward suggesting near-horizontal ice surfaces.
Future measurements of carbon dioxide by manometry should make ice thickness
estimates more accurate and meaningful.
ORAL
CORRESPONDING AUTHOR: B.I. Cameron
3
Overview of Quaternary Volcanism in Western Canada
[*John J. Clague*] (Department of Earth Sciences, Simon Fraser University, Burnaby, BC,
Canada, V5A 1S6; ph: 1-604-291-4924; fax: 1-604-291-4198; e-mail: jclague@sfu.ca)
Volcanic eruptions in Canada during the Quaternary Period are limited to British Columbia
and Yukon Territory. Although many eruptions have occurred in the past several thousand
years, none is historic; the youngest eruption probably happened in northwestern B.C.
sometime in the 1800s. Eruptions have occurred in the Holocene at shield volcanoes,
stratovolcanoes, and cinder cones. Most volcanoes have been sites of a single pulse of
activity during which one or more small cones were built or basalt erupted to form local
blocky flows. In contrast, other volcanoes are sites of recurrent explosive activity that has
produced pyroclastic flows, lahars, and ash clouds. The largest eruption of this type occurred
about 1000 years ago near the Alaska-Yukon boundary, depositing a layer of ash over an area
of tens of thousands of square kilometers in southern Yukon and adjacent District of
Mackenzie. Quaternary volcanoes in B.C. and Yukon occur in several clusters and linear
belts that can be related to plate interactions off the west coast of North America. A belt of
volcanoes in southwest B.C. marks the north end of the Cascade volcanic arc, which is a
product of subduction of the Juan de Fuca plate beneath North America. A large belt of
volcanoes in northwest British Columbia and southern Yukon is related to rifting caused by
crustal extension east of the Pacific-North America plate boundary. A smaller group of
volcanoes in west-central British Columbia may have formed above a hot spot in the mantle.
An interesting aspect of late Cenozoic volcanism in British Columbia is the association of
some of it with the ice sheet that repeatedly covered the province during latest Pliocene and
Pleistocene time. Ice-contact flows, tuyas, and other subglacial volcanic landforms are
common in several parts of B.C. Ice-contact volcanism is probably associated with glacioisostatic deformation of the crust during periods of ice-sheet decay and, perhaps, during icesheet growth.
INVITED ORAL
CORRESPONDING AUTHOR: John J. Clague
How subglacial drainage systems work
[*Garry K. C. Clarke *] (Earth & Ocean Sciences, University of British Columbia,
Vancouver, Canada; ph: 1-604-822-3602; e-mail: clarke@eos.ubc.ca)
An inescapable problem of studying ice–volcano interactions is that subglacial drainage
systems and volcanic systems are complicated in themselves. What we know about subglacial
drainage systems draws from proglacial, supraglacial and subglacial measurements, the
morphology of deglaciated surfaces and the application of hydraulic theory. In this
presentation I shall review the present state of knowledge and our ability to encode this
knowledge in computational models of “normal” and volcanically-forced systems.
CORRESPONDING AUTHOR: Gary Clarke (clarke@eos.ubc.ca)
INVITED ORAL
4
Extreme Life in Subglacial Volcanoes: Implications for Astrobiology on Mars
[*Claire Cousins*] (University College London, Gower Street, London, WC1E 6BT, UK; ph:
02076792577; email: c.cousins@ucl.ac.uk; (Supervisors: A. Jones, I. Crawford, J. Ward).
Mars has been a prime contender in the search for extraterrestrial life in the past decade; with
many suggestions as to what Martian environments could be viable habitats. Subglacial
volcanic systems are a promising extreme environment, as they cover such a broad spectrum
of environmental conditions. These environments range from the cold glacial environment to
the high temperature subsurface, and have the potential to support an equally varying range
of microbial life forms. The key element of this environment is the presence of a
hydrothermal system, which forms during a subglacial eruption. Hydrothermal systems on
the Earth’s terrestrial surface and at Mid-ocean ridges are well known to harbour diverse
forms of microbial life, and so there is reason to believe subglacial hydrothermal systems will
be much the same. Additionally, this subsurface environment is shielded from many
destructive forces, such as UV radiation and heavy meteor bombardment and is independent
of light and oxygen.
The study of these systems will be focused primarily in Iceland. Iceland provides the perfect
locality for studying such an environment as subglacial volcanism predominates and
sampling can take place across several different volcanic systems. Many locations in Iceland,
particularly the NE have been long thought of as Martian analogues. Seasonal sampling will
take place here over the next 3 years to characterise a range of subglacial environments, as a
whole and their components individually. Collected samples are cultured, identified and
understood, in terms of biochemistry and metabolism. This will be achieved by culturing
microbes directly from samples, isolating their DNA for genetic identification, locating
microfossils within volcanic glass, olivine and pyroxene, and isolating bacterial viruses. In
conjunction with this, biomarkers will be characterised and related to the future ESA
Exomars mission.
POSTER
CORRESPONDING AUTHOR: Claire Cousins (c.cousins@ucl.ac.uk)
FIRST AUTHOR IS A STUDENT.
The use of TGA-MS in the Analysis of Perlites and Subglacially Erupted Volcanic Glass
From Iceland
Jo Denton (Department of Environmental Science, Lancaster University, Lancaster LA1
4YQ; ph +44 (0)1524 593975; e-mail j.denton1@lancaster.ac.uk); [*Hugh Tuffen*]
(Department of Environmental Science, Lancaster University, Lancaster LA1 4YQ; ph +44
(0)1524 593571; e-mail h.tuffen@lancaster.ac.uk); Jennie S Gilbert (Department of
Environmental Science, Lancaster Universtiy, Lancaster LA1 4YQ; ph +44 (0)1524 593022;
e-mail J.S.Gilbert@lancaster.ac.uk)
Perlitised lavas are commonly found in subglacially erupted rhyolitic deposits. Perlite is
silicic glass that contains abundant, intersecting, arcuate and gently curved cracks
surrounding cores of intact glass, generally less than a few millimetres across [McPhie et al.,
1993]. Knowledge of how perlite forms is required for us to better understand how meltwater
and lava interact during and after subglacial eruptions. Perlite formation is currently poorly
understood and there are two contrasting models detailed in the literature. Friedman et al.
[1966] state that diffusion of water into the glass modifies/increases the volume of the
material. The resultant stresses generate the cracks. Marshall [1961] argues that the cracks
appear first and are a mechanical effect of the rapid cooling of the material. These cracks act
5
as conduits through which water can enter from the environment and hydrate the glass. The
formation of perlite is of interest to many branches of subglacial volcanism, for example, the
interaction of meltwater with magma as well as the potential use of degassing patterns as a
palaeo-climatic indicator. In this project, analysis of perlitic samples using TGA-MS will be
combined with a study of perlitic textures to shed new light on how perlite is generated.
A thermogravimetric analyser (TGA) allows the weight of a sample to be measured as a
function of temperature while being subjected to a controlled heating program. The addition
of a mass spectrometer (MS) allows the user to identify the volatiles being exsolved at
specific temperatures/temperature ranges. TGA analyses at Lancaster University can be
carried out on samples that weigh up to 100 mg over a temperature range of 0-1500 ºC. The
balance sensitivity is 0.1 µg.
The TGA-MS has been used previously by the authors to successfully establish a suitable
heating program through which all the volatiles from a sample are driven off. Preliminary
results are presented of analyses of perlitised rhyolitic glass from Bláhnúkur, Iceland. The
temperature at which the volatile species are exsolved and the degassing intensity will reflect
the original thermal history of the material and the relative timing of cooling and fracturing of
the lava.
Results are also presented of analyses of a lava lobe, from Bláhnúkur, which contains discrete
textural zones. These zones range from central microcrystalline rhyolite to an obsidian
carapace and reflect different cooling rates, degassing patterns and/or varying degrees of
meteoric water ingress. The lobe has been described in detail in Tuffen et al. [2001; 2002].
The degassing patterns (e.g. temperature of water exsolution) produced by TGA-MS analysis
reveal information on the structure of the glass which, in turn, reflects the temperature and
overlying pressure of the eruptive environment. This information will address issues
surrounding the degassing and emplacement of subglacial rhyolite lavas.
References
Friedman, I., Smith, R.L., Long, W.D., 1966. Hydration of natural glass and formation of
perlite. Geol Soc Am Bull 77:323-327.
Marshall, R.R., 1961. Devitrification of natural glass. Geol Soc Am Bull 72:1493-1520.
McPhie, J., Doyle, M., Allen, R.L., 1993. Volcanic Textures: A guide to the interpretation of
textures in volcanic rocks. University of Tasmania. Centre for Ore Deposit and Exploration
Studies.
Tuffen, H., Gilbert, J., McGarvie, D., 2001. Products of an effusive subglacial rhyolite
eruption: Bláhnúkur, Torfajökull, Iceland. Bull Volcanol 63:179-190.
Tuffen, H., Pinkerton, H., McGarvie, D.W., Gilbert, J.S., 2002. Melting of the glacier base
during a small-volume subglacial rhyolite eruption: evidence from Bláhnúkur, Iceland. Sed
Geol 149:183-198.
POSTER
CORRESPONDING AUTHOR: H TUFFEN
FIRST AUTHOR IS A STUDENT
6
Physiographic Controls on Glaciovolcanism and the Cordilleran Ice Sheet in the
northern Cordilleran Volcanic Province, Western Canada
[*B. Edwards*] (Dept of Geology, Dickinson College, Carlisle, PA, 17013, USA, ph: 717254-8934; edwardsb@dickinson.edu); J. Osborn (Dept Geo and Geophys, U. Calgary,
Calgary, AB); J.K. Russell (Dept Earth & Ocean Science, U.B.C., Vancouver, BC); I.P.
Skilling (Dept of Geology and Planetary Science, 200 SRCC, University of Pittsburgh, PA);
C. Evenchick (Geological Survey of Canada, Vancouver, BC); I. Spooner (Acadia
University, Halifax, Nova Scotia); K. Simpson (Geological Survey of Canada, Vancouver,
BC); B. Cameron (Dept of Geology, University of Wisconsin-Milwaukee, Milwaukee)
Glaciovolcanism in the northern Cordilleran volcanic province of western Canada
occurs in three different physiographic settings, each displaying unique relationships between
topography, volcanism and glacial ice. The first is represented by the type tuya locality, Tuya
Butte, located on the Tanzilla Plateau in northern British Columbia. This region comprises
broad flatlands with comparatively low-lying hills, and may have been one of the centers of
ice accumulation for the Cordilleran ice-sheet (CIS). Pleistocene volcanism in this area is
dominated by basaltic glaciovolcanic eruptions. The most common glaciovolcanic landforms
are individual tuyas, which seldom range in height more than a few hundreds of meters.
Other edifice morphologies are also present, some of which are primary and others that are
possibly the products of erosion. The subdued relief on the Tanzilla Plateau strongly
influenced edifice morphologies, favoring broad-based tuyas. The hydrological conditions in
these areas likely were controlled by relatively uniform ice thicknesses and gentle
topography. Glaciovolcanic processes and landforms conform to Icelandic and Antarctic
examples described in detail by many authors.
The second type of interaction is found mainly to the east, south and southwest of the
Tanzilla Plateau, in the Cassiar Mountains and the Skeena and Boundary ranges. Volcanism
in these areas is also predominantly mafic, but the greater local relief of these areas had a
much more dramatic influence on glaciation and glaciovolcanism. Alpine-style glaciation
dominated when the CIS was absent, and even when overwhelmed by the CIS basal ice
movement was strongly influenced by the deep, pre-glacial drainages. Accordingly, the
erosional remnants of glaciovolcanic centers tend to be smaller and discontinuous. In the
Skeena Ranges, the isolated outcrops of pillow lava and volcanic breccia are often located
along the tops of arête-like ridges, with presumably temporally-associated volcaniclastic
rocks having collected downslope.
The third type of interaction is geographically limited to three volcanic structures:
Level Mountain, Mount Edziza, and Hoodoo Mountain. All three are within the eastern part
of the Boundary Ranges physiographic province and are large enough to support ice caps that
probably influenced local ice flow, while still being influenced by the CIS at different times.
Mount Edziza and Level Mountain have benches of older lava 1 km+ in elevations, and have
been sites of volcanism long enough that their geothermal outputs may have had an influence
on CIS dynamics much as the modern day Grímsvötn caldera is an important heat source
beneath Vatnajökull in Iceland. At Mount Edziza most of the glaciovolcanic products are
located on top of the main volcanic plateau, which is now elevated ~1000m above the
surrounding stream valleys. Mount Edziza comprises a variety of interesting mafic
glaciovolcanic products, but more uniquely, along with Hoodoo Mountain and possibly Level
Mountain, comprises some of the largest deposits of peralkaline felsic glaciovolcanic
products known. At Mount Edziza and Level Mountain, glacier hydrology of the CIS
probably was controlled by a complex interplay between drainage on the flat plateaus beneath
relatively thinner ice and drainage within adjacent, steep valleys filled with much thicker ice.
ORAL
7
Convoluted Columns and Pillow Piles: Constraints on Lava Cooling from Field, CSD
and Conduction Models
[*B.R. Edwards*] and A. Lloyd, Department of Geology, Dickinson College, Carlisle, PA
17013, USA, edwardsb@dickinson.edu; J.K. Russell (Volcanology and Petrology
Laboratory, Department of Earth and Ocean Sciences, University of British Columbia,
Vancouver, BC); A. Mathews and I. Renfrew (School of Environmental Sciences, University
of East Anglia, Norwich, Norfolk, UK); I. Skilling (University of Pittsburgh, Pittsburgh, PA).
Accurately constraining rates of lava cooling is critical for understanding how fast heat is
released during glaciovolcanic eruptions. Rates of heat loss should at least partly control the
dimensions of lava bodies (e.g. pillow diameters) as well as the sizes and thicknesses of
prismatic joints. Variations in lava morphology and jointing also reflect variations in the
intrinsic physico-chemical properties of the lavas (e.g., thermal diffusivity, thermal
conductivity, density, heat capacity and viscosity). Here, we explore how the time-scales of
cooling vary as a function of composition and geometry. Constraints on cooling rates are
derived from field observations of pillow dimensions, thickness of glassy pillow rinds,
thickness of lava collonades and entabulatures, and crystal size distributions (CSD) in pillow
lava. We show that differences in published values of thermal diffusivity (k) produce
significantly different predictions for time-scales of cooling when used in analytical models
for different lava geometries (e.g., sheet, sphere, cylinder). For example, estimates of k for
basalt lava vary from 10-6 to 3 x 10-7 m2/s. This range of values allows for a 100% increase or
decrease in cooling rates derived from analytical solutions of cooling of spherical pillow
lavas. We will also discuss results of numerical models designed to explore cooling when
values of k are allowed to vary as a function of temperature. The range of cooling time-scales
may have important implications for predicting the rate at which ice is melted during effusive
eruptions beneath ice, and hence may also be important for predicting hazards from
associated floods.
ORAL
Overview of the IAVCEI Volcano-Ice Working Group Online ICON Image Database
[*B.R. Edwards*] (Dept of Geology, Dickinson College, Carlisle, PA, 17013, USA, ph: 717254-8934; edwardsb@dickinson.edu); J. Smellie (British Antarctic Survey, Natural
Environment Research Council, High Cross, Madingley Road, Cambridge, CB3 0ET, UK);
B. Landis (Information Technology, Dickinson College, Carlisle, PA)
As interest in planetary volcano-ice interactions increases, the scientific community has a
growing need for ready access to images documenting all aspects of glaciovolcanism for
teaching and other presentations. The IAVCEI Volcano-Ice Working Group database was
created in ICON to fulfil that need. ICON is an online image database designed to allow users
to sort efficiently through large archives of images, create slide shows, download images, and
access information about the images. Whilst commercial use of the images is prohibited
unless you are the owner of those images, the easy availability and (we hope) wide noncommercial use of an on-line photographic library will help to publicise glaciovolcanic issues
and research much better than was previously possible. At present the database contains 90
8
images, from seven different scientists, with geographic coverage from Antarctica, British
Columbia, Chile, Iceland and Washington State. The purpose of this presentation is to
announce publicly the existence of the database, to demonstrate its uses and to show how
further images can be contributed since, the larger and more inclusive the database, the more
frequently it will get used, thus publicising and enhancing the scientific profile of
glaciovolcanism.
The database can be entered view the VIWG website
(http://volcanoes.dickinson.edu/VIWG/database.html) or at http://icon.dickinson.edu. Users
can login in by clicking the ‘Guest User’ button, and then selecting the ‘Volcano-Ice
Interactions’ database. The database can be searched by several methods. The ‘Keyword
Search’ works for searching metadata in the ‘open text’ fields, including photographer’s
name, image description, and references. Only one word should be entered at a time using
this search. Searches using the dropdown menus, including ‘Type of Eruption’, ‘Primary’ and
‘Secondary Features’, ‘Country’, ‘Type of Volcano’, and ‘Lava Composition’. Searching
without selecting any of the above results displays all of the images in the database. Once
your search results are displayed, they can be sorted by all of the dropdown menu items as
well as by ‘Name’, ‘Photo Date’, and ‘Photographer’. Selecting ‘Full view and description’
produces an image window with a small navigation bar, as well as zoom features. From this
view the image can be downloaded and the full metadata for the image can be viewed. The
data is automatically downloaded as a text file when an image is downloaded.
More advanced users can request a username and password by emailing
edwardsb@dickinson.edu. With login permission users can store image information in ‘My
Favorites’ or create a slideshow, the information for which is then saved in the ICON user
database for easy future access. Administrative login status allows users to add they own
images and metadata as well as edit their own contributions. Alternatively people wishing to
contribute images can send a CD/DVD to Brenda Landis (see address above), who can
upload large numbers of images simultaneously. Users will then only need to edit the image
metadata.
Further good quality images of any aspects of glaciovolcanism are always required, and we
encourage all working in the topic to contribute to the database. The database will provide an
efficient means for people to share important images related to volcano-ice interactions and
hopefully will promote the broader scientific community’s awareness of glaciovolcanism and
its characteristic features.
POSTER
Differences in the Visible Through Short Wave Infrared Reflectance Spectra of Tuff
Ring and Tuff Cone Palagonite Tuffs Versus Tuffs from Subglacial Eruptions:
Relevance for the Earth and Mars
[*Wiliam H. Farrand*] (Space Science Institute, 4750 Walnut Street, #205 Boulder, Colorado
80301, Phone: 720-974-5825, Fax: 720-974-5837, E-mail: farrand@spacescience.org)
Palagonite tuffs from a set of tuff rings and tuff cones and from a smaller set of subglacially
erupted tephras have been examined and their visible to short-wave infrared reflectance (0.4
to 2.5 micrometers) has been examined. Hydrovolcanic features that have been examined
include the Pavant Butte tuff cone in Utah, the Cerro Colorado tuff cone in the Pinacates
volcanic field in Sonora, Mexico, the Ft. Rock tuff ring in Oregon, and Koko Crater tuff
9
cone in Hawaii. Subglacially erupted samples included a piece of palagonite tuff float from
Iceland and sampled collected from the Lone Butte and Crazy Hills complex in southern
Washington.
The palagonite tuff samples from tuff rings and tuff cones display significant 1.4 and
1.9 μm absorption bands and also have minor absorption bands at 2.3 μm (in the Pavant Butte
and Cerro Colorado samples) and 2.2 μm (in the Ft. Rock and Koko Crater samples). In the
subglacially erupted Lone Butte, Crazy Hills and Iceland tuff samples, the 1.4 and 1.9 μm are
present but with weaker band depths and there is only an extremely weak 2.2 μm absorption.
The 1.4 and 1.9 μm absorption bands are caused by vibrational overtones of bound hydroxyl
and water molecules. The greater band depths of these features in the tuff ring and tuff cone
spectra indicate a greater level of hydration than in the subglacially erupted tuffs. The 2.2 μm
absorption is caused by an Al-OH overtone and, given the weak nature of the band, a specific
mineral assignment is difficult, but it could be caused by illite or montmorillonite. The 2.3
μm absorption is caused by either a Fe-OH overtone from nontronite or a Mg-OH overtone
from a mineral such as saponite. These absorption features are better developed in the tuff
ring and tuff cone palagonite tuff spectra and thus indicate a greater degree of clay mineral
development than in the Lone Butte and Crazy Hills samples.
These results are valid only for tephras that have undergone extensive palagonitic
alteration and, given the small number of subglacially erupted tuffs that were examined, must
be considered tentative. Many tuff ring beds do not display extensive palagonitization and
some can be largely pristine. The 0.4 to 1.0 μm reflectance of water-magma and ice-magma
produced tephras can also be compared to a putative tuff ring, the "Home Plate" feature,
examined by the Spirit rover within Gusev crater on Mars. The Home Plate spectra indicate a
low level of oxidation more consistent with a largely pristine set of tephras. The lack of
palagonitization in the Home Plate layers indicates that, if the feature is indeed a tuff ring,
that steam and hot water were almost immediately separated from the tephra beds after their
deposition. In contrast, highly palagonitized materials such as those from the studied tuff ring
and tuff cone and also the subglacially erupted palagonite tuffs became palagonitized due in
large part to deposition of steam and hot water with the tephras and rapid palagonitization.
POSTER
CORRESPONDING AUTHOR: W. Farrand
Subglacial Lakes and Life at the Volcano-Ice Interface
[*E. Gaidos*], (Department of Geology and Geophysics and NASA Astrobiology Institute,
University of Hawaii, Honolulu, Hawaii 96822 USA, gaidos@hawaii.edu), T. Thorsteinsson
(National Energy Authority, Reykjavik, Iceland), Tomas Johannesson (Icelandic
Meteorological Office, Reykjavik, Iceland), Andri Stefánsson (Institute of Earth Sciences,
University of Iceland, Reykjavik, Iceland), Viggó Marteinsson (MATIS-Prokaria, Reykjavik,
Iceland) Brian Glazer (Department of Oceanography, University of Hawaii, Honolulu,
Hawaii USA), Mark Skidmore (Department of Earth Sciences, Montana State University,
Bozeman, Montana USA), and Brian Lanoil (Department of Environmental Sciences,
University of California, Riverside, California USA)
Heat from subglacial volcanoes melts ice, and meltwater will accumulate at minima in the
subglacial hydrological potential, creating lakes. There are at least three such lakes under the
Vatnajökull ice sheet in Iceland: Grímsvötn and the western and eastern Skaftárkatlar lakes.
These lakes range from 1-20 square kilometers in extent and are covered by 300-400 meters
of ice. We have carried out a hydrological, chemical and biological reconnaissance of two of
10
the lakes; Grímsvötn in 2002 and the western Skaftárkatlar lake in 2006. We used specialized
hot-water drilling equipment to penetrate the overlying ice, make temperature profiles in the
water column, and retrieve samples of the lake water for chemical and biological analysis.
Preparations for an investigation of the eastern lake in June are underway.
The two lakes represent contrasting physical, chemical, and biological states. Grímsvötn,
which had experienced a jökulhlaup drainage episode a few months before sampling, was at
the freezing point, was oxic, and had the chemistry of glacial meltwater with no indication of
geothermal fluids. The lake contained a relatively low abundance of cold-adapted microbes
with diverse phylotypes the closest relatives of which were isolates or other phylotypes from
cold environments. In contrast, the western Skaftárkatlar water column was near 4 deg. C.,
strictly anoxic with greater than 1 mM sulphide, and contained an abundant community of
bacteria with extremely low diversity. Anaerobic metabolisms based on conversion of carbon
dioxide and hydrogen to acetate, and reduction of sulphur to sulphide were inferred from
molecular data. The inverted temperature profile of the Skaftárkatlar water column suggests a
mechanism of bottom water formation and lake circulation in which glacial meltwater mixes
with geothermally heated water.
ORAL
CORRESPONDING AUTHOR: Eric Gaidos
Passage Zones, Tuyas and the Stability of Englacial Lakes
[*Magnús T. Gudmundsson*]
(Institute of Earth Sciences, University of Iceland, Sturlugata 7, 101 Reykjavík, Iceland; ph.
+354-525 5867; fax: +354-562 9767; email: mtg@raunvis.hi.is)
It is generally accepted that basaltic tuyas form in eruptions within large glaciers or ice
sheets. Essentially identical landforms are created in sustained eruptions in lakes or the
ocean, with the island of Surtsey, formed off the south coast of Iceland in 1963-1967, being a
good example. Observed eruptions within glaciers have not built tuyas. These eruptions were
relatively short-lived and did not make the necessary transformation from the
phreatomagmatic phase to the effusive phase. Although some variations exist, a striking
feature of most tuyas is the largely horizontal passage zone, indicating a semi-stable level of
the englacial lake in which a tuya grew through the progression of lava-fed deltas. The
apparent stability of these lakes is enigmatic; since ice-dammed lakes in present-day glaciers
are for the most part characterized by gradual rise in lake level, followed by rapid drainage in
jökulhlaups, resulting in a fast drop in lake level sometimes exceed 100 m. It has recently
been suggested that supraglacial drainage may provide an explanation for semi-stable
meltwater lakes during tuya-forming eruptions (Smellie, 2006). However, the mechanism by
which supraglacial drainage may be established and remain stable has not been resolved in a
satisfactory manner. It is likely that tuyas are mainly formed in long-lived low-discharge
eruptions similar to the presently on-going eruption at Kilauea. Most basaltic eruptions start
off with an initial high-discharge phase. A subglacial eruption of this kind will in its early
phases melt large volumes of ice with the meltwater draining away subglacially. If the
eruption continues for an extended period of time, magma discharge is likely to drop.
Considering the thermal regime within the growing tuya, and that the subglacial meltwater
tunnel may close, even if only temporarily, a supraglacial drainage path may form along the
depression in the ice surface overlying the initial subglacial drainage path. The rate of
incision of the supraglacial channel will depend on the thermal regime of the glacier,
11
meltwater temperature and flow rate. At the moderate meltwater flow rates expected for low
magma discharge, the development of a supraglacial channel and the stability of the water
level at the eruption site will depend on the temperature of the upper layers of the glacier. In a
temperate glacier incision will lead to gradual lowering of water level and is unlikely to result
in a semi-stable water level. However, in a thick polythermal glacier, surface ice
temperatures well below zero will slow down or stop incision. This suggests that conditions
for stable water level with supraglacial drainage can develop in polythermal glaciers, possibly
explaining the horizontal passage zones in tuyas. This conforms to the fact that known tuyas
in Iceland were formed during glacial periods, within glaciers that must have had equilibrium
lines at much lower elevations than seen in present-day glaciers.
Reference
Smellie, J.L. 2006. The relative importance of supraglacial versus subglacial meltwater
escape in basaltic tuya eruptions: An important unresolved conundrum. Earth Science
Reviews, 74, 241-268.
ORAL
CORRESPONDING AUTHOR: MT Guðmundsson
Melting Rates and Jökulhlaup Potential in Flank Eruptions at Ice-Covered
Stratovolcanoes
[*Magnús T. Gudmundsson*] and Thórdís Högnadóttir
(Institute of Earth Sciences, University of Iceland, Sturlugata 7, 101 Reykjavík, Iceland; ph.
+354-525 5867; fax: +354-562 9767; email: mtg@raunvis.hi.is, disah@raunvis.hi.is)
Eruptions on radial fissures are common at stratovolcanoes and on concentric volcanic
fissures at caldera faults. When volcanoes are ice-covered, eruptions usually melt ice causing
jökulhlaups and lahars. The magnitude of such floods is of prime importance in hazard
assessment and insight into the processes operating may throw new light on magma-water
interaction in such settings. Two principal end member cases can be defined, “thick ice” and
“thin ice”. In the thick-ice case the glacier responds to melting and drainage predominantly in
a ductile manner, with inflow of ice being of major importance. This type of activity mainly
occurs within ice sheets and large glaciers. At stratovolcanoes ice thicknesses are limited and
the thin-ice case is more common and the ice predominantly behaves in a brittle manner since
confining pressures are small due to the limited thickness. In the observed thin-ice eruptions
(e.g. Deception Island 1969, Grímsvötn 1998, 2004) chasms with vertical ice walls have
formed around the volcanic fissures. The rate of widening is fast at first but then slows down
quickly. An order of magnitude estimate of melting rate can be determined from a simple
empirical model based on this characteristic behavior. The model predicts maximum melting
rates similar to 1000 m3/s for every kilometer of ice-covered volcanic fissure for 100 m thick
ice. The model can be used to assess the potential of flooding from flank eruptions. It seems
to perform reasonably well for ice thicknesses of 50-200 m and has been applied in hazard
assessment for glacier-covered volcanoes in South Iceland where it yielded estimates of
maximum meltwater discharge of 103-104 m3/s, with the values for individual ice-covered
flank segments depending on possible fissure length and ice thickness.
ORAL
CORRESPONDING AUTHOR: MT Guðmundsson
12
Rootless Cone Archetypes and Their Relation to Lava Flow Emplacement Processes
C.W. Hamilton (University of Hawaii at Manoa, Hawaii Institute of Geophysics and
Planetology, 1680 East-West Road, POST 504, Honolulu, HI 96822, USA, ph: +1 808-9569503, fax: 1-808-956-6322, e-mail: christopher@higp.hawaii.edu); [*S.A. Fagents*]
(University of Hawaii at Manoa, Hawaii Institute of Geophysics and Planetology, 1680 EastWest Road, POST 504, Honolulu, HI 96822, USA, ph: +1 808-956-3163, fax: +1 808-9566322, e-mail: fagents@hawaii.edu); T. Thordarson (School of Geosciences, University of
Edinburgh, Grant Institute, The Kings Buildings, West Mains Road, Edinburgh EH9 3JW;
ph: +44 (0) 131-650-8526 fax: +44 (0) 131 668 3184; e-mail: thor.thordarson@ed.ac.uk)
Rootless cones are common volcanic structures within mafic lava flows in Iceland and on
Mars. Rootless eruptions are generated by explosive interactions between molten lava and
water-bearing substrate and they represent one of the most hazardous aspects of lava flow
emplacement. On Mars, low-latitude rootless cone groups are of particular interest because
they imply the presence of groundwater (or ice) in the equatorial region, which is generally
thought to be devoid of water. Although volcanic rootless cones have long since been known
as the products of explosive lava-water interactions, the specific mechanisms of their
formation remain poorly understood. To constrain the controls on the initiation of rootless
eruptions, we have conducted detailed surveys of rootless cone groups within the 1783-84
Laki lava using Differential Global Positioning System (DGPS) measurements (decimeterscale accuracy in kinematic mode) and ArcGIS. In this study we identify five distinct rootless
cone morphologies: (1) simple symmetrical cones with a single central explosion crater; (2)
half-cone pairs with two symmetrical halves separated by a V-shaped depression; (3) simple
cone chains with explosion centers aligned along a single axis; (4) parallel cone chains with
cones aligned in two rows that are separated by a V-shaped depression; and (5) complex
rootless cones containing numerous explosion centers that are distributed over a broad area
without quasi-linear alignments. Morphological and structural evidence suggests that: (1)
Simple cones are formed by localized explosive lava-water interactions that initiate below a
lava tube. (2) Half-cone pairs form where explosions are localized below the center of a lava
channel. (3) Simple cone chains form when a cylindrical lava pathway (usually a tube)
subsides into a compactable substrate. The subsidence causes maximum strain accumulation
along flow axis, which results in fracturing of the pathway floor (analogous to formation of
an axial cleft in an elongate tumulus). The fracture allows molten lava to physically mix with
the underlying water-bearing substrate to produce a rootless eruption. The explosions also
disrupt the lava pathway and prevent further transportation of lava down-flow. The blockage
may then promote enhanced lava inflation in the up-flow region, increase subsidence rates
due to localized mass loading, trigger additional fracturing episodes, and cause the active
explosion sites to migrate up-flow to form a quasi-linear chain of rootless cones. (4) Parallel
cone chains form along lava pathways that have a broad rectangular form rather than a
cylindrical shape. In this instance, the subsiding pathway will concentrate strain within hinge
zones that run along the margin of the slab (analogous to the monoclinal fractures that bound
lava-rise plateaus). Failure along the hinge zones produces explosion sites that are aligned
along the margins of the pathway. Once the rootless eruption disrupts the pathway at one
locality, the restriction will promote up-flow lava accumulation, increased subsidence rates,
additional pathway failures, and an up-flow shift in the location of active explosion sites.
Lastly, (5) complex rootless cones are generally not associated with a discrete lava pathway
such as a tube, but rather with broad flow of lava. Basal failure at one locality within the
broad flow does little to restrict the overall down-flow motion of the lava and so systematic
13
alignments of crater centers do not form due to pathway effects. These rootless cone
archetypes provide useful insight into how explosive lava-water interactions initiate and to
what extent lava emplacement conditions affect cone group morphology--both of which are
important considerations for interpreting the significance of rootless cones on Earth and
Mars.
POSTER
CORRESPONDING AUTHOR: C.W. Hamilton
FIRST AUTHOR IS A STUDENT
14
The Correlation Between Geothermal and Volcanic Activity at Grímsvötn, Iceland
[*Thórdís Högnadóttir*] and Magnús T. Gudmundsson
(Institute of Earth Sciences, University of Iceland, Sturlugata 7, 101 Reykjavík, Iceland; ph.
+354-525 5862; fax: +354-562 9767; email: disah@raunvis.hi.is, mtg@raunvis.hi.is)
Grímsvötn is located in the western part of Vatnajökull glacier and has the highest eruption
frequency of all volcanoes in Iceland with over 60 known eruptions in the last 800 years. It is
in most parts covered by 200-600 m thick ice and all known eruptions have been basaltic.
Two eruptions have occurred in the Grímsvötn caldera over the last 10 years (1998 and 2004)
and are believed to mark the end of an unusually long low-activity period that began at the
end of the 1930s. Both were typical Grímsvötn eruptions of moderate size, dominated by
surtseyan activity. Since 1997 annual mapping of the glacier surface in Grímsvötn has been
carried out, using submeter DGPS or kinematic GPS, mounted on snowmobiles that traversed
the area along a network of survey lines. In areas not accessible for ground vehicles, the
measurements have been done with airborne ground-clearance radar altimetry, and traverses
on foot. By taking into account variations in the level of the subglacial Grímsvötn lake, the
annual maps that are compiled from the survey data have been used to measure ice mass
changes, revealing the rate of basal melting. The results show that ice cauldrons deepened in
the year before the eruption in December 1998. The eruption was accompanied with a sharp
peak in geothermal power. The heat output declined gradually after the eruption and reached
a minimum in 2002. An increase was again detected in 2003, followed by the 2004 eruption.
A similar increase to the one observed after the 1998 eruption occurred after 2004. Thus,
fluctuations in geothermal power correlate well with the eruptions. In both cases increases in
geothermal activity precede the eruptions by about 1 year and the eruptions sparked off large
increases in geothermal activity that seem to recede over a few years. According to
deformation measurements at a nunatak on the caldera rim, both eruptions were preceded by
ground inflation (Sturkell et al., 2006). The observed correlations in Grímsvötn indicate that
monitoring of geothermal activity through repeated ice surface mapping holds considerable
potential as a method to study the state of subglacial volcanoes.
Reference:
Sturkell, E, and 9 others (2006) Volcano geodesy and magma dynamics in Iceland. J.
Volcanol. Geotherm. Res., 150, 14-34.
POSTER
CORRESPONDING AUTHOR: ÞÓRDÍS HÖGNADÓTTIR (disah@raunvis.hi.is)
Tephra Deposition on Glaciers and Ice Sheets on Mars: Implications for Glacial Debris
Content and Glacial Melting
[*J. W. Head*] (Geological Sciences Dept., Brown University, Providence, RI 02912, USA;
ph: +1-401-863-2526; email: James_Head@brown.edu); L. Wilson (Environmental Science
Dept., Lancaster University, Lancaster, LA1 4YQ, UK; ph: +44-1524-593889; email:
L.Wilson@Lancaster.ac.uk); D. R. Marchant (Department of Earth Sciences, Boston
University, Boston, MA 02215, USA; ph: 617-353-3236; email: marchant@bu.edu)
Late Amazonian cold-based tropical mountain glacier deposits have been observed along the
northwest flanks of the Tharsis Montes. Facies include concentric ridged deposits (drop
moraines), hummocky deposits (sublimation tills), and lobate deposits (debris-covered glacier
remnants). These facies represent the long-term behavior of an extensive ice sheet, including
15
its advance, retreat, collapse and reactivation, and consist of sediment that is influenced by
subsequent eolian activity. Candidate sources of sediment include global airborne dust,
volcano summit and flank explosive tephra-forming eruptions, country rocks from
phreatomagmatic eruptions, and rocky debris derived from alcoves and scarps near the
accumulation area. Abundant evidence suggests that the glaciers were cold-based and thus
incorporation of subglacial debris into the glacial ice was minimal.
Subglacial volcanic eruptions occurring during the time of the presence of the glaciers show
evidence for dike, sill, subglacial flow, móberg ridge, and cone-like activity. Cone-like
activity may represent eruptions that produce local to regional tephra blankets that could be a
substantial source of glacial debris and serve as agents of direct heating and melting (hot
airfall deposits) or albedo-induced melting (blanketing dark tephra). We model a hypothetical
set of cones based on composite observations. We locate the cones along a 40 km long ridge
beneath the Tharsis Montes tropical mountain glacier ridged facies, interpreted to be drop
moraines at the outer, thinner part of the glacial system. The ridge is interpreted to be the
subglacial manifestation of a dike; the cones are interpreted to represent eruptive centers
forming along wide places in the dike. We model the cones as occupying a 6 km long
segment of the ridge and ranging from 600-1200 m in width, 600-1800 m in length, and 150280 m in height. Such cones are interpreted to represent eruptions into a melted cavity in the
ice or onto the glacier surface. Widths of such cones are related to the ranges of ejected
pyroclasts and imply explosive eruption speeds of ~ 40 m/s and magma water contents of
~0.1 wt %. If they were emplaced from dikes releasing magma at 1-10 cubic meters per
second per meter along strike (a common rate on Earth), emplacement times would have been
a few hours to 1-2 days. Away from the cones, the ridge is modelled as ~ 60 m high and ~250
m wide; here effusion rates would be much smaller, ~0.004 cubic meters per second per
meter from a dike ~0.1 m wide. These values and this difference are well within the range
observed for narrow and wider places in basaltic dikes feeding eruptions along the East rift
zone of Kilauea.
This treatment can be used to investigate the potential contribution of a layer of tephra
deposits to: 1) the sediment load in the glacier, 2) the amount of melting that might have
occurred due to hot tephra emplacement, and 3) the effect of a dispersed tephra layer on
albedo-induced melting. We find that fine-grained (sub-mm sized) tephra would have been
dispersed in a ~20 km high plinian-like cloud under a normal range of eruption conditions. If
this were deposited only over an ~100,000 square km area of the glacial deposits it would
produce a layer of cooled pyroclasts ~3-4 mm deep. More likely the ambient winds would
have dispersed it over a 1-2 million square km area to a mean depth of ~200 microns. This
could not have provided a significant contribution to the drop moraines, but would have
greatly influenced the thermal stability of the ice surface by dramatically changing its albedo.
Thus, there may be links between subglacial eruptions, surface tephra emplacement and ice
sheet behavior.
ORAL
CORRESPONDING AUTHOR JW Head
contact info: J. W. Head, Geological Sciences Dept., Brown University, Providence, RI
02912, USA; ph: +1-401-863-2526; email: James_Head@brown.edu
16
Subglacial Eruptions and Polythermal Glaciation on Mars: The Pavonis Mons Tropical
Mountain Glacier Steep-Sided Subglacial Flows
[*J. W. Head*] (Geological Sciences Dept., Brown University, Providence, RI 02912, USA;
ph: +1-401-863-2526; email: James_Head@brown.edu); L. Wilson (Environmental Science
Dept., Lancaster University, Lancaster LA1 4YQ, UK; ph: +44-1524-593889; email:
L.Wilson@Lancaster.ac.uk); D. E. Shean (Department of Earth Sciences, Boston University,
Boston, MA 02215, USA; ph: 617-353-3236; email: dshean@bu.edu); D. R. Marchant
(Department of Earth Sciences, Boston University, Boston, MA 02215, USA; ph: 617-3533236; email: marchant@bu.edu)
Fan-shaped deposits on the northwest flanks of the Tharsis Montes have been interpreted to
be formed by Late Amazonian cold-based tropical mountain glaciers. Several examples of
subglacial eruption deposits and landforms have been documented beneath and within these
deposits, including low ridges interpreted to be radial and non-radial dikes, rimmed plateaulike lobate deposits interpreted to be steep-sided flows, linear mounds and low ridges
interpreted to be móberg-like ridges, and circular to elongate aligned and often cratered cones
interpreted to represent subglacial tephra cones.
Subglacial and englacial magma in contact with glacial ice can cause melting and produce
meltwater. Here we examine the heating associated with such subglacial eruptions in order to
assess their relative ability to melt glacial ice and to produce sufficient meltwater to
potentially cause a transition from cold-based to wet-based glacial conditions. We find that
among the range of subglacial eruption features, steep-sided subglacial lava flows are the
most efficient in producing meltwater. Due to their sill-like nature and the continuing supply
of overlying ice replacing draining meltwater, water volumes of up to 450 km3 can be
generated from a typical 100 km long, 15 km wide, and 500 m thick subglacial flow
emplaced at a plausible effusion rate of ~ 10,000 m3/s over a 2.5 year period. This volume of
meltwater is equivalent to a layer averaging 4.5 m deep below an area equivalent to the
maximum extent of the Pavonis tropical mountain glacier but more importantly is equivalent
to an average water layer thickness of 45 m beneath the specific zone between the flows and
the distal part of the lobe.
This analysis suggests that there may have been some predictable influences from the
generation of this volume of meltwater. First, thermal calculations show that a transient
subglacial lake would form down-slope of the advancing lava flow. Shallow lakes would
rapidly refreeze, but the volumes generated here could lead to additional consequences. For
example, volumetrically significant local subglacial magmatic melting (such as calculated
here) could produce sufficient meltwater to cause a local transition from cold-based to wetbased conditions, producing local wet-based surging of a portion of the otherwise cold-based
glacier. Evidence for such activity might be manifested in anomalous arcuate lobes of
moraines downslope of the subglacial flows. In addition, drainage of the meltwater from
beneath the glacier out into the surrounding terrain could form local fluvial zones ranging in
scale from small channels up to the scale of jökulhlaups. Rapid drainage would be consistent
with the survival of the released water against rapid freezing in the low-temperature, lowatmospheric pressure Martian environment. In summary, we find that local sill-like subglacial
eruptions can produce sufficient meltwater to cause localized transitions from cold-based to
wet-based glacial behavior, and produce related, and perhaps catastrophic, drainage of
meltwater. Studies are underway to assess the relationships between the specific steep-sided
lobate features interpreted to be subglacial sill-like lava flows, local arcuate perturbations in
patterns of moraines, and fluvial features located along the distal margin of the glacier.
ORAL
17
CORRESPONDING AUTHOR JW Head
contact info: J. W. Head, Geological Sciences Dept., Brown University, Providence, RI
02912, USA; ph: +1-401-863-2526; email: James_Head@brown.edu
Arsia Mons Cold-Based Tropical Mountain Glacier: Subglacial Eruptions, Polythermal
Glaciation, and Distal Drainage of Meltwater
[*J. W. Head*] (Geological Sciences Dept., Brown University, Providence, RI 02912, USA;
ph: +1-401-863-2526; email: James_Head@brown.edu); L. Wilson (Environmental Science
Dept., Lancaster University, Lancaster LA1 4YQ, UK; ph: +44-1524-593889; email:
L.Wilson@Lancaster.ac.uk).
Late Amazonian fan-shaped deposits on the northwest flanks of the Tharsis Montes
volcanoes are interpreted to have formed as a result of cold-based tropical mountain
glaciation. Deposits and landforms interpreted to have formed during subglacial eruptions
have been documented beneath and within these deposits, and at Pavonis Mons, sufficient
meltwater appears to have been generated during the intrusion of steep-sided sill-like
subglacial flows to produce subglacial lakes. The presence of subglacial lakes could have led
to local wet-based conditions, local tongue-like wet based glacial surges, and to the release of
meltwater from the margin of the glacier in fluvial drainage channels or jökulhlaups.
At Arsia Mons, the tropical mountain glacier deposits contain features and structures
interpreted to represent subglacial and englacial eruptions. These include low ridges
interpreted to be dikes, lobate deposits interpreted to be steep-sided flows, linear mounds and
low ridges interpreted to be móberg-like ridges and cones, and elongated depressions and
trough-like features interpreted to be the result of subglacial and englacial phreatomagmatic
eruptions. Here we describe a series of features that together are interpreted to represent a
linear subglacial eruption that caused the production of sufficient meltwater to form eskers, a
local wet-based glacial surge that formed anomalously lobate moraines, and a distal series of
channels emerging from the edge of the glacial deposit and flowing downslope into the
surrounding terrain that is interpreted to represent subglacial drainage following the eruption.
A series of NW-trending preglacial lava flows extend down the flanks of Arsia and the fanshaped glacial deposits are superposed. Facies in the NW part of the Arsia deposit include
concentric ridged deposits interpreted to be drop moraines and hummocky deposits
interpreted to be sublimation tills. Two generally parallel graben-ridge systems trend for
several hundred km across the NW part of the Arsia deposit and are separated by about 25-30
km. Graben, pits and móberg-like ridges characterize the eastern ridge. The western structure
contains a lobe-shaped plateau and crater near the edge of the hummocky facies, and forms
subaerial cones along its northern extension, between the inner and outer ridged facies, and
again outside the glacial deposits to the north. The lobe-shaped plateau (~9 km long and 6 km
wide) extends downslope and the adjacent crater is ~4 km wide and ~100 m deep. The
plateau is ~130-150 m high and extends from the base of a ridge that is ~350 m high. We
interpret the ridge, located along the strike of the linear trend, to be a sub-glacial móberg-like
ridge, and the elongate plateau to be a subglacial, sill-like lava flow extending from the vent.
Superposed on the lobate plateau, and extending downslope and out onto the subjacent lava
flows, is a sinuous ridge that is generally continuous for ~14 km; we interpret this to be an
esker draining subglacial eruption-induced meltwater. Adjacent to and downslope from the
ridge/lobate plateau the configuration of the drop moraines bows outward for a distance of
18
~5-10 km. The most prominent drop moraine occurs at the distal (downslope) edge of this
inner set of drop moraines and several fluvial channels emerge from its base and extend at
least 5-7 km into the surrounding terrain.
We interpret this configuration to be the result of subglacial volcanism and volcanically
induced meltwater generation and drainage. Sufficient meltwater appears to have been
generated to cause a local transition from cold-based to wet-based conditions, producing local
wet-based surging of a 30-40 km wide portion of the otherwise cold-based glacier. In
addition, drainage of the meltwater from beneath the glacier out into the surrounding terrain
formed local fluvial channels at the glacier margin.
INVITED ORAL
CORRESPONDING AUTHOR JW Head
contact info: J. W. Head, Geological Sciences Dept., Brown University, Providence, RI
02912, USA; ph: +1-401-863-2526; email: James_Head@brown.edu
Emplacement of Sinuous Basaltic Pillow Ridges on the Flanks of the Mount Edziza
Volcanic Complex, British Columbia, Canada
[*Jefferson D. G. Hungerford*] (Department of Geology and Planetary Science, University
of Pittsburgh, 4107 O’Hara St. Rm 200 SRCC, Pittsburgh PA, 15260; ph: 1-412–624-7988;
fax: 1-412-624-3914; email: jdh53@pitt.edu); Ian Skilling (Department of Geology and
Planetary Science, University of Pittsburgh, 4107 O’Hara St. Rm 200 SRCC, Pittsburgh PA,
15260; ph: 1-412–624-5873; fax: 1-412-624-3914; email: skilling@pitt.edu); Benjamin
Edwards (Department of Geology, Dickinson College, 5 N. Orange St., Carlisle, PA 17013,
U.S.A.; ph: 1-717-254-8934; fax: 1-717-245-1971; email: edwardsb@dickinson.edu
The Mount Edziza Volcanic Complex (MEVC) is a northeast trending, 75km long volcanic
edifice in northern British Columbia. Erupting cyclically over the last 7.5 Ma., volcanic
centers along the complex have erupted alkaline basalts, rhyolites and a range of intermediate
alkaline lavas. Several episodes of sub-ice or ice-contact volcanism are recorded since 3Ma
(Souther, 1992), including a basaltic volcaniclastic sequence and associated lava flows at
Tennena Cone 4 km south-southwest of the summit of Mount Edziza. Tennena Cone (TC) is
a 200m high volcanic pile dominated by pillow breccia, but also comprising lenses of
coherent pillow lava intruded by a dike(s). Pillow and subaqueous lobate or sheet-like lava
flows originating from the TC vent flowed west for ~3km towards the Sezill Creek valley..
These subaqueous basaltic lavas overlie topography of varying steepness created by 2 to 7
meter thick flows of the Ice Peak Formation (IPF). Distally, the Tennena flows overlie both
glaciogenic sediments and steeply-dipping IPF flows. Proximal to TC, lava flows were
emplaced as laterally extensive sheet-like pillow lavas on a plateau of underlying IPF lavas.
Further west, as the underlying topography steepens, the flows form sinuous, hemispherical
pillow ridges 5 to 10 meters wide, 5 meters high and up to a few hundred meters long. The
volume of proximal pillow lava sheets are an order of magnitude greater than these distal
sinuous flows. We interpret the narrow, sinuous pillow lava flows as esker-like Röthlisberger
channel fill. The creation of abundant meltwater from the emplacement of flows proximal to
the vent at or near the TC pile either overwhelmed the subglacial drainage system of the wet
based glacier thermally excavating new sub-ice channels, or expanded the radius of existing
drainage channels by convective heat transfer from meltwater at elevated temperatures. The
sinuous pillow lava flows occupied these new or newly excavated Röthlisberger channels
19
following volcaniclastic sediment laden meltwater. Fragments of Tennena pillow basalts are
abundant in the diamict sediments immediately below the sinuous pillow pile, indicating flow
of sediment laden meltwater in established drainage channels immediately prior to the
emplacement of the sinuous pillow pile.
Ice thickness is a particularly useful tool in interpreting paleoclimate conditions. Initial
volatile analyses indicate TC lavas were emplaced under a minimum of 500m of ice; a
reasonable value given a thickness of roughly 750 meters of flat surfaced regional ice
calculated from the substrate topography. Interpreting the original ice thickness over these
sinuous distal flows is preferable because of the small volume of lava relative to proximal
sheet flows. Since a unit volume of basalt has the capacity to melt approximately 10 times the
volume of ice (Hoskuldsson & Sparks, 1997), thermal degradation of ice overlying the
sinuous distal flows would have been considerably lower.
ORAL
CORRESPONDING AUTHOR JDG Hungerford
FIRST AUTHOR IS A STUDENT
Products of Subsurface Magma-Ice Interaction on Mars, and Implications for Life
N.M. Jacques (Department of Earth Sciences, The University of Western Ontario, London,
ON, Canada, N6A 5B7; ph: 1-519-661-3187; fax: 1-519-661-3198; email:
nmjacque@uwo.ca); [*D.T. Lescinsky*] (Department of Earth Sciences, The University of
Western Ontario, London, ON, Canada, N6A 5B7; ph: 1-519-661-2111 x.86063; fax: 1-519661-3198; email: dlescins@uwo.ca); P.J. Stooke (Department of Geography, The University
of Western Ontario, London, ON, Canada, N6A 5C2; ph: 1-519-661-2111 x.85022; fax: 1519-661-3750; email: pjstooke@uwo.ca)
Volcano/magma–volatile interactions on Mars have been a subject of significant research as
there is diverse and widespread evidence for its occurrence [1]. On Earth, intrusive igneous
activity commonly results in the production of features such as volcanic necks or plugs,
which manifest themselves in distinctive surface morphologies when they are exposed due to
erosion of the surrounding terrain. While similar situations are thought to have occurred on
Mars [2], the hypothesized presence of a global cryosphere [3] implies that there are
significant differences in the properties of the subsurface environment, which are likely to
result in distinctly different features than those that we observe on Earth. Previous Martian
studies have focused mainly on the surface manifestation of subsurface magma–cryosphere
interactions, such as outflow channels and outburst floods, collapse features, and explosive
products from shallow interactions [4]. Our research, however, focuses on the processes
occurring at the site of the interactions, in the Martian subsurface, and the resulting
subsurface morphologic products.
A better understanding of the processes occurring at the site of subsurface magma–
cryosphere interactions is needed in order to quantify the effect of intrusions on the
subsurface environment on Mars. Intrusions are extremely important for the transport of heat
into the cryosphere, causing melting of the ice, creating slurries into which magma can then
intrude and further interact, and setting up local hydrothermal circulation systems. With
appropriate modifications, the principles used to analyze subaerial eruptions and intrusions
on Earth are applied to eruptions into a cryosphere on Mars, just as they have been applied to
20
eruptions into water, ice, permafrost, or saturated sediments on Earth, in order to characterize
the environment that such interactions would create. Both analytical and numerical
techniques are used to characterize the thermal evolution of the subsurface environment.
While there are no direct terrestrial analogues to such interactions, terrestrial interactions that
most closely approximate the processes involved (having similar physical parameters and
environmental conditions) are then considered. Several different settings give rise to these
potential analogue terrestrial eruptions: volcanic regions covered by ice produce eruptions
beneath ice sheets, while volcanic activity on the ocean floors produce eruptions into water
and saturated sediments. Both theoretical modelling and direct comparison with potential
terrestrial analogues are undertaken in order to gain better insight into interactions in the
Martian environment and to predict the geomorphologic outcome of these interactions.
Finally, an examination of current and past Mars mission data is undertaken to look for
evidence of the predicted interaction products exhumed to the Martian surface.
Since the interplay of water and sources of heat is most likely the crucial factor involved in
the evolution of life on extraterrestrial planets such as Mars, it is expected that the best places
to look for evidence of life on other planets is where we find indications of such interactions.
Subsurface magma–cryosphere interaction environments possess excellent potential as sites
for long–lived hydrothermal activity, and, consequently, for demonstrating strong
exobiological potential. On Earth, colonies of microbial life have been identified at
significant depth (kilometers) within the crust, and interest in the deep biosphere is increasing
[5]. Because these environments are arguably the most likely havens for the evolution of life
on Mars, they warrant considerable attention. As such, there is great potential and importance
for investigating the occurrence and behaviour of subsurface magma–cryosphere interactions
on Mars, both analytically/numerically and in the context of potential terrestrial analogue
environments.
References: [1] Wilson, L. and J. W. Head (2002) Lunar Planet. Sci. 33, Abstract #1275. [2]
Head, J. W., Wilson, L., Dickson J. and G. Neukum (2006) Geology 34, 285-288. [3]
Clifford, S. M. (1993) J. Geophys. Res. 98, 10,973-11,016. [4] Head, J. W. and L. Wilson
(2002) In: Smellie, J. L and Chapman, M. G. (Eds.), Volcano-Ice Interaction on Earth and
Mars. Geological Society of London, Special Publication No. 202, pp. 27-57. [5] Amend, J.
P. and A. Teske (2005) Palaeogeogr. Palaeocl. 219, 131-155.
POSTER
CORRESPONDING AUTHOR: N.M. Jacques
FIRST AUTHOR IS A STUDENT
Volcano-ice Interaction at Mount Rainier and Mount Baker, Washington: Lessons and
Examples
[D.T. Lescinsky] (Dept. of Earth Sciences, University of Western Ontario, London, ON,
Canada, N6A 5B7; ph: 1-519-661-2111 ext. 86063; fax: 1-519-661-3198; email:
dlescins@uwo.ca)
Mount Rainier and Mount Baker of the Cascades volcanic range, Washington are large
stratovolcanoes that have undergone extensive glaciation during recent ice ages and still
retain large volumes of ice. In addition, both volcanoes have experienced abundant and
diverse styles of eruptive activity. Not surprisingly, Mount Rainier and Mount Baker display
21
numerous examples of volcanic interaction with glacial ice. During the past ten years, these
volcanoes have been extensively studied (e.g., Sisson et al., 2001; Hildreth et al., 2003)
revealing how geologically dynamic glaciated volcanoes can be.
Evidence of volcano-ice interaction ranges from fairly obvious to fairly ambiguous due to
formation processes, and aspects of preservation and erosion. Ice contact features in lava
flows offer compelling evidence of glacial interaction. Features include: glassy, closely
spaced polygonal joints, with and without water penetration fractures, and having curving to
subhorizontal orientations. Spectacular examples of these features are visible at the terminus
of Burroughs Mountain lava flow and along the margins and at the terminus of Mazama
Ridge lava flow at Mount Rainier, and near the end of Kulshan Ridge at Mount Baker. The
fracture orientations are indicative of a vertical cooling surface and the great thickness of
these flows is attributed to ponding against ice. The Mazama Ridge lava flow is notable for
multiple “table-like” portions, where the advancing lava intersected and ponded against
tributary glaciers.
Exceptionally thick lava flows (>100 m thick) are scattered around Mount Rainier and Mount
Baker. While their great thickness is related to ponding, at times it is unclear whether the
flows have been emplaced within valleys with rock walls or with ice walls (or one of each).
The inverse topography commonly associated with these lava flows can be produced in either
situation. Extensive erosion caused by glacial reoccupation of drainage valleys and ice sheet
advance during subsequent glaciations, can remove the margins of the flows and any
evidence of ice contact. Hildreth et al. (2003) note examples of stranded lava flow remnants,
such as Lookout Mountain, where a gorge >500 m deep separates the lava flow from its
likely source. The rapid incision rates required highlight the difficulty in evaluating older
ridge-capping and valley-wall lava flows.
The exceptionally thick lava flows at Mount Rainier and Mount Baker correspond primarily
to discrete periods of high effusion rates. Lava flows erupted outside of these periods are
thinner, less voluminous, and number in the hundreds. These lava flows mantle the upper
slopes of both volcanoes and show localized evidence of interaction with ice: glassy textures,
and small diameter polygonal joints. Small block-and-ash flows were formed by collapse of
such lavas on the steep upper slopes travelled out onto the glacial ice. These pyroclastic flows
commonly travelled out onto the glacial ice and are likely responsible for a large number of
debris flows and several of the non-vesicular tephra layers found at Mount Rainier.
Both Mount Rainier and Mount Baker have experienced sector collapse and the generation of
large volume clay-rich debris flows (i.e., the Osceola Mudflow at Mount Rainier). Detailed
study of the debris flow deposits indicates that they tend not to be associated with large
eruptions, but instead with small phreatomagmatic eruptions or seismic events (Sisson et al.,
2001). In fact, Mount Rainier and Mount Baker have experienced few large explosive events.
However, the record of activity may be somewhat influenced by a preservation bias towards
lava over volcaniclastics. This is highlighted by the removal of almost the entire proximal
deposit of the Kulshan caldera ignimbrite (1.15 ma; Mount Baker) by glacial erosion.
References
Hildreth, W., J. Fierstein, and M. Lanphere (2003), Eruptive history and geochronology of
the Mount Baker volcanic field, Washington, Geol. Soc. Amer. Bull., 115 (6), 729-764.
Sisson, T. W., J. W. Vallance, and P. T. Pringle (2001), Progress made in understanding
Mount Rainier's hazards, Eos Trans. AGU, 82 (9), 113, 118-120.
INVITED ORAL
CORRESPONDING AUTHOR: DT Lescinsky
22
Insight on the timing of the last deglaciation in Iceland from surface exposure ages of
subglacial and postglacial volcanic features
[*Joseph M. Licciardi*] (University of New Hampshire, 56 College Rd., Durham, NH, USA,
03824; ph: (603) 862-3135; Fax: (603) 862-2649; email: joe.licciardi@unh.edu); Mark D.
Kurz (Woods Hole Oceanographic Institution, MS #25, Clark Laboratory 421, Woods Hole,
MA, USA, 02543; ph: (508) 289-2888; Fax: (508) 457-2193; email: mkurz@whoi.edu);
Joshua M. Curtice (Woods Hole Oceanographic Institution, MS #25, Clark Laboratory 421,
Woods Hole, MA, USA, 02543; ph: (508) 289-2618; Fax: (508) 457-2193; email:
jcurtice@whoi.edu)
The latest Pleistocene and Holocene deglacial history of Iceland is based largely on
radiocarbon and tephra dating of glacial landforms and deposits (e.g., Kirkbride and
Dugmore, 2001; Norðdahl and Pétursson, 2005). However, radiocarbon dating is often
limited due to the scarcity of preserved organic material, and many tephra ages are of coarse
resolution. Surface exposure dating using cosmogenic nuclides offers an alternative tool for
developing high-resolution glacial and volcanic chronologies in Iceland (Principato et al.,
2006; Licciardi et al., 2007). New local calibrations of cosmogenic 3He and 36Cl production
rates increase the attainable accuracy of exposure dating in Iceland (Licciardi et al., 2006, in
review).
Here we report new constraints on the deglacial history of the Western Volcanic Zone
(WVZ) of Iceland provided by 3He exposure ages of subglacially formed table mountains
(tuyas) and postglacial lava flows. Our strategy is to compare ages of the youngest table
mountains with ages of the oldest postglacial volcanic units in the same vicinity. Owing to
their subglacial origin, the age of the youngest table mountain in a region provides a
minimum age for the presence of an ice sheet. The age of the oldest postglacial volcanic unit
in a region indicates the maximum age of final ice withdrawal. It follows that the difference
in age between table mountains and postglacial lava flows should bracket the time interval
corresponding to deglaciation.
We have measured cosmogenic 3He concentrations in olivine phenocrysts from the surfaces
of subaerially erupted basaltic lava caps of table mountains and from postglacial basaltic lava
flows in the WVZ to develop an exposure age chronology of their formation. The resulting
3
He exposure ages in the WVZ comprise >30 individual ages from six table mountains
(reported in Licciardi et al., 2007) and six different postglacial lava flow units (unpublished).
Mean exposure ages from table mountains range from 12.6 ± 0.4 ka to 8.2 ± 1.0 ka. The 8.2ka age of Hlöðufell conflicts with radiocarbon and 3He exposure ages of nearby postglacial
lava flows and is considered anomalously young, perhaps due to past cover by snow and ice
(Licciardi et al., 2007). The next-youngest table mountains are Högnhöfði and Skriða. At
face value, the spatial and temporal distribution of table mountain ages suggests persistence
of ice in the interior region until ca. 10.5-10.2 ka.
The mean ages of the six dated postglacial lava flow units in the WVZ range from 11.2 ± 0.8
ka (Brunnar, n = 2) to 7.8 ± 0.5 ka (Sköflungur, n = 2). Exposure ages of the postglacial lava
flow units are broadly consistent with stratigraphic relationships and inferred ages compiled
by Sinton et al. (2005). The oldest exposure-dated postglacial lava units (Brunnar and
Selvogsheiði) are older than but within 1-sigma age error of the oldest 14C-dated postglacial
lavas (Þingvallahraun, 10.3 ± 0.1 cal ka; and Hellisheiði A, 10.4 ± 0.2 cal ka). The exposure
age of Geitlandshraun is consistent with its 14C-based age (~8.9 cal ka).
23
Comparison of table mountain ages and postglacial flow ages suggests extremely rapid
deglaciation in southwestern Iceland, consistent with previous work in this region. The
spatial pattern of early postglacial lava flow ages is consistent with the distribution of table
mountain ages, within error limits, although some ages (e.g., Hlöðufell) are difficult to
reconcile. The new 3He data demonstrate the great potential for developing exposure ages of
as-yet undated volcanic features elsewhere in Iceland.
References
Kirkbride, M.P., Dugmore, A.J., 2001. Timing and significance of mid-Holocene glacier
advances in northern and central Iceland. J. Quat. Sci. 16, 145-153.
Licciardi, J.M., Kurz, M.D., Curtice, J.M., 2006. Cosmogenic 3He production rates from
Holocene lava flows in Iceland. Earth Planet. Sci. Lett. 246, 251-264.
Licciardi, J.M., Kurz, M.D., Curtice, J.M., 2007. Glacial and volcanic history of Icelandic
table mountains from cosmogenic 3He exposure ages. Quat. Sci. Rev., in press.
Licciardi, J.M., Denoncourt, C.L., Stone, J.O.H., Finkel, R.C., in review. Cosmogenic 36Cl
production rates from Ca spallation in Iceland. Earth Planet. Sci. Lett.
Norðdahl, H., Pétursson, H.G., 2005. Relative sea-level changes in Iceland: New aspects of
the Weichselian deglaciation of Iceland. In: Caseldine, C., Russell, A., Harðardóttir, J.,
Knudsen, Ó. (Eds.), Iceland - Modern Processes and Past Environments: Developments
in Quat. Sci. 5. Elsevier, Amsterdam, pp. 25-78.
Principato, S.M., Geirsdóttir, A., Johannsdóttir, G., Andrews, J.T., 2006. Late Quaternary
glacial and deglacial history of eastern Vestfirdir, Iceland, using cosmogenic isotope
(36Cl) exposure ages and marine cores. Journal of Quaternary Science 21, 271-285.
Sinton, J.M., Grönvold, K., Sæmundsson, K., 2005. Postglacial eruptive history of the
Western Volcanic Zone, Iceland. Geochem., Geophys., Geosys. 6, Q12009.
POSTER
CORRESPONDING AUTHOR: JOE LICCIARDI
Sheet Fractures and Cooling Histories at Lava-Ice Contacts: Case Studies at South
Sister, OR, and Mount Rainier, WA.
[*R.W.D. Lodge*] (M.Sc. Student, Department of Earth Sciences, University of Western
Ontario, London, ON, Canada, N6A 5B7; ph: 1-519-709-1271; fax: 1-519-661-3198; email:
rlodge@uwo.ca); D.T. Lescinsky (Department of Earth Sciences, University of Western
Ontario, London, ON, Canada, N6A 5B7; ph: 1-519-661-2111 x86063; fax: 1-519-661-3198;
email: dlescins@uwo.ca)
Cooling lava commonly develops of polygonal joints that form equant hexagonal columns.
After emplacement of the lava flow, thermal contraction produces an isotropic tensional
stress regime resulting in fracturing directed perpendicular to the cooling surface. These
fractures continue to propagate through the body until it is fully solidified, producing
columns. This model has been shown to apply to many systems including flood basalts and
lava lakes, however, certain fracture patterns observed at glaciated volcanoes do not appear to
fit this model. These features include sheet fractures and fractures that crosscut multiple
columns along a single plane perpendicular to the column axis. Despite the relatively
common occurrence of these fractures, their significance and mode of formation have not
been fully explored. Although the surfaces of the fractures show evidence of incremental
24
growth similar to regular columns, sheet fractures show preferred orientations suggesting that
cooling is not the only stress acting on the lava. This study investigates the stress-regimes
responsible for producing these unique fractures and their significance to interpreting cooling
histories at lava-ice contacts.
Sheet fractures are long and largely parallel fractures with perpendicular connecting
fractures that result in rectangular columns. The connecting fractures vary locally from
primary fractures (associated with cooling toward the flow interior) to secondary fractures
(associated with cooling by water infiltration). Because sheet fractures have been exclusively
observed at lava flows emplaced in glacial environments, it is no surprise that stress regimes
would not be purely isotropic due to thermal contraction. Lava flows are commonly
emplaced and solidified on slopes and/or in confinement due to either valley walls or
glaciers. Flows that are confined by glaciers develop a characteristic geometry consisting of a
zone of horizontal columns at their margins where sheet fractures are found. In these
environments, there are a number of potential influences on the stress regime of solidifying
lava flow. However, since there is no evidence of shear on the sheet fractures, the nonisotropic component of the total stress regime must be primarily compressional or tensional.
To test these possibilities, we studied fracture patterns at Kokostick Butte dacite flow at
South Sister, OR, and Mazama Ridge andesite flow at Mount Rainier, WA. Both flows have
well developed sheet-like fractures and abundant evidence of ice-contact during
emplacement.
Data collected at Kokostick Butte and Mazama Ridge show a distinct relationship between
the flow geometry and the orientation of the fractures. The direction of incremental fracture
growth is toward the flow interior and perpendicular to flow margins indicating that thermal
stress is the major component of the total stress regime. In addition to thermal stresses,
horizontal column-bounding fractures will also be influenced by gravitational compression
thus preferentially inhibiting fracture growth in the perpendicular to the direction of
compression. As a result, fractures parallel to gravitational stress would dominate the final
pattern. However, in cases where sheet fractures are observed in vertical column-bounding
fractures, gravitational stress would be directed parallel to fracture growth and would not
influence the final fracture pattern. The source of non-isotropic stress causing these sheet
fractures can be attributed to flow margin inflation. Flow margin inflation would result in
margin parallel extension and fractures preferentially oriented perpendicular to the margin.
Sheet fracturing at steep to moderate angles to the flow margin may be caused by both
mechanisms acting simultaneously. Patterns of sheet fractures produced during
compressional and extensional stress regimes using starch-water analog experiments support
our observations at natural lava flows.
ORAL
CORRESPONDING AUTHOR RWD Lodge
FIRST AUTHOR IS A STUDENT
25
Silicic Glaciovolcanism in Iceland
[*Dave McGarvie*] (Department of Earth Sciences, The Open University, UK, MK7 6AA;
e-mail d.mcgarvie@open.ac.uk)
Iceland contains an abundance and diversity of silicic glaciovolcanic edifices. For the past
nine years a multi-faceted and multi-institutional UK-based research programme has been
unravelling the key processes responsible for silicic glaciovolcanic edifice formation in
Iceland. Five themes (below) summarise and contextualise the programme, and provide
forward directions for future work.
(1) Tuyas.
These ice-confined edifices are 300-900 m high and form during sustained eruptions into
thick ice. Individual tuyas can be up to 3.5 km3 in volume, but 0.5-1.5 km3 is more typical
(McGarvie et al., 2006). An initial phreatomagmatic phase builds a tephra pile up to 350 m
high within a well-drained vault (Tuffen et al., 2002; Stevenson, 2004). Gradual upwards
increases in highly-inflated clasts points to decreasing meltwater interaction as the tephra pile
grows (Stevenson, 2004). A final effusive phase creates a lava cap up to 250 m thick.
(2) Effusion-dominated edifices.
Typified by the small-volume (<0.1 km3) drape of Bláhnúkur (Tuffen at al., 2002) and the
larger volume (0.6 km3) edifice of Prestahnúkur (McGarvie et al., in press). Bláhnúkur is
unusual because uprising magma was dispersed into multiple subsurface conduits which then
produced distinctive conical lava lobes at the edifice-ice interface. Prestahnúkur is unusual
because there is good evidence of substantial magma-water interactions at the start of the
eruption. Despite volume differences three common features of Bláhnúkur and Prestahnúkur
are: (a) clast textures indicate that fragmentation was dominated by quenching; (b) ice
completely covered the edifices throughout the eruption; (c) space was episodically created at
the ice-edifice interfaces, allowing (for example) extrusion of substantial lava bodies.
(3) Volcano-ice interactions through time.
After establishing a sound physical volcanology footing, an additional investigative technique
was added – using Ar/Ar dating to provide timelines. Three carefully-selected volcanic
centres were studied with the aim of evaluating glaciovolcanism through time: Torfajökull
(McGarvie et al., 2006); Kerlingarfjöll (Flude, 2005); and Ljósufjöll (Flude et al., under
review). At Torfajökull there is a pleasing correlation between the eruption ages of rhyolite
tuyas and cold periods in the oxygen isotope record (McGarvie et al., 2006).
(4) Eruptions at stratovolcanoes.
Iceland largest stratovolcano (Öraefajökull) is a 2,119 m high glaciovolcanic edifice with its
upper c.1000 m ice-covered, and a c.4 km diameter summit caldera filled with ice up to 500
m thick (Björnsson, 1998). Silicic rocks occur as nunataks and as substantial flank
outpourings. The first-ever volcanological study, by Stevenson et al., (2006), discovered
substantial ice thicknesses changes (of up to 500 m) between eruptions. Recent fieldwork has
extended this work, and future plans involve blending physical volcanology and Ar/Ar dating
to unravel the evolution of this massive glaciovolcanic edifice.
(5) The Undiscovered Country.
The future of silicic glaciovolcanism looks rosy, if the number of unresolved problems
provides a measure of its health. Three examples are given. Firstly, the relative ease and
higher precision with which silicic rocks can be Ar/Ar dated currently gives them the edge
over basalts regarding useful information that can be provided on past ice sheet thicknesses
26
during specific glacial periods. Secondly is the much higher diversity of silicic tuya-marginal
formations (relative to mafic counterparts), in which unusual volcano-ice interactions are
preserved (e.g. Tuffen et al., in press). And thirdly, there is the challenge of understanding
silicic volcano-ice interactions at ice-capped stratovolcanoes such as Öraefajökull, where
early indications are that new models to explain volcano-ice interactions may need to be
developed.
References
Björnsson, H. (1988) Hydrology of Ice Caps in Volcanic regions. Soc. Sci. Islandica. Volume
45, 139 pp, Reykjavik.
Flude, S. (2005) Rhyolite volcanism in Iceland: timing and timescales of eruption.
Unpublished PhD thesis, University of Manchester, UK, 257pp.
Flude, S., R. Burgess and D.W. McGarvie (under review). Eruptive History of silicic
volcanism at Ljósufjöll Volcano, Iceland.
McGarvie DW, Burgess R, Tindle AG, Tuffen H, and Stevenson JA (2006). Pleistocene
rhyolitic volcanism at the Torfajökull central volcano, Iceland: eruption ages,
glaciovolcanism, and geochemical evolution. Jökull 56: 57-75.
McGarvie DW, Stevenson JA, Burgess R, Tuffen H and Tindle AG (2007). Volcano-ice
interactions at Prestahnúkur, Iceland: rhyolite eruption during the last interglacial-glacial
transition. Annals of Glaciology. In press.
Stevenson, J.A. 2005. Volcano-ice interaction at Öraefajökull and Kerlingarfjoll, Iceland.
Unpublished PhD thesis, The Open University, Milton Keynes, 330 pp.
Stevenson, J.A., D.W. McGarvie, J.L. Smellie and J.S. Gilbert. 2006. Subglacial and icecontact volcanism at the Öraefajökull stratovolcano, Iceland. Bull. Volcanol., 68, 737-752.
Tuffen, H. 2001. Subglacial rhyolite volcanism at Torfajökull , Iceland. Unpublished PhD
thesis, The Open University, 381 pp.
Tuffen, H., J.S. Gilbert and D.W. McGarvie 2001. Products of an effusive subglacial rhyolite
eruption: Bláhnúkur, Torfajökull , Iceland. Bull. Volcanol 63, 179-190.
Tuffen, H., D.W. McGarvie, J.S. Gilbert and H. Pinkerton 2002. Physical volcanology of a
subglacial-to-emergent rhyolitic tuya at Rauðfossafjöll, Torfajökull , Iceland. Geological
Society, London, Special Publications, 202, 213-236.
Tuffen, H., D.W. McGarvie., H. Pinkerton., J.S. Gilbert., J.A. Stevenson and R. Brooker.
(2007). An explosive-intrusive subglacial rhyolite eruption at Dalakvisl, Raudufossafjöll,
Iceland. Bulletin of Volcanology (in press).
INVITED ORAL
CORRESPONDING AUTHOR: DAVE McGARVIE
Volcano-Ice Interactions at Öraefajökull, Iceland
[*Dave McGarvie*] (Department of Earth Sciences, The Open University, UK, MK7 6AA;
e-mail d.mcgarvie@open.ac.uk); John Stevenson (Department of Earth Sciences, The Open
University, UK, MK7 6AA; e-mail johnalexanderstevenson@yahoo.co.uk).
Iceland has few stratovolcanoes, but its largest (Öraefajökull) is considered Europe’s secondlargest after Etna (Thorarinsson, 1958). Its prominence is further emphasised by a rhyolite
nunatak on the caldera rim, which at 2,119 m elevation constitutes Iceland’s highest land.
The upper c.1,000 m of the stratovolcano is mostly ice covered, whilst the summit area
comprises a 2.5 km diameter caldera filled with ice up to 500 m thick (Björnsson, 1998).
Nothing is known about the eruption(s) that formed the caldera.
27
Rhyolites are locally abundant at Öraefajökull, and form nunataks that surround the summit
caldera and pierce the ice-clad upper slopes, as well as substantial lava outcrops on the lower
flanks – some of which reach the foot of the edifice. All outcrops that have been examined
show evidence of interactions between erupting magmas and ice/snow.
The stratovolcano’s flanks are incised by numerous valley glaciers and one of these
(Kviárjökull) is the location of the only detailed volcanological study undertaken at
Öraefajökull (Stevenson et al., 2006). Volcano-ice interactions at Kviárjökull tell a tale of
substantial changes in ice thicknesses (of up to 500 m) in the valley glacier. At Kviárjökull a
well-preserved feeder dyke provides unequivocal evidence for flank eruption, while there is
also evidence of ridge-bounded flows similar to those reported at Mount Rainier by
Lescinsky and Sisson (2000). An unusual feature of some subglacial rhyolite eruptions at
Kviárjökull (compared to North American stratovolcanoes) is the preservation of abundant
tephra, although the significance of this has yet to be investigated.
Reconnaissance fieldwork was carried out in summer 2006 as a precursor to a new project
which will establish a timeline for the glaciovolcanic and geochemical evolution of
Öraefajökull by combining Ar/Ar dating with physical volcanology and geochemistry. This is
a methodology that has proved successful at other Icelandic volcanic systems (e.g. McGarvie
et al., 2006; McGarvie et al., in press).
The reconnaissance fieldwork has highlighted the need for a more thorough investigation of
those lava outcrops on Öraefajökull’s lower flanks which may represent a new type of
glaciovolcanic formation. These substantial lava outcrops, some of which reach the foot of
the edifice, may have originated as simple ridge-bounded flows at high elevations where ice
was thicker, but when they encountered a combination of thinner ice and steeper topography
and/or drainage channels in distal locations and at lower elevations, they became laterally
unconfined and were then able to intrude ice-edifice interfaces as sill-like sheets and lobes.
References
Björnsson, H. (1988) Hydrology of Ice Caps in Volcanic regions. Soc. Sci. Islandica. Volume
45, 139 pp, Reykjavik.
Lescinsky, D.T. and J.H. Fink. (2000). Lava and ice interaction at stratovolcanoes: Use of
characteristic features to determine past glacial extents and future volcanic hazards. J.
Geophys. Res., 105, B10, 23711-23726.
McGarvie DW, Burgess R, Tindle AG, Tuffen H, and Stevenson JA (2006). Pleistocene
rhyolitic volcanism at the Torfajökull central volcano, Iceland: eruption ages,
glaciovolcanism, and geochemical evolution. Jökull 56: 57-75.
McGarvie DW, Stevenson JA, Burgess R, Tuffen H and Tindle AG (2007). Volcano-ice
interactions at Prestahnúkur, Iceland: rhyolite eruption during the last interglacial-glacial
transition. Annals of Glaciology. In press.
Stevenson, J.A., D.W. McGarvie, J.L. Smellie and J.S. Gilbert. 2006. Subglacial and icecontact volcanism at the Öraefajökull stratovolcano, Iceland. Bull. Volcanol., 68, 737752.
ORAL
CORRESPONDING AUTHOR: DAVE McGARVIE
28
The Phreatomagmatic Origin of Home Plate, GUSEV Crater
[*James W. Rice, Jr.*] (School of Earth and Space Exploration, Arizona State University, PO
Box 876305, Tempe, AZ 85287-6305, Ph. 480-965-3205, FAX 480-965-1787,
jrice@asu.edu); Nathalie Cabrol (NASA Ames Research Center, Space Science Division, MS
245-3, Moffett Field, CA. 94035-100); Timothy McCoy (Dept. of Mineral Sciences, National
Museum of Natural History, Smithsonian Institution, 10th and Constitution Ave, NW,
Washington, D.C. 20560-0119); Mariek Schmidt (Dept. of Mineral Sciences, National
Museum of Natural History, Smithsonian Institution, 10th and Constitution Ave, NW,
Washington, D.C. 20560-0119); Steven W. Squyres (Dept. of Astronomy, Space Sciences
Bldg, Cornell University, Ithaca, NY 14853); R. Aileen Yingst (Space Grant Center,
Department of Natural and Applied Sciences, University of Wisconsin-Green Bay, 2420
Nicolet Drive, Green Bay, WI 54311-7001); and the MER Athena Science Team
Terrestrial phreatomagmatic eruptions occur when ascending magma contacts ground water,
ice and or wet sediments resulting in an explosion and forming one of the following volcanic
edifices; tuff cones, tuff rings and maars. Tuff cones and tuff rings form by shallow
explosions and tend to have finer grained deposits with better sorting than maars which are
formed by deeper more powerful eruptions [Fisher and Schmincke, 1984]. Deposits of
phreatomagmatic eruptions are characterized by well developed beds typically ranging in
thickness from a few millimeters to several tens of centimeters, however most are less than
ten centimeters thick. This profusion of numerous thin beds is the result of a large number of
short eruptive pulses. Bedding varies from plane parallel to cross bedded. Lapilli and bomb
sags are also commonly associated with phreatomagmatic deposits.
The Mars Exploration Rover Spirit continues to investigate a layered feature called Home
Plate. These deposits may contain the first known example of extraterrestrial
phreatomagmatic eruptions. Home Plate, a roughly circular shaped plateau structure 2 to 4 m
high and 90 m diameter, is located on the floor of the Inner Basin of the Columbia Hills
[Squyres et al, 2007]. The Inner Basin is an amphitheatre-shaped lowland which opens
westward toward the basaltic plains of Gusev crater. Spirit reached Home Plate on sol 744
and found this feature to be composed of inward dipping layered rock capped with
scoriaceous basaltic rocks. Home Plate is composed of two major rock units. The lower most
unit called Barnhill is a coarser grained wavy undulating, laminated to massive, ribbed rock
with alternating coarse and fine layers. Home Plate was found to have a composition very
similar to nearby scoriaceous basalts, but with markedly higher abundances of Cl, Br, Ge,
and Zn [Schmidt, et al this meeting]. The Barnhill unit contains sub-rounded to rounded
coarse granules up to several mm in size and rounded voids. These granules are interpreted to
be accretionary lapilli, spherical balls of volcanic ash that form from a wet nucleus falling
through a volcanic ash cloud, and the voids are interpreted to be vesicles formed by the
entrainment of vapor in wet ash or possibly cavities created when lapilli eroded out of the
rock. However, the most impressive feature in this lower unit is a bomb sag. The sag was
created by the ballistic emplacement of a clast into sediments with a volume of pore water
~15-20% based on terrestrial observations of similar features [Waters and Fisher, 1971]. All
of these features suggest that Home Plate may have been created by hydrovolcanic
explosions forming a tuff ring/maar. The Barnhill unit represents base surge deposits from
these events. The upper unit called Rogan is a finer grained, moderately sorted, finely
laminated, matrix supported cross bedded clastic rock. Rogan has been interpreted to have
been reworked by aeolian processes or may represent the distal facies of a base surge deposit.
References
Fisher, R.V. and Schmincke, H.U., 1984, Pyroclastic Rocks, 472p.
29
Schmidt, M. et al, 2007, Magma-brine interaction to produce Home Plate, Gusev Crater,
Volcano-Ice Interactions on Earth and Mars abstract (this meeting).
Squyres, S.W. et al., 2007 in revision, Mars Exploration Rover Results at Home Plate, Gusev
Crater, Science.
Waters, A. C. and Fisher, R. V., 1971, Base surges and their deposits: Capelinhos and Taal
volcanoes, J. Geophys. Res., 76:5596-5614.
ORAL
CORRESPONDING AUTHOR: JAMES W. RICE, JR.
The 60th Anniversary of the Tuya: a BC perspective
J.K. Russell (Volcanology & Petrology Laboratory, Earth & Ocean Sciences, University of
British Columbia, Vancouver, Canada, ph: 1-604-822-2703, fax: 1-604-822-6088, e-mail
krussell@eos.ubc.ca); B.R. Edwards (Department of Environmental Science, University of
East Anglia, Norwich, Great Britain); K.A. Simpson (Geological Survey of Canada,
Vancouver, Canada).
Sixty years ago, W.H. Matthews published a series of landmark papers describing the
stratigraphy and morphology of basaltic volcanoes in the Tuya-Teslin region of northwestern
British Columbia [1, 2]. There, he encountered numerous, small, apparently young, volcanic
edifices hosting a variety of enigmatic features. Most of the volcanoes are basaltic in
composition but are steep-sided and commonly have “flat” tops. Later Mathews recognized
that these volcanic edifices shared common stratigraphic elements, including: pillow lavas
and breccias, massive to bedded deposits of fragmented glassy basalt (hyaloclastite), and
capping massive basalt lava. He proposed the term ‘tuya’ for these volcanoes and interpreted
their morphology and attendant volcanic lithofacies as indicative of volcanic eruptions from
beneath and within late Pleistocene glacial ice sheets.
Recent research shows that the Tuya Region tuyas are part of the northern Cordilleran
volcanic province (NCVP), a region of extension-related mafic alkaline magmatism that was
active when the Cordilleran Ice Sheet (CIS) inundated the Cordillera. Modern studies of five
tuyas (Mathews Tuya, Tuya Butte, South Tuya, Ash Mountain, Blackfly tuya) show that the
stratigraphy at each centre is unique but all show evidence that eruption, transport and
deposition of volcanic deposits transitioned from being entirely subaqueous to entirely
subaerial, resulting in a morphology and stratigraphy consistent with the classic “tuya”
formation. Recognising the transition from subaqueous to subaerial volcanism at multiple
volcanic centres provides constraints on regional water depth and by association ice
thickness. Volcano morphology likely reflects primary differences in eruptive style in the
latter subaerial stages of volcanism (explosive vs effusive).
Mathews later pioneering work [3, 4, 5] involved volcanic deposits in the Garibaldi
volcanic belt (GVB) (e.g., Canadian Cascades). This work provided a first description of an
andesitic tuya and identified a variety of morphologies (esker-confined lavas), structures
(vertical lava fronts) and features (horizontal and radiating patterns of columns) indicative of
ice-contact volcanism. The recent work of Kelman et al. [6] at Mount Cayley has shown the
GVB to have a distinctive expression of glaciovolcanism. The Mount Cayley volcanic field is
situated at high elevation and in very high relief and, thus, the ice masses associated with
most eruptions were probably relatively thin and highly permeable, promoting meltwater
escape. This has resulted in glaciovolcanic landforms, which are virtually devoid of pillows
30
and hyaloclastite. The strongest evidence for glaciovolcanic events is the occurrence of: a)
flow-dominated tuyas (e.g., Table; [3]); b) steep to vertical walls of lavas perched on valley
walls at high elevations, and c) the pervasive presence of intense columnar jointing oriented
horizontally or in radiating masses.
The linkages between glaciation and volcanism in the BC cordillera were first explored
by Grove [7]. He posed the question whether unloading of the crustal lithosphere triggered
the onset of volcanism. Edwards and Russell [8] reviewed this question and argued that the
age-relationships between volcanism and glacial loading and unloading events are not
sufficiently well known to test this idea in a meaningful way. The question of coupling
between glaciation and cordilleran volcanism remains one of the seminal issues in the
Canadian cordillera. The answer will require a concerted effort at obtaining ages for both
volcanic and glacial deposits.
References
[1] Watson, K.D., Mathews, W.H. 1944. The Tuya-Teslin Area, Northern British Columbia, British
Columbia Department of Mines, Bulletin 19, 52 pp.
[2] Mathews, W.H. 1947. ‘‘Tuyas,’’ flat-topped volcanoes in northern BC. Am J Sci 245, 560–570.
[3] Mathews, W.H. 1951. The Table, a flat-topped volcano in southern BC. Am J Sci 249, 830–841.
[4] Mathews, W.H. 1952. Mt Garibaldi, a supraglacial Pleistocene volcano in SW BC. Am J Sci 250,
81–103.
[5] Mathews, W.H. 1958. Geology of the Mount Garibaldi map-area, southwestern British Columbia,
Canada: Part II. Geomorphology and Quaternary volcanic rocks. Geol Soc Am Bull 69, 179–198.
[6] Kelman, M.C., Russell, J.K., and Hickson, C.J. 2002. Effusive intermediate glaciovolcanism in the
Garibaldi Volcanic Belt, SW BC. In: Smellie, J.L., & Chapman, M.G. (eds.) Volcano-Ice
Interaction on Earth & Mars. Geol Soc Lond, Sp. Pub, 202, 195-211.
[7] Grove, E.W., 1974. Deglaciation — a possible triggering mechanism for recent volcanism. Proc.
IAVCEI, Santiago, Chile, pp. 88–97.
[8] Edwards, B.R. & Russell, J.K. 1999. The northern Cordilleran volcanic province. Geol, 27, 243246.
Magma-Brine Interaction to Produce Home Plate, Gusev Crater
[*Mariek E. Schmidt*] (Dept. of Mineral Sciences, National Museum of Natural History,
Smithsonian Institution, 10th and Constitution Ave, NW, Washington, D.C. 20560-0119, Ph.
202-633-1799, FAX 202-357-2476, schmidtm@si.edu); Nathalie Cabrol (NASA Ames
Research Center, Space Science Division, MS 245-3, Moffett Field, CA. 94035-100);
Timothy McCoy (Dept. of Mineral Sciences, National Museum of Natural History,
Smithsonian Institution, 10th and Constitution Ave, NW, Washington, D.C. 20560-0119);
James Rice (Arizona State University, PO Box 871404, Tempe, AZ 85287-1404; Steven W.
Squyres (Dept. of Astronomy, Space Sciences Bldg, Cornell University, Ithaca, NY 14853);
R. Aileen Yingst (Space Grant Center, Department of Natural and Applied Sciences,
University of Wisconsin-Green Bay, 2420 Nicolet Drive, Green Bay, WI 54311-7001); and
the MER Athena Science Team
The most exciting discovery to date made by the Mars Exploration Rover Spirit is
Home Plate, a light-toned ~80 m. diameter, ~2 m tall platform of layered and cross-bedded
rock in the Inner Basin of the Columbia Hills of Gusev Crater (Squyres et al, in revision).
Spirit examined the Barnhill section at the north end of Home Plate from Sols 746-751. Three
rocks, as float or outcrop, were analyzed by the instruments on the rover’s arm or IDD
(Instrument Deployment Device), which include the Microscopic Imager, Alpha Particle XRay Spectrometer, and Mossbauer. Home Plate was found to have a composition very similar
31
to nearby scoriaceous basalts, but with markedly higher abundances of Cl, Br, Ge, and Zn.
Elevated Ge concentrations (as high as 70 ppm) at Home Plate are not correlated with high
Ni concentrations (300 to 400 ppm). If the source of the Ge were meteoritic, we would expect
up to 5 wt% Ni, indicating a volcanic source for the Ge at Home Plate. Micro- and
macroscopic textures at Home Plate including a bomb sag and possible accretionary lapilli
are consistent with a phreatomagmatic origin (Rice et al, this meeting).
Volcanic textures at Home Plate and its enrichment in halogen and volatile metal
elements points towards volcanic interaction with external briny groundwater. Nearby
hydrated sulphate salt soils have chemical characteristics akin to nearby rocks (Yen et al,
2007) and may be the exhalents of acidic hydrothermal vapor that separated from a Cl-rich
brine at depth. Separation of a S-rich vapor from a Cl-rich brine is a common process in
hydrothermal systems on Earth, such as at Yellowstone (Fournier, 1986). A hydrothermal
system such as this requires persistent water that may have ultimately originated from the
devolatilization of a large magma system, the flooding event at Gusev Crater that carved out
the Ma'adim Vallis canyon, or some other unknown process. The cold, dry environment of
Mars makes it likely that water resided for some time as ground ice or porous salty slush in
the near surface. Subsequently, this ground ice was melted during volcanic activity in the
Inner Basin of the Columbia Hills that produced scoriaceous basalts and the tephra that make
up Home Plate.
References
Fournier, R.O., 1989, Geochemistry and dynamics of the Yellowstone National Park
hydrothermal system, Annual Review of Earth and Planetary Science, vol. 17, p. 1353.
Rice, J. et al, 2007, The Phreatomagmatic origin of Home Plate, Volcano-Ice Interactions
on Earth and Mars abstract (this meeting).
Squyres, S.W. et al., in revision, Mars Exploration Rover Results at Home Plate, Gusev
Crater, Science.
Yen, A. et al., 2007, Composition and Formation of the Paso Robles Class Soils, Gusev
Crater, LPSC abstract #2030.
ORAL
CORRESPONDING AUTHOR M Schmidt
Testing the Links Between Deglaciation and Magma Evolution in Iceland Using
Geochemical Signatures From Basaltic Table Mountains
[*Kerri C. Schorzman*] (University of New Hampshire, 56 College Rd., Durham, NH, USA,
03824; ph: (603) 862-1718; Fax: (603) 862-2649; email: k.schorzman@unh.edu); Joe
Licciardi (University of New Hampshire, 56 College Rd., Durham, NH, USA, 03824; ph:
(603) 862-1718; Fax: (603) 862-2649; email: joe.licciardi@unh.edu); Julie Bryce (University
of New Hampshire, 56 College Rd., Durham, NH, USA, 03824; ph: (603) 862-1718; Fax:
(603) 862-2649; email: julie.bryce@unh.edu)
Basaltic table mountains in the neovolcanic zones of Iceland preserve a unique history of the
interplay between glaciation and hot spot volcanism. Geochemical signatures in the eruptive
units that comprise these subglacially erupted landforms provide insight into a variety of
geologic processes associated with rift-related volcanism including trends in melt production
rates, variations in magma composition, and the influence of deglaciation on mantle
32
processes. To address how glacial loading and unloading may have affected mantle processes
beneath Iceland, we measured major and trace element concentrations in samples from
lithostratigraphic units of thirteen table mountains in the northern (NVZ) and western (WVZ)
volcanic zones. The eruptive ages of the sampled table mountains were recently inferred from
cosmogenic 3He surface exposure dating of their subaerially erupted summit lavas (Licciardi
et al., 2007), affording an opportunity to evaluate both spatial and temporal trends in the
geochemical data.
Several previous studies have linked glacier dynamics in Iceland with changing eruption rates
and marked differences in magma compositions (cf. Sinton et al., 2005). For example, Slater
et al. (1998) and Maclennan et al. (2002) attributed temporal variations of incompatible trace
element concentrations to increased mantle melting rates during deglaciation. This
mechanism is supported by the theoretical model of Jull and Mackenzie (1996) which
showed that rapid glacial unloading can stimulate increased melt generation in the upper
mantle. In contrast, Gee et al. (1998) argued that temporal variations in geochemistry could
arise entirely from magma chamber processes related to crustal instability during ice removal.
Few studies have documented compositional trends in lithostratigraphic units within
Icelandic table mountains (e.g., Moore and Calk, 1991; Werner et al., 1996), and the absence
of age control has prevented previous studies from examining the geochemical data in the
context of a chronology of table mountain formation.
Major and trace element compositions have now been measured in samples collected from
the subaerial cap lavas of all thirteen table mountains exposure-dated by Licciardi et al.
(2007). Geochemical data have also been obtained from the pillow lava bases of five of the
table mountains, enabling a base-summit comparison of geochemical signatures. All samples
are tholeiitic basalt, ranging in MgO composition from 7 to 10 wt %. Preliminary analyses
indicate that for the five table mountains with paired base-summit samples, basal pillow lavas
are consistently less evolved than their subaerial cap lava counterparts. Results also suggest
geographic controls on variations in magma sources between table mountains in the NVZ and
WVZ. Ongoing analyses are focusing on correcting for crustal-level processes so that we
may properly evaluate trends in geochemical signatures sensitive to the degree and/or
pressure of melting. By combining age constraints from exposure dating with geochemical
modeling of melting processes and source-region characteristics, we will test hypotheses that
link ice unloading with changing mantle melting conditions.
References
Gee, M.A., Taylor, R.N., Thirlwall, M.F., Murton, B.J. 1998. Glacioisotasy controls chemical
and isotopic characteristics of tholeiites from the Reykjanes Peninsula, SW Iceland.
Earth Planet. Sci. Lett. 164, 1–5.
Jull, M., McKenzie, D., 1996. The effect of deglaciation on mantle melting beneath Iceland.
J. Geophys. Res. 101, 21,815–21,828.
Licciardi, J.M., Kurz, M.D., and Curtice, J.M., 2007. Glacial and volcanic history of
Icelandic table mountains from cosmogenic 3He exposure ages: Quat. Sci. Rev., in
press.
Maclennan, J., Jull, M., McKenzie, D., Slater, L., Grönvold, K., 2002. The link between
volcanism and deglaciation in Iceland. Geochem. Geophys, Geosys. 3, 1062,
doi:1029/2001GC000282.
Moore, J.G., Calk, L.C., 1991. Degassing and differentiation in subglacial volcanoes, Iceland.
J. Volcanol. Geoth. Res. 46, 157–180.
33
Sinton, J.M., Grönvold, K., Sæmundsson, K., 2005. Postglacial eruptive history of the
Western Volcanic Zone, Iceland. Geochem. Geophys, Geosys. 6, Q12009,
doi:10.1029/2005GC001021.
Slater, L., Jull, M., McKenzie, D., Grönvold, K., 1998. Deglaciation effects on mantle
melting beneath Iceland: Results from the Northern Volcanic Zone. Earth Planet. Sci.
Lett. 164, 151–164.
Werner, R., Schmincke, H.-U., Sigvaldason, G., 1996. A new model for the evolution of table
mountains: volcanological and petrological evidence from Herdubreid and
Herdubreidartögl volcanoes (Iceland). Geol. Rundsch. 85, 390–397.
ORAL
CORRESPONDING AUTHOR: Kerri C. Schorzman
FIRST AUTHOR IS A STUDENT
Using Lavas at the Mount Edziza Volcanic Complex (British Columbia, Canada) to
Infer Ice Contact and Processes of Syn-Eruptive Subglacial Water Drainage
[*I.P. Skilling*] (Dept of Geology and Planetary Science, 200 SRCC, University of
Pittsburgh, PA 15260, USA, ph: 412-624-5873; fax: 512-624-3914; email:
skilling@pitt.edu); J.D.G. Hungerford (Dept of Geology and Planetary Science, 200 SRCC,
University of Pittsburgh, PA 15260, USA, ph: 412-624-5873; fax: 512-624-3914; email:
jdh53@pitt.edu); B. Edwards (Dept of Geology, PO Box 1773, Dickinson College, Carlisle,
PA 17013, USA, ph: 1-717-254-8934; fax: 1-717-245-1971; email:
edwardsb@dickinson.edu); K. LaMoreaux (Dept of Geology and Planetary Science, 200
SRCC, University of Pittsburgh, PA 15260, USA, ph: 412-624-5873; fax: 512-624-3914;
email: kal46@pitt.edu); B. Cameron (Dept of Geology, PO Box 413, University of
Wisconsin-Milwaukee, Milwaukee, WI 53201, USA, ph: 1-414-229-3136, fax: 1-414 2295452, email: bcameron@uwm.edu)
Understanding the behavior of both ice and meltwater during eruptions at glaciovolcanic
centers is important for inferring former ice conditions and for modeling hydrological
hazards at active centers. Both coherent lavas and volcaniclastic rocks have been used to infer
the presence and thickness of former ice, but coherent lavas have better preservation potential
and typically preserve less equivocal evidence of ice contact. Using coherent lava to infer
former ice presence and thickness relies on several criteria, including the interpretation of (1)
gross morphology, especially unusually high aspect ratios for a particular magma
composition, (2) marginal morphology and structures, especially narrow cooling columns
perpendicular to subvertical flow margins, (3) interbedded or laterally associated
contemporaneous glaciogenic clastic rocks and (4) elevated H2O and CO2 content of glass
margins, given supporting evidence for a sub-ice paleoenvironment from the other criteria.
Recent studies in northern British Columbia at the Mount Edziza Volcanic Complex (MEVC)
demonstrate that inferring former ice conditions from basaltic lavas using the first two of
these criteria is more straightforward than for more viscous lavas, namely trachytes in this
case. This is because basaltic lava aspect ratios, structures resulting from emplacement and
cooling, and volatile solubilities are better known than for all other compositions. In addition,
trachytes at the MEVC and elsewhere commonly form steep elevated landforms, where
erosion rates are higher, and also, at the MEVC at least, trachytic glass is scarce or absent.
However, Triangle Dome at the MEVC is interpreted as a sub-ice lava dome(s) based on the
occurrence of near-vertical margins and well-preserved curvi-columnar fanning joint sets
34
along its margins The maximum preserved thickness of the dome suggests emplacement
beneath ice of at least 120m thickness.
Basaltic lava flows erupted from Tennena Cone at the MEVC provide less equivocal
evidence of ice contact and evidence for ponding and drainage of syn-eruptive meltwater.
The lavas comprise both pillowed and more massive sheet-like forms. The pillow lavas
locally occur as sinuous (meandering) narrow, <10m high ridges with near vertical margins,
interpreted as R-channel fills, and they also rest on glaciofluvial sediments which contain
basaltic pillow lava clasts. Lava in these channels was fed through a steep-sided notch atop a
sub-ice cliff from a voluminous area of dominantly massive sheet-like lavas closer to the
vent(s). The sheet-like lavas are associated with pillow lava mounds that locally preserve
very steep margins, and both are also interpreted as having been emplaced in a sub-ice
environment. This change in lava gross morphology with distance from the vent is interpreted
as reflecting a change from larger volume proximal ponded syn-eruptive meltwater to a
marginal area of low volume leakage, which occurred only or dominantly along pre-eruptive
bedrock channels. Several factors may have influenced ponding in this area include
(obviously) proximity to the site of meltwater generation (i.e. vent), bedrock topography, ice
surface topography (?) and perhaps obstruction to drainage of water and lava on encountering
an ice wall (i.e. thicker ice) above the drop in bedrock topography at the sub-ice cliffs. The
nature or existence of the pre-eruption sub-ice meltwater drainage in the vent area is not
clear, but the transition described above does suggest the (not surprising) importance of
locally enlarged meltwater volumes as a consequence of the eruption. Initial H20 analysis of
the glass from several Tennena Cone pillow rinds from lavas in the R-channels suggests a
minimum ice thickness of about 500m.
ORAL
CORRESPONDING AUTHOR IP SKILLING
Using Cooling-Contraction Joints in a Trachyte Lava Dome to Infer Ice Contact and
Water Ingress: Mount Edziza Volcanic Complex, British Columbia, Canada
[*I.P. Skilling*] (Dept of Geology and Planetary Science, 200 SRCC, University of
Pittsburgh, PA 15260, USA, ph: 412-624-5873; fax: 512-624-3914; email:
skilling@pitt.edu); K. LaMoreaux (Dept of Geology and Planetary Science, 200 SRCC,
University of Pittsburgh, PA 15260, USA, ph: 412-624-5873; fax: 512-624-3914; email:
kal46@pitt.edu); B. Edwards (Dept of Geology, PO Box 1773, Dickinson College, Carlisle,
PA 17013, USA, ph: 1-717-254-8934; fax: 1-717-245-1971; email:
edwardsb@dickinson.edu)
The pattern and dimensions of cooling-contraction joints in lavas at glaciovolcanic centers
can often provide important evidence for former ice contact and minimum ice thickness. In
particular, narrow columns perpendicular to subvertical flow margins are commonly
interpreted as evidence for contact with steep ice. Interpretation of joint patterns in such lavas
can be complex and several factors need to be considered, including the topography of the
contacted ice, water levels, ingress of water and steam along earlier joints, bedrock (including
earlier flows) topography, lava effusion rates and rheology, lava intrusion (endogenous
growth), ice unloading, recent freeze-thaw processes etc. This presentation focuses on the
influence of water ingress and the shape of the ice cavity on jointing patterns preserved at a
trachytic lava dome at the Mount Edziza Volcanic Complex (MEVC), British Columbia,
Canada, and also briefly discusses the possible role of ice unloading and endogenous dome
growth on joint patterns at this dome
35
Triangle Dome (TD), MEVC is a 120m thick trachytic dome that is interpreted as having
been emplaced in a cavity beneath ice of at least this thickness. TD preserves a complex
pattern of curvi-columnar, planar columnar and hackly/blocky jointing, but can broadly be
divided into an upper and lower zone at about a maximum of 60m above its exposed base.
The upper zone is dominated by hackly to poorly developed columnar joints, in addition to a
prominent jointing on the 10’s centimeters to 2m scale, that dips at <20 degrees into the core
of the dome, and local sill and dike-like areas of well-developed columnar joints. This upper
zone lies with a subhorizontal to gently dipping contact on the lower zone, where welldeveloped 10cm to 3m-wide curvi-columnar and planar columnar jointing dominate. In the
lower zone the columns are near-vertical, planar and 75cm-3m wide in the interior of the
complex, but grade in a fan-like manner to curved subhorizontal <75cm-wide columns in the
outer 20m of the margins of the lower zone. The upper zone is interpreted as an
“entablature”, i.e. an area where initial jointing was overprinted by cooling joints due to
water ingress. The origin of the prominent jointing that dips into the core is unclear, but may
reflect ice unloading during or after dome emplacement. Some of these dipping surfaces may
also reflect flow contacts, but were not accessible. The local areas with well-developed
columnar jointing in the upper zone may reflect endogenous growth by late-stage dike and
intrusive lobe emplacement, or may also be areas where water ingress was less efficient. The
lower zone, with its better-developed columnar joints, is interpreted as a thick “lower
colonnade” with slower cooling and less water ingress during cooling. The fan-like columns
in the outer margins of the lower zone reflect cooling dominated by direct contact with
curved margins of the ice cavity. The absence of an upper colonnade may be due to removal
by erosion, or it may never have developed because of water ponding, as opposed to direct
lava-ice contact, in the upper part of the cavity
ORAL
CORRESPONDING AUTHOR IP Skilling
Glaciovolcanic Studies: a Pivotal Role in Characterising Past ice Sheets
[*John Smellie*] (British Antarctic Survey, High Cross, Madingley Road, Cambridge CB3 0ET,
UK; ph: +44 1223 221418; fax: +44 1223 362616; email: jlsm@bas.ac.uk)
Glaciovolcanic sequences can be used as proxies for a uniquely wide range of palaeo-ice
parameters. They are thus potentially highly useful archives of palaeo-environmental
information, particularly for pre-Quaternary periods, and the parameters derived can be
incorporated by climate and ice sheet modellers in the same way as other environmental proxies.
However, despite their unique potential, there are still few published environmental studies using
glaciovolcanic methodology. On Earth, at least nine different types of terrestrial subglacial
volcanic successions can be identified using landform characteristics, lithofacies and sequence
architecture. They include mafic tuyas, felsic tuyas (2 types), pillow mounds/ridges, pillow
sheets, tephra mounds/ridges, felsic lobes/domes, and subglacial sheet-like sequences (2 types).
When tectonic influences are removed (e.g. eruptions from fissures), the different landforms are
relatively well distinguished on morphometric diagrams using height versus width (i.e. aspect
ratio). The potential for each landform and/or succession type for yielding useful environmental
information is variable and is poorly known for several. Conversely, because of the abundance of
terrestrial basaltic volcanism, the only mature palaeo-environmental studies published so far are
those that have focused on products of mafic magmas. This is illustrated here by using the results
of a five-year investigation of the Antarctic Ice Sheet using information from basaltic sequences
36
in northern Antarctic Peninsula. It is the most detailed palaeo-environmental glaciovolcanic
study conducted to date.
The James Ross Island Volcanic Group is a large (c. 6000 km2) basaltic volcanic field situated in
northern Antarctic Peninsula and dominated by the long-lived (> 6.25 m.y.) Mount Haddington
stratovolcano. It is in a pivotal position (northerly latitude, low elevation) to record the dynamics
of the adjacent Antarctic Peninsula Ice Sheet. At least 50 eruptive phases have been identified
and Mt Haddington is now the most intensively isotopically dated volcanic centre in Antarctica.
Eruptions were mainly of large-volume (individually tens of km3) lava-fed deltas, which gave
the volcano its low-profile shield-like form. Most of the deltas show well preserved and
beautifully displayed structural features diagnostic of eruption in association with erosive, wetbased ice, but a few deltas and several tuff cones were also probably constructed in the sea
during interglacial periods, particularly during the early Pliocene. More than 90% of eruptions
took place in association with a glacial cover. However, because of high errors associated with
dating young K-poor basalts, it is currently impossible to determine at what stage(s) in any
glacial cycle eruptions took place. The lack of a precise dating method is now a major factor
limiting further environmental research in these and similar sequences worldwide. However, the
overwhelming evidence for a former ice cover affecting most eruptions suggests that both glacial
and interglacial conditions are being preserved. Thus, the so-called interglacial periods are best
regarded as ice-poor rather than ice-free, and may have been similar in appearance to today. The
thickness of the glacial cover was typically just 200-300 m, with maximum thicknesses of about
700 m achieved rarely. The ice sheet would thus have had a low profile, with a large dome
centred on Mount Haddington dominating the local morphodynamics. There is no evidence for a
giant ice sheet at any stage during the eruptive period.
These results are the first evidence for the morphology, thickness and thermal regime of the
Antarctic Peninsula Ice Sheet for pre-Quaternary periods. The study illustrates the utility and
importance of the new glaciovolcanic methodology as an important palaeo-environmental proxy.
It is uniquely capable of yielding more quantifiable parameters of past ice sheets than any other
methodology currently used.
INVITED ORAL
CORRESPONDING AUTHOR JL Smellie
Jökulhlaups from Snæfellsjökull: Evidence from the Geomorphological and
Sedimentological Record
[*K. T. Smith*] (Institute of Earth Sciences, University of Iceland, Sturlugata 7, Reykjavík,
IS – 108, Iceland, ph: +354-894-9068; email: kate@raunvis.hi.is)
Snæfellsjökull is an ice-capped stratovolcano (1446 m high) with a 200 m deep summit crater
infilled with ice. Three major phreatic (plinian) eruptions are known of in the Holocene
which produced lava and airfall tephra, most recently c.1750 BP. Geological maps indicate
that over 25 eruptions have occurred in the volcanic system in the last 10.000 years.
Snæfellsjökull is central to a national park. A future eruption there could place villages,
tourist centres, farms and transport infrastructure at risk.
In terms of Holocene history and hazard perception, Snæfellsjökull has often been
overshadowed by the likes of Hekla and Katla, more active volcanoes where eruptions have
been witnessed, but as we know from the tragic examples of Öræfajökull and Nevado del
37
Ruiz, the eruptions of stratovolcanoes can be dramatic and are worthy of assessment. The
hazards they pose are no less significant because when they erupt these volcanoes
produce jökulhlaups, lahars, pyroclastic flows and lava flows along short, steep paths towards
nearby communities. This poster presents new research into Holocene jökulhlaups / lahars.
Geomorphological mapping, analysis of 36 sediment profiles around Snæfellsjökull and
geochemical analysis of tephra and putative flood deposits using an electron microprobe
show that there is evidence for major postglacial flooding events on the northern, western and
southern slopes of the volcano, along at least five different flood routes. Fluvial landforms
radiating from the volcano include rock-cut and sedimentary channels, polished microchannels, potholes, cataracts and aligned and imbricated boulders. Many of these landforms
cut into or are deposited on top of porous postglacial lava flows across which normal
streamflow is very limited.
Sedimentary investigations outside of present-day active (spring melt) channels show
extensive overbank deposits of rounded, fluvially-transported volcanogenic material with
pumice clasts up to 20 cm (c-axis) within narrow stratigraphic windows, not repeated
throughout the profiles. This indicates that the depositional events were rare and larger in
discharge than normal precipitation or snowmelt floods. These deposits are matrix-supported
with coarse, well-rounded pale pumice clasts in a sand to granule-sized tephra-dominated
matrix. Structures are minimal but include crude bedding and both fining upwards and
coarsening upwards sequences. These characteristics are found in hyperconcentrated flows
and types of fluidal flows. Geochemical analysis of pumice within these putative flood
deposits to the south of the volcano shows marked similarity with published analyses of
tephra from the youngest Snæfellsjökull central volcano eruption. This data all placed
together points towards Holocene jökulhlaup activity at Snæfellsjökull in channels radiating
from the volcano. No sign has been found to date of extremely extensive flooding over whole
flanks of the volcano.
These overbank but spatially limited deposits are more similar to lahar and jökulhlaup
deposits found in confined valley systems rather than the well studied sandur deposits
associated with the largest historical jökulhlaup events in Iceland. This reflects the limited
source of water from the smaller ice cap of Snæfellsjökull compared to Mýrdalsjökull and
Vatnajökull and also topography which constrains flows to narrow zones rather than allowing
spreading over large expanses immediately they leave the glacier margin.
This poster presents new evidence of postglacial floods of volcanogenic material, most likely
associated with eruptions within the Snæfellsjökull volcanic system, particularly the youngest
eruption of the Snæfellsjökull central volcano c.1750 BP. This is important for assessing
future hazards here and lessons learned about the nature of palaeo-jökulhlaup evidence at
Snæfellsjökull can be applied elsewhere in the world. Importantly these results highlight the
variability of evidence left behind by jökulhlaups in the geological record resulting from
differences in volcano morphology and nature, extent of ice cover and topographic
conditions.
POSTER
CORRESPONDING AUTHOR: KT Smith
38
Holocene Sublacustrine Basalt Lava Flow at Mount Baker, North Cascades,
Washington
[*David S. Tucker*] (Geology Department, Western Washington University, 516 High St.,
Bellingham Washington, 98225; ph: 360-734-9743; email: DaveTucker@mbvo.wwu.edu );
Kevin M. Scott (Geology Department, Western Washington University, 516 High St.,
Bellingham Washington, 98225)
The ca. 8800 14C years BP Sulphur Creek basalt (51-52% SiO2) lava flowed across the floor
of 27-km-long Glacial Lake Baker in the Baker River valley east of Mount Baker,
Washington. From the vent at the Schreibers Meadow cinder cone, on the south flank of
Mount Baker, lavas flowed eastward 12 km down Sulphur Creek valley. Tephra set SC
preceded the lava eruption and built the cone. After leaving the confined creek valley, lava
flows spread out into the Baker River valley, eventually forming a 2-km-wide flow front.
Lava entered, filled and crossed the glacial lake, which had drowned the Baker River. The
glacial lake, which was ca. 2-km-wide and 60-m-deep, was impounded behind deposits of the
Vashon Glaciation. The river valley is again submerged, now beneath the Baker Lake
reservoir (220 a.s.l.). Drawdown of up to 13 m of the modern lake in winter and early
spring permits fieldwork in the area of the subaqueous lava. At these times, lava exposures
extend for 2.5 km along the steep west bank near Horseshoe Cove, and are found on the
east bank for nearly 1 km. Lava that crossed the valley is found at 230 m a.s.l., or ca. 62 m
above the pre-reservoir river bottom (168 m a.s.l.). The volume of the sublacustrine lava
may have been as high as 240 x 106 m3. Prior to reservoir inundation, the Baker River at
Horseshoe Cove lay in a deep horseshoe bend canyon cut into, and floored by, basalt. In
contrast, the valley above the lava was wide, straight and braided.
The lava invaded, assimilated, and displaced laminated lake clay. Some sediment
layers were domed gently upward 1-2 m by burrowing lava; displaced clay blocks were
rafted along on the lava surface. Peperite clast morphologies are predominantly close-packed,
irregular and blocky, though fine-grained, rounded fluidal margins occur. Baked clay adheres
to some lava blocks; red-brown baking fronts extend a few cm into the host clay. Peperite
sills invade bedding planes, and resemble beds of basalt breccia conformable with lake
sediments. These may thicken toward the margins of the lava flow.
Basalt exposed below the level of Glacial Lake Baker is glassy, usually sparsely
vesicular, and pervasively fractured. Hackly fracture dominates, though pseudopillow
fractures a few cm apart are also found; both reflect quenching and fracturing by lake water
along penetrative thermal fronts. No hyaloclastite has been found. In the lower reaches
of Sulphur Creek, drowned by Glacial Lake Baker, the flow front broke up and flowed along
the lake floor as a block and ash flow. It was not over-ridden by any subsequent lava flow,
though covered with later lake clays.
The lava formed a thick plug across the lake bottom. Subaerial lava is found just west
of the modern lake, within the bounds of Glacial Lake Baker, so must have at least partly
filled the lake, reaching above its surface. Lava formed at least an emergent dam as Glacial
Lake Baker eventually drained: shallow water delta deposits rich in remobilized SC lapilli
and volcanic sediment were deposited above the lava at Horseshoe Cove; Mazama ash (6800
14
C years BP) is found in lake sediments upstream of the lava. River gravels are deposited
on an eroded surface of the lava at 218 m a.s.l. at Horseshoe Cove. Therefore, the Baker
River cut a 50-m-deep lava gorge at Horseshoe Cove since deposition of the ash, facilitated
by the shattered structure of the basalt. Drilling by Howard Coombs and Harold Stearns
during construction of the dam in 1958 indicate that at least 76 m of fractured lava are found
just west of the modern lake.
ORAL
CORRESPONDING AUTHOR: DS TUCKER
39
The Eruption of a Rhyolitic Dyke Into Shallow Ice: Hrafntinnuhryggur, Krafla, Iceland
[*Hugh Tuffen*] (Department of Environmental Science, Lancaster University, Lancaster,
UK, LA1 4YQ; phone: +44-1524-593-571; fax: +44-1524-593-985; email:
h.tuffen@lancaster.ac.uk ); Jonathan Castro (Division of Petrology and Volcanology,
Smithsonian Institution, Washington, DC 20560-0119, USA, phone: +1-202-633-1810; fax:
+1-202-357-2476; email: castroj@si.edu).
Hrafntinnuhryggur (Obsidian Ridge) at Krafla volcano, northern Iceland was generated when
a rhyolitic dyke erupted into shallow ice towards the end of the last glacial period. Several
small-volume lava bodies budded from the feeder dyke, forming a 2.5 km long ridge that is
200-400 m wide and 30-80 m in height (total volume ~0.01 km3). Although the bulk of the
ridge is made up of variably spherulitic obsidian lava, tuffaceous hyaloclastites are also
present and were probably formed at the onset of the eruption. The dyke is locally exposed at
one end of the ridge, where it cut older basaltic hyaloclastite and fed a lava flow.
Lava bodies on the ridge crest have columnar-jointed sides, characteristic of chilling
against steeply-inclined ice walls, but display remarkably well-preserved pumiceous flow
tops. Lava exposed at lower elevations close to the base of the ridge is mostly strongly
perlitised and may have intruded waterlogged hyaloclastite. It thus appears that the initially
subglacial eruption melted through ~80 m of ice and lava effusion then occurred in iceconfined but subaerial conditions.
Lava bodies on the ridge display striking textural zonations, with an outer perlitised,
hackly-fractured zone surrounding concentric zones of collapsed foam, welded obsidian
breccia and lithophysae-rich obsidian, which envelop a platy-fractured core of pervasively
spherulitised lava. These textural zones record the penetration of meltwater into the margins,
together with variations in the rate of strain and cooling of the lava with distance from the
margin. In some locations deep etching of obsidian surfaces at lava margins indicates the
action of particularly corrosive hydrothermal fluids.
In addition to reconstructing the eruption mechanisms and patterns of magma-ice
interaction at Hrafntinnuhryggur, the following studies have been carried out:
1. Measurement of water contents in the obsidian. These have shown that the lavas are
predominantly degassed (0.10-0.15 wt %), consistent with field evidence for an ice-confined
subaerial eruption. The water content of lava in the feeder dyke is significantly higher,
consistent with incomplete degassing at higher confining pressures. Small-scale
heterogeneities in water content are being measured with synchrotron FTIR and used to
reconstruct the timescale of diffusion around bubbles and spherulites. This is helping to place
constraints on the timescale of emplacement and cooling.
2. Characterisation of the major element chemistry of the lava. Despite the broad spectrum of
textures in the obsidian, its compositions is very homogenous, being a tholeiitic rhyolite with
~75.2 wt % SiO2 and minor proportions of clinoferrosilite and Fe/Ti oxide microlites.
3. Experimental determination of the high-temperature fracture mechanics of the lava. The
compressive shear strength of the lava has been measured at a range of temperatures and
strain rates. The exceptionally high strength of flawless aphyric obsidian shows the
importance of crystals, bubbles and cooling cracks in concentrating stresses and weakening
the magma.
4. Documentation of brittle-ductile deformation structures and foam collapse textures within
the obsidian lava bodies. These textures record the brittle-ductile response of the lava to
40
stresses over a variety of timescales and mechanisms of degassing from the lava, which are
key controls on the behaviour of rhyolitic eruptions.
These studies are being combined in order to reconstruct the timescale of lava emplacement
and cooling, patterns of degassing and interaction with meteoric water and ice. The eruption
at Hrafntinnuhryggur is therefore of interest both from the perspective of volcano-ice
interaction and, more broadly, as an example of an effusive rhyolitic eruption.
ORAL
CORRESPONDING AUTHOR H Tuffen
What Controls the Explosivity of Eruptions Beneath Ice Sheets? Field Evidence, Model
Results and Some Challenging Questions
[*Hugh Tuffen*] (Department of Environmental Science, Lancaster University, Lancaster,
UK, LA1 4YQ; phone: +44-1524-593-571; fax: +44-1524-593-985; email:
h.tuffen@lancaster.ac.uk )
The explosivity of eruptions beneath ice sheets largely determines the rate of melting and
thus is an important control on the hazards posed by syn-eruptive jökulhlaups. Ancient
deposits indicate a range of explosivity in both basaltic and rhyolitic sequences (e.g. in
basalts: effusive pillow lavas vs hyalotuffs; in rhyolites: lava lobes vs phreatomagmatic ash).
Explosions can be triggered by violent magma-water interaction or driven by magma
vesiculation, meaning that the confining pressure, ability of magma and water to mix and
volatile content of the magma are all important factors. As such eruptions occur within
cavities melted into the base of ice sheets, explosive magma-water interaction can only occur
if there is space in the cavity for meltwater to collect. This creates an additional complication
as the space in cavities depends upon the relative rates of melting by magmatic heat and
closure driven by ice deformation, creating complex feedback processes between eruption
mechanisms and ice sheet response.
In this presentation I will present field evidence for contrasting eruption mechanisms
in basaltic and rhyolitic sequences. The results of simplified models of eruptions beneath ice
sheets will then be used to argue that the space within cavities is critical to eruption
mechanisms, with explosive eruptions favoured by high magma discharge rates and thin ice
sheets. Some of the challenges and uncertainties in attempting to formulate generalised
models will then be discussed, as other factors such as the hydrology of the vent area are also
significant.
References
Tuffen, H. (2007) Models of ice melting and edifice growth during subglacial basaltic
eruptions. Journal of Geophysical Research 112, B03203, doi:10.1029/2006JB004523.
Tuffen, H., Gilbert J.S. and McGarvie, D.W. (2007) Will subglacial rhyolite eruptions be
explosive or intrusive? Some insights from analytical models. Annals of Glaciology, 45, in
press.
Tuffen, H., McGarvie, D.W., Pinkerton, H., Gilbert, J.S., Stevenson J.A., Brooker R. (2007)
An explosive-intrusive subglacial rhyolite eruption at Dalakvísl, Rauðufossafjöll, Iceland.
Bulletin of Volcanology, in press.
ORAL
CORRESPONDING AUTHOR H Tuffen
41
Emplacement of a Silicic Lava Dome Through the Crater Glacier of Mount St. Helens,
USA, 2004-2007
[Joseph S. Walder] (USGS, Cascades Volcano Observatory, 1300 SE Cardinal Ct., Bldg. 10,
Suite 100, Vancouver, Washington 98683, USA; ph: 1-360-993-8948; fax: 360-993-8981;
email: jswalder@usgs.gov); Richard G. LaHusen (USGS, Cascades Volcano Observatory,
1300 SE Cardinal Ct., Bldg. 10, Suite 100, Vancouver, Washington 98683, USA; ph: 1-360993-8924; fax: 360-993-8981; email: rlahusen@usgs.gov); James W. Vallance (USGS,
Cascades Volcano Observatory, 1300 SE Cardinal Ct., Bldg. 10, Suite 100, Vancouver,
Washington 98683, USA; ph: 1-360-993-8959; fax: 360-993-8981; email:
vallance@usgs.gov); Steve P. Schilling (USGS, Cascades Volcano Observatory, 1300 SE
Cardinal Ct., Bldg. 10, Suite 100, Vancouver, Washington 98683, USA; ph: 1-360-993-8939;
fax: 360-993-8981; email: sschilli@usgs.gov)
The process of lava-dome emplacement through a glacier was observed in detail for the first
time after Mount St. Helens reawakened in September 2004. The horseshoe-shaped glacier
that had grown in the crater since the cataclysmic 1980 eruption was split in two by the new
lava dome. The two parts of the glacier were successively squeezed against the crater wall,
and the termini have advanced by several hundred meters. Photography, photogrammetry and
geodetic measurements document glacier deformation of an extreme variety, with strain rates
of extraordinary magnitude as compared to those of normal alpine glaciers. Mechanical
consideration of the glacier-squeeze process leads to an estimate of about 1.3 MPa for the
driving pressure applied by the growing lava dome on the glacier.
All previous measurements on temperate glaciers (glaciers at the melting point) reveal speedup at the beginning of the ablation season and marked diurnal speed fluctuations during the
ablation season. These phenomena result from development of a meltwater drainage system
that conveys water along the glacier bed to the terminus. The drainage system modulates the
stress state at the glacier bed and thus the rate of glacier sliding. The Mount St. Helens crater
glacier is unique in manifesting neither spring speed-up nor diurnal speed fluctuations. Thus
there is evidently no slip of the glacier over its bed. The most reasonable explanation for this
anomaly is that meltwater penetrating the glacier drains into a thick layer of coarse rubble at
the bed and then enters the volcano’s groundwater system, rather than flowing through a
drainage network along the bed. The rubble was derived mostly from rock avalanches shed
from the crater walls before the glacier began to grow.
Volcano/glacier interactions are commonly thought to result in rapid meltwater generation
either as magma contacts the glacier bed or as lava or pyroclasts are erupted onto the glacier
surface. At Mount St. Helens, however, glacier melt associated with dome emplacement has
been minor, with no more than 10% of the glacier mass being lost so far. The reason is twofold: the solidified exterior of the lava dome, and the talus aprons surrounding it, greatly
reduce the rate of heat transfer from the dome’s interior to the glacier; furthermore, hot
material has been emplaced on the glacier surface only rarely as rock avalanches have been
shed from the dome.
INVITED ORAL
CORRESPONDING AUTHOR: JS Walder
42
Exploring for Ancient Hydrothermal Systems on Mars: A Case Study from southern
Iceland
[N.H. Warner] (School of Earth and Space Exploration, Arizona State University,
P.O. Box 871404, Tempe, AZ 85276-1404; nicholas.warner@asu.edu); J.D. Farmer
(School of Earth and Space Exploration, Arizona State University, P.O. Box
871404, Tempe, AZ 85276-1404; jack.farmer@asu.edu)
The identification of the hydrothermal alteration products on Mars is of particular importance
because of implications for past surface or sub-surface water, habitable environments and life
(Farmer, 2000). In this study, we investigate the hydrothermal alteration materials deposited
on sandur plains of southern Iceland during subglacial catastrophic outfloods. Such deposits
may be process analogs for proposed Martian catastrophic outflood systems, whose origins
may be tied to geothermal melting of the Martian cryosphere (Carr, 1987). Several active
basaltic volcanic centers are located beneath the ice sheets in southern Iceland. Geothermal
heating and melting of the ice has produced several sub-glacial lakes that catastrophically
drain along the ice cap margin due to hydraulic over-pressure and ice dam breakage
(Björnsson and Kristmannsdottir, 1984; Steinthorsson et al. 2000). In 1996, an eruption at the
Gjálp fissure produced a high discharge, sediment-laden catastrophic outflood (jökulhlaup)
that inundated Skeidararsandur, an ice marginal outwash plain located south of the
Vatnajökull ice sheet. Approximately 100 km west of Skeidararsandur, the Mýrdalsandur
outwash plain was the site of a similar magnitude catastrophic outflood following the 1918
eruption of Katla volcano beneath Mýrdalsjökull .
For the present study, we collected outflood sediments from surface and outcrop localities at
Myrdalsandur and Skeidararsandur. The 1996 Skeidararsandur outwash surface is
lithologically diverse. The > 2 mm size fraction is dominated by moderately vesicular
basaltic and intermediate volcanic clasts. Vesicles are commonly infilled with secondary
silica, zeolites, or clay minerals. Approximately 2-5% of the > 2 mm fraction is comprised of
yellowish-tan palagonite clasts. Cobble-sized clasts of finely bedded cryptocrystalline silica
and coarsely crystalline (vein) calcite have also been identified. The < 2 mm size matrix
material contains basaltic glass, altered palagonite glass, plagioclase (albite), fine grained
silica, and calcite. The 1918 Myrdalsandur deposits exhibit better sorting and greater
compositional homogeneity. Approximately 97% of the surface is composed of gravel, to
granule-size basaltic scoria. Yellowish-tan palagonite and dense basaltic clasts make up 2 –
5% of the gravel and granule-size surface materials. The finer matrix (< 2 mm) contains
abundant basaltic glass and minor amounts of palagonitized glass and plagioclase.
Short wave infrared (1.6 – 2.4 microns) reflectance data from the Advanced Spaceborne
Thermal Emission and Reflection Radiometer (ASTER) show significant absorption features
at 2.2 and 2.3 microns for the surface of both sandur plains. The spectral shape and
absorptions are consistent with hydrated aluminum silicates, including saponite, kaolinite,
illite, montmorillonite, natrolite, mesolite, and scolecite. Spectra suggestive of albite,
pyroxene, and goethite were also identified on the sandur surfaces. Calcite was not identified
remotely. Powder x-ray diffraction (XRD) analysis of homogenized surface materials from
both sandurs confirmed the presence of zeolite minerals and clays (smectite, illite, and
kaolinite) within the palagonite clasts. Albite, pyroxene, and quartz were also identified in the
basaltic scoria and palagonite clasts. Oriented powder x-ray diffractograms (< 2 micron
fraction) for palagonite clasts revealed randomly ordered smectite (R < 1), mixed layer
smectite/illite with 30 – 40% illite concentration, zeolites, quartz, and albite. The hydrated
aluminum silicates identified in the remote analysis likely correspond with the clay and
zeolite minerals identified in the palagonite clasts by thin section petrography and XRD. The
43
mineral associations present in the palagonite clasts suggest alteration at temperatures near
100 C.
Terrestrial analog studies that use remote, field, and laboratory techniques provide a
framework for future work at Mars. The sandur plains of southern Iceland are ideal analogs
for Martian fluvial systems whose origins may be tied to geothermal activity (Carr, 1987).
Remote identification of hydrothermal mineral assemblages within the Martian outflood
channels may provide important information regarding the hydrologic, thermal, and
(potentially) biologic evolution of the planet.
References
Björnsson, H., H. Kristmannsdottir, 1984, The Grímsvötn geothermal area, Vatnajökull,
Iceland, Jökull, v. 34, p. 25 – 50.
Carr, M.H., 1987, Water on Mars, Nature (London), v. 326, p. 30 – 35
Farmer, J.D. 2000, Hydrothermal systems: Doorways to early biosphere evolution.
GSA Today, v. 10, p. 1-9.
Steinthorsson, S., B.S. Hararson, R.M. Ellam, G. Larsen, 2000, Petrochemistry
of the Gjalp-1996 eruption, Vatnajökull, SE Iceland, Journal of Volcanology and
Geothermal Research, v. 98, p. 79 – 90.
POSTER
CORRESPONDING AUTHOR: N. H. Warner
Volcano-Glacier Interactions during Recent Volcanic Unrest at Aleutian Arc Volcanoes
[*C.F. Waythomas*] (USGS, Alaska Volcano Observatory, 4230 University Drive, Suite
201, Anchorage, AK, USA, 99508; ph: 907-786-7122; fax: 907-786-7150; email:
chris@usgs.gov); C.A. Neal (USGS, Alaska Volcano Observatory, 4200 University Drive,
Anchorage, AK, USA, 99508; ph: 907-786-7458; fax: 907-786-7425; email:
tneal@usgs.gov); R.G. McGimsey (USGS, Alaska Volcano Observatory, 4200 University
Drive, Anchorage, AK, USA, 99508; ph: 907-786-7432; fax: 907-786-7150; email:
mcgimsey@usgs.gov); K. Bull (Alaska DGGS, Alaska Volcano Observatory, 3354 College
Rd., Fairbanks, AK, USA, 99709; ph: 907-451-5055; fax: 907-455-6879; email:
kate_bull@dnr.state.ak.us); R. Wessels (USGS, Alaska Volcano Observatory, 4200
University Drive, Anchorage, AK, USA, 99508; ph: 907-786-7492; fax: 907-786-7425;
email: rwessels@usgs.gov); M.L. Coombs (USGS, Alaska Volcano Observatory, 4200
University Drive, Anchorage, AK, USA, 99508; ph: 907-786-7403; fax: 907-786-7425;
email: mcoombs@usgs.gov, K.L. Wallace (USGS, Alaska Volcano Observatory, 4230
University Drive, Suite 201, Anchorage, AK, USA, 99508; ph: 907-786-7109; fax: 907-7867150; email: kwallace@usgs.gov); C. Huggel (Department of Geography, University of
Zurich, Winterthurerstrasse 190, 8057 Zürich, Switzerland; ph: +41 44 635 51 75; fax: +41
44 635 68 41; email:chuggel@geo.unizh.ch
Recent unrest at Aleutian arc volcanoes has provided a number of diverse examples of the
interaction between volcanic processes and glaciers, ice and snow. The case studies described
44
below, illustrate how these interactions may lead to potentially hazardous outcomes and how
knowledge of these phenomena may help mitigate hazards posed by volcano-glacier
interactions elsewhere.
The 1989-90 eruption of Redoubt volcano (60.485 N, 152. 744 W) was an explosive domebuilding eruption that physically removed the upper one third of Drift Glacier, a 14 km2
valley glacier on the northeast flank of the volcano. Pyroclastic flows and surges ravaged the
lower reaches of Drift Glacier and led to lahars and floods that inundated the Drift River
valley, including an oil storage and transfer facility near the mouth of the Drift River about
40 km downstream of the volcano. This eruption highlights pyroclastic flow interaction with
ice and snow and the subsequent generation of destructive lahars.
The 1983-84 eruption of Veniaminof volcano (56.195 N, 159.39 W) was a Strombolian
eruption from a vent within the 8-km-diameter ice-filled caldera that characterizes the
volcano. Lava flows produced during the eruption created ice cauldrons that evolved to open
melt pits with ephemeral lakes. Although catastrophic release of water and flooding did not
occur, larger eruptions that produce more extensive lava flows could lead to outburst floods
from the caldera ice field. This eruption highlights lava-flow interaction with glacier ice.
Augustine Volcano (59.363 N, 153.435 W) erupted explosively in January- February 2006.
Pyroclastic flows generated during the explosive phase of the eruption swept across steep,
extensively snow-covered slopes on all flanks of the volcano. The snow blanket caused the
flows to inflate, spread out, and form thin, sheet-like pyroclastic-flow deposits. Pyroclastic
flows erupted later, that did not interact with snow, followed drainages and produced deposits
with more classic morphology, including blocky, lobate margins and levees. Lahars and
mixed avalanches of snow, water, and pyroclastic debris formed beyond the pyroclastic flows
that swept across snow. Flow morphologies, lithologies, and depositional contacts indicate
that the lahars and mixed avalanches evolved from these pyroclastic flows. These deposits
highlight the interaction between snow and pyroclastic debris.
Volcanic unrest at Mt. Spurr volcano (61.298 N, 152.253 W) from 2004 to present has
resulted in the development of a 200 x 200 m diameter, lake bearing melt pit at the 3374 m
high summit of the volcano, and several episodes of unusual water release. Although no
eruption of the Mt. Spurr summit vent has occurred, the melt pit lake remains open and
partially ice free. This unrest highlights low-level geothermal interaction with ice and snow.
On September 17, 2006, a major steam emission occurred at Fourpeaked Volcano (58.769 N,
153.674 W) that produced a plume reaching about 6000 m above sea level. Observations
made after the event indicated that a small debris flow and several melt pits developed on the
north flank of the volcano. The debris flow was the product of an outburst event that may
have initiated the collapse of an ice vault above the newly reinvigorated hydrothermal vent
system. The collapsing ice vault is thought to have brought glacier ice and water into rapid
contact with hot gasses and rock leading to a phreatic eruption that produced the steam plume
observed on September 17, 2006. This unrest highlights explosive water-ice-hydrothermal
interaction.
These examples are representative of the diverse range of volcano-glacier interactions in
Alaska during recent periods of unrest. The high frequency and range of interactions offers
many possibilities for further study and illustrates the importance of this region as a natural
laboratory for study of volcano-ice interactions.
POSTER
CORRESPONDING AUTHOR: C Waythomas
45
The Effect of Tephra Deposition on Equatorial Glacial and Polar Ice on Mars
[*L. Wilson*] (Environmental Science Dept., Lancaster University, Lancaster, LA1 4YQ,
UK; ph: +44-1524-593889; email: L.Wilson@Lancaster.ac.uk); J. W. Head (Geological
Sciences Dept., Brown University, Providence, RI 02912, USA; ph: +1-401-863-2526; email:
James_Head@brown.edu)
The deposition of tephra from explosive volcanic eruptions on Mars can influence the
stability of ice by: 1) direct melting from hot tephra, 2) changing thermal inertia, 3) changing
surface albedo, 4) changing atmosphere temperature, and 5) modifying rheological
properties. The Tharsis Montes, themselves formed of a mixture of products of effusive and
explosive eruptions, contain cold-based glacial deposits on their NW flanks that imply ice
cover over areas of order 100,000 km2 to a depth of up to ~2-3 km. Summit magma
reservoirs are estimated to have vertical extents of up to 10 km and diameters of up to 120
km (Arsia Mons). The volumes of typical eruptions discharging ~1% of the reservoir magma
could be up to ~1000 km3. If this volume were erupted exclusively as tephra and distributed
uniformly over the volcano surface, the deposit depth would be ~3 m, or ~6 m if distributed
preferentially into the glacial deposit sector. A fall deposit from a summit eruption would
consist entirely of clasts that had reached ambient temperature (~200 K) by the time that they
landed, and would produce no direct thermal effects. A pyroclastic density current could have
a temperature of ~700 K; about half of the heat content would be transferred to the
underlying ice, and could melt an ice thickness approximately equal to the thickness of the
pyroclast layer, i.e. up to ~6 m, a very small fraction of the ~2-3 km estimated maximum
thickness of cold-based glacial ice. A subglacial phreatomagmatic eruption creating a graben
of the scale of Aganippe Fossa on the north-western flank of Arsia Mons could produce
sufficient country rock ejecta to form a layer ~4 m thick averaged over the entire glacial
deposit. A second consequence of volcanic mantling deposits is their effect on the long-term
stability of the underlying ice. Both fall and pyroclastic flow deposits dominated by juvenile
magma on Mars should consist of generally sub-mm sized clasts and have a low thermal
inertia and high tortuosity, thus tending to protect underlying ice from sublimation; however,
they would also have a lower albedo than the ice, thus reaching higher daytime surface
temperatures. This effect would compete with the protective properties of the deposit. In
contrast, phreatomagmatic country rock deposits should be coarse-grained, having higher
thermal inertias that might encourage significant local melting. Furthermore, the deposition
of extensive phreatomagmatic ejecta on top of the proximal/accumulation region of the
tropical mountain glacier system could have major effects on the dynamics, stability,
evolution and ultimate fate of the ice sheet.
Calculations using the sub-mm grain size expected for juvenile pyroclasts on Mars show that
fall deposits from the highest mass eruption rate events from these equatorial volcanoes can
be dispersed planet-wide; 1000 km3 distributed uniformly over the entire surface of Mars
represents a layer ~1 cm deep. Such a layer over the polar ice deposits would have negligible
protective ability but could significantly enhance sublimation, and would contribute to the
lower-albedo layers seen in the polar layered terrain.
ORAL
CORRESPONDING AUTHOR L. Wilson
46
Intrusion of Dikes into a Late Amazonian Tropical Mountain Glacier: Evidence for a
Phreatomagmatic Origin of the Aganippe Fossae System, Western Arsia Mons, Mars
[*L. Wilson*] (Environmental Science Dept., Lancaster University, Lancaster, LA1 4YQ,
UK; ph: +44-1524-593889; email: L.Wilson@Lancaster.ac.uk); J. W. Head (Geological
Sciences Dept., Brown University, Providence, RI 02912, USA; ph: +1-401-863-2526; email:
James_Head@brown.edu)
Late Amazonian cold-based tropical mountain glaciers (TMG) have been documented on the
western flanks of the Tharsis Montes, which themselves involved contemporaneous effusive
and explosive eruptions. Synglacial volcanism appears to be largely related to the
emplacement of dikes into the shallow crust that predominantly produce narrow graben
across the surface in non-glacial areas, but that intrude into glacial ice when they encounter
the TMGs. Dike-related synglacial eruptions produce dike-like ridges, sill-like intrusions,
steep-sided flows, tephra cones, and móberg-like ridges. Here we document a new type of
synglacial TMG subglacial eruption: linear troughs produced by dike-related
phreatomagmatic eruptions.
The Aganippe Fossae system consists of two major troughs that extend in a NNW direction
across the northwestern flank of Arsia Mons and are very closely associated with TMG
deposits. The fossae typically range from ~2-5 km in width and are discontinuous in nature,
having been modified by portions of the fan-shaped deposit. At the southern margin of the
fan-shaped deposit, the wide western trough abruptly changes its morphology into a narrow
graben, which is eventually covered along strike by later lava flows. Although the fossae
have clearly been modified by subsequent events, the widest and most distinctive structures
generally occur in the interior of the fan-shaped deposits. In this analysis we addressed the
questions of why the fossae are confined to the fan-shaped deposit, why their morphology
changes in relation to proximity to the deposit margins, their stratigraphic position in relation
to the cold-based tropical mountain glacier deposits, and candidate modes of origin.
Superposition relationships show that the fossae have been modified by the glacial processes
associated with the Late Amazonian TMG. Smooth facies interpreted to represent late-stage
debris-covered glaciers completely cover portions of the troughs and debris-covered glacial
features on the trough floors clearly indicate that troughs have been modified and widened by
glacial processes. In some places the troughs appear to be buried by the knobby facies,
thought to represent the collapse and vertical downwasting of the TMG, while in others the
margins of the troughs appear sharp, but are likely to have been modified by post-formation
mass-wasting and glacial modification of the trough walls. Stratigraphic relationships suggest
that the troughs formed during the emplacement of the Arsia TMG. The occurrence in the
western distal part of the Arsia TMG of dike-related subglacial volcanism strengthens the
likelihood that subglacial eruptions might also have occurred in the thicker proximal portion
of the ice sheet. As overlying ice becomes thicker (estimates range up to 2-3 km), dikes will
intrude further into the ice sheet causing melting of marginal ice and the production of
meltwater both in the crustal cryosphere and the glacier itself. We have modelled the
intrusion of dikes into this configuration and find that 1) sufficient meltwater is produced to
create hydromagmatic interaction and explosive phreatomagmatic eruptions, and 2) that
further explosiveness occurs when the glacial ice is breached and exsolved magmatic gas
undergoes catastrophic decompression to ambient conditions. For example, a typical 40 m
wide dike emplaced into this configuration can create a phreatomagmatic eruption that will
form an elongate depression ~2.7 km wide and ~2000 m deep. These phreatomagmatic
eruptions would have produced ejecta composed of a combination of chilled fragmental
magma, fragmented country rock, and aqueously altered tephra. Estimates of the initial
volumes of the troughs suggest that as much as 600 km3 of country rock ejecta might have
47
been produced, forming a layer averaging ~4 meters thickness over the entire TMG deposit,
potentially influencing glacial dynamics and stability. Subsequent glacial and mass-wasting
activity has altered the fossae to their present configuration.
ORAL
CORRESPONDING AUTHOR L Wilson
48