This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
ARTICLE IN PRESS
Planetary and Space Science 57 (2009) 579–596
Contents lists available at ScienceDirect
Planetary and Space Science
journal homepage: www.elsevier.com/locate/pss
Mesoscale raised rim depressions (MRRDs) on Earth: A review of the
characteristics, processes, and spatial distributions of analogs for Mars
Devon M. Burr a,b,, Barbara C. Bruno c,d, Peter D. Lanagan e, Lori S. Glaze d,f, Windy L. Jaeger g,
Richard J. Soare h,i,j, Jean-Michel Wan Bun Tseung k, James A. Skinner Jr.g, Stephen M. Baloga d
a
Earth and Planetary Science Department, University of Tennessee, 1412 Circle Drive, Knoxville, TN 37996-1410, USA
SETI Institute, 515 North Whisman Road, Mountain View, CA 94043, USA
c
School of Ocean and Earth Science and Technology (SOEST), University of Hawaii, 1000 Pope Road, MSB 205, Honolulu, HI 96822, USA
d
Proxemy Research, 20528 Farcroft Lane, Gaithersburg, MD, 20882, USA
e
700 College Dr., Mail Code H3C, College of Southern Nevada, Henderson, NV 89002, USA
f
Goddard Space Flight Center, Greenbelt, MD 20771, USA
g
Geological Survey, Astrogeology Team, 2255 N. Gemini Dr., Flagstaff, AZ 86001, USA
h
Department of Geography, Planning and Environment, Concordia University, 1455 De Maisoneuve W., Montreal, Quebec, Canada H3G 1M8
i
Department of Geography, University of Calgary, 2500 University Drive NW, Calgary, Alberta, Canada T2N 1N4
j
Department of Earth Sciences & Physics and Astronomy, University of Western Ontario, 1151 Richmond Street, London, Canada, N6A 5B7
k
Department of Geography, Dawson College, 3040 Sherbrooke St. W., Montreal, Canada, H3Z 1A4
b
a r t i c l e in fo
abstract
Article history:
Received 13 November 2008
Accepted 25 November 2008
Available online 6 December 2008
Fields of mesoscale raised rim depressions (MRRDs) of various origins are found on Earth and Mars.
Examples include rootless cones, mud volcanoes, collapsed pingos, rimmed kettle holes, and basaltic
ring structures. Correct identification of MRRDs on Mars is valuable because different MRRD types have
different geologic and/or climatic implications and are often associated with volcanism and/or water,
which may provide locales for biotic or prebiotic activity. In order to facilitate correct identification of
fields of MRRDs on Mars and their implications, this work provides a review of common terrestrial
MRRD types that occur in fields. In this review, MRRDs by formation mechanism, including
hydrovolcanic (phreatomagmatic cones, basaltic ring structures), sedimentological (mud volcanoes),
and ice-related (pingos, volatile ice-block forms) mechanisms. For each broad mechanism, we present a
comparative synopsis of (i) morphology and observations, (ii) physical formation processes, and (iii)
published hypothesized locations on Mars. Because the morphology for MRRDs may be ambiguous, an
additional tool is provided for distinguishing fields of MRRDs by origin on Mars, namely, spatial
distribution analyses for MRRDs within fields on Earth. We find that MRRDs have both distinguishing
and similar characteristics, and observation that applies both to their mesoscale morphology and to
their spatial distribution statistics. Thus, this review provides tools for distinguishing between various
MRRDs, while highlighting the utility of the multiple working hypotheses approach.
& 2008 Elsevier Ltd. All rights reserved.
Keywords:
Mars
Surface
Terrestrial analogs
Spatial analysis
1. Introduction
Fields of mesoscale raised rim depressions (MRRDs) of various
origins are found on Earth and Mars. We define MRRDs morphologically, as 100-m-scale-diameter forms of circular, elongate, or
irregular plan view shape, having closed or mostly closed positiveelevation rims encircling lower elevation terrain. Examples of
fields of Martian MRRDs are shown in Fig. 1. Correctly identifying
fields of MRRDs by type (origin) on Mars is important because
different types of MRRDs have widely different geologic and/or
climatic implications. Identification is useful in the search for
Corresponding author.
E-mail address: [email protected] (D.M. Burr).
0032-0633/$ - see front matter & 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.pss.2008.11.011
biotic or prebiotic activity on Mars because MRRDs are often
associated with volcanism and/or water, which are constituents of
such activity on Earth (Farmer, 1998 and references therein).
As terrestrial experience commonly guides extra-terrestrial
investigations, terrestrial MRRD types have been commonly
hypothesized to explain MRRDs on Mars. These terrestrial MRRD
types include phreatomagmatic volcanic constructs, massive
surficial or ground ice forms, and mud volcanoes.
This work reviews terrestrial MRRD types that most commonly
occur in fields and that have been hypothesized to explain fields of
MRRDs on Mars. The approach is two-fold: (1) to provide generic
observations of individual MRRDs from the literature down to
meter-scale, a size scale commensurate with current spacecraft
data of Mars, and (2) to quantify the spatial distribution of
MRRDs in these fields. In this work, we focus on particularly
Author's personal copy
ARTICLE IN PRESS
580
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
Fig. 1. Example MRRDs on Mars, showing a variety of morphologies and settings. Multiple MRRD morphologies are visible in each image. (a) Subset of MOC image M0700373, located in western Amazonis Planitia, near 12N 164W. (b) Subset of MOC image E22-00813, located in Acidalia Planitia, near 44N 28W. (c) Subset of MOC image R0100365, in Athabasca Valles, near 10N 204W.
well-researched, well-formed, or well-preserved examples in
order to present criteria that characterize fields of MRRDs on
Earth. These criteria in turn provide tools for assessing fields of
MRRDs on Mars. This review is limited to those types of terrestrial
MRRDs that have been hypothesized in the Mars literature to be
analogs for Martian fields of MRRDs.
The intent of this work is to aid in correctly identifying MRRDs
by type on Mars and in understanding their implications. For each
type of terrestrial MRRD, we (i) describe the meter-scale morphology and other relevant observations, (ii) summarize the physical
processes of formation, and (iii) give locations on Mars where such
MRRDs have been hypothesized to exist. As an additional tool for
distinguishing among the possible origins of fields of MRRDs on
Mars, we also quantify the spatial distribution of MRRDs that occur
in fields on Earth. This includes both summarizing previous work
and adding new statistical measures for additional MRRD types,
doubling the total number of types for which spatial analyses have
been published. Two tables summarizing these morphological and
statistical comparisons are provided.
The various MRRDs are grouped according to three fundamental origins: hydrovolcanic, sedimentological, and ice-related.
Other origins are possible for MRRDs on Mars but fall outside the
focus of this work. Spring deposits (Crumpler, 2003a, b) and sand
diapirs (Mahaney et al., 2004 and references therein) may form
MRRDs and have been hypothesized to explain MRRDs on Mars.
However, they do not commonly occur in fields and so are not
amenable to spatial distribution analyses as given here. Secondary
impact craters form fields of MRRDs on Mars (e.g., McEwen et al.,
2005). We have not performed new analyses of impact craters on
Earth but have included the results of previous spatial analysis of
secondary and pedestal impact craters in Table 2.
2. Hydrovolcanic MRRDs
Hydrovolcanic MRRDs are constructs emplaced through interaction of lava with surficial or near-surface water. These
constructs include ‘rootless’ cones (eruptive cones formed by
tube- and channel-fed lavas) and basaltic ring structures.
2.1. Rootless cones (RCs)
2.1.1. RCs: morphology and observations
2.1.1.1. Icelandic tube-fed rootless cones. Tube-fed rootless cones
have been studied in Iceland by a number of workers (see a review
in Fagents and Thordarson, 2007). Thorarinsson (1953) first
classified volcanic conical features in lacustrine or riverine
environments that lack direct connection to primary volcanic
vents as pseudocraters, and identified these structures as resulting
from surficial lavas heating groundwater to generate steam. In
recent years, the term pseudocrater has been largely replaced by
the term rootless cone since these are real features and therefore
the prefix pseudo- is inapplicable. (The term rootless cone also
sidesteps an issue of confusion among planetary scientists who
associate craters with meteorite impacts.)
Phreatomagmatic rootless cones in the Mývatn and Raudhólar
areas are located around the shorelines of lakes (Thorarinsson,
1953; Greeley and Fagents, 2001, Figs. 1 and 5, see their Fig. 4 for
locations of the 4 Icelandic rootless cone fields). These cones have
basal diameters of 3–320 m, heights of up to 31 m, summit-crater
to basal diameter ratios of 0.45–0.54, and exterior slopes of
20–251 (Thorarinsson, 1953; Frey and Jarosewich, 1982; Greeley
and Fagents, 2001). Both of these rootless cone populations show
a close-packed arrangement with overlapping of excavated
material (Greeley and Fagents, 2001). The Raudhólar cones tend
to cluster near the edges of their host lava flows and appear to
transition upwards in size from hornitos composed of spatter to
larger rootless scoria cones. According to Fagents and Thordarson
(2007), coexistence of different types of cone morphology
connotes variation in eruption intensity, which is in turn a
function of variation in the lava–water interaction. The Raudhólar
cones also incorporate xenoliths of compositions similar to those
of the underlying lavas. Although phreatovolcanic rootless cones
are not the source of lava flows, their spatter can coalesce into
rheomorphic flows that may appear similar to small, vent-source
flows, and small amounts of lava are associated with peripheral
cones of the Raudhólar group. A geospatial analysis of cone
alignment within the Raudholar group revealed no strong linear
alignments (Bruno et al., 2004).
Rootless cone fields in the Álftaver and Landbrot (Fig. 2)
districts are located on the margins of glacial river outwash plains
(see Greeley and Fagents, 2001, Figs. 6–9). Compared to the
Mývatn and Raudhólar cones, the outwash plains rootless
cones are more widely spaced. Some have distinctive summit
depressions, whereas others have convex-up slopes with summit
peaks. Cones reach up to 20 m in height, with maximum basal
diameters of generally 60 m, although some large cones are as
wide as 100 m (Greeley and Fagents, 2001; Fagents et al., 2002).
The cones have convex-upward slopes, suggesting post-formational modification by glacial floods. The structure of these cones
Author's personal copy
ARTICLE IN PRESS
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
581
Fig. 2. (a) Aerial photograph and (b) high resolution image showing well-formed cones in the Landbrot rootless cone field, southern Iceland. Illumination is from the upper
right. Cone alignments are discussed by Bruno et al. (2004).
Fig. 4. Image of two littoral cones, southwest Kilauea, Hawaii. To the west (left) is
Pu’u Kaimu’uwala, showing a segmented morphology; white arrows point out a
lava channel, consistent with the model proposed for segmented cones (see Fig. 5).
To the east (right) is Pu’u Waimanalo, showing a classic circular morphology.
Image credit: Google Earth.
Fig. 3. Spatter at the Landbrot cone field, showing the molten state of the lava
upon eruption.
is similar to those in Mývatn and Raudhólar. Linear alignment
detection analyses (Bruno et al., 2004) found strong linear
alignments within the Landbrot cone group (Fig. 2). The difference
in cone morphology and spacing may be due in part to differences
in availability of water at the time of formation; the rootless cones
associated with lake basins presumably had a more abundant
water supply than those formed along glacial rivers (Greeley and
Fagents, 2001). As in the Mývatn and Raudhólar rootless cone
fields, the scoria clasts comprising the Landbrot cones show taffylike textures (Fig. 3), indicative of semi-molten material.
The Landbrot cones are embayed in some locations by the 1783
Lakagı́gar flow. This more recent flow also contains welldeveloped rootless cones, which retain near-pristine morphology.
See Lanagan et al. (2001), Fig. 2b, for an image of these cones.
2.1.1.2. Hawaiian littoral cones. Littoral cones, the best-studied
examples of which are in Hawaii, are hydrovolcanic constructs
that form when lavas encounter ocean water. Littoral cones are
typically comprised of a few cone segments, although they may
also form complete cones (Fig. 4). Multiple clusters of littoral
cones exist along the southwest coast of Hawaii. A series of
littoral cones located at Nā Pu’u a Pele (The Hills of Pele) shows
the typical segmented morphology of littoral cones. Several
kilometers to the southeast, a cluster of five complete littoral
cones is located on the Pohue Bay flow and the Kolo flow, with
individual cone diameters of 200–300 m (Jurado-Chichay et al.,
1996a). Pu’u Kı̄ is the largest of these cones, which are usually
referred to by that name. In this cluster, Three-rim Cone is a
truncated a 300-m-wide crescent composed of pyroclastics and
opening seaward. Two concentric rims interior to Three-rim Cone
composed of spatter are also truncated. The north littoral cone of
the Pu’u Ki cluster exhibits an outer rim of ‘‘lapilli-, bomb-, and
block-sized ejecta surrounding inner deposits of dense spatter
and ponded lavas’’ (Jurado-Chichay et al., 1996a). Approximately
10 km further to the southwest, Pu’u Kaimu’uwala is a traditional
segmented cone, and Pu’u Waimanalo has a classic circular
morphology (Fig. 4). Southeast of these cones, Pu’u Hou (New Hill)
is another prominent segmented littoral cone, shown in Greeley
and Fagents (2001).
Littoral cones may be emplaced either by channel-fed lavas
(e.g., Macdonald, 1972; Macdonald et al., 1983) or by tube-fed
lavas (Fisher and Schmincke, 1984). As described below, the
difference in cone morphology—segmented versus complete—has
been inferred to be a function of the emplacement mechanism.
2.1.2. Rcs: formation mechanism
Observations of Icelandic rootless cones suggest that they
formed where tube-fed lavas overrode marshy surfaces (Morrissey
and Thordarson, 1991; Thordarson, 2000). Many of these cones sit
on top of lava flows that show no evidence of deformation around
the cones, indicating that some time elapsed between the
emplacement of the flow and the cone-forming event. Fieldwork
has shown that these cones are primarily composed of scoria and
tephra interfingering with beds of lacustrine muds overlain by
agglutinate (Morrissey and Thordarson, 1991; Thordarson, 2000).
Distinct horizons of lacustrine muds indicate that multiple
explosions are necessary to form a rootless cone. If these
explosions vary in strength and/or location, they may result in
nested and overlapping cones (Thorarinsson, 1953). The scoria
clasts which comprise these cones often exhibit fluidal, taffy-like
textures (e.g., Fig. 3), indicating that the material was molten
or semi-molten during cone formation. Grain sizes generally
increase with height within the cone, indicating decreasing
Author's personal copy
ARTICLE IN PRESS
582
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
explosive power over the course of the eruption. Nearest neighbor
and linear alignment detection analyses of the spatial distributions of rootless cones indicate that they have a clustered
distribution and may exhibit preferred alignments, hypothesized
to result from lava tubes and/or groundwater locations (Bruno
et al., 2004).
Using geologic constraints based on Icelandic and Hawaiian
fieldwork, Keszthelyi (2005) calculated steam accumulation rates
within waterlogged soils beneath surficial lavas. For reasonable
permeabilities, the water has to be less than a few centimeters
under the surface to allow enough pressure to build to produce a
cone before the steam escapes laterally (Keszthelyi, 2005).
Models for rootless cone formation date back to the early
1980s (e.g., Wohletz and Sheridan (1983)). More recently Greeley
and Fagents (2001) suggested a model for rootless cone formation
based on Icelandic rootless cones as follows: (1) Lava is emplaced,
which heats the substrate and volatilizes the ground water,
building subsurface gas pressure. (2) Vapor pressure eventually
exceeds a pressure threshold (which is equal to the weight and
strength of overlying lava) and results in rapid gas expansion
which excavates overlying lava and perhaps some amount of
substrate. (3) The inflow of lava and other debris seals the void. (4)
Finally, more gas is generated to repeat the cycle.
The first two steps of this model may be controlled by
fuel–coolant interaction (FCI). This process and its application to
geology are reviewed by Morrissey et al. (2000), from which the
following summary is taken. FCI involves heat transfer that
converts thermal energy to kinetic energy on short time scales
(o1 ms). The ‘fuel’ is a hot fluid with a solidus temperature above
the nucleation temperature of the coolant (i.e., the coolant is a
volatile). The explosivity of the interaction depends on the rate of
heat transfer from fuel to coolant, which in turn depends on the
surface area over which the fuel contacts the coolant. The contact
surface area may be increased by fragmentation of the fuel or
through intimate mixing of coolant within the fuel. Heat transfer
is also enhanced by the thermal expansion of the coolant, which
further increases the contact surface. Morrissey et al. (2000) list
four steps in explosive FCI processes: (1) initial contact and
mixing of fuel and coolant with stable vapor film conditions; (2)
vapor film collapse, leading to thermal and mechanical coupling
of fuel and coolant; (3) episodic heat transfer from fuel to coolant,
leading to rapid expansion of the coolant, subsequent elastic
rebound, and explosive fragmentation of the fuel; and (4)
volumetric expansion of the fuel–coolant mixture due to the
phase change of superheated water to steam. Volcanologically, the
‘fuel’ is the lava and the ‘coolant’ is the water. This hydrovolcanic
process generally results in cones with summit craters, although
littoral cones are not necessarily complete circles and some
rootless cones lack summit craters.
During littoral cone formation, the lava is quenched as it
enters the seawater, forming a lava delta (Mattox and Mangan,
1997). As this delta expands, it may collapse so that seawater
penetrates the delta and contacts the flowing lava, resulting in a
steam explosion. Pyroclastic material from the explosion is
ejected to form the littoral cone. The cycle continues as long as
the lava flow remains active. If the lava flow is channel-fed, such
that the surface of the flow in the channel continues to move
during the explosion, the ejected material that lands on the
moving surface will be carried away, leaving two arcuate
segments of a cone (Fig. 5a) (Jurado-Chichay et al., 1996a; Mattox
and Mangan, 1997; for an oblique aerial view of a Hawaiian
channel-fed littoral cone, see Greeley and Fagents, 2001, Fig. 2). If
the lava flow is tube-fed, so that the surface of the flow is solid,
the ejection of material will form a circular construct (Fig. 5b).
Erosion of the lava bench subsequent to the formation of a littoral
cone may erode the cone.
Fig. 5. Diagrams showing the formation of littoral cones. (a) Formation of a
‘typical’ bisected littoral cone, resulting from channel-fed lava (modified from
Fisher and Schmincke, 1984); and (b) formation of a circular littoral cone, resulting
from tube-fed lava (modified from Jurado-Chichay et al., 1996b).
2.1.3. RCs: hypothesized locations on mars
Numerical modeling indicates that the lower gravity and
atmospheric pressure on Mars should produce hydrovolcanic
rootless cones that are wider than those observed on Earth
(Greeley and Fagents, 2001; Fagents et al., 2002). Structures
hypothesized to be Martian rootless cones include mounds in
Chryse Planitia (Greeley and Theilig, 1978), Deuteronilus Mensae
(Lucchitta, 1978), Acidalia Planitia (Frey et al., 1979; Allen, 1980),
Isidis Planitia (Frey and Jarosewich, 1982), Elysium Planitia
(Mouginis-Mark, 1985), the Cerberus and Amazonian plains
(Greeley and Fagents, 2001; Lanagan et al., 2001; Lanagan,
2004), and Athabasca Valles (Jaeger et al., 2007). However,
observations that the Isidis cones have similar crater to basal
diameter ratios as Icelandic rootless cones have been used to
argue against them being rootless cones, since Martian gravity
and atmospheric density should result in broader features
(Bridges et al., 2003). Farrand et al. (2005) argue that the low
thermal inertias of the cones in Acidalia weigh against their being
rootless cones, since the component spatter is expected to have
relatively high thermal inertia.
2.2. Basaltic ring structures (BRSs)
Basaltic ring structure (BRS) is a generic descriptive term
referring to a rimmed topographic depression within basaltic lava
flow. BRSs appear in plan view as circular or elliptical structures
with raised rims and may have interior concentric ridges
composed of locally derived dike material. They are genetically
related to rootless cones in that they involve lava–water
interaction. However, some additional or different process is
apparently involved, since concentric basalt dikes are not
observed in rootless cones as they are in some BRSs.
2.2.1. BRSs: morphology and observations
Two morphological populations of BRSs are found in the
Channeled Scabland flood tract, set into the Columbia River
Basalts in Washington state, USA (Hodges, 1978; see also Figs. 4.16
and 4.17 in Baker, 1978, Table 1 in McKee and Stradling, 1970).
Although these two populations have slight morphological
differences as discussed below, we have grouped them together
because of their similar appearance in plan view and their close
mutual proximity. A possible but unverified field of BRSs might
also exist on the Serra Geral lavas of the Parana Basin, Brazil
(Araújo, 1982).
Odessa BRSs. The more distinctive set of BRSs is found near
Odessa, Washington, USA (Fig. 6a). The Odessa BRS’s are 50–500 m
Author's personal copy
Table 1
Summary of characteristics of terrestrial and Martian MRRDs.
Type of MRRD
Hydrovolcanic
Rootless cones (RC)
Material, context
Size, spatial density
Characteristics
Phreatovolcanism (explosive
lava–water interaction)
Lava;
RC: 3–320 m basal dia.
Do not source lava flows, but spatter may coalesce
to form short rheomorphic flows. Icelandic rootles
cones (‘pseudocraters’) do not always have summit
depressions, may be nested. Hawai’ian littoral
cones on channel fed lavas are bisected.
RC: where water accumulates in
near surface (e.g., shallow lake
basins, river channels)
LC: at ocean margin
Lava; form where water has
produced phreatovolcanism,
followed by erosion (in this case, by
catastrophic flooding)
LC: p300 m
Littoral cones (LC)
Basaltic ring structures (BRSs)
Phreatovolcanism (+inflation?),
then erosion
May have interior mesas and partial concentric
interior rings. Might have exterior moats.
Diapirism due to density inversion
( ¼ overpressurization of fine
sediments)
Sediments; commonly occur in
compressional tectonic regimes,
also occur on large river deltas
pseveral hundred m high.
Can have large range of heights. Morphology varies
from domes with summit craters to flat ‘pies.’ May
show mudflows, sedimentary ejecta.
ptens of km basal dia
density varies widely
Ice-related
Pingos
Groundwater pressure+freezing,
then collapse
Sediments (few reported hydraulic
pingos in bedrock)b
p50 m high
p1000 m basal dia.
Hydrostatic
Hydraulic
Show continuous range of morphologies.
Hydrostatic: If formed within drained thaw
lake basins, are often associated with thermal
contraction polygons.
Hydraulic: If formed on slopes, may align
Ice-block forms (IBFs)
Deposition of ice-blocks with
various amounts of sediment, then
melting
Sediments
p100/km2
porder 10 m high
Glaciofluvial: Boulder rings/rimmed kettle
holes are found preferentially on high points
(e.g., bars) in the channel.
Glaciolacustrine: Ice floe grounding pits form
in water depths o5–500 m.
Glacial: CMFs may align with slope or with
direction of glacial movement.
porder 100 m basal dia.
p70/km2
a
b
Hodges (1978).
Burr et al. (2008).
ARTICLE IN PRESS
Odessa: pfew/km2
Tokio: pfew tens/km2
Odessa field
Tokio field
Sedimentary
Mud volcanoes
50–500 m basal diaa
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
Mechanism
583
Author's personal copy
ARTICLE IN PRESS
584
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
Fig. 6. (a) USGS Digital Orthophotograph in the Channeled Scabland, near Odessa,
Washington, USA, centered near 47.4N 188.7W. The two BRS shown in Fig. 5b are
indicated by black arrows (lower left); other BRS as denoted by Hodges (1978) are
indicated by white arrows. Image credit: NASA World Wind. (b) Oblique air photo
of two BRSs showing concentric rings. Two-lane road across middle of image
provides scale. Image credit: Hideaki Miyamoto.
in diameter and have discontinuous concentric ridges (Fig. 6b).
These concentric ridges are comprised largely of autointrusive
dikes, which differ from the surrounding lava flow in the plunge
direction of their lava columns (McKee and Stradling, 1970;
Hodges, 1978; Jaeger et al., 2003a). These ridges may surround
either mesas or depressions. Gravity data over three of the Odessa
BRSs has been reported as showing a weak negative anomaly of
1 mgal (Parks and Banami, 1971). The implications of this finding
are somewhat ambiguous, because it does not have a uniquely
identifiable cause; however, it does suggest that some of the
material comprising the BRSs is less dense than the surrounding
material.
Tokio BRSs. A second population of BRSs is found near Tokio,
Washington, USA. In comparison with the Odessa BRSs, the Tokio
BRSs are smaller (40–150 m in diameter) and do not have visible
interior concentric ridges (dikes) or mesas (McKee and Stradling,
1970; Hodges, 1978). The interiors of the Tokio BRSs are flatfloored and partially infilled by loess, possibly obscuring any
interior concentric dikes.
2.2.2. BRSs: formation mechanism
BRSs were originally hypothesized as purely lava structures,
formed by the foundering of a solid crust into a still molten
interior, with subsequent lava extrusion into the cracks in the
crust from the interior (McKee and Stradling, 1970). Subsequent
investigations found palagonite within three BRS centers (Hodges,
Fig. 7. ‘Water escape structures’ (four vertical white arrows) of different sizes in
Bacon Butte, Snake River Plains basalts. Pillow basalt (horizontal white arrow) is
visible at the base of the flow. Person at lower right (horizontal black arrow) gives
sense of scale.
1978). Narrowly jointed basalt with high glass content was
also found within a hypothesized crypto-BRS, which is
unexcavated in plan view but visible in cross-section in a cliff
face (McKee and Stradling, 1970; Jaeger et al., 2003b, 2005).
The palagonite and jointing morphology both suggest rapid
quenching of the lava most likely by water at the interior of the
BRSs. Thus, BRSs likely reflect phreatovolcanism (Hodges, 1978;
Jaeger et al., 2003a, b). The basaltic dikes within Odessa BRSs may
be a result of inflation of the surrounding lava flow causing tensile
stress and concentric cracking of the crust. Molten lava from the
flow interior would have then filled in the cracks, forming
the observed dikes. Structural data (trends and plunges) of the
basaltic columns show that the Odessa BRSs formed within
the lava flow 15 Ma (Jaeger et al., 2005). Subsequently, the
BRSs were likely exposed through plucking by macroturbulent
kolks (vertical vortices in turbulent water) during the Channeled
Scabland flooding 15 ka (Baker, 1978). In contrast to common
erosion by sediment abrasion, plucking excavates whole
columns of basalt. Plucking by kolks during the Channeled
Scabland flooding may have acted preferentially on the more
vertically jointed country rock; in contrast, the more narrowly or
horizontally jointed ring dike material may have been less
susceptible.
Author's personal copy
ARTICLE IN PRESS
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
According to the model above, shallow groundwater is
required for the initial formation of BRSs within lava flows, and
some form of extensive erosion (e.g., plucking by catstrophic
floods) is required for their planview exposure. Although an order
of magnitude smaller in plan view diameter, ‘water escape
structures’ in the Snake River Plains (SRP) lavas (Fig. 7) may form
by a similar phreatovolcanic process. These structures are
subvertical conduits on the order of 1 m in diameter, spaced
5–15 m apart, and partially filled with basaltic rubble. The upper
surface of the lava flow above the largest of these conduits forms a
depression 10 m in diameter and 1.5 m deep and the columnar
jointing changes plunge from vertical along the base of the flow to
horizontal adjacent to the conduits (Shervais et al., 2005). The
host SRP lava in which these structures are found fills paleodrainage channels, and water interaction is indicated by the glassy
texture, underlying pillows (Fig. 7), and hyaloclastites (Shervais
et al., 2005). However, basaltic dike structures such as are found in
Odessa BRSs are not evident, perhaps either because they are not
present within the SRP lavas or because they have not been
exposed by plucking during catastrophic flooding. These ‘water
escape structures’ in the Snake River Plains may represent the
smallest end-member of a continuum that includes the larger
Tokio structures, which do not have dikes, and the largest
structures found near Odessa, which do have basalt dikes.
2.2.3. BRSs: hypothesized locations on mars
On Mars, BRSs have been hypothesized to explain MRRDs in
Athabasca Valles (Jaeger et al., 2003a, b, 2005). Athabasca Valles is
a young catastrophic outflow channel hypothesized to have
experienced multiple water floods interleaved with or subsequently modified by lava (Burr et al., 2002; Werner et al., 2003;
Lanagan, 2004; Jaeger et al., 2007). Thus, the channel has the
elements necessary for formation of BRSs, namely, an initial water
flood to provide saturated sediments, subsequent lava flows to
produce phreatovolcanism, and finally another water flood
causing prolonged erosion to expose that phreatovolcanism in
the form of BRSs. However, subsequent interpretation with higher
resolution data infers these forms to be more similar to rootless
cones formed in deflating, not inflating, lavas (Jaeger et al., 2007).
2.3. Summary: hydrovolcanic MRRDs
Hydrovolcanic MRRDs include rootless cones and basaltic ring
structures. These MRRDs are found on lava plains or, in the case of
Hawaiian littoral cones, in locations where the flows have entered
an ocean. Rootless cones typically show summit craters, although
decreasing explosivity during eruption or post-eruption erosion
may fill them in to create simple mounds or subsequent erosion
may form convex-up slopes. Simple rootless cones require more
than a single episode of explosivity. A large number of eruptions
that vary in strength and/or location may produce nested rootless
cones. Neither rootless cones nor BRSs are sources of lava flows,
but rootless cones may have associated rheomorphic flows. BRSs
may contain (parts of) concentric interior rings and central mesas,
or just central depressions. Both rootless cones and BRSs require
surface or subsurface water for their genesis, and BRSs require
extensive flood erosion for their exposure to observation.
3. Sedimentary MRRDs
Sedimentary MRRDs are formed by the movement and/or
emplacement of sediment. They result from confined upward
transport of subsurface material that is deposited in a ring or
mound on the surface. Sedimentary MRRDs include both mud
585
Fig. 8. (a) Image from Google Earth, showing three mud volcanoes in Azerbaijan
(boxes). Airantekyan and Kotturdag mud volcanoes are labeled. (b) Mud flow from
the Airantekyan mud volcano, vent to the upper right. (c) Typically circular vent of
the Kotturdag mud volcano, located along the same structural ridge as Airentekyan
(images b and c are panchromatic images, res. 0.6 m/px, illumination from the
north).
volcanoes and sand diapirs. However, because sand diapirs do not
form in fields but rather in isolation, and because they are poorly
understood, we do not address them here. Guhman and Pederson
(1992) provide a discussion of mechanism and Mahaney et al.
(2004) address the possibility of sand diapirs on Mars.
3.1. Mud volcanoes (MVs)
3.1.1. MVs: morphology and observations
Mud volcanoes (or sedimentary volcanoes) vary greatly both in
morphology and in size (Fig. 8). Morphologically, they are
generally circular or oval in plan form (Fig. 8c), but can also have
elongate ‘mud ridge’ morphology or irregular shapes. In the third
dimension, mud volcano morphology varies not only with the
amount of material extruded, but with the degree of material
fluidity (Brown, 1990, see also, e.g., Cifci and Ergun, 2004). Muds
with low porosity and/or high cohesion give rise to cones or
domes with summit calderas. As their name suggests, mud
volcanoes can appear similar to magmatic volcanoes, and these
conical mud volcanoes often exhibit flows of sediment, extending
from their summits (Fig. 8b and c) or flanks (Kopf, 2002; see also
the cover pages of Geology, v. 32, no. 7, GSA Today v. 17, no. 2.)
They can erupt explosively (e.g., White, 1955), resulting in fields of
boulders with diameters on the order of meters. Mud extrusions
can also give rise to ‘mud pies,’ flat (slope angle o51) features
composed of pooled mud with higher porosity or lower cohesion
(e.g., Henry et al., 1990; Lance et al., 1998; see Kopf, 2002, Fig. 3).
Mud volcanoes can have negative relief; such basins or ‘pockmarks’ are attributed to subsidence due to high fluid discharge
through a conduit. The mud volcano field in the Gulf of Cadiz off
the Iberian coast provides examples of pockmarks (Baraza and
Ercilla, 1996), where submarine fluid venting has been sighted
(Van Rooij et al., 2005). A model of Lusi, a mud volcano in East Java
which may have resulted from drilling, suggests that surficial sag
zones develop as a result of subsurface normal faulting around the
conduit (Davies et al., 2007). Multiple morphologies are often
present within a single local region.
Author's personal copy
ARTICLE IN PRESS
586
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
Fig. 9. Margin of the Lokbatan flow, Azerbaijan. Image credit: Ronnie Gallagher (BP
Exploration, Angola).
Subaerial mud volcanoes range in size from meters to tens of
kilometers in diameter and from sub-meter to hundreds of meters
in height (Kholodov, 2002; Kopf, 2002 and references therein).
Submarine mud volcanoes are more commonly found on active
than passive margins, and appear to have an even wider range of
morphologies than subaerial mud volcanoes, with submarine
‘mud pie’ diameters ranging up to greater than 30 km (Kopf et al.,
2001). A morphometric (shape plus size) classification of mud
volcanoes is proposed by Kholodov (2002), namely: (1) diapir
structures with a single vent, indicating a significant feeder
channel; (2) mud volcanoes with multiple edifices or vents
located on a cone-shaped accumulation of mud breccia; (3) large
(tens of kilometers in diameter) ‘swamped’ areas located at the
same elevation as the surrounding topography and having
scattered breccia mounds and liquid mud pools; and (4) mud
accumulation in depressed areas, which result from tectonic
subsidence.
Collectively, mud volcanoes can be associated with other
positive relief features, such as anticlinal ridges several kilometers
in length, which are ascribed to faulting (Jakubov et al., 1971;
Kopf, 2002). This scenario applies to the extensive mud volcanism
in Azerbaijan (Jakubov et al., 1971, see also Skinner and Tanaka,
2007). Conversely, mud volcano fields can sit in structural
troughs, which host thick sedimentary deposits conducive to
diapirism; for example, on the floor of the Black Sea, mud
volcanoes sit within the Sorokin Trough, which is pressurized on
either side by buried structural highs (Woodside et al., 1997;
Bohmann et al., 2002).
The sedimentary deposits forming mud volcanoes can be
highly viscous (Fig. 9). They are generally comprised of clasts and
rock fragments in a clay mineral-rich matrix. The clasts are often
incorporated from the conduit wall rock, whereas the homogeneous muds are thought to be related to the final phase of an
eruptive cycle, when loose wall rock has already been removed.
This ratio of clasts and rock fragments to fine-grained matrix
ranges from clast-supported debris to clast-free muds (Kopf, 2002
and references therein).
3.1.2. MVs: formation mechanisms
Mud volcanoes result from density inversions or overpressurization of fine-grained sediments, i.e., sediments with greater
porosity than that expected from their overlying lithostatic load
(Kopf, 2002). This overpressurization, or buoyancy, can be caused
by various processes (see Wangen, 2001 for a quantitative review).
The most common process is disequilibrium compaction, which
results from the inability of low permeability sediments to expel
fluids. Another process, fluidization, occurs due to generation or
incorporation of gases or other fluids; the gas is commonly
methane, although carbon dioxide and nitrogen have also been
produced (Dimitrov, 2002; Etiope et al., 2004). Consequently, the
sediments undergo extrusion, either in the form of diapirs or
diatremes. Diapirism is driven by density contrasts between an
overpressured mud and a denser overburden, and is commonly
used to describe a slow migration (Kopf, 2002). Diatremes are
pipe-conduits that result from fluidization of a sedimentary mass
due to passage of pore fluids, and may be more forceful. Though
very similar in appearance and often co-located—indeed, diatremes may originate from diapirs—diapirs and diatremes may be
distinguished by the mechanical stress states of the resultant mud
intrusion (see Brown (1990) and references therein). Following
the primary or predisposing cause of overpressurization, secondary or triggering causes for mud volcanism include removal of
overlying material, earthquakes, or other modifications of the
stress field.
As summarized by Kopf (2002), terrestrial mud volcanoes are
most commonly found in convergent tectonic settings, where
fluid-rich sediments accumulate in deep-sea trenches, or accretionary wedges, and are subducted. The high pressure and
temperature in the ‘subduction factory’ increase the buoyancy of
the fluid-rich sediments. As a result, the overpressurized sediments ascend through the lithologic overburden, extruding at the
surface both on- and off-shore. The fluids may arise from several
kilometers’ depth. Mud volcanoes are also found on some of the
Earth’s great river deltas, such as the Orinoco River delta (Brown
and Westbrook, 1988), the Mississippi River delta (Morgan et al.,
1968), and the Copper River basin in Alaska (Nichols and Yehle,
1961). In such settings, the overpressurization is attributed to an
increase in the weight of deposited material over time, such that
less dense, higher porosity material is capped by heavier, less
porous sediments. In these settings, the fluids are more likely to
originate from shallow source regions. More than 1000 subaerial
and subaqueous mud volcanoes have been found on Earth; see
Kholodov (2002), Fig. 1, or Kopf (2002), Fig. 1, for plots of global
distribution.
3.1.3. MVs: hypothesized locations on Mars
Mud volcano-type processes have been suggested for Mars
(e.g., Komar, 1991; Bradak and Kereszturi, 2002), and mud
volcanoes have been proposed for numerous locations in the
northern plains, including in Chryse and Acidalia Planitiae
outwash plains (Tanaka, 1997), on the north polar deposits
(Tanaka, 2005), and in southern Utopia Planitia (Tanaka et al.,
2003a; Skinner and Tanaka, 2007). Dewatering of debris flow
material from Simud/Tiu Valles has been hypothesized to have
produced ‘mud eruptions’ that formed low, p5-km-diameter,
pancake-like shields (Tanaka, 1999). Mud volcanoes have also
been hypothesized for various smooth-topped mounds or ‘bulges’
in impact craters (Ori and Baliva, 1999), for various pits and
sinkhole-like structures (Ori, 1999), and for curvilinear ridges in
Isidis Planitia (Davis and Tanaka, 1995).
Terrestrial mud volcanoes are highly variable in appearance
and size, spanning the greatest size range of the MRRDs
discussed here. For use in identifying mud volcanoes on Mars,
Author's personal copy
ARTICLE IN PRESS
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
587
morphometric information (Kopf et al., 2001; Kopf, 2002;
Kholodov, 2002) provides some guidelines, namely, that those
with the greater diameters tend to be flatter or even depressed in
elevation. Mud volcanoes can also be identified by association
with anticlinal ridges and other evidence for compressional
tectonics. Although plate tectonics is limited or absent on Mars,
large volcanic edifices and wrinkle ridges suggests compressional
tectonic regimes (e.g., Hall et al., 1986; Withers and Neumann,
2001). Terrestrial mud volcanoes are also found in sedimentary
settings, where variation in deposition led to density inversions
and consequent diapirism. Impact craters provide basins for
sediment accumulation on Mars, as evidenced by the fans or
fluvial deltas that are found therein (Malin and Edgett, 2003;
Pondrelli et al., 2005; Bhattacharya et al., 2005; Mangold and
Ansan, 2006), and may have experienced similar variation in
deposition.
On Earth, mud volcanoes are an important factor in the fluid
back flux from the lithosphere to the hydrosphere and support
active extremophile communities (Kopf, 2002 and references
therein). If mud volcanoes in compressional tectonic regimes or
impact basins on Mars similarly incorporate deep crustal water
and/or gas into their erupted material, they may provide
samples of associated biotic or prebiotic processes on Mars’
surface. Subsurface models for mud-volcanic processes, including
chemosynthetic biotic activity, are discussed by Ori et al. (2001).
4. Ice-derived MRRDs
Ice-derived MRRDs result from either the growth of massive
ground ice to form pingos or from the surficial deposition and
partial or complete burial of block ice.
4.1. Pingos
4.1.1. Pingos: morphology and observations
Resulting from the growth of massive ground ice, pingos are
the largest of a range of ice-cored mounds. Terrestrial pingos may
be several tens to hundreds of meters in basal diameter, and range
up to 50 m height (Mackay, 1998). This is an empirical height
limit that depends on the amount of available groundwater, the
amount of overburden, and other factors, and should not be
influenced by gravity (see discussion in Cabrol et al., 2000,
pp. 101–105; Burr et al., 2005, p. 61). In their initial and growth
state, pingos are simple mounds. If the tensional strength of the
overburden is exceeded by ice-core growth, dilational cracking
may occur. Rupture of the sub-pingo water lens or diffusion over
time also contributes to pingo collapse. The cause of the collapse
influences rim morphology, which ranges from irregular or jagged
to smooth and flat. Eventually, a pingo collapses entirely, leaving
behind a raised rim or rampart, up to several meters in height,
known as a pingo scar (Flemal, 1976; De Gans, 1988). (See Burr
et al., 2005, Fig. 5, for four-part image representing this sequence.)
Individual pingo genesis and collapse are not necessarily
synchronous. Thus, a range of pingo heights and morphologies,
from incipient pingos to pingo scars, may be observed in one
region at any one time (Mackay, 1979). Of the 1350 hydrostatic
pingos on the Tuktoyaktuk Peninsula, only 50 are currently
growing (Mackay, 1973). Pingo scar densities can reach 100/km2
(Flemal, 1976).
4.1.2. Pingos: formation mechanisms
Pingos result from the pressurization and then freezing of
groundwater (Washburn, 1980; French, 1996), although small
amounts of additional ice may derive from atmospheric sources
Fig. 10. Images of Ibyuk, a collapsing hydrostatic pingo, Tuktoyaktuk Peninsula,
Northwest Territories, Canada. (a) Nadir-looking air photo (July 1993), north is to
the top; surrounding terrain is a drained thaw lake basin; (b) ground photo (May
2004), looking roughly north; snowmobile tracks on the side of the pingo provide
scale; and (c) summit crater, showing the topographic relief. The far edge of a
frozen pond, produced by melting and refreezing of the ice core, is visible; coauthors at left provide scale.
(Mackay, 1985). They are genetically classified according to their
groundwater pressurization mechanism. See Burr et al. (2008) for
further discussion of pingo formation mechanisms.
4.1.2.1. Hydrostatic (closed system) pingos. The groundwater may
be pressurized through propagation of a subsurface freezing front,
as documented for many of the 1350 pingos located in thermokarst lakes on the Tuktoyaktuk Peninsula, Northwest Territories, Canada (Fig. 10a) (Mackay, 1979, 1998). The lakes occur in
deep, continuous permafrost. However, the heat capacity of the
lake water keeps the sediments immediately beneath the lake
unfrozen, and this unfrozen zone comprises a ‘talik’ (Mackay,
1979, 1998). When the lakes lose their water by drainage and/or
evaporation, the saturated lake bottom sediments are exposed to
Author's personal copy
ARTICLE IN PRESS
588
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
seasonal freezing temperatures. A freezing front propagates both
from the edges of the lake basin inward, reducing the size of the
talik, and from the lake basin floor downwards. Where residual
water forms (a) pond(s) at the low point(s) in the basin, permafrost growth is locally minimized. As pore water is expelled ahead
of the inward-propagating freezing front, it collects beneath the
site(s) of minimum permafrost thickness and freezes. Constrained
by continuous permafrost at depth, the pressurized water domes
up the overlying permafrost as it freezes, forming a pingo. The
doming is primarily due to subsurface water flow inward,
and only negligibly due to the 9% volume expansion during
phase change. The resulting ice-cored mounds are referred to as
hydrostatic pingos.
4.1.2.2. Hydraulic (open system) pingos. Alternatively, the water
pressurization mechanism may be hydraulic, supplied by topography (Müller, 1959) or by the ascent of deeply seated groundwater through local secondary faults (O’Brien, 1971). In either
case, groundwater is driven upward through the permafrost by
artesian pressure and domes up the overlying sediments.
As it approaches the surface, the groundwater freezes to create a
near-surface ice-core (Holmes et al., 1968; French, 1996). As with
hydrostatic pingos, this doming occurs at sites of least permafrost
development, which are commonly river valleys or terraces.
Hydraulic pingos occur in the very thin permafrost of the Yukon
Territory, Canada (Hughes, 1969), in river valleys in Alaska
(e.g., Holmes et al., 1968), and in near-coastal areas of Greenland
(e.g., Müller, 1959; Bennike, 1998). The most commonly studied
hydraulic pingos are found in Alaska and Greenland, although
they are more easily viewed in Greenland due to lack of vegetation
(see Bennike, 1998, Fig. 3A–F, for oblique aerial and ground photos
of Greenland hydraulic pingos). Some pingos present the appearance of fitting into neither of these mechanisms. Hence, a
third classification of pingos, called polygenetic pingos, has been
proposed (Gurney, 1998), although a mechanism was not clearly
defined.
4.1.3. Pingos: hypothesized locations on Mars
Pingos and collapsed pingos have been hypothesized to exist in
a variety of potentially periglacial and/or paleo-lacustrine environments on Mars. These include within Gusev crater (Cabrol et al.,
2000), on the Cerberus plains (Nussbaumer, 2005; Page and
Murray, 2006), on flood saturated sediments in the Athbasca
Valles outflow channel (Burr et al., 2005), within high-northernlatitude impact craters (Soare et al., 2005; Sakimoto, 2005a, b;
Bacastow and Sakimoto, 2006), and on the Amazonian-aged
hummocky member (AHvh) in the northern plains (Tanaka et al.,
2003b). The most recent studies of pingo-like forms on Mars have
centered on Utopia Planitia (Soare et al., 2005; Osinski and Soare,
2007; Dundas et al., 2008; de Pablo and Komatsu, 2008). See Burr
et al. (2008) for further discussion of hypothesized pingos on
Mars.
4.2. Ice-block forms (IBFs)
Ice-block MRRDs result from the deposition of block ice, partial
of complete burial of those blocks by synchronous or subsequent
sedimentation, and eventual melting of the blocks to produce
mesoscale depressions. These kettle hole depressions may have
raised rims under some circumstances.
4.2.1. IBFs: morphology and observations
In plan view, ice-block forms may be circular, oblong, or
irregularly shaped where multiple ice-blocks have been deposited
together. The existence, size, and depth of a central depression
depend on the amount of sediment in the block, as well as its
depth of burial. Laboratory work shows sediment-poor and/or
deeply buried ice blocks form normal kettle holes, depressions
with no rims or only thin rims (Maizels, 1977, 1992). Examples of
such normal kettle holes (see e.g., Chapman et al., 2003, Fig. 13b)
can range up to 2 km in diameter and a couple of tens of meters
deep (Flint, 1971). However, increasing sediment-load and/or
decreasing burial depth produces ‘crater kettles,’ with more
significant rims, or ‘till-fill’ kettles, which are sediment mounds
(Maizels, 1992). The amount of sediment and degree of burial are
influenced by the emplacement mechanism.
4.2.2. IBFs: formation mechanisms
Ice blocks may be emplaced through glaciofluvial, glaciolacustrine, and glacial/ablational processes (Parizek, 1969).
4.2.2.1. Glaciofluvial. The glaciofluvial type of ice-block MRRD
derives from breaking up of glacier margins by sub-, en-, or
supraglacial flooding and transport and deposition of the derived
ice blocks by the floodwaters in the proglacial environment.
Glaciofluvial ice-block production and transport is commonly
observed in glaciated regions such as Iceland (e.g., Russell and
Knudsen, 2002; Fay, 2002b), Greenland (Sugden et al., 1985;
Russell, 1993), and Norway (Longva and Thoresen, 1991).
The 1918 glacial flood over Myrdalsandur in Iceland (Tómasson, 1996) produced over 150 ‘rimmed’ kettle holes and mounds,
which were attributed to meltout of ice-blocks with various
amounts of sediment loads and burial depths (Maizels, 1992).
These ‘boulder ring structures’ are comprised of individual
boulders up to 1.6 m in diameter. They tend to be higher on their
downstream side, from which a sediment ‘tail’ (or obstacle
shadow) may extend. Boulder ring structures are found preferentially on elevated streamlined bar surfaces, where they grounded
first during waning flood flow. They are associated with
hummocky outwash plains, and fields of individual boulders and
boulder-sediment mounds (Maizels, 1992; cf. Fay, 2001, 2002a).
Icelandic examples have a maximum measured density of 70/km2.
Some of the boulder rings are contiguous (Maizels, 1992); others
may show distinct streamlining (obstacle shadow ridges) from
flow around the grounded ice-block during waning flow (cf. Fay,
2001, 2002a; see also Chapman et al., 2003, Fig. 13b). They are
described as looking like pingos (Maizels, 1977). However, their
relative topographic location on elevated bars or channel margins
is distinct from that of hydrostatic and hydraulic pingos, both of
which form in local lows.
4.2.2.2. Glaciolacustrine. Glaciolacustrine MRRD production results from calving of glaciers that terminate in lakes and subsequent deposition of these ice-blocks on the lake bottom during
lake drainage. If moving during deposition, the ice blocks produce
linear ‘scour marks’ on the lake bed, which have raised sides or
lateral berm piles due to plowing up of the lake bed sediment
(Eden and Eyles, 2002). If stationary during deposition,
the ice blocks form ‘ice floe pits.’ In their study of Kasei Valles and
Echus Chasma, Woodward-Lynas and Guigné (2003) imply
that these pits also have raised berms, although an explanation
for these raised berms was not given. In his study of ice-contact
rings, Parizek (1969, p. 54) implied that the raised berms
around circular or irregularly shaped pits can form where
lake sediments are deposited over stagnant ice. Glaciolauctrine
MRRDs have been identified on Earth in the bed of paleo-Glacial
Lakes Agassiz (Mollard, 1983) and Iroquois (Gilbert et al., 1992),
Author's personal copy
ARTICLE IN PRESS
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
589
although these attributions were somewhat speculative and not
exclusive.
Grizzaffi and Schultz, 1989). For MRRDs in those areas, CMFs may
be a viable hypothesis.
4.2.2.3. Glacial/ablational. Ice-block MRRDs may also result from
final deglaciation under cold-based conditions. Circular moraine
features (CMFs) are an ablational ice-block form that is hypothesized to result from ice-block deposition by stagnating cold-based
glaciers (Ebert and Kleman, 2004). The dead-ice blocks, with
varying amounts of till (sediment), form circular ‘dirt cones’
during melting. As observed on Himalayan glaciers, these ‘dirt
cones’ are conical ice blocks up to 85 m high covered with a layer
of glacial sediment (Swithinbank, 1950). As the ice cone melts, the
sediment slides down the slopes of the cones. The resultant circular features may be mound, sheet, or ring shaped in cross-section, depending on the amount and distribution of sediment
within the erstwhile ice block and the length of melt time. This
aspect of the hypothesis agrees with what Maizels (1992) found in
her boulder ring laboratory experiments.
CMFs are 20–170 m in diameter and 0.5–10 m high. Particularly
good examples of CMFs are found on the Varanger Peninsula,
northern Norway, where they are readily distinguished by having
a different color sediment than the underlying boulder cover
(Ebert and Kleman, 2004 and references therein). In some cases,
the CMF boulder size is larger than that of the surrounding
boulder cover; in other cases it is smaller. At least 50% of CMFs
have ‘tails’ of sediment that taper in the direction of paleo-ice
flow; the origin of these tails is not understood (Ebert and Kleman,
2004). The Varanger Peninsula CMFs are found on plateaus
associated with other glacial features, and may have formed in
association with a polythermal glacial cover. According to this
hypothesis, sediment was entrained in the warm-based part of the
glacier base and transported onto the plateaus; during deglaciation under cold-based conditions, this sediment was let down
onto the underlying surface, draping dead-ice blocks from the
glacier (Ebert and Kleman, 2004).
4.3. Summary: ice-derived MRRDs
4.2.3. IBFs: hypothesized locations on Mars
‘Kettle-like features’ in the circum-north-polar region have
been interpreted as delineating a former extent of the ice cap
(Fishbaugh and Head, 2000, 2001). Features in the circum-Chryse
outflow channel Maja Valles have been hypothesized to be kettle
holes and associated obstacle marks (Chapman et al., 2003).
Rimmed kettle holes (‘gravity craters or kettle craters’) could
explain ‘circular, elliptical, or peanut-shaped rings’ on the floor of
the Athabasca Valles outflow channel (Gaidos and Marion, 2003).
Hoffman and Tanaka (2002) proposed a variation on this theme:
that ‘‘the chains of low, pitted cones’’ in Athabasca Valles resulted
from deposits of volatile blocks from a debris flow, where the
volatile was either cold CO2 or warm H2O. After the cessation of
flooding and consequent drop of pressure, this volatile explodes,
mimicking a phreatic cone. In its appeal to the deposition of
CO2–ice blocks, the process is analogous to the formation of
H2O–ice kettle holes. In its scenario of exploding volatiles, the
process mimics phreatomagmatic rootless cones.
Glaciolacustrine ice-block forms have been hypothesized for
features in Kasei Valles and parts of Echus Chasma on Mars
(Woodward-Lynas and Guigné, 2003). This hypothesis applies to
the irregular to subcircular forms outlined by ridges, which are
attributed to ice floes.
To our knowledge, CMFs per se have not been proposed to
explain MRRDs on Mars. However, large stagnating ice blocks
have been hypothesized for various locales on Mars, including
cold-based glaciers on volcanoes (e.g., Neukum et al., 2004; Shean
et al., 2005), the meltback region of polar ice (e.g., Fishbaugh,
2004), and volatile-rich debris layers in impact basins (e.g.,
This class of MRRDs results either from formation of massive
ice-cored pingos through groundwater flow, or from ice-block
deposition in pro- or sub-glacial environments. Pingos may form
through either hydraulic or hydrostatic mechanisms (see Burr
et al. 2008 (this volume) for details). The ice-block MRRDs may be
aligned, either by water or glacial flow. Boulder rings tend to be
higher on their downstream side, and may have obstacle shadows.
Circular moraine features tend to have tails on the down-flow (not
necessarily the down slope) side.
5. Spatial analysis
Morphological differences are helpful in determining MRRD
origin, but significant morphological overlap among MRRD types
often precludes conclusive categorization. In order to strengthen
our ability to identify MRRDs by type, we have applied two
independent, non-morphological statistical methods for analyzing
MRRD spatial distribution. Before we discuss these results, we
describe the data.
5.1. MRRD data
Icelandic RC groups (RCGs) were previously analyzed by Bruno
et al. (2004). Where we repeated this dimensionless analysis
(discussed in Section 5.2 and summarized in Table 1), we
incorporated the values from the RCGs used in this previous
work, excluding the smallest and largest population groups.
Where we conducted new dimensional analysis (discussed in
Section 5.3 and summarized in Table 2), we re-analyzed the
same five Icelandic datasets. These data consist of air photos
at 1:37,000 scale on which the RCs were identified by an
experienced Icelandic geologist. (See Bruno et al., 2004 for more
information.)
Likewise, the pingo data were analyzed previously by Bruno
et al. (2006). They include hydrostatic pingos from the Tuktoyaktuk Peninsula, Northwest Territories, Canada, hydraulic pingos
from central Alaska, and low broad mounds of uncertain origin at
Prudhoe Bay, on the Alaskan North Slope. (See Bruno et al., 2006
for more information).
As discussed above, the BRSs fall into two groups located near
Tokio and Odessa, Washington, USA. Our data for the Tokio group
come from Landsat 7 imagery available via the NASA WorldWind
server-client software (worldwind.arc.nasa.gov), which was compared with oblique aerial and ground photographs to mark the
BRSs. The data for the Odessa group come from a mapping of BRSs
provided in the popular science book by Mueller and Mueller
(1997), some of which were also visited by some of the authors of
this work.
Kettle holes were analyzed from aerial photographs of the
Skeijarár outwash plains in southern Iceland photographs
produced in 1997 by the National Road Authority of Iceland. We
marked and analyzed the locations of kettle holes at two sites,
both affected by the 1996 jökulhlaup (glacier outburst flood) from
the Skeijarár glacier. One location was the ‘ice embayment’ outlet
at the eastern part of the glacier front, and the second location was
the Gı́gjukvı́sl outflow channel farther to the west where the
channel crosses Route One. These are locations that some of us
have previously visited after the 1996 jökulhlaup and for which
other workers produced maps of the 1996 jökulhlaup deposits,
Author's personal copy
ARTICLE IN PRESS
590
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
Table 2
Summary of spatial analysis of MRRDs that occur in fields.
Type of MRRD
Location
Rootless conesa
Iceland
Hnuta (North)
Hnuta (South)
Trollhamar
Stakafell
Raudholar
Pingo formsa
Hydrostatic
Tuktoyaktuk Pen., Canada
Western
Central
Eastern
Tanacross, Alaskaf
Prudhoe Bay, Alaskag
Steep
Broad (E)
Broad (W)
Odessab
Hydraulic
Unknown
Basaltic ring
structures
(BSRs)
Mud volcanoes
Glaciofluvial
kettle holesd
a
Impact craters
Secondaries
Pedestal
Tokioc
Absheron Pen., Azerbaijan
Iceland
Total in sample
set
|c|e
Average NN
distance (km)
R
357
301
131
192
198
0.016
0.013
0.051
0.020
0.054
0.83
0.74
0.59
0.66
0.94
6.15
8.52
8.66
8.78
1.56
282
665
213
152
1.6
1.8
2.0
2.6
0.80
0.85
0.83
0.51
6.28
7.31
4.53
11.15
376
112
287
143
1.3
0.93
0.97
0.7
0.70
0.78
0.78
0.76
10.94
4.14
7.02
5.26
28
196
0.1
3.4
1.12
0.67
1.00
8.59
0.81
0.58
9.89
32.51
Conclusion
Clustered
Clustered
Skeijarársandur
Gı́gjuvı́sl
Mars
Acidalia Planitia
E. Acidalia Planitia 1
728
1661
67
257
0.3
2.5
1.11
1.00
1.54
0.07
E. Acidalia Planitia 2
98
2.6
1.07
1.28
Clustered
Clustered
Inconclusive due to data
limitations (see footnotes)
Clustered
Clustered
Inconclusive
Not significant at the 0.05
level
Bolded results indicate significance at the 0.05 level.
a
Data from Bruno et al. (2006).
b
Data from Mueller and Mueller (2005), p. 114. Additional analysis strongly suggests that the presence of roads is a dominant factor in the apparent clustering. See text
for details.
c
Data derived from Landsat 7 imagery in comparison with oblique aerial photos. The low number of features in this population renders these results inconclusive. See
text for details.
d
Data derived from aerial photographs from the National Road Authority, Iceland, in comparison with Russell and Knudsen (2002), Fay (2002a, b), and ground
photographs.
e
The sign of c is dependent on the value of R. If R41.00, then c will be positive; if Ro1.00, then c will be negative.
f
Data from Holmes et al. (1968) Fig. 4 and Table 1.
g
Data from Walker et al. (1985).
including ice kettle deposits (Russell and Knudsen, 2002; Fay,
2002b).
The mud volcano data were from the Absheron Peninsula,
Azerbaijan. These data were derived from co-registering an
industry document detailing the geologic features of the area
with a Landsat 7 image. The locations of a subset of the known
features were then used to correct for a rotational shift and slight
offset. Only the subaerial mud volcanoes were mapped. The offshore vents were more widely spaced, which could be an effect of
(lower resolution) submarine mapping, so those mud volcanos
were not used in this work.
The secondary and pedestal impact crater datasets were those
from Acidalia Planitia previously derived and analyzed by Bruno
et al. (2004).
5.2. Dimensionless NN analysis
Our first approach was to apply a dimensionless ‘‘nearest
neighbor’’ (NN) analysis to the MRRDs discussed here. This
analysis, based on the NN analysis of Clark and Evans (1954),
had been previously applied to terrestrial pingos, terrestrial
rootless cones, terrestrial tumuli, and Martian impact craters
(see Bruno et al., 2004, 2006; Baloga et al., 2007). This approach
involves calculation of two statistics for the purpose of (1)
quantifying the spatial distribution of features within a field, and
(2) estimating the statistical confidence in this assessment. The
statistic R is the dimensionless ratio of the mean actual distance
between a feature and its nearest neighbor to the distance
expected in a spatially random distribution. Following conventional nearest-neighbor analysis (Clark and Evans, 1954), this
work assumed a Poisson probability distribution for its spatially
random distribution. This form of randomness would suggest that
there are no physical interactions during the formation of the
independent geologic features (i.e., MRRDs). Departures from
Poisson randomness would suggest that the formation of one
feature influenced the formation of other features, for example,
due to resource depletion (Baloga et al., 2007). To apply this
analysis technique, each MRRD is represented as a single point.
R1 is associated with a (Poisson) random distribution. R values
significantly less than or greater than 1 suggest that non-random
processes are controlling spatial distribution (Bruno et al., 2004;
Baloga et al., 2007). A confidence estimate, c, is calculated to
assess the significance of the departure from randomness; a value
of c outside the region defined by 1.5oco3.3 indicates a
significant departure from randomness at the 0.05 significance
level (Baloga et al., 2007; see Clark and Evans, 1954 for
mathematical derivations for R and c).
Rootless cones and collapsed pingos have similar R and c
values. Thus, skewness and kurtosis of the pairwise NN data were
Author's personal copy
ARTICLE IN PRESS
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
Fig. 11. Plot illustrating the effect of resolution on skewness and kurtosis.
examined as possible discriminators between these two MRRD
types (Bruno et al., 2006). Given that the images used in these
prior analyses have a range of resolutions (3–74 m/px), we reexamined those data to assess the influence of resolution. Six
selected feature fields were analyzed at varying degrees of image
resolution (degradation), with the R, c, skewness, and kurtosis
computed at each stage. Although R and c values varied with
resolution, in no case did that variation effect the final
interpretation (Bruno, 2006). Neither was these two statistics
significantly affected by the trimming technique. Trimming is a
standard procedure in exploration data analysis (e.g., Hoaglin et
al., 1983; Sheshkin, 1997) to delete outliers prior to analysis. The
procedure is designed to expedite the evaluation of the key
features of the distribution. We used a 5% asymmetric (high-end)
trim (cf., Bruno et al., 2006).
Whereas R and c values did not vary significantly with
resolution or trimming technique, the skewness and kurtosis
values for the images we used did vary with resolution. Higher
resolution data tended to give lower skewness and kurtosis values
than lower resolution data (Fig. 11), and this variation was
significant enough to affect the final interpretation. For data at
the same or similar resolutions, skewness and kurtosis may prove
useful for distinguishing among population types. However, the
data used here range over a factor of 20 in resolution. The
conditions under which resolution could affect skewness and
kurtosis are presently at issue and have not yet been fully
investigated. Therefore, this avenue of analysis is not discussed
further in this work.
Here we calculate R and c from the datasets discussed above
for terrestrial populations of BRSs, mud volcanoes, and glaciofluvial kettle holes. Our results are compiled with those from
previous publications and presented in Table 2. We note the
following:
The kettle holes data were collected at the ice embayment
outlet of the Skeijarár glacier, before the water spread out over
the more distal Skeijarár outwash plains. These kettle holes
may therefore be more closely spaced than the boulder ring
structures identified farther down the outwash plains.
Conversely, the ice blocks forming the boulder ring structures
were observed to have grounded preferentially where the flow
was stagnant and/or the water was shallow, such as on
streamlined bars. This tendency would also produce clustering.
Our second dataset of kettle holes from the Gı́gjukvı́sl channel
provides a check on this possible clustering influence and
resulted in a similar R value. Thus, the result that R is
591
significantly less than one for kettle holes (Table 2) is expected,
regardless of the location of the scene, and we do not attribute
it to a biased data sample.
The Odessa BRSs have a low R value with |c| 43.3 (Table 2).
However, this value may be an artifact of the roads running
through the population set, which may have influenced the
‘observation’ of these features. We developed and applied a
spatial analysis approach to determine the effect of linear
features on point features in order to assess the effect of the
roads on the BRS spatial distribution. Our specific question was
if the roads could be the cause of the apparent departure from
spatial randomness for that population. To perform the
evaluation, we compared the statistical distribution of actual
distances from each BRS to the nearest road with the distances
expected for a hypothetical Poisson-based distribution. The
results were distinctly dissimilar, with a log-normal result for
the distances of the actual population versus a uniform result
for the distances for a hypothetical randomly distributed
population. This distinct dissimilarity strongly suggests that
accessibility or other factors represented by the roads could be
dominating the non-random test results. Further ground
truthing or additional data (e.g., from aerial photographs) are
needed to provide definitive results.
A minimum of 30–40 data points was used in previous work, as
synthetic images with less than this number produced
inconsistent results (Bruno et al., 2004). Thus, the Tokio BRS
(n ¼ 28) results, though presented for completeness, are
considered inconclusive.
In summary, our results for rootless cones, pingo forms, mud
volcanoes, and glaciofluvial kettle holes were all similar. Thus, the
spatial analysis statistics R and c alone cannot distinguish among
these groups. In comparison, secondary and pedestal impact
crater populations do not show clustering (Table 2; Bruno et al.,
2006). Thus, this technique can differentiate (endogenic) surficial
MRRDs from (exogenic) impact MRRDs. The results for the BRSs
were not conclusive due to limited data (for the Tokio field) and
separate statistical analysis indicating a dominant correspondence between the BRSs and the roads (for the Odessa group). We
have not used skewness and kurtosis in this work due to the effect
of resolution on the results.
5.3. Dimensional NN analysis
Given this somewhat ambiguous result, we investigated
additional means to distinguish among MRRDs types. Noting that
some MRRD types exhibit distinctive scales in both feature size
and distance between neighboring features, we therefore applied
a dimensional NN analysis to our data. The following discussion,
therefore, presents an analysis of differences in characteristic
nearest neighbor distances as determined for each MRRD type.
To assess the significance of these differences, we have used
the standard statistical inference techniques commonly referred
to as analysis of variance, or ANOVA. There are a variety of
hypothesis tests that can be used within ANOVA to establish the
statistical uniqueness of mean values when several (more than
two) samples are being compared (Sheshkin, 1997). These
hypothesis tests result in the estimation of confidence intervals
around each mean value, where the reduced degrees of freedom
resulting from comparison of multiple samples are accounted for.
The confidence intervals are then used to infer whether or not the
individual samples are statistically unique (i.e., samples with
confidence intervals that overlap are indistinguishable, whereas
those that do not overlap are different). Each of the individual
hypothesis tests result in slight differences in the magnitude of
Author's personal copy
ARTICLE IN PRESS
592
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
Fig. 12. (a) Geometric mean NN distances by individual MRRD population, and
corresponding 95% confidence intervals. Colored boxes indicate MRRD type. (b)
Geometric mean NN distances by MRRD type. However, the pingo MRRDs are
given by population, as each pingo population exhibits different NN distances,
possibly corresponding to difference formation mechanisms (see text). From left to
right, MRRD types shown in order of increasing geometric mean NN distances.
Fig. 13. Plot showing (a) an example of a raw distribution, and (b) its
corresponding transformed distribution.
the associated confidence intervals. For this study we have chosen
the Bonferroni–Dunn (Sheshkin, 1997) intervals, which generally
lie somewhere between some of the other more extreme choices.
In this study, our samples sizes are large enough that the choice of
interval is inconsequential. For both Fig. 12a and b, we have used
95% confidence intervals. In other words, there is a 95%
probability that these intervals overlap with the true mean value
for each sample.
Fig. 12a is a plot of the geometric mean NN distance and
corresponding confidence intervals for each group of MRRDs in
Table 2. In some cases the confidence intervals are so narrow they
appear as a single black horizontal line through the center of the
data point. Each group in Table 2 is represented by at least one
example. Colored boxes indicate specific types of MRRD, as
indicated in the legend on the figure.
Fig. 12b is a plot of the geometric mean nearest neighbor
distances as determined by averaging distances for all features in
each MRRD type. For example, the RCGs contain NN distances for
Raudholar, Hnuta (north and south), Trollhamer, and Stakafell. The
only exceptions are the groups of pingos. As discussed above, the
three pingo locations exhibit different formation mechanisms
—hydrostatic, hydraulic, and an uncertain origin. Because these
data yield statistically different NN distances (i.e., confidence
intervals do not overlap), we have plotted them as separate
populations on Fig. 12b.
One of the primary assumptions of all the ANOVA hypothesis
tests is that data in each of the individual samples come from a
Guassian (Normal) probability distribution. Like many other types
of geologic data, the nearest neighbor distances for each of the
samples in Table 2 are very strongly skewed. Fortunately, this type
of distribution can easily be transformed by taking the natural log
of each NN distance (e.g., Glaze et al., 2002, 2005; Bulmer et al.,
2005; Martin et al., 2007). The resulting transformed distribution
is, in most cases, very nearly normal, and the ANOVA techniques
can be applied. Fig. 13 provides an example of this transformation
applied to data from the rootless cone population at Raudholar,
Iceland.
In both Fig. 12a and b, all data have been transformed by taking
the natural log before performing the Bonferroni–Dunn hypothesis test. Afterward, the mean values and associated confidence
limits are transformed back to the original units (km) by taking
the exponential. As a note, the mean values shown in both figures
are, in fact, the mean value of the transformed data (mean of the
normal distributions). When transformed back by taking the
exponential, the resulting value is effectively the geometric
mean value. This value differs slightly from the arithmetic mean
(Fig. 13a), but better represents the peak value in a distribution
that is asymmetric.
Fig. 12a shows that, according to the Bonferroni–Dunn test,
most of the individual samples within each MRRD type are virtually
Author's personal copy
ARTICLE IN PRESS
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
indistinguishable from each other. This provides confidence that
those MRRD types with similar geometric mean NN distances—
namely kettleholes, rootless cones, and Adidalia Planitia pedestal
craters—, the NN statistical measure is robust. Pingos and basaltic
ring structures are exceptions to this. In the case of the pingos,
there are statistically three distinct groups: (1) Tuktoyaktuk (West
and Central), (2) Tuktoyaktuk East and Holmes, and (3) Prudhoe
Bay. Other than the fact that the Tuktoyaktuk and Holmes pingo
NN distances appear similar, these groupings are generally
consistent with the formation process distinctions of hydrostatic,
hydraulic, and unkown, respectively. The BRS results may be due
to influence on the observations by the roads and the low number
of Tokio BRS, as discussed above. Only a single population was
available for mud volcanoes and for secondary impact craters.
593
Odessa Crater on their property. Robin Raddi of the Inuvialuit
Nation guided us during field research in the Tuktoyaktuk
Peninsula, NWT, Canada, through permits granted by the Aurora
Research Institute (Inuvik), Parks Canada (western Canada office),
the Inuvialuit Land Administration (Tuktoyaktuk), and the
Tuktoyaktuk Hunters and Trappers Association (Tuktoyaktuk).
Árni Snorrason, head of the Hydrological Service, National Energy
Authority of Iceland, provided logistical support and scientific
information during field work in Iceland. Travel to Iceland for
DMB was sponsored by the Fulbright Foundation. Financial
support for BRS field work was provided through the USGS by
Interagency Agreement W-10,265 issued through the NASA
Planetary Geology and Geophysics Program. Funding for manuscript preparation was provided to DMB through the Mars Data
Analysis Program Grant NNG05GP31G and Mars Fundamental
Research Program Grant NNG06GH12G.
6. Summary and implications
References
Equifinality is a similarity in final morphology that results from
different preceding processes (Chorley et al., 1984; Beven, 1996).
For this reason, features are best assessed through genetic
hypotheses formulated as multiple working hypotheses, which
can be a greater advantage in planetary science than better
resolution data (Zimbelman, 2001). Our aim in this paper has been
to provide information to aid in developing possible hypotheses
for MRRDs, for testing among them, and for understanding their
implications. Various morphologic characteristics combined with
statistically different geometric mean NN distances provide
means for distinguishing among some MRRDs by origin. At the
same time, morphological similarity and overlapping spatial
distribution measures among other MRRD populations reinforce
the importance of using multiple working hypotheses (Gilbert
1886; Chamberlin 1897) to determine the origin of any give field
of Martian MRRDs.
Table 1 provides a summary of some distinguishing morphological characteristics of terrestrial MRRDs, discernable at the
resolution of high-resolution Mars Orbiter Camera (Malin et al.,
2004) or HiRISE (McEwen et al., 2007) imagery. These provide
means to evaluate the origin(s) of observed MRRDs on Mars. Table 1
also summarizes physical characteristics that would be shared by
disparate types of MRRDs. For example, flows may be associated
with rootless cones (i.e., rheomorphic flows), mud volcanoes (i.e.,
mud flows), and pingos (i.e., sedimentary slumping).
Table 2 provides the results of our nearest neighbor analyses.
The results show that dimensionless NN statistics are similar for
the surficial processes discussed here, although they can be used
to distinguish between endogenic and exogenic MRRDs. The
statistical measures of R and c are not biased by image resolution
(Bruno, 2006) so the results represented here should be applicable
to data of varying resolutions. In contrast to dimensionless NN
statistics, dimensional NN statistics could be helpful in distinguishing certain classes of MRRDs. Dimensional statistics alone
would provide ambiguous results for distinguishing rootless cones
from kettle holes or pedestal craters from hydraulic pingos.
However, our terrestrial datasets yield order of magnitude
differences in the (dimensional) spacing among (1) rootless cones
and kettle holes, (2) secondary craters and basaltic ring structures,
and (3) pingos, pedestal craters and mud volcanoes. Thus, using
the NN dimensional statistic could be quite useful for distinguishing among these origins for MRRD populations.
Acknowledgements
Many persons assisted in research that contributed to this
manuscript. Michael and Terry Connelly allowed us to investigate
Allen, C.C., 1980. Volcano–Ice Interactions on the Earth and Mars. In: Advances in
Planetary Geology, NASA TM-81979, pp. 161–264.
Araújo, J.S., 1982. Estruturas, Circulares de Água Vermelha. Dissertac-ão de
Mestrado, IG-USP, S.Paulo, p. 79.
Bacastow, A.L., Sakimoto, S.E.H., 2006. Martian north polar crater morphology:
implications for an aquifer. Lunar and Planetary Science Conference XXXVII
(abstract 2239).
Baker, V.R., 1978. Paleohydraulics and Hydrodynamics of Sabland Floods. In: Baker,
V.R., Nummedal, D. (Eds.), The Channeled Scabland. National Areonautics and
Space Administration, Washington D.C., pp. 59–79.
Baloga, S.M., Glaze, L.S., Bruno, B.C., 2007. Nearest-neighbor analysis of small
features on Mars: applications to tumuli and rootless cones. Journal of
Geophysical Research 112, E3,E03002.
Baraza, J., Ercilla, G., 1996. Gas-charged sediments and large pockmark-like
features on the Gulf of Cadiz slope (SW Spain). Marine and Petroleum Geology
13 (2), 253–261.
Bennike, O., 1998. Pingos at Nioghalvfjerdsfjorden, eastern North Greenland.
Geology of Greenland Survey Bulletin 180, 159–162.
Beven, K., 1996. Equifinality and uncertainty in geomorphological modeling. In:
Rhoads, B.L., Thorn, C.E. (Eds.), The Scientific Nature of Geomorphology,
Proceedings of the 27th Binghamton Geomorphology Symposium. Wiley, New
York, pp. 289–313.
Bhattacharya, J.P., Payenberg, T.H.D., Lang, S.C., Bourke, M., 2005. Dynamic river
channels suggest a long-lived Noachian crater lake on Mars. Geophysical
Research Letters 32 (10), L10201.
Bohmann, G., Schenck, S., cruise participants, 2002. RV METEOR Cruise Report
M52/1 MARGASCH, GEOMAR Report no. 108, p. 191.
Bradak, B., Kereszturi, A., 2002. Application of mud volcanism for rocky and icy
planetary bodies. 34th COSPAR Scientific Assembly, The Second World Space
Congress, Houston, TX, USA.
Bridges, J.C., Seabrook, A.M., Rothery, D.A., Kim, J.R., Pillinger, C.T., Sims, M.R.,
Golombek, M.P., Duxbury, T., Head, J.W., Haldemann, A.F.C., Mitchell, K.L.,
Muller, J.P., Lewis, S.R., Moncrieff, C., Wright, I.P., Grady, M.M., Morley, J.G.,
2003. Selection of the landing site in Isidis Planitia of Mars probe Beagle 2.
Journal of Geophysical Research 108.
Brown, K.M., 1990. The nature and hydrogeologic significance of mud diapirs and
diatremes for accretionary systems. Journal of Geophysical Research 95 (B6),
8969–8982.
Brown, K., Westbrook, G.K., 1988. Mud diapirism and subcretion in the Barbados
ridge accretionary complex: the role of fluids in accretionary processes.
Tectonics 7 (3), 613–640.
Bulmer, M.H., Glaze, L.S., Shockey, K.M. Anderson, S., 2005. Using grain size
distributions to improve understanding of emplacement mechanisms. J
Geophys Res, 110, B01201, doi:10.1029/2003JB002841.
Bruno, B.C., 2006. A closer look at the nearest neighbor technique to study
geospatial distributions: effect of image resolution. Eos Transactions AGU 87
(52) (Fall Meet. Suppl., Abstract P23C-0080).
Bruno, B.C., Fagents, S.A., Thordarson, T., Baloga, S.M., Pilger, E., 2004. Clustering
within rootless cone groups on Iceland and Mars: effect of nonrandom
processes. Journal of Geophysical Research 109, E07009.
Bruno, B.C., Fagents, S.A., Hamilton, C.W., Burr, D.M., Baloga, S.M., 2006.
Identification of volcanic rootless cones and ice mounds on Earth and Mars:
using spatial distribution as a remote sensing tool. Journal of Geophysical
Research 111 (E6), E06017.
Burr, D.M., Grier, J.A., McEwen, A.S., Keszthelyi, L.P., 2002. Repeated aqueous
flooding from the Cerberus Fossae: evidence for very recently extant, deep
groundwater on Mars. Icarus 159 (1), 53–73.
Burr, D.M., Soare, R.J., Wan Bun Tseung, J-M., 2005. Young (late Amazonian), nearsurface, ground ice near the equator, Athabasca Valles, Mars. Icarus 178 (1), 56–73.
Burr, D.M., Tanaka, K.L., Yoshikawa, K., 2008. Pingos on Earth and Mars. Planet.
Space Sci. doi:10.1016/j.pss.2008.11.003.
Author's personal copy
ARTICLE IN PRESS
594
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
Cabrol, N.A., Grin, E.A., Pollard, W.H., 2000. Possible frost mounds in an ancient
Martian lake bed. Icarus 145, 91–107.
Chamberlin, T.C., 1897. The method of multiple working hypotheses. J. Geol. 5,
837–848.
Chapman, M.G., Gudmundsson, M.T., Russell, A.J., Hare, T.M., 2003. Possible
Juventae Chasma sub-ice volcanic eruptions and Maja Valles ice outburst
floods, Mars: Implications of MGS crater densities, geomorphology, and
topography. Journal of Geophysical Research 108 (E10), 5113.
Chorley, R.J., Schumm, S.A., Sugden, D.E., 1984. Geomorphology. Methuen and Co.
Ltd., London.
Cifci, G., Ergun, M., 2004. Image characteristics of the Mediterranean Sea
and the Black Sea mud volcanoes and sea bed features. Second Balkan
Geophysical Congress and Exhibition, Balkan Geophysical Society, Bucharest,
Romania.
Clark, P.J., Evans, F.C., 1954. Distance to nearest neighbor as a measure of spatial
relationships in populations. Ecology 35, 445–453.
Crumpler, L.S., 2003a. Spring deposits on Mars: physical processes from terrestrial
analogs. Lunar and Planetary XXXIV (Abstract No. 2002, Lunar and Planetary
Institute, Houston).
Crumpler, L.S., 2003b. Physical characteristics, geologic setting, and possible
formation processes of spring deposits on Mars based on terrestrial analogs.
Sixth International Conference on Mars, Abstract No. 3228.
Davies, R.J., Swarbrick, R.E., Evans, R.J., Huuse, M., 2007. Birth of a mud volcano:
East Java, 29 May 2006. GSA Today 17 (2).
Davis, P.A., Tanaka, K.L., 1995. Curvilinear ridges in Isidis Planitia, Mars—the result
of mud volcanism? Lunar and Planetary XXXVI, 321–322 (Lunar and Planetary
Institute, Houston).
De Gans, W., 1988. Pingo scars and their identification. In: Clark, M.J. (Ed.),
Advances in Periglacial Geomorphology. Wiley, New York, pp. 299–322.
de Pablo, M.A., Komatsu, G., 2008. Possible pingo fields in the Utopia Basin, Mars:
geological and climatic implications. Icarus.
Dimitrov, L.I., 2002. Mud volcanoes—the most important pathway for degassing
deeply buried sediments. Earth Science Reviews 59, 49–76.
Dundas, C.M., Mellon, M.T., McEwen, A.S., Lefort, A., Keszthelyi, L.P., Thomas, N., the
HiRISE Team, 2008. HiRISE observations of fractured mounds: possible Martian
pingos. Geophysical Research Letters.
Ebert, K., Kleman, J., 2004. Circular moraine features on the Varanger Peninsula,
northern Norway, and their possible relation to polythermal ice sheet
coverage. Geomorphology 62, 159–168.
Eden, D.J., Eyles, N., 2002. Case study of a relict iceberg scour exposed at
Scarborough Bluffs, Toronto, Ontario: implications for pipeline engineering.
Canadian Geotechnical Journal 39 (3), 519–534.
Etiope, G., Feyzullayev, A., Baciu, C.L., Milkov, A.V., 2004. Methane emission from
mud volcanoes in eastern Azerbaijan. Geology 32 (6), 465–468.
Farmer, J., 1998. Thermophiles, early biosphere evolution, and the origin of life on
Earth: Implications for the exobiological exploration of Mars. Journal of
Geophysical Research 103 (E12), 28457–28461.
Fagents, S.A., Thordarson, Th., 2007. Rootless volcanic cones in Iceland and on
Mars. In: Chapman, M.G. (Ed.), The Geology of Mars: Evidence from EarthBased Analogs. Cambridge University Press, Cambridge, pp. 151–177.
Fagents, S.A., Lanagan, P.D., Greeley, R., 2002. Rootless cones on Mars: a
consequence of lava-ground ice interaction. In: Smellie, J.L., Chapman, M.G.
(Eds.), Volcano–Ice Interaction on Earth and Mars, vol. 202. Special Publications, Geological Society, London, pp. 295–317.
Farrand, W.H., Gaddis, L.R., Keszthelyi, L.P., 2005. Pitted cones and domes on Mars:
observations in Acidalia Planitia and Cydonia Mensae using MOC, THEMIS, and
TES data. Journal of Geophysical Research 110.
Fay, H., 2001. The role of ice blocks in the creation of distinctive proglacial
landscapes during and following the November 1996 glacier-outburst flood
jökulhlaup, Skeidararsandur, southern Iceland. Eos Transactions Fall Meet.
Suppl., Abstract IP52A-0755.
Fay, H., 2002a. Formation of kettle holes following a glacial outburst flood
(jökulhlaup), Skeijarársandur, southern Iceland. In: Snorrason, Á., Finnsdóttir,
H., Moss, M., (Eds.), The Extremes of the Extremes: Extraordinary Floods
(Proceedings of a symposium held at Reykjavik, Iceland, July 2000). International Association of Hydrological Sciences Publication Number 271, pp.
205–210.
Fay, H., 2002b. Formation of ice-block obstacle marks during the November 1996
glacier-outburst flood (jökulhlaup), Skeijarársandur, southern Iceland. In:
Special Publication Number 32 of the International Association of Sedimentologists, pp. 85–97.
Fishbaugh, K.E., 2004. Geologic history of the north polar region of Mars from the
Late Hesperian to the present: stratigraphy, melting and retreat. Ph.D. Thesis.
Brown University, p. 209.
Fishbaugh, K.E., Head III, J.W., 2000. North polar region of Mars: topography of
circumpolar deposits from Mars Orbiter Laser Altimeter (MOLA) data and
evidence for asymmetric retreat of the polar cap. Journal of Geophysical
Research 105 (E9), 22455–22486.
Fishbaugh, K., Head, J.W., 2001. Characterization of features associated with
volatile removal in the martian north circumpolar region. Lunar and Planetary
XXXII (Abstract No. 1426, Lunar and Planetary Institute, Houston).
Fisher, R.V., Schmincke, H-U., 1984. Pyroclastic Rocks. Springer, Berlin.
Flemal, R.C., 1976. Pingos and pingo scars: their characteristics, distribution, and
utility in reconstructing former permafrost environments. Quaternary Research 6, 37–53.
Flint, R.F., 1971. Glacial and Quaternary Geology. Wiley, New York, 892 pp.
Frey, H., Jarosewich, M., 1982. Subkilometer Martian volcanoes: properties and
possible terretrial analogs. Journal of Geophysical Research 87, 9867–9879.
Frey, H., Lowry, B.L., Chase, S.A., 1979. Pseudocraters on Mars. Journal of
Geophysical Research 84, 8075–8086.
French, H.M., 1996. The periglacial environment, 2nd ed. Longman, London, 341 pp.
Gaidos, E., Marion, G., 2003. Geological and geochemical legacy of a cold early
Mars. Journal of Geophysical Research 108 (E6), 5055.
Gilbert, G.K., 1886. Inculcation of the scientific method. Am. J. Sci. 31, 284–299.
Gilbert, R.J., Handford, K.J., Shaw, J., 1992. Ice scours in the sediments of glacial Lake
Iroquois, Prince Edward County, eastern Ontario. Géographie Physique et
Quaternaire 46, 189–194.
Glaze, L.S., Stofan, E.R., Smrekar, S.E., Baloga, S.M., 2002. Insights into corona
formation through statistical analyses. J. Geophys Res, 107, E12, doi:10.1029/
2002JE001904.
Glaze, L.S., Anderson, S.W., Stofan, E.R., Baloga, S.M., Smrekar, S.E., 2005. Statistical
distribution of tumuli on pahoehoe flow surfaces: Analysis of examples in
Hawaii and Iceland and potential applications to lava flows on Mars. J Geophys
Res, 110, B08202, doi:10.1029/2004JB003564.
Greeley, R., Theilig, E., 1978. Small volcanic constructs in the Chryse Planitia region
of Mars, NASA TM-79729, p. 202.
Greeley, R., Fagents, S.A., 2001. Icelandic pseudocraters as analogs to some volcanic
cones on Mars. Journal of Geophysical Research 106, 20527–20546.
Grizzaffi, P., Schultz, P.H., 1989. Isidis basin: site of ancient volatile-rich debris
layers. Icarus 77, 358–381.
Guhman, A.I., Pederson, D.T., 1992. Boiling sand springs, Dismal River, Nebraska;
agents for formation of vertical cylindrical structures and geomorphic change.
Geology 20 (1), 8–10 (doi:10.1130/0091-7613(1992)020/0008:BSSDRNS2.3.
CO;2).
Gurney, S.D., 1998. Aspects of the genesis and geomorphology of pingos: perennial
permafrost mounds. Progress in Physical Geography 22 (3), 307–324.
Hall, J.L., Solomon, S.C., Head, J.W., 1986. Elysium Region, Mars: tests of lithospheric
loading models for the formation of tectonic features. Journal of Geophysical
Research 91 (B11), 11,377–11,392.
Henry, P., Le Pichon, X., Lallemant, S., Foucher, J.P., Westbrook, G., Hobart, M., 1990.
Mud volcano field seaward of the Barbado accretionary complex: a deep-towed
side-scan sonar survey. Journal of Geophysical Research 95, 8917–8929.
Hoaglin, D.C., Mosteller, F., Tukey, J.W., 1983. Understanding robust and exploratory
data analysis. Wiley, New York, p. 447.
Hodges, C.A., 1978. Basaltic ring structures of the Columbia Plateau. Geological
Society of America Bulletin 89 (9), 1281–1289 (doi: 10.1130/00167606(1978)89/1281:BRSOTCS2.0.CO;2).
Hoffman, N., Tanaka, K., 2002. Co-existing ‘‘flood’’ and ‘‘volcanic’’ morphologies in
Elysium as evidence for cold CO2 or warm H2O outbursts. Lunar and Planetary
Science XXXIII (abstract 1505).
Holmes, G.W., Hopkins, D.M., Foster, H.L., 1968. Pingos in central Alaska. US
Department of Interior Geological Survey. Bulletin 1241-H, p. 40.
Hughes, O.L., 1969. Distribution of open system pingos in central Yukon Territory
with respect to glacial limits. Paper 69-34. Ottawa: Geological Survey of
Canada.
Jaeger, W.L., Keszthelyi, L.P., Burr, D.M., McEwen, A.S., Baker, V.R., Miyamoto, H.,
Beyer, R.A., 2003a. Ring dike structures in the Channeled Scabland as analogs
for circular features in Athabasca Valles, Mars. Lunar and Planetary Science
XXXIV (abstract 2045).
Jaeger, W.L., Burr, D.M., Keszthelyi, L.P., Emery, J., McEwen, A.S., Baker, V.R.,
Miyamoto, H., Beyer, R.A., 2003b. Basaltic ring structures in the Channeled
Scabland reexamined: implications for Mars. Geological Society of America
Annual Meeting, Seattle, WA, November. Abstract 53-7.
Jaeger, W.L., Keszthelyi, L.P., Burr, D.M., Emery, J.P., Baker, V.R., McEwen, A.S.,
Miyamoto, H., 2005. Basaltic ring structures as an analog for ring features in
Athabasca Valles, Mars. Lunar and Planetary XXXVI (Abstract No. 1886, Lunar
and Planetary Institute, Houston).
Jaeger, W.L., Keszthelyi, L.P., McEwen, A.S., Dundas, C.M., Russell, P.S., 2007.
Athabasca Valles, Mars: a lava-draped channel system. Science 317, 1709–1711.
Jakubov, A.A., Ali-Zade, A.A., Zeinalov, M.M., 1971. Mud Volcanoes of the Azerbaijan
SSR. Academy of Science of Azerbaijan, Baku, Azerbaijan. p. 257.
Jurado-Chichay, Z., Rowland, S.K., Walker, G.P.L., 1996a. The formation of circular
littoral cones from tube-fed pahoehoe: Mauna Loa, Hawaii. Bulletin of
Volcanology 57, 471–482.
Jurado-Chichay, Z., Urrutia-Fucugauchi, J., Rowland, S.K., 1996b. A paleomagnetic
study of the Pohue Bay flow and its associated cones, Mauna Loa volcano,
Hawaii: constraints on their origin and temporal relationships. Physics of the
Earth and Planetary Interiors 97, 269–277.
Keszthelyi, L.P., 2005. Some simple models for rootless cone formation on Mars.
Lunar and Planetary XXXVI (Abstract No. 1914, Lunar and Planetary Institute,
Houston).
Kholodov, V.N., 2002. Mud volcanoes, their distribution regularities and genesis:
communication 1. Mud volcanic provinces and morphology of mud volcanoes.
Lithology and Mineral Resources 37 (3), 197–209 (Translated from Litologiya I
Poleznye Iskopaemye, No. 3, (2002), p. 227–241).
Komar, P.D., 1991. Mud volcanoes on Mars? In: Reports of Planetary Geology and
Geophysics Program 1990. NASA, Washington, D.C., pp. 539–541.
Kopf, A.J., 2002. Significance of mud volcanism. Reviews of Geophysics 40 (2),
1005.
Kopf, A., Klaeschen, D., Mascle, J., 2001. Extreme efficiency of mud volcanism in
dewatering accretionary prisms. Earth and Planetary Science Letters 189,
295–313.
Author's personal copy
ARTICLE IN PRESS
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
Lanagan, P.D., 2004. Geologic history of the Cerberus Plains, Mars (Ph.D. thesis).
University of Arizona, Tuscon, p. 142.
Lanagan, P.D., McEwen, A.S., Keszthelyi, L.P., Thordarson, T., 2001. Rootless cones on
Mars indicating the presence of shallow equatorial ground ice in recent times.
Geophysical Research Letters 28 (12), 2365–2368.
Lance, S., Henry, P., Le Pichon, X., lallemant, S., Chamley, H., Rostek, F., Faugeres, J.C., Gonthier, E., Olu, K., 1998. Submersible study of mud volcanoes seaward of
the Barbados accretionary wedge: sedimentology, structure, and rheology.
Marine Geology 145, 255–292.
Longva, O., Thoresen, M.K., 1991. Iceberg scours, iceberg gravity craters and current
erosion marks from a gigantic Preboreal flood in southeastern Norway. Boreas
20, 47–62.
Lucchitta, B.K., 1978. Geologic map of the Ismenius Lacus quadrangle, Mars, scale
1:5,000,000, US Geol. Surv. Misc. Invests. Map I-1065.
Macdonald, G.A., 1972. Volcanoes. Prentice Hall, Englewood Cliffs, New Jersey, 510
pp.
Macdonalt, G.A., Abbot, A.T., Peterson, F.L., 1983. Volcanoes in the Sea: the Geology
of Hawaii, 2nd ed. Univeristy Press of Hawaii, Honolulu, 517 pp.
Mackay, J.R., 1973. The growth of pingos, western arctic coast, Canada. Canadian
Journal of Earth Sciences 10 (6), 979–1004.
Mackay, J.R., 1979. Pingos of the Tuktoyaktuk Peninsula area, Northwest Territories.
Géographie physique et Quaternaire 23 (1), 3–61.
Mackay, J.R., 1985. Pingo ice of the western Arctic coast, Canada. Canadian Journal
of Earth Sciences 22, 1452–1464.
Mackay, J.R., 1998. Pingo growth and collapse, Tuktoyaktuk Peninsula area, western
arctic coast, Canada: a long-term field study. Géographie physique et
Quaternaire 52 (3), 271–323.
Mahaney, W.C., Milner, M.W., Netoff, D.I., Malloch, D., Dohm, J.M., Baker, V.R.,
Miyamoto, H., Hare, T.M., Komatsu, G., 2004. Ancient wet aeolian environments on Earth: clues to presence of fossil/live microorganisms on Mars. Icarus
171, 39–53.
Maizels, J., 1977. Experiments on the origin of kettle holes. Journal of Glaciology 18,
291–303.
Maizels, J., 1992. Boulder ring structures produced during jökulhlaup flows: origin
and hydraulic significance. Geografiska Annalar 74A (1), 21–33.
Malin, M.C., Edgett, K.S., 2003. Evidence for persistent flow and aqueous
sedimentation on early Mars. Science 302, 1931–1934.
Malin, M., Edgett, K., Cantor, B., Caplinger, M., Davis, S., Jensen, E., Lipkaman, L.,
Nixon, B., Posiolova, L., Sandoval, J., Supulver, K., Williams, R., Zimbar, R., 2004.
Mars global surveyor Mars orbiter camera in the extended mission: the MOC
toolkit. Lunar and Planetary Science XXXV (Abstract No. 1189, Lunar and
Planetary Institute, Houston).
Mangold, N., Ansan, V., 2006. Detailed study of an hydrological system of valleys, a
delta, and lakes in the Southwest Thaumasia region, Mars. Icarus 180, 75–87.
Martin, P., Stofan, E.R., Glaze, L.S., Smrekar, S., 2007. Coronae of Parga Chasma,
Venus. J Geophys Res, 112, E04S03, doi:10.1029/2006JE002758.
Mattox, T.N., Mangan, M.T., 1997. Littoral hydrovolcanic explosions: a case study of
lava-seawater interaction at Kilauea Volcano. Journal of Volcanology and
Geothermal Research 75, 1–17.
McEwen, A.S., Preblich, B.S., Turtle, E.P., Artemieva, N.A., Golombek, M.P., Hurst, M.,
Kirk, R.L., Burr, D.M., Christensen, P.R., 2005. The rayed crater Zunil and
interpretations of small impact craters on Mars. Icarus 176, 351–381.
McEwen, A.S., Eliason, E.M., Bergstrom, J.W., Bridges, N.T., Hansen, C.J., Delamere,
W.A., Grant, J.A., Gulick, V.C., Herkenhoff, K.E., Keszthelyi, L., Kirk, R.L., Mellon,
M.T., Squyres, S.W., Thomas, N., Weitz, C.M., 2007. Mars Reconnaissance
Orbiter’s High Resolution Imaging Science Experiment (HiRISE). Journal of
Geophysical Research 112.
McKee, B., Stradling, D., 1970. The sag flowout: a newly described volcanic
structure. Geological Society of America Bulletin 81, 2035–2044.
Mollard, J.D., 1983. The origin of reticulate and orbicular patterns on the floor of
the Lake Agassiz basin. In: Teller, J.T., Clayton, L., (Eds.), Glacial Lake Agassiz.
Geological Association of Canada Special Paper 26, pp. 356–374.
Morgan, J.P., Coleman, J.M., Gagliano, S.M., 1968. Mudlumps: diapiric structures in
Mississippi Delta sediments. In: Braunstein, J., O’Brien, G.D. (Eds.), Diapirism
and Diapirs. AAPG Memoir 8. Tulsa, OK, USA, pp. 145–161.
Morrissey, M.M., Thordarson, T., 1991. Origins and occurrences of pseudocrater
fields in S. Iceland. Eos, Transactions of the American Geophyscial Union 72,
556.
Morrissey, M., Zimanowski, B., Wohletz, K., Buettner, R., 2000. Phreatomagmatic
fragmentation. In: Sigurdsson, H., Houghton, B., McNutt, S.R., Rymer, H., Stix, J.
(Eds.), Encyclopedia of Volcanoes. Academic Press, San Diego, pp. 431–445.
Mouginis-Mark, P.J., 1985. Volcano/Ground Ice Interactions in Elysium Planitia,
Mars. Icarus 64, 265–284.
Mueller, M., Mueller, T., 1997. Fire, Faults, and Floods: a Road and Trail Guide
Exploring the Origins of the Columbia River Basin. University of Idaho Press,
Moscow, Idaho, p. 288.
Müller, F., 1959. Observations on pingos. Meddelelser om Grønland 153 (3), 127
(Translated by D.A. Sinclair, National Research Council of Canada, Technical
Translation 1073).
Neukum, G., Jaumann, R., Hoffmann, H., Hauber, E., Head, J.W., Basilevsky, A.T.,
Ivanov, B.A., Werner, S.C., van Gasselt, S., Murray, J.B., McCord, T., The HRSC CoInvestigator Team, 2004. Recent and episodic volcanic and glacial activity on
Mars revealed by the high resolution stereo camera. Nature 432, 971–979.
Nichols, D.R., Yehle, L.A., 1961. Mud volcanoes in the Copper River Basin, Alaska. In:
Geology of the Arctic, vol. II. University of Toronto Press, Toronto, Ontario,
Canada, pp. 1063–1087.
595
Nussbaumer, J., 2005. Extent and further characteristics of former glaciated terrain
in Elysium Planitia, Mars. Lunar and Planetary Science XXXVI (Abstract No.
1949, Lunar and Planetary Institute, Houston).
O’Brien, R., 1971. Observations on pingos and permafrost hydrology in Schuchert
Dal, N.E. Greenland. Meddelelser om Grønland 195 (1), 1–20.
Ori, G.G., 1999. Gas-related (methane?) features on the Martian surface: evidence
and identification of biogenic and abiogenic subsurface reservoirs. Eos
Transactions Fall Meet. Suppl. 80 (Abstract B42B-12).
Ori, G.G., Baliva, A., 1999. Large bulges at the center of impact craters on Mars.
Lunar and Planetary Science XXX (Abstract No. 1758, Lunar and Planetary
Institute, Houston).
Ori, G.G., Komatsu, G., Ormo, J., Marinangeli, L., 2001. Subsurface models for the
formation of mound-like morphologies on Mars. Lunar and Planetary Science
XXXII (Abstract No. 1539, Lunar and Planetary Institute, Houston).
Osinski, G.R., Soare, R.J., 2007. Circular structures in Utopia Planitia, Mars: impact
V. Periglacial origin and implications for assessing ground-ice content. Lunar
and Planetary Science Conferrence XXXVIII CD-ROM (abstract 1609).
Page, D.P., Murray, J.B., 2006. Stratigraphical and morphological evidence for pingo
genesis in the Cerberus plains. Icarus 183 (1), 46–54.
Parizek, R.R., 1969. Glacial ice-contact rings and ridges. Geological Society of
America Special Paper 123, pp. 49–102.
Parks, H., Banami, A.M., 1971. Gravity survey of Columbia Plateau ring dikes. Eos
Transactions 52 (5), 433.
Pondrelli, M., Baliva, A., Di Lorenzo, S., Marinangeli, L., Rossi, A.P., 2005. Complex
evolution of paleolacustrine systems on Mars: an example from the Holden
crater. Journal of Geophysical Research 110 (E4), E04016.
Russell, A.J., 1993. Obstacle marks produced by flow around stranded ice blocks
during a glacier outburst flood (jökulhlaup) in west Greenland. Sedimentology
40, 1091–1111.
Russell, A.J., Knudsen, Ó., 2002. The effects of glacier-outburst flood flow dynamics
on ice-contact deposits: November 1996 jökulhlaup, Skeijarársandur, Iceland.
In: Special Publication Number 32 of the International Association of
Sedimentologists, pp. 67–83.
Sakimoto, S.E.H., 2005a. Central mounds in Martian impact craters: assessment as
possible perennial permafrost mounds (pingos). Lunar and Planetary Science
XXXVI (Abstract No. 2009, Lunar and Planetary Institute, Houston).
Sakimoto, S.E.H., 2005b. Martian polar craters: possible polar materials apparent
effects and post-impact modification by apparent perennial permafrost mound
(pingo). Role of Volatiles and Atmospheres on Martian Impact Craters (abstract
3036).
Shean, D.E., Head, J.W., Marchant, D.R., 2005. Origin and evolution of a cold-based
tropical mountain glacier on Mars: the Pavonis Mons fan-shaped deposit.
Journal of Geophysical Research 110 (E5), E05001.
Shervais, J.W., Kauffman, J.D., Gillerman, V.S., Othberg, K.L., Vetter, S.K., Hobson,
V.R., Zarnetske, M., Cooke, M.F., Matthews, S.H., Hanan, B.B., 2005. Basaltic
volcanism of the central and western Snake River Plain: a guide to field
relations between Twin Falls and Mountain Home, Idaho. In: Pederson, J.,
Dehler, C.M., (Eds.), Interior Western United States: Geological Society of
America Field Guide 6, pp. 27–52, doi:10.1130/2005.fld006(02).
Sheshkin, D.J., 1997. Handbook of Parametric and Nonparametric Statistical
Procedures. CRC Press, Boca Raton, FL, p. 719.
Skinner Jr., J.A., Tanaka, K.L., 2007. Evidence for and implications of sedimentary
diapirism and mud volcanism in the southern Utopia highland–lowland
boundary plain, Mars. Icarus 186 (1), 41–59.
Sugden, D.E., Clapperton, C.M., Knight, P.G., 1985. A jökulhlaup near Søndre
Strømfjord, west Greenland, and some effects on the ice-sheed margin. Journal
of Glaciology 31 (109), 366–368.
Soare, R.J., Burr, D.M., Wan Bun Tseung, J.-M., 2005. Possible pingos and a
periglacial landscape in northwest Utopia Planitia. Icarus 174, 373–382.
Swithinbank, C., 1950. The origin of dirt cones on glaciers. J. Glaciol. 1, 461–465.
Tanaka, K.L., 1997. Sedimentary history and mass flow structures of Chryse
and Acidalia Planitae, Mars. Journal of Geophysical Research 102 (E2),
4131–4149.
Tanaka, K.L., 1999. Debris-flow origin for the Simud/Tiu deposit on Mars. Journal of
Geophysical Research 104 (E4), 8637–8652.
Tanaka, K.L., 2005. Geology and insolation-driven climatic history of Amazonian
north polar materials on Mars. Nature 437, 991–994.
Tanaka, K.L., Carr, M.H., Skinner, J.A., Gilmore, M.S., Hare, T.M., 2003a. Geology of
the MER 2003 ‘‘Elysium’’ candidate landing site in southeastern Utopia
Planitia, Mars. Journal of Geophysical Research 108 (E12), 8079.
Tanaka, K.L., Skinner, J.A., Hare, T.M., Wenker, A., 2003b. Resurfacing history of the
northern plains of Mars based on geologic mapping of Mars global surveyor
data. Journal of Geophysical Research 108 (E4), 8043.
Thorarinsson, S., 1953. The crater groups in Iceland. Bulletin of Volcanology 14,
3–44.
Thordarson, Th., 2000. Rootless eruptions and cone groups in Iceland: products of
authentic explosive water to magma interactions. In: Volcano–Ice Interactions
on Earth and Mars Conference. Reykjavı́k, p. 48.
Tómasson, H., 1996. The jökulhlaup from Katla in 1918. Annals of Glaciology 22,
249–254.
Van Rooij, D., Depreiter, D., Bouimetarhan, I., De Boever, E., De Rycher, K., Foubert,
A., Huvenne, V., Réveillaud, J., Staelens, P., Vercruysse, J., Versteeg, W., Henriet,
J-P., 2005. A first sighting of active fluid venting in the Gulf of Cadiz. Eos
Transactions, American Geophysical Union 86 (49), 509–511.
Walker, D., Walker, M., Everett, K., Webber, P., 1985. Pingos of the 1293 Prudhoe Bay
Region, Alaska. Arctic Alpine Research 17, 321–336.
Author's personal copy
ARTICLE IN PRESS
596
D.M. Burr et al. / Planetary and Space Science 57 (2009) 579–596
Wangen, M., 2001. A quantitative comparison of some mechanisms generating
overpressure in sedimentary basins. Tectonophysics 334, 211–234.
Washburn, A.L., 1980. Geocryology. Wiley, New York, p. 406.
Werner, S.C., van Gasselt, S., Neukum, G., 2003. Continual geologic activity in
Athabasca Valles, Mars. Journal of Geophysical Research 108 (E12) (ROV 22-1).
White, D.E., 1955. Violent mud volcano eruption of Lake City hot springs, northeast
California. Geological Society of America Bulletin 66, 1109–1130.
Withers, P., Neumann, G.A., 2001. Enigmatic northern plains of Mars. Nature 410,
651.
Wohletz, K.H., Sheridan, M.F., 1983. Hydrovolcanic explosions II. Evolution of
basaltic tuff rings and tuff cones. American Journal of Science 283, 385–413.
Woodside, J.N., Ivanov, M.K., Limonov, A.F., 1997. Neotectonics and fluid flow
through seafloor sediments in the Eastern Mediterranean and Black Sea—Part I
and II, IOC Technical Series no. 48, UNESCO.
Woodward-Lynas, Guigné, 2003. Ice keel scour marks on Mars: evidence for
floating and grounding ice floes in Kasei Valles. Oceanography 16 (4), 90–97.
Zimbelman, J.R., 2001. Image resolution and evaluation of genetic hypotheses for
planetary landscapes. Geomorphology 37, 179–199.