Origin of the Medusae Fossae Formation, Mars:
Insights from a synoptic approach
Kathleen E. Mandt,1,2 Shanaka L. de Silva,3 James R. Zimbelman,4 and David A. Crown5
Received 29 January 2008; revised 17 June 2008; accepted 19 August 2008; published 20 December 2008.
[1] The geologic origin of the Medusae Fossae Formation (MFF) has remained a
mystery despite three decades of research. To better constrain its formation, an in-depth
analysis of observations made in the literature was combined with a new survey of over
700 Mars Orbiter Camera narrow-angle images of the MFF to identify morphologic
characteristics and material properties that define this formation as a whole. While
previous work has identified clear agreement on some characteristics, our analysis
identifies yardangs, collapse features, and layering as pervasive features of the MFF.
Whereas collapse features and layering may implicate several different physical and
chemical processes, yardangs provide vital information on material properties that
inform about mechanical properties of the MFF lithology. Aspect ratios of
megayardangs range from 3:1 to 50:1, and slope analyses reveal heights of up to 200 m
with cliffs that are almost vertical. Other yardangs show lower aspect ratios and
topographic profiles. These characteristics coupled to the presence of serrated margins,
suggest that MFF lithology must be of weakly to heavily indurated material that lends
itself to jointing. The characteristics and properties of the MFF are inconsistent with
those of terrestrial pyroclastic fall deposits or loess, but are in common with large
terrestrial ignimbrites, a hypothesis that explains all key observations with a single
mechanism. Yardang fields developed in regionally extensive ignimbrite sheets in the
central Andes display morphologic characteristics that correlate with degree of
induration of the host lithology and suggest an origin by pyroclastic flow for the MFF.
Citation: Mandt, K. E., S. L. de Silva, J. R. Zimbelman, and D. A. Crown (2008), Origin of the Medusae Fossae Formation, Mars:
Insights from a synoptic approach, J. Geophys. Res., 113, E12011, doi:10.1029/2008JE003076.
1. Introduction
[2] The Medusae Fossae Formation (MFF) is an enig-
matic deposit located along the equator of Mars that
stretches between 170 and 240
E (120?190
W) in the
Amazonis Planitia region and lies between the Tharsis and
Elysium volcanic centers (Figure 1). The MFF has been a
focus of nearly three decades of research because its age,
distribution, and material properties have important impli-
cations for Martian history. It is considered to be one of
the youngest deposits on Mars, largely Amazonian in age,
but deposition may have begun as early as the Hesperian
Period [Scott and Tanaka, 1986; Greeley and Guest, 1987;
Tanaka, 2000; Head and Kreslavsky, 2001, 2004]. How-
ever, despite general agreement about the relatively young
age and fine-grained nature of the MFF materials, its origin
is still debated and a variety of disparate hypotheses have
been proposed, including deposition by pyroclastic flows
[Scott and Tanaka, 1982, 1986; Tanaka, 2000] or falls
[Tanaka, 2000; Bradley et al., 2002; Hynek et al., 2003],
aeolian deposits [Greeley and Guest, 1987; Scott and
Tanaka, 1986; Head and Kreslavsky, 2004], polar deposits
[Schultz and Lutz, 1988], exhumed faulted rocks [Forsythe
and Zimbelman, 1988], carbonate platform [Parker, 1991],
rafted pumice [Mouginis-Mark, 1993], lacustrine deposits
[Malin and Edgett, 2000], and as deposits from meteorite
impact into a subsurface aquifer [Nussbaumer, 2005].
2. Evaluation
[3] To bring some clarity to this issue, we have compiled
and analyzed all the published information on the MFF in
search of an approach that may help advance the debate. We
note that some hypotheses have not been published in the
peer reviewed literature. Several hypotheses have been
eliminated on the basis of subsequent work (Table 1). For
instance, the demonstrably Amazonian age, if correct, of the
MFF [Scott and Tanaka, 1986; Greeley and Guest, 1987;
Tanaka, 2000; Head and Kreslavsky, 2001, 2004] is incon-
sistent with the paleopolar deposit, exhumed faulted rocks,
and lacustrine deposit hypotheses because each of these
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 113, E12011, doi:10.1029/2008JE003076, 2008
Click
Here
for
Full
Article
1Department of Space Studies, University of North Dakota, Grand
Forks, North Dakota, USA.
2Now at Space Science and Engineering Division, Southwest Research
Institute, San Antonio, Texas, USA.
3Department of Geosciences, Oregon State University, Corvallis,
Oregon, USA.
4Center for Earth and Planetary Sciences, National Air and Space
Museum, Smithsonian Institution, Washington, District of Columbia, USA.
5Planetary Science Institute, Tucson, Arizona, USA.
Copyright 2008 by the American Geophysical Union.
0148-0227/08/2008JE003076$09.00
E12011 1 of 15
requires the MFF to have a Noachian or Early Hesperian
formation age. Improved resolution provided by Mars
Global Surveyor (MGS) and topographic data from the
Mars Orbiter Laser Altimeter (MOLA) have ruled out the
carbonate platform hypothesis. The type of layering
expected in a carbonate platform is not seen in the MFF
[Bradley et al., 2002]. The rafted pumice deposit hypothesis
was contradicted by the same data because no wave-cut
cliffs were seen on Olympus Mons [Malin, 1999]. We find
that, of the original nine hypotheses, only three are still
Figure 1. Map of the Medusae Fossae Formation overlain on MOLA shaded relief 128 pixel/
 digital
elevation model (DEM) (shown in context with MOLA shaded relief centered around 180
E and 0
N).
MFF subdivisions of upper, middle, and lower are based on mapping by Scott and Tanaka [1986] and
Greeley and Guest [1987].
Table 1. Hypotheses for the MFF
Hypothesis Basis Source Ruled Out By
Ignimbrite/ash flow tuff Similarities to ignimbrites
in the Pancake Range of Nevada
Scott and Tanaka [1982]
Aeolian deposit location along dichotomy boundary Greeley and Guest [1987]
Paleopolar deposits similarities to current polar deposits Schultz and Lutz [1988] MFF is same age as current polar deposits,
and polar wander is limited on the basis
of fault patterns [Tanaka and Leonard, 1998]
Morphological differences in slopes and
valleys of MFF compared to current polar
deposits [Bradley et al., 2002]
Exhumed fault rocks as support for proof of dynamic
tectonic activity
Forsythe and Zimbelman [1988] MFF must be Noachian for this hypothesis
[Scott and Tanaka, 1986; Greeley and Guest, 1987]
Carbonate platform similar morphology to terrestrial
carbonate platforms
Parker [1991] topography too complex and layering not appropriate
[Sakimoto et al., 1999; Bradley et al., 2002] and
Thermal Emission Spectrometer (TES)
data [Chapman, 2002]
Rafted pumice deposits signs on Olympus Mons
of suboceanic volcanism
Mouginis-Mark [1993] layering not like a shoreline deposit
and no wave-cut cliffs in Olympus Mons
[Malin, 1999; Bradley et al., 2002]
Lacustrine deposits outcrops resembling
sedimentary rock
Malin and Edgett [2000] age must be Noachian for this hypothesis
[Scott and Tanaka, 1986;
Greeley and Guest, 1987]
Ashfall tuff draped over underlying terrain Tanaka [2000] Bradley et al. [2002]
Hynek et al. [2003]
Subsurface aquifer evidence of glacial activity,
ground moraines, and sublimation
Nussbaumer et al. [2005] supporting evidence cited is
contradictory
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
2 of 15
E12011
T
ab
le
2.
A
na
ly
si
s
of
L
it
er
at
ur
ea
C
at
eg
or
y
F
ea
tu
re
E
xt
ri
ns
ic
In
ti
ns
ic
P
ro
pe
rt
ie
s
P
ro
ce
ss
es
T
ra
ns
po
rt
at
io
n
M
ec
ha
ni
sm
s
E
ro
si
on
al
D
ep
os
it
io
na
l
A
eo
li
an
S
ub
te
ra
ni
an
S
ub
li
m
at
io
n
F
lu
vi
al
G
la
ci
al
F
in
e
G
ra
in
ed
Jo
in
te
d
In
du
ra
te
d
F
ri
ab
le
L
ar
ge
S
ca
le
V
ol
um
e
(l
ar
ge
vo
lu
m
e
of
m
at
er
ia
l)
3,
12
,
25
T
hi
ck
(u
p
to
3
km
in
so
m
e
pl
ac
es
)
8,
9,
10
,
25
,
28
In
cr
ea
se
d
re
li
ef
to
w
ar
d
T
ha
rs
is
11
T
hi
ck
ne
ss
tr
en
ds
in
cr
ea
se
to
w
ar
d
T
ha
rs
is
12
Y
ou
ng
ba
se
d
on
st
ra
ti
gr
ap
hi
c
re
la
tio
ns
2,
8,
9,
10
,
28
O
ut
ly
in
g
ya
rd
an
g
fi
el
ds
,
pe
de
st
al
cr
at
er
s
an
d
m
ou
nd
s
w
ith
in
cr
at
er
s
3,
9,
10
,
12
,
17
9,
12
,
17
10
16
3
C
om
pl
ex
st
ra
ti
gr
ap
hy
in
di
ca
ti
ng
m
ul
ti
pl
e
ep
is
od
es
of
de
po
si
ti
on
9
9
U
nd
er
ly
in
g
te
rr
ai
n
no
t
m
od
if
ie
d
12
,
16
,
23
,
29
12
16
16
In
te
rw
ea
vi
ng
of
la
va
an
d
M
F
F
de
po
si
ti
on
16
M
es
as
3,
30
30
S
te
ep
cl
if
fs
18
,
25
18
,
25
S
er
ra
te
d
sc
ar
ps
8,
26
8,
26
T
ea
rd
ro
p-
sh
ap
ed
hi
ll
s
se
pa
ra
te
d
by
la
rg
e
va
ll
ey
s
9,
10
F
is
h-
sc
al
e-
li
ke
kn
ob
s
an
d
cu
rv
il
in
ea
r
tr
ou
gh
s
19
19
R
ou
gh
,
ha
ck
ly
su
rf
ac
e
25
25
L
ac
k
of
bo
ul
de
rs
al
on
g
er
os
io
na
l
co
nt
ac
ts
12
,
18
12
,
18
12
,
18
W
id
e
ra
ng
e
of
ro
ug
hn
es
s
ch
ar
ac
te
ri
st
ic
s
7
7
7
Ir
re
gu
la
r
to
po
gr
ap
hy
27
27
S
m
oo
th
su
rf
ac
es
w
it
h
br
oa
dl
y
cu
rv
ed
m
ar
gi
ns
25
25
U
nc
on
fo
rm
ab
ly
ov
er
li
es
di
ch
ot
om
y
bo
un
da
ry
an
d
au
re
ol
e
de
po
si
ts
9,
10
B
ou
nd
ar
y
te
rr
ac
es
fl
at
,
no
t
co
nf
or
m
in
g
to
to
po
gr
ap
hy
25
L
in
ea
r
Y
ar
da
ng
s
2,
3,
4,
8,
12
,
18
,
23
,
25
,
26
,
27
,
28
,
30
2,
3,
4,
8,
12
,
18
,
23
,
25
,
26
,
27
,
28
,
30
20
18
,
27
18
,
24
,
25
3,
4,
23
,
26
,
27
,
30
B
id
ir
ec
ti
on
al
ya
rd
an
gs
4,
25
4,
25
4,
25
25
4
T
ro
ug
hs
9,
10
9,
10
10
R
id
ge
s
an
d
gr
oo
ve
s
8,
17
,
26
8,
17
,
26
C
hu
te
s
or
da
rk
sc
re
e
sl
op
es
12
12
12
12
12
F
lu
te
s
12
,
30
12
,
30
C
or
ru
ga
te
d
te
xt
ur
e
19
19
24
19
In
te
rs
ec
ti
ng
bl
oc
ky
la
ye
rs
4
3,
4
4
H
um
m
oc
ky
te
rr
ai
n
20
20
S
ur
fa
ce
cu
es
ta
s
an
d
un
du
la
ti
on
s
6
6
W
in
d
st
re
ak
s
30
30
V
is
co
us
fl
ow
fe
at
ur
es
19
,
21
21
19
R
ai
se
d
cu
rv
il
in
ea
r
fe
at
ur
es
5,
9,
10
5,
20
,
21
5
5,
9,
10
,
20
,
21
5,
20
,
21
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
3 of 15
E12011
T
ab
le
2.
(c
on
ti
nu
ed
)
C
at
eg
or
y
F
ea
tu
re
E
xt
ri
ns
ic
In
ti
ns
ic
P
ro
pe
rt
ie
s
P
ro
ce
ss
es
T
ra
ns
po
rt
at
io
n
M
ec
ha
ni
sm
s
E
ro
si
on
al
D
ep
os
it
io
na
l
A
eo
li
an
S
ub
te
ra
ni
an
S
ub
li
m
at
io
n
F
lu
vi
al
G
la
ci
al
F
in
e
G
ra
in
ed
Jo
in
te
d
In
du
ra
te
d
F
ri
ab
le
C
ir
cu
la
r
Im
pa
ct
cr
at
er
de
pt
h-
di
am
et
er
ra
ti
os
1
1
L
ar
ge
se
co
nd
ar
y
cr
at
er
s
24
E
xt
en
si
ve
se
co
nd
ar
y
fi
el
ds
ar
ou
nd
pr
im
ar
y
cr
at
er
s
24
D
ou
bl
e-
ri
ng
ed
ex
hu
m
ed
cr
at
er
s
9,
10
9
10
27
C
ra
te
rs
be
tw
ee
n
la
ye
rs
11
,
12
,
27
27
Im
pa
ct
cr
at
er
s
on
re
m
na
nt
m
es
as
17
17
P
ed
es
ta
l
cr
at
er
s
9,
10
,
14
,
24
,
25
,
30
24
,
25
,
30
10
,
14
14
L
ow
ri
m
s
of
su
pe
rp
os
ed
cr
at
er
s
25
25
K
no
bs
3,
24
,
25
20
,
21
24
,
25
20
,
21
20
,
21
24
,
25
S
em
ic
ir
cu
la
r
C
re
sc
en
t-
sh
ap
ed
du
ne
s
24
24
24
H
or
se
sh
oe
w
in
d
st
re
ak
s
34
34
34
D
ep
re
ss
io
ns
V
-s
ha
pe
d
de
pr
es
si
on
s
3
3
3
Ir
re
gu
la
r
fl
at
-f
lo
or
ed
de
pr
es
si
on
s
24
24
24
F
lu
vi
al
ch
an
ne
ls
vi
si
bl
e
th
ro
ug
h
de
po
si
t
3,
4,
33
E
ro
si
on
al
pi
tt
in
g
22
22
C
ol
la
ps
e
fe
at
ur
es
19
25
19
,
25
L
ay
er
in
g
T
op
og
ra
ph
ic
sl
op
e
br
ea
ks
co
in
ci
de
nt
w
it
h
al
be
do
sh
if
ts
3,
4,
23
3,
4
3,
4
C
ha
ng
in
g
ya
rd
an
g
or
ie
nt
at
io
n
30
,
31
,
32
3,
12
,
25
3,
30
,
32
31
3,
25
P
at
te
rn
s
vi
si
bl
e
be
ne
at
h
pa
rt
ly
er
od
ed
co
ve
r
25
S
ta
ir
st
ep
s
18
C
li
ff
be
nc
he
s
18
B
an
de
d
te
rr
ai
n
18
M
es
a-
to
-m
es
a
co
m
m
on
el
ev
at
io
n
ra
ng
e
23
L
ow
er
al
be
do
w
he
re
li
ne
ar
ri
dg
es
ar
e
vi
si
bl
e
25
,
26
H
or
iz
on
ta
l
la
ye
r
in
cl
if
f
w
it
h
lo
w
er
te
m
pe
ra
tu
re
32
A
lb
ed
o
sh
if
ts
6
6
U
ni
fo
rm
,
gr
ad
ua
l
sl
op
es
(n
o
la
ye
ri
ng
)
2
R
em
ot
e
se
ns
in
g
G
am
m
a
R
ay
S
pe
ct
ro
m
et
er
m
ea
su
re
m
en
ts
13
,
15
V
ik
in
g
in
fr
ar
ed
th
er
m
al
m
ap
pe
r
an
d
T
E
S
(l
ow
th
er
m
al
in
te
ri
a)
12
,
28
12
,
28
L
it
tle
to
no
ra
da
r
si
gn
at
ur
e
16
,
22
16
,
22
a
N
um
be
rs
in
th
e
bo
dy
of
th
e
ta
bl
e
re
fe
r
to
th
e
fo
ll
ow
in
g
re
fe
re
nc
es
:
1,
B
ar
lo
w
[1
99
3]
;
2,
B
ra
dl
ey
et
al
.
[2
00
0]
;
3,
B
ra
dl
ey
an
d
Sa
ki
m
ot
o
[2
00
1]
;
4,
B
ra
dl
ey
et
al
.
[2
00
2]
;
5,
B
ur
r
et
al
.
[2
00
6]
;
6,
F
or
sy
th
e
an
d
Z
im
be
lm
an
[1
98
8]
;
7,
F
re
y
et
al
.[
19
98
];
8,
G
re
el
ey
an
d
G
ue
st
[1
98
7]
;
9,
H
ea
d
an
d
K
re
sl
av
sk
y
[2
00
1]
;
10
,H
ea
d
an
d
K
re
sl
av
sk
y
[2
00
4]
;
11
,H
yn
ek
et
al
.[
20
02
];
12
,H
yn
ek
et
al
.[
20
03
];
13
,I
va
no
v
et
al
.[
20
05
];
14
,
K
ad
is
h
an
d
B
ar
lo
w
[2
00
6]
;1
5,
K
el
le
r
et
al
.[
20
06
];
16
,L
an
ag
an
an
d
M
cE
w
an
[2
00
3]
;1
7,
M
al
in
et
al
.[
19
98
];
18
,M
al
in
an
d
E
dg
et
t[
20
00
];
19
,M
cC
ol
le
y
et
al
.[
20
05
];
20
,N
us
sb
au
m
er
et
al
.[
20
03
];
21
,N
us
sb
au
m
er
[2
00
5]
;2
2,
P
ar
ke
r
[1
99
1]
;2
3,
Sa
ki
m
ot
o
et
al
.[
19
99
];
24
,S
ch
ul
tz
an
d
L
ut
z
[1
98
8]
;2
5,
Sc
ot
ta
nd
Ta
na
ka
[1
98
2]
;2
6,
Sc
ot
t
an
d
Ta
na
ka
[1
98
6]
;2
7,
Ta
ka
gi
an
d
Z
im
be
lm
an
[2
00
1]
;2
8,
Ta
na
ka
[2
00
0]
;2
9,
Z
im
be
lm
an
et
al
.
[1
99
7b
];
30
,
Z
im
be
lm
an
an
d
P
at
el
[1
99
8]
;
31
,
Z
im
be
lm
an
et
al
.
[2
00
0]
;
32
,
Z
im
be
lm
an
[2
00
3]
;
33
,
Z
im
be
lm
an
et
al
.
[2
00
3]
;
an
d
34
,
J.
R
.
Z
im
be
lm
an
(p
er
so
na
l
co
m
m
un
ic
at
io
n,
20
07
).
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
4 of 15
E12011
considered to be viable at present: the deposits of pyroclas-
tic flow or fall, and aeolian deposition.
[4] On Earth, deposits associated with these three pro-
cesses are distinguished by characteristic material and
deposit properties resulting from distinct depositional mech-
anisms and environments. For instance, although all might
produce widespread sheet-like deposits that are generally
fine-grained, pyroclastic flows of regional scale will inun-
date and bury topography and show variable jointing,
induration, and welding [e.g., Ross and Smith, 1961; Baker,
1981; Criswell and Elston, 1981; Selby et al., 1988; de
Silva, 1989]. Moreover, while pyroclastic fall deposits and
aeolian deposits may show small-scale induration and
welding, they will not be pervasively indurated, welded,
or jointed on a regional scale and will drape topography.
Thus, through analogy with terrestrial deposits it should be
possible in principle to distinguish between Martian pyro-
clastic flow deposits and those of fall origin such as ashfall
or loess. The key then is to identify the deposit-wide
characteristics of the MFF and what they reveal about its
material properties.
[5] The apparent regional-scale consistency of the MFF
material, fine-grain size, geologically young age, erosional
features, and considerable vertical roughness (unevenness of
the surface features derived from MOLA optical pulse width
data), suggests that a dominant process likely produced most
if not all of the deposit [Scott and Tanaka, 1982, 1986;
Greeley and Guest, 1987]. Thus identifying the process that
best explains the general characteristics of the MFF is one
way of making some progress. Our approach has therefore
been to take a synoptic perspective that focuses on character-
istics that occur throughout the deposit because they repre-
sent the MFF as a whole. With this in mind, this contribution
is based on the following three-pronged approach:
[6] 1. We have reviewed the available literature on the
MFF and attempted to organize and classify all the
disparate observations made about the MFF, and interpre-
tations made thereof, into groups that inform about process
(erosional or depositional), transport mechanism (aeolian,
fluvial, glacial etc), and material properties (fine grained,
indurated, jointed, friable). The entire analysis is summa-
rized in Table 2, with further details in Table 3.
[7] 2. We have independently analyzed 713 Mars
Orbiter Camera (MOC) images across the MFF to assess
previous observations and to conduct our own synoptic
survey of the deposit-wide characteristics (see auxiliary
material).1 This image analysis, combined with the liter-
ature review, allowed us to identify purely local versus
deposit-wide characteristics.
[8] 3. We have evaluated the deposit-wide characteristics
through a comparison with terrestrial analogs.
[9] Our main conclusions are that while yardangs, layer-
ing, and collapse features are found deposit wide, yardangs
hold the most promise to inform about the origin of the
MFF. These erosional features point to a process that
produced friable regional-scale deposits that are layered,
differentially indurated, and often jointed. The morphology
of the yardang fields throughout the MFF is strongly
analogous to those seen in terrestrial yardangs fields devel-
oped in extensive ignimbrites and may implicate a pyro-
clastic flow origin for much of the MFF.
[10] The MFF is located between two major volcanic
centers. The eastern portion of the MFF is found on the
upslope of the Tharsis rise while the western portion is
located about 1000 km south of the Elysium volcanic center
[Scott and Tanaka, 1986; Greeley and Guest, 1987]. The
eastern portion of the MFF is partially overlain by a part of
the ??stealth region,?? an area of little to no radar signal
Table 3. Characteristics of Yardangs on Earth and the MFF
Location Lithology Yardang Characteristics Aspect Ratio (length:width)
Iran, Lut Deserta silty clays, gypsiferous sands
moderately indurated
10s of km long, 60?80 m tall
Peru, Pisco Formationb marine sediments
China, Lop Nora,c lake beds 100s of meters long, 20 m high up to 10:1
Rogers Lake, CA, USAd gravel, sand, silt, and clay
moderately consolidated
5:1
Egypt, Western Desertc,e silicified limestone, lake beds, chalk 10?100 m long, 2?5 m tall
Chade indurated sandstone 20?30 km long, 1 km wide;
spacing 500 m to 2 km
20:1?30:1
Argentina, Payun Matruf ignimbrites and lavas 2?10 km long; lavas result in broad troughs;
Ignimbrites yardangs
N. Saudi Arabiac Cambrian sandstone 40 m high, 100s of meters long
Namibiac igneous and metamorphic basement 8?0 km long; spacing 300?350 m up to 20:1
Bolivia, Argentina,
central Andesg
ignimbrite variably indurated, welded 100 m tall (indurated),
10 m tall (nonindurated);
vertical to steep slopes
20:?40:1 (indurated),
5:1?10:1 (nonindurated)
Mars, MFFh,i Medusa Fossae Formation vertical to steep slopes;
>100 m tall (range 5?700 m)
5:1 minimum to
20:1?50:1 maximum
aMcCauley et al. [1977b].
bMcCauley [1973].
cGoudie [2007].
dWard and Greeley [1984].
eBreed et al. [1989].
fInbar and Risso [2001].
gBailey et al. [2007] and S. L. de Silva (manuscript in preparation, 2008).
hWard [1979].
iThis work.
1Auxiliary materials are available in the HTML. doi:10.1029/
2008JE003076.
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
5 of 15
E12011
thought to be lightweight material of volcanic origin [Edgett
et al., 1997]. In all places where they are in contact, the
southern portion of the MFF overlies the hemispheric
dichotomy boundary [Scott and Tanaka, 1986; Greeley
and Guest, 1987; Bradley et al., 2000; Head and Kreslavsky,
2001, 2004; Hynek et al., 2003]. The MFF has been divided
into three members (upper, middle, and lower) on the basis
of morphology, stratigraphy, albedo and apparent age [Scott
and Tanaka, 1986; Greeley and Guest, 1987]. Though
reevaluation of this mapping was advised when early Mars
Orbiter Laser Altimeter data was analyzed [Sakimoto et al.,
1999], a deposit-wide reevaluation has not yet been pub-
lished. Though the MFF stretches for over 5000 km and
covers an elevation range from 
2000 to +2000 m, it is not
continuous over that distance, but broken up into segments
of 1000?2000 km, separated by up to 1000 km in the
western portion of the deposit. Albedo varies from low to
high within the deposit [Scott and Tanaka, 1982] and
boulders are not seen at erosional contacts [Malin and
Edgett, 2000; Hynek et al., 2003].
3. Analysis of the Literature
[11] During an exhaustive review of the literature we have
found that over one hundred separate observations have
been used to support the conclusions of 33 published studies
of the MFF (Table 2). We have classified these observations
into extrinsic, those that inform about process (transport,
emplacement, erosion) and intrinsic, those that inform about
the lithology, morphology, and material properties of the
deposit. Examples of the former are bedding [Malin and
Edgett, 2000] that implies an intermittent accumulation, as
well as draping [Bradley and Sakimoto, 2001] and an
apparent thickening toward Tharsis [Hynek et al., 2002,
2003] that have been used to suggest a pyroclastic fall
origin. Intrinsic attributes include raised curvilinear features
[Burr et al., 2006], stair steps [Malin and Edgett, 2000],
knobs [Scott and Tanaka, 1982; Schultz and Lutz, 1988;
Bradley and Sakimoto, 2001; Nussbaumer et al., 2003;
Nussbaumer, 2005], and yardangs [Scott and Tanaka,
1982; Bradley et al., 2002]; these all speak to the nature
of the MFF material.
[12] Our complete analysis of the literature on the MFF is
given in Table 2. It is clear from this literature review that a
strong consensus is shown for the following attributes of the
MFF:
[13] 1. Large aerial extent (2.1?2.5  106 km2) and
volume (1?1.9  106 km3) [Scott and Tanaka, 1982;
Tanaka, 2000; Bradley et al., 2002; Head and Kreslavsky,
2001, 2004; Hynek et al., 2003].
[14] 2. Amazonian age is based on the following evi-
dence. The upper member of the MFF overlies the youngest
aureole deposits [Head and Kreslavsky, 2001], though
interweaving with some younger lava flows has been
observed where the MFF borders the Cerberus Plains
[Lanagan and McEwen, 2003]. Otherwise, the MFF appears
to overlie all other geologic formations in its region. These
stratigraphic relations have determined that most of the
MFF was emplaced in the Amazonian Period [Scott and
Tanaka, 1982, 1986; Greeley and Guest, 1987; Bradley and
Sakimoto, 2001; Head and Kreslavsky, 2001, 2004; Hynek
et al., 2003]. Head and Kreslavsky [2001, 2004] suggest
that the complex stratigraphy described above indicates that
deposition may have begun as early as the Hesperian.
[15] 3. Multiple episodes of deposition as evidenced by
changing yardang orientations [Scott and Tanaka, 1982;
Bradley and Sakimoto, 2001; Hynek et al., 2002, 2003],
complex stratigraphy [Head and Kreslavsky, 2001, 2004], and
craters between layers [Tanaka and Leonard, 1998,
Sakimoto et al., 1999; Hynek et al., 2002, 2003].
[16] 4. Fine-grain size based on the lack of boulders along
erosional contacts [Malin and Edgett, 2000; Hynek et al.,
2003], different crater depth-diameter ratios than areas
known to be caprock [Barlow, 1993], erosional morphology
(yardangs, chutes or dark scree slopes and pedestal craters)
[Hynek et al., 2003; Malin and Edgett, 2000; Schultz and
Lutz, 1988], low thermal inertia and radar signals [Hynek et
al., 2003].
[17] 5. Heavily eroded by aeolian processes [Bradley et
al., 2002; Head and Kreslavsky, 2001, 2004; Hynek et al.,
2003; Malin et al., 1998; Zimbelman et al., 1997a; Greeley
and Guest, 1987; Scott and Tanaka, 1986; Frey et al., 1998;
Takagi and Zimbelman, 2001; Schultz and Lutz, 1988] with
an estimated original aerial extent of 5  106 km2 [Bradley
et al., 2002].
[18] Beyond this, our analysis has also allowed us to
eliminate specific local features such as draping [see also
Watters et al., 2007], and has resulted in the identification of
a set of deposit-wide features that have been ubiquitous in
studies of the MFF: yardangs, collapse features, and layering
(Figure 2). The yardangs hold the most promise for provid-
ing insight into material properties. The collapse features and
layering are seen deposit wide, but are equivocal and will
require more investigation with higher-resolution data to be
useful.
3.1. Yardangs
[19] Yardangs are aerodynamically shaped landforms that
represent the remnants of deposits eroded by the wind and
are the most commonly observed features in the MFF
[Ward, 1979]. They occur across the extent of the deposit
and within various members at different elevations. Though
most publications refer to the yardangs, few provide detail
on their texture or appearance. Focus has primarily been on
their wide variety of orientations [Scott and Tanaka, 1982;
Zimbelman and Patel, 1998; Zimbelman et al., 2000;
Figure 2. Erosional features of the MFF that were the
focus of this study. (a) Portion of THEMIS visible image
V19253002 centered at 177.4
E, 10.2
S, showing large-
scale yardangs in the MFF. (b) Portion of THEMIS daytime
IR image I01280001 centered at 189
E, 7.24
S, showing
layered terrain in the MFF. (c) Portion of MOC image E03-
01084 centered at 178.8
E, 2.2
N, showing collapse
features in the MFF.
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
6 of 15
E12011
Bradley et al., 2002; Hynek et al., 2003; Zimbelman, 2003]
and their proof of extensive aeolian erosion. All members of
the MFF display yardangs, so they are clearly a formation-
wide characteristic. Two studies have focused on yardangs
formed in MFF material.
[20] Ward [1979] observed that Martian yardangs had
aspect ratios (length:width) up to 50:1 and suggested that
they are much larger than those known on Earth, and they.
He suggested this to be the result of a lack of secondary
processes (e.g., fluvial erosion) to subdue their forms,
thicker formations on Mars than on Earth, and that trough
extension was much faster than streamlining. He suggested
that the lithology of the MFF had to be something that
weathered to a granular substance that could be efficiently
deflated.
[21] Bradley et al. [2002] observed bidirectional yardangs
in 23% of the 155 MOC Narrow Angle images available at
the time of their survey of the MFF. They interpreted the
consistent angle of intersection (Figure 3c) to be a sign of
columnar jointing, which is in agreement with other
researchers [Scott and Tanaka, 1982, 1986; Zimbelman et
al., 1997b]. Bradley et al. [2002] posited that this apparent
jointing was developed through cooling and compaction on
the basis of terrestrial analog studies. They noted that joint
patterns are not found throughout the entire MFF, and
suggest that this is either due to wind direction being the
dominant force on yardang orientation, or because jointing
is not found throughout the formation.
3.2. Collapse Features
[22] In a study of the middle member of the MFF located
between 180 and 192
E (168?180
W) longitude and 2
N?
4
S latitude, McColley et al. [2005] noted several kilometer-
scale rectilinear troughs (see Figure 2c). They interpreted
these depressions to be created by subsurface collapse or
movement on scales ranging from one to tens of kilometers,
and to be the result of volatile release. On the basis of the
literature these features appear to be local, but the image
analysis described in the next section shows otherwise.
Similar features have been observed elsewhere on Mars,
but generally on a much larger scale (observable in THEMIS
(Thermal Emission Imaging System) images) and with a
more linear appearance. The frequency of these features in
areas showing regional extension or local fissuring led to
their interpretation to be the result of dilational faulting
[Wyrick et al., 2004].
3.3. Layering
[23] Many authors have observed signs of layering
throughout the deposit [Scott and Tanaka, 1982, 1986;
Zimbelman et al., 1997a; Zimbelman and Patel, 1998;
Forsythe and Zimbelman, 1988; Sakimoto et al., 1999; Malin
and Edgett, 2000; Bradley and Sakimoto, 2001; Bradley
et al., 2002; Hynek et al., 2003], although only a few
provide specific details characterizing the layering. Three
layers at the scale of hundreds of meters were observed
in MFF material in the region of the Gordii Dorsum
[Zimbelman et al., 1996; Zimbelman and Patel, 1998].
Hundreds of meter-scale layers were observed using
MOLA data near Apollineras Patera and south of Elysium
Mons [Bradley et al., 2002]. Layering at the scale of 10s
of meters cannot be observed using MOLA data, but
some evidence of such layering was seen in the 2002
survey of 155 MOC images [Bradley et al., 2002]. The
area of the layering is uncertain from the literature.
4. Image Analysis
[24] We have examined 713 MOC Narrow Angle (1.5?
3 m per pixel for the images surveyed) images of the
MFF (Figure 4) that were obtained from the NASA
Planetary Data System at http://ida.wr.usgs.gov/. Each
Figure 3. Examples of MFF yardang characteristics. (a) Portion of MOC E10-03225, centered at
221.1
E, 4.27
S, showing smooth-textured curvilinear yardangs. (b) Portion of MOC M09-04901
centered at 147.37
E, 6.45
S, showing rough-textured curvilinear yardangs. (c) Portion of MOC M08-
01561 centered at 7.42
N, 197.11
E, showing bidirectional yardangs and an exhumed crater. (d) Portion
of THEMIS infrared I17167016 centered at 209.742
E, 2.241
N, showing a large-scale example of a
serrated or feathered scarp. Scarp is 3 km high. (e) Portion of MOC E04-00932 centered at 210.63
E,
3.7
N, with arrow pointing to a small-scale example of the serrated or feathered scarp.
Figure 4. Footprints of MOC images overlain on a portion
of the Viking-based geologic map of Mars showing the
MFF [Scott and Tanaka, 1986; Greeley and Guest, 1987].
Map range is 
15 to +15
N latitude and 135?245
W
longitude with 5
 grid spacing. The MFF units are shown in
dark green, light green, and teal.
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
7 of 15
E12011
Figure 5. (a) Locations of MOC images showing rough and smooth-textured curvilinear yardangs.
(b) Locations of MOC images showing streamlined yardangs. (c) Locations of MOC images showing
bidirectional yardangs like those seen in Figure 2c. (d) Locations of MOC images showing erosional
patterns interpreted as collapse features. (e) Locations of MOC images showing layering at the scale of
10s of meters.
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
8 of 15
E12011
individual image was analyzed in conjunction with the
Mars Spaceflight Facility at Arizona State University?s
online Geographical Information System (GIS): Java Mis-
sion Planning and Analysis for Remote Sensing (JMARS)
http://jmars.asu.edu [Christensen et al., 2007]. The JMARS
day and night global THEMIS layers were used to provide
medium resolution (230 m/pixel) context for the MOC
images. Additionally, the JMARS MOLA 16-bit elevation
data, at 128 pixels/
, were used to create topographic
profiles to evaluate large-scale layering, and heights and
slopes of the larger MFF yardangs. Our observations
focused on the three main features we have identified as
characteristic of the MFF as a whole: yardangs, collapse
features and layering. The footprints of each image in our
database (available as a digital addendum to this paper)
containing these features are shown in Figure 4.
4.1. Yardangs
[25] Yardangs occur as extensive fields, some with areas
of up to 40,000 km2, and are clearly visible at THEMIS
resolution. Of the 713 MOC images of MFF, 78% showed
yardangs in some form.
[26] Two types of yardang shape were seen in the MOC
images: curvilinear, characterized by curving lines or edges
(55% of all images surveyed) and streamlined, having a
smooth even shape (6% of all images surveyed). The MFF
yardangs are predominantly curvilinear in form. The
texture of these yardangs ranges from smooth surfaced
(Figure 3a) to rough, with a platy or sharp-edged appear-
ance (Figure 3b). 54% of the curvilinear yardang images
were smooth and the remainder rough. The locations of
images showing these yardangs are mapped in Figure 5a,
indicating that yardangs with similar textures tend to be
grouped together. Smooth yardangs appear to be more
common in the middle and upper members whereas rough
yardangs are more common in the middle and lower
members. Aspect ratios (length:width) of the curvilinear
yardangs range from 5:1 to 20:1, but there is no clear
trend between aspect ratio and texture or spatial distribu-
tion. Streamlined yardangs (Figure 3c) appear to be
relatively rare in the MFF (mapped in Figure 5b). They
also coincide with areas where curvilinear yardangs are
common. The aspect ratios of the streamlined yardangs
seen in 40 MOC images range from 3:1 to as high as 50:1.
[27] Bidirectional yardangs (Figure 3c) appear to trend in
two directions with an angle of intersection between 30 and
45
 [Bradley et al., 2002]. These yardangs are only about
10?15 m wide and 30?40 m long and are seen in 44% of
the MOC images surveyed (mapped in Figure 5c). They are
too small to be seen at THEMIS resolution, and MOC
resolution is insufficient to discern clearly the shape or
texture of these yardangs. Of the 713 images, 554 showed
yardangs in some form: 204 curvilinear only, 125 bidirec-
tional only, 186 curvilinear and bidirectional, 31 stream-
lined, 8 streamlined and bidirectional. Of the images
showing both curvilinear and bidirectional yardangs, 157
(84%) have smooth curvilinear yardangs suggesting a
strong relationship between the smooth curvilinear and
bidirectional yardangs.
4.2. Collapse Features
[28] We have found features that resemble the pit chains
and rectilinear troughs observed by McColley et al. [2005]
and Wyrick et al. [2004] in 14% of the MOC images
surveyed (mapped in Figure 5d). They are not as prevalent
in the MFF as the yardangs, but they are seen in all
geographic areas and all three members of the formation.
The majority of these features are seen at the edges of the
deposit where relief is limited to less than 250 m.
[29] The pits that the chains and troughs form from
(Figure 6a) are about 50?100 m in diameter and are circular
or elongate in shape. Troughs range from 100 to 250 m wide
and can extend several kilometers in length (Figure 6b). The
troughs form in many different directions and intersect at a
variety of angles, creating varied geometric patterns in the
MFF material (Figure 6c).
4.3. Layering
[30] We have found very limited evidence for layering in
the MFF at the scale of 100s of meters beyond the evidence
cited by previous workers. Only five MOC images showed
cliffs greater than 400 m high with benches resulting from
probable layers in the cliffs (Figure 7a). Moreover, the
MOLA shot size on Mars is 160 m and the shot
spacing is typically 330 m (endnote 12 of Smith et
al. [1999]), which limits the ability to discern layering at
this scale.
[31] Features that appear to represent layering at a scale of
10s of meters can be noted in some MOC images with
caprock layers over friable material and ??onion skin??
topography (Figures 7b and 7c), which is a common aeolian
erosional feature found in terrestrial material with layering
Figure 6. Examples of characteristics of collapse features. (a) Portion of MOC E03-01084 centered at
178.8
E, 2.2
N, showing collapse pits in center with troughs on the perimeter. (b) Portion of MOC E05-
02396 centered at 178.14
E, 2.31
N, showing troughs formed from collapse pits. (c) Portion of MOC
R04-02077 centered at 172.85
E, 1.68
S, showing geometric pattern commonly seen in MFF collapse
troughs.
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
9 of 15
E12011
at this scale. Figure 5e shows the locations of MOC images
which evidence layering.
[32] The layering is present in each of the mapped
members of the formation so it could be described as
formation wide, but is not abundant enough to be described
as extensive. At this stage we cannot determine whether the
observed layering represents primary bedding, grading, a
superimposed property such as differential welding or
induration or some combination of these. Higher-resolution
data (e.g., HiRISE) will play an important role in further
defining the character of layering in the MFF.
5. Discussion
[33] In an attempt to resolve between the three most likely
origins for the MFF, we have taken a synoptic view of the
formation and tried to identify the deposit-wide character-
istics as a means of addressing the origin of the formation as
a whole.
[34] As recognized by previous workers, the MFF is a
deposit of large aerial extent and volume that shows certain
deposit-wide characteristics like its young (Amazonian)
age, evidence for multiple episodes of deposition, a perva-
sively fine-grained nature, and erosion by aeolian processes.
However, these general characteristics alone do not inform
about a specific process of origin and our effort has been to
try and identify features in the MFF that are a function of
the deposit properties, as we believe these will vary as a
function of depositional mechanism. For instance, by anal-
ogy with Earth, deposition by pyroclastic flows will result
in induration and welding characteristics and cooling and
compaction related to jointing and fracturing that are not
typical of primary fall or aeolian deposits. Welding in
ashfall deposits is local and limited to less than 10 km from
the source [e.g., Cas and Wright, 1992], and these are
otherwise completely friable, as is loess directly after
deposition. Any induration process for air fall deposits on
the surface of Mars would be secondary and is therefore
another degree removed from a primary mechanism. Given
the large area of the MFF and our focus on deposit-wide
characteristics, it is difficult to envision a consistent and
pervasive deposit-wide induration process that could take
place under Amazonian conditions that is not linked to its
formation. The properties of deposits from these three
mechanisms (pyroclastic flow, pyroclastic fall and loess)
should therefore be quite distinct.
[35] Our analysis has identified three features that are
characteristic of the MFF across its extent; yardangs,
collapse features, and layering. Of these three, we find that
collapse features and layering do not provide enough
information at this time to allow us to differentiate between
the viable hypotheses of origin of the MFF materials; layer-
ing could result from any of the three processes, and the
resolution of the data is insufficient to differentiate between
styles of layering. In turn, collapse features are enigmatic.
Surface erosion alone cannot fully explain the collapse
features, so some sort of subsurface activity must be involved
in their formation. Whether this is due to volatile release
[McColley et al., 2005], faulting [Wyrick et al., 2004] or some
other process is difficult to discern at this time.
[36] Ultimately, the yardangs appear to hold the most
promise for inferences regarding formation. The form and
texture of wind-eroded forms is strongly dependant on rock
structure and lithology because these control the erosional
effectiveness of the wind [Breed et al., 1989] and the
character of the ridges and slopes of the forms [Selby et
al., 1988]. On Earth, extensive yardang fields are found in a
variety of deposits ranging from lake deposits (Iran), sand-
stones (Peru), and ignimbrites and lava flows (Andes),
among others, as summarized in Table 3. This compilation
shows that although there is wide-ranging variability in
yardang morphology, there is a strong relationship between
aspect ratio, height, and slopes. These are largely a function
of the degree of induration of the deposits, with poorly
indurated material forming small, stubby yardangs, and
indurated deposits supporting long, high, steep-sided yard-
angs. This discussion now focuses on the properties of
yardangs and their implications for the origin of the MFF.
5.1. Yardangs in the MFF
[37] Ward [1979] observed that yardangs in the MFF on
Mars were considerably larger and more elongate (aspect
ratios up to 50:1) than mature terrestrial yardangs (aspect
ratio 4:1 [Ward and Greeley, 1984]). He suggested that the
Figure 7. Examples of features that characterize layering in the MFF. (a) Portion of MOC M12-02293
centered at 148.95
E, 2.79
S, showing three distinct benches in a 600 m high cliff. Illumination is from
top. (b) Portion of MOC E05-01356 centered at 147.45
E, 3.32
N, showing a layer of caprock breaking
off as the less indurated layer below is eroded. (c) Portion of MOC M02-02978 centered at 142.09
E,
0.25
S, showing an example of onion skin topography.
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
10 of 15
E12011
size, slopes, and forms of Martian yardangs implied an in-
durated lithology that weathers to a granular substrate. The
other pervasive lithologic property attributed to the MFF
lithology is jointing, which Bradley et al. [2002] attributed
to cooling and compaction [see also Scott and Tanaka, 1982].
[38] We have attempted to further these analyses by
establishing more detailed morphometric parameters for
the MFF yardangs. The height and slope of yardangs that
are wider than one kilometer can be evaluated with the
MOLA topographic profile tool in JMARS (Figure 8).
Although these megayardangs only represent about 7% of
the aerial coverage of the yardangs in the formation, this
analysis gives some insight into these particular yardangs.
The slopes of these megayardangs are on average 75?80

and they range in height from 5 to 700 m, with the majority
being over 100 m tall (Figure 8). Their width ranges from
one to five kilometers. Aspect ratios (length:width) vary
widely but can be greater than 50:1, with yardangs that are
several tens of kilometers long. Because slope, height and
aspect ratio are largely a function of the degree of induration
of the deposits, these measurements suggest heavily indu-
rated material.
5.2. Inferred Material Properties of the MFF and a
Likely Analog Lithology
[39] Our analysis of the MFF is consistent with a layered
lithology that supports a range of yardang morphologies and
develops jointing. At one extreme, the steep slopes and
height of the megayardang ridges in the MFF point to a
strong lithology, as the ability to support such forms is a
characteristic of strong rocks [Selby et al., 1988]. However,
the variable size, forms and textures of yardangs seen at
high resolution suggest that some are made from weaker
materials. The material must weather to a granular substance
that is easily deflated by the wind [Ward, 1979]. By analogy
with Earth, these properties suggest variably cemented or
indurated deposits of clastic material, either sediments or
pyroclastic rocks. The potential presence of jointing is
important. Direct evidence for jointing is difficult to find,
and we note that Bradley et al. [2002] largely inferred this
to explain bidirectional yardangs. We note that some of the
MFF yardang fields form perpendicular to a scarp or slope
greater than 45
, leaving a serrated edge in the scarp or
feathering of the slope (Figures 3d and 3e). The images
showing streamlined forms are among areas with abundant
curvilinear yardangs, so it is likely that these were originally
curvilinear in form and eroded into more streamlined
features. Thus we interpret no difference between the
lithology of the different yardang types.
[40] Of the potential analogs for the MFF on Earth,
neither volcanic air fall deposits nor loess (suspended silt)
deposits display characteristics consistent with the MFF.
These deposits are rarely indurated or jointed on a regional
scale. Ancient loess deposits on Earth (Paleozoic) may be
indurated and behave as strong rocks after burial and
metamorphism [e.g., Tramp et al., 2004] but young loess
deposits are rarely consolidated and behave as weak to
strong soils once subjected to surface pedogenesis [Gu?nster
et al., 2001]. These deposits collapse as granular flows as
during the 1556 Shaanxi Earthquake [People?s Republic of
China, 1982, p.100]. No examples of regional extensive
indurated pyroclastic fall deposits are known to the authors.
We also do not know of extensive yardang fields that have
developed in these deposits. Thus, if pyroclastic fall and/or
loess deposits form the MFF, induration would have to be
the result of secondary processes. It would likely be a low
temperature hydrous/chemical process like that seen in loess
deposits [Gu?nster et al., 2001] but would have to work from
the top down to give the strength profile required (top
stronger than bottom inferred from the discontinuous but
pervasive layering characteristics observed at high resolu-
tion). Such induration processes are rarely pervasive, typi-
cally produce only incipient induration and do not produce
systematic jointing on a large scale.
[41] On the contrary, ignimbrites, the depositional prod-
uct of pyroclastic flows, display the range of characteristics
that have been attributed to the MFF. In fact of all the
terrestrial yardang fields examined by us (Table 3), the
erosional pattern is most pervasive where ignimbrites are
the host lithology. Ignimbrites, uniquely in our opinion,
develop the range of properties found in the MFF because
of their depositional environment, and we show below that
this is probably the most likely lithology in the MFF.
5.3. Yardangs in Terrestrial Ignimbrites
[42] As part of an ongoing study of extensive ignimbrites
of the Central Andes [de Silva, 1989; de Silva et al., 2006],
we have found that a wide range of yardang morphologies is
produced by the variable material properties that characterize
Figure 8. Examples of topographic profiles derived from
MOLA 128 pixels/
 DEM, created using JMARS to
evaluate the height, width, and slope of large-scale yard-
angs. (a) Yardangs in the western portion of the MFF
formed along a steep slope facing into the inferred primary
wind direction on the basis of yardang orientation
[McCauley et al., 1977a]. The profile line is centered at
150.28
E, 0.79
N. (b) Yardangs in the central portion of the
MFF. In this case the topographic slope is perpendicular to
the wind that formed the yardangs, so the yardangs formed
perpendicular to the slope. The profile line is centered at
176.64
E, 10.57
S.
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
11 of 15
E12011
extensive regional ignimbrite sheets. The ignimbrites of the
Central Andes are among the largest on Earth, with volumes
in excess of 1000 km3 and areas of over 10,000 km2. At the
scale of remotely sensed data, layering can be distinct in
plateaus consisting of several stacked sheets, but can also be
found within individual sheets, although this is rarely
pervasive. Ignimbrites are fragmental (volcaniclastic) rocks,
and large ignimbrites like these typically show two main
facies: an upper massive, indurated and jointed facies, the
result of postdepositional induration by escaping heat and
hot gasses [Roever, 1966; de Silva, 1989] and a lower weak
to poorly indurated, massive ash- and pumice-rich facies
(Figure 9 (top)). These facies have a welding profile super-
imposed on them, but this is not typical and is only local in
occurrence. The two main facies may occur individually, and
may be locally vertically associated, or regionally grade into
each other laterally (the result of heat loss by the pyroclastic
flow as it travels). The two facies have quite different
mechanical properties; the upper facies behaves as strong
rock, fails by block collapse and supports steep/vertical
cliffs, whereas the lower facies has deeply weathered soils
and forms gentle slopes [Selby et al., 1988; Crown et al.,
1989]. These two facies produce yardangs of quite different
character (Figure 9). The presence of the indurated facies
results in massive, extremely elongate, high aspect ratio
(20:1 to 40:1) yardangs, that form tall (100 m), thin ridges
with steep to vertical walls (Figure 9 (top and middle)). This
type of yardang demonstrates the importance of differential
induration and mechanical weathering in producing tall,
steep-sided, megayardangs with broad shallow-sloped bases.
Weakly indurated ignimbrites are carved into stubby and
chaotic yardangs with aspect ratios of 5:1 to 10:1 and heights
rarely exceeding 20 m (Figure 9 (bottom)). The indurated
(not necessarily welded) facies is commonly pervasively
jointed [de Silva, 1989], and the development of yardangs in
these areas clearly demonstrates that a persistent, strong
unidirectional wind is the dominant parameter controlling
yardang orientation [Inbar and Risso, 2001; Bailey et al.,
2007]. In many locations, we have found serrated margins
quite like those in the MFF to be the result of oblique
intersections of jointing and yardangs or scarps of ignimbrite
sheets (Figure 10).
[43] We suggest that these terrestrial ignimbrites provide
the strongest analog to much of the MFF and we concur
with previous workers such as Scott and Tanaka [1982,
1986] that a large part of the MFF resulted from the
deposition by pyroclastic flows. If so, it is interesting to
speculate about possible source structures for these massive
eruptions. Studies of Martian plume stability suggest that
pyroclastic flows could extend for hundreds of kilometers
from the source [Wilson and Head, 1994], though Glaze
and Baloga [2002] set limits on eruption conditions that
would allow flows up to only 200 km from the source.
Previous analyses of the energetics of explosive volcanic
eruptions in conjunction with the Martian highland paterae
suggested that their flanks lengths (e.g., over 400 km for
Hadriaca Patera) were too great to result from air fall
deposits (given required eruption column heights) and
attributed the formation of Hadriaca and Tyrrhena Paterae
to pyroclastic flow emplacement driven by either magmatic
or hydromagmatic explosive eruptions [Greeley and Crown,
1990; Crown and Greeley, 1993]. Models of cooling during
Figure 9. Yardangs in ignimbrites in the central Andes.
(top) Megayardang in the 3.6 Ma Tara ignimbrite plateau in
northern Chile. Wind is from the left. The yardang is 100 m
tall and consists of a white poorly indurated base forming a
shallow sloped platform to the vertical walled ??blade?? of
the vapor phase indurated facies. (middle) Aerial view of
the same yardang used with permission of Google Earth
imagery# Google Inc., map data copyright 2007 DMapas/
El Mercamio, and image copyright 2007 DigitalGlobe. Note
the columnar structure of the blade produced by pervasive
jointing in the indurated facies and formation of an
extensive block field at the base resulting from collapse.
This yardang demonstrates the importance of differential
induration and mechanical weathering in producing tall,
steep-sided, megayardangs with broad shallow sloped
bases. Image centered at 22
51043.1500S, 67
18009.5900W.
(bottom) Aeolian erosion of the poorly indurated Campo
Piedra Pomez ignimbrite (70,000 years old) in the Puna of
Argentina. Elevation is 4000 m. Yardangs are relatively
small ranging up to 20 m high and 100 m long, with aspect
ratios of 5:1?10:1. Note weak layering within the
ignimbrite seen in the foreground. There are no massive
vertical walls and tall feature is minimized through block
collapse. Gravel ripples and dunes are found in the troughs.
Wind direction is from the right.
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
12 of 15
E12011
emplacement indicate that welding within pyroclastic flows
could occur at large distances from a source vent on Mars
and thus could explain layering within extensive volcanic
deposits [Crown and Greeley, 1993]. Given the limits on the
distribution of volcanic materials, multiple local sources
within and buried by the deposit are necessary for either
type of deposit. Scott and Tanaka [1982] identified three
collapse features to be possible source vents, and Tanaka
[2000] suggested that the Gordii and Eumenides Dorsa and
an unnamed dorsum, all located in areas of 1?3 km relief,
were possible source fissures. Eruption of major ignim-
brites from fissures, faults or volcano-tectonic features is
increasingly being recognized on Earth [Aguirre-Diaz and
Labarthe-Hernandez, 2003; de Silva et al., 2006] and may
explain the lack of evidence for distinct ??caldera?? struc-
tures that could have sourced the MFF.
6. Conclusion
[44] Researchers are in agreement about some character-
istics of the MFF such as its young (Amazonian) age, the
evidence for multiple episodes of deposition, its perva-
sively fine-grained nature, and the evidence for heavy
erosion by aeolian processes. However, none of these
characteristics can discriminate between the three viable
origins for the MFF; pyroclastic flow, pyroclastic fall and
aeolian (loess). A survey of MOC Narrow Angle images
identified three features that are seen deposit wide: yard-
angs, collapse features, and layering. The collapse features
are in themselves enigmatic but suggest some postdeposi-
tional processes and need further investigation before they
can inform about the origin or material properties of the
deposit. The evidence for layering is not abundant, but
occurs in all geographic areas of the MFF. Higher-resolution
data can better characterize the layering to provide insight
into origin. Yardangs provide the greatest insight into
material properties of the MFF. With aspect ratios of up
to 50:1 and nearly vertical cliffs, yardangs in the MFF
suggest a lithology that displays a wide range of indura-
tion, in some areas strong enough to support steep cliffs.
Jointing is inferred from the bidirectional yardangs and the
serrated margins, but requires higher-resolution data to test
directly.
[45] The similarity of the morphological features dis-
played by the MFF yardangs and those of the terrestrial
ignimbrites is striking. The depositional processes and
resulting lithologic characteristics of large ignimbrites in
the Central Andes provide the most viable analog to the
MFF lithology. We recognize that large terrestrial ignim-
brites are typically felsic in composition, and that large-
scale silicic volcanism is not known on Mars [Francis and
Wood, 1982]; however, we note that basaltic ignimbrites on
Earth, although considerably smaller in volume, do display
similar facies variations [Freundt and Schmincke, 1995;
Giordano et al., 2006]. We would anticipate that pyroclastic
eruptions on Mars, even if they are mafic, would produce
deposits with a similar range of facies and therefore display
physical properties similar to the large terrestrial ignimbrites
discussed here. An ignimbrite is the only hypothesis pre-
sented to date that explains all of the major observations
with a single mechanism.
Figure 10. Formation of a serrated scarp. (top) Aeolian
erosion and yardang formation in the indurated 5.6 Ma
Chuhuihuilla ignimbrite in southwest Bolivia. Oblique view
from west to east used with permission of Google Earth
imagery # Google Inc. and image copyright 2007 Terra-
Metrics. Prevailing wind is from NW to SE. The longest of
these ridges is about 2 km long and aspect ratios of 20:1 are
common here. Note the discordance between pervasive
jointing and wind direction producing serrated margins and
rhomboid blocks. (middle) Serrated edges in mature yard-
angs (aspect ratio 4:1) in the 3.6Ma Tara ignimbrite produced
by pervasive jointing oblique to wind direction (from right).
The yardangs have grown by deepening and widening
ancestral fluvial channels. Image used with permission of
Google Earth imagery # Google Inc. and image copyright
2007 DigitalGlobe. Image center: 22
5801100S 67
1705800W.
(bottom) Portion of aerial photo 22,1378 (15 April 1961) of
Cerro Escorial region in Argentina (10 km across). North is to
the top. Prevailing wind direction is NW-SE (from top left to
bottom right), but local channeling by topography produces a
weak sinuosity to the yardang orientation. Several stages of
yardang formation and surface erosion can be seen from
incipient, mature through to stripped. Aspect ratios vary from
4:1 to about 10:1. Serrated edges are seen where fluvial
erosion exposes jointing in ignimbrite.
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
13 of 15
E12011
[46] Acknowledgments. K.M. acknowledges research support from a
2006 North Dakota View Consortium Grant. S.D.S. thanks North Dakota
NASA Space Grant and Oregon NASA Space Grant Consortium for
support to conduct some of this work. Portions of this work were supported
by NASA PGG grant NNX07AP42G to J.R.Z. for the investigation of MFF.
The detailed review by an anonymous reviewer and the editorial handling
by Robert Carlson improved the form and content of this paper consider-
ably and are gratefully acknowledged.
References
Aguirre-Diaz, G. J., and G. Labarthe-Hernandez (2003), Fissure ignim-
brites: Fissure-source origin for voluminous ignimbrites of the Sierra
Madre Occidental and its relationship to basin and range faulting, Geol-
ogy, 31(9), 773?776, doi:10.1130/G19665.1.
Bailey, J. E., S. Self, L. K. Wooller, and P. J. Mouginis-Mark (2007),
Discrimination of fluvial and eolian features on large ignimbrite sheets
around La Pacana Caldera, Chile, using Landsat and SRTM-derived
DEM, Remote Sens. Environ., 108(1), 24 ? 41, doi:10.1016/j.rse.
2006.10.018.
Baker, M. C. W. (1981), The nature and distribution of upper Cenozoic
ignimbrite centres in the Central Andes, J. Volcanol. Geotherm. Res., 11,
293?315.
Barlow, N. G. (1993), Increased depth diameter ratios in the Medusae
Fossae Formation deposits of Mars (abstract), Proc. Lunar Planet. Sci.
Conf., 24, 61?62.
Bradley, B. A., and S. E. H. Sakimoto (2001), Relationships between the
Medusae Fossae Formation (MFF), fluvial channels, and the dichotomy
boundary southeast of Nicholson Crater, Mars (abstract), Proc. Lunar
Planet. Sci. Conf., 32, 1335.
Bradley, B. A., E. B. Grosfils, and S. E. H. Sakimoto (2000), Boundaries of
the Medusae Fossae Formation and Elysium Basin materials using Mars
Orbiter Laser Altimeter (MOLA) data (abstract), Proc. Lunar Planet. Sci.
Conf., 31, 2055.
Bradley, B. A., S. E. H. Sakimoto, H. Frey, and J. R. Zimbelman (2002),
Medusae Fossae Formation: New perspectives from Mars Global Sur-
veyor, J. Geophys. Res., 107(E8), 5058, doi:10.1029/2001JE001537.
Breed, C. S., J. F. McCauley, and M. I. Whitney (1989), Wind erosion forms,
in Arid Zone Geomorphology, edited by D. S. G. Thomas, pp. 284?307,
Belhaven, London.
Burr, D. M., R. M. E. Williams, J. Nussbaumer, and J. R. Zimbelman
(2006), Multiple, distinct (glacio?) fluvial paleochannels throughout the
western Medusae Fossae Formation, Mars (abstract), Proc. Lunar Planet.
Sci. Conf., 36, 1367.
Cas, R. A. F., and J. V. Wright (1992), Volcanic Successions, Modern and
Ancient: A Geological Approach to Processes, Products and Successions,
Chapman Hall, London.
Christensen, P., N. Gorelick, S. Anwar, S. Dickenshied, C. Edwards, and
E. Engle (2007), New insights about Mars from the creation and ana-
lysis of Mars global data sets, Eos Trans. AGU, 88(52), Fall Meet.
Suppl., Abstract P11E?01.
Criswell, C. W., and W. E. Elston (1981), Morphological characteristics of
terrestrial ash-flow tuff (ignimbrite) and related pyroclastic deposits,
Tech. Mem. TM-84211, NASA, Washington, D. C.
Crown, D. A., and R. Greeley (1993), Volcanic geology of Hadriaca Patera
and the eastern Hellas region of Mars, J. Geophys. Res., 98, 3431?3451.
Crown, D. A., R. Greeley, M. F. Sheridan, and R. Carrasco (1989), Analysis
of an ignimbrite plateau in the Central Andes using LANDSAT thematic
mapper data: Implications for the identification of ash deposits on Mars
(abstract), Proc. Lunar Planet. Sci. Conf., 20, 206?207.
de Silva, S. L. (1989), Geochronology and stratigraphy of the ignimbrites
from the 21
300S to 23
300S portion of the central Andes of northern
Chile, J. Volcanol. Geotherm. Res., 37(2), 93?191, doi:10.1016/0377-
0273(89)90065-6.
de Silva, S. L., G. Zandt, R. Trumbull, J. Viramonte, G. Salas, and
N. Jimenez (2006), Large-scale silicic volcanism in the Central Andes?
A tectonomagmatic phenomenon, Geol. Soc. Spec. Publ., 269, 47?64.
Edgett, K. S., B. J. Butler, J. R. Zimbelman, and V. E. Hamilton (1997),
Geologic context of the Mars radar ??Stealth?? region in southwestern
Tharsis, J. Geophys. Res., 102, 21,545?21,567, doi:10.1029/97JE01685.
Forsythe, R. D., and J. R. Zimbelman (1988), Is the Gordii Dorsum escarp-
ment on Mars an exhumed transcurrent fault?, Nature, 336(6195),
143?146, doi:10.1038/336143a0.
Francis, P.W., and C. A.Wood (1982), Absence of silicic volcanism onMars:
Implications for crustal composition and volatile abundance, J. Geophys.
Res., 87(B12), 9881?9889, doi:10.1029/JB087iB12p09881.
Freundt, A., and H.-U. Schmincke (1995), Eruption and emplacement of
basaltic welded ignimbrite during caldera formation on Gran Canaria,
Bull. Volcanol., 56, 640?659.
Frey, H. V., S. E. H. Sakimoto, and J. H. Roarke (1998), MOLA topography
and stratigraphy of geologic units at the dichotomy boundary: The Me-
dusae Fossae Formation (abstract), Proc. Lunar Planet. Sci. Conf., 29,
1619.
Giordano, G., A. A. De Benedetti, A. Diana, G. Diano, F. Gaudioso,
F. Marasco, M. Miceli, S. Mollo, R. A. F. Cas, and R. Funicello
(2006), The Colli Albani mafic cladera (Roma, Italy): Stratigraphy,
structure, and petrology, J. Volcanol. Geotherm. Res., 155, 49?80,
doi:10.1016/j.jvolgeores.2006.02.009.
Glaze, L. S., and S. M. Baloga (2002), Volcanic plume heights on Mars:
Limits of validity for convective models, J. Geophys. Res., 107(E10),
5086, doi:10.1029/2001JE001830.
Goudie, A. S. (2007), Mega-yardangs: A global analysis, Geogr. Compass,
1, 65?81.
Greeley, R., and D. A. Crown (1990), Volcanic geology of Tyrrhena Patera,
Mars, J. Geophys. Res., 95, 7133?7149.
Greeley, R., and J. E. Guest (1987), Geologic map of the eastern equatorial
region of Mars, U.S. Geol. Surv. Misc. Invest. Map, I-1802-B.
Gu?nster, N., P. Eck, A. Skowronek, and L. Zoller (2001), Late Pleistocene
loess and their paleosols in the Granada Basin, Southern Spain, Quat.
Int., 76?77, 241?245, doi:10.1016/S1040-6182(00)00106-3.
Head, J. W., and M. A. Kreslavsky (2001), Medusae Fossae Formation as
volatile-rich sediments deposited during high obliquity: An hypothesis
and tests, paper presented at Conference on the Geophysical Detection
of Subsurface Water on Mars, Lunar and Planet. Inst., Houston, Tex.,
6?10 Aug.
Head, J. W., and M. A. Kreslavsky (2004), Medusae Fossae Formation: Ice-
rich airborne dust deposited during periods of high obliquity? (abstract),
Proc. Lunar Planet. Sci. Conf., 35, 1635.
Hynek, B.M., R. E. Arvidson, and R. J. Phillips (2002), Explosive volcanism
from Tharsis: Global evidence in the Martian geologic record (abstract),
Proc. Lunar Planet. Sci. Conf., 33, 1408.
Hynek, B. M., R. J. Phillips, and R. E. Arvidson (2003), Explosive volcan-
ism in the Tharsis region: Global evidence in the Martian geologic record,
J. Geophys. Res., 108(E9), 5111, doi:10.1029/2003JE002062.
Inbar, M., and C. Risso (2001), Holocene yardangs in volcanic terrains in
the southern Andes, Argentina, Earth Surf. Processes Landforms, 26,
657?666, doi:10.1002/esp.207.
Ivanov, M. A., A. S. Kozyrev, M. L. Litvak, and I. G. Mitrofanov (2005),
The Medusae Fossae-Elysium Region Mars: Geologic characteristics of
the depression of epithermal-neutron flux based on HEND measurements
onboard the Mars Odyssey spacecraft, Sol. Syst. Res., 39(1), 1?21.
Kadish, S. J., and N. G. Barlow (2006), Pedestal crater distribution and
implications for a new model of formation, (abstract), Proc. Lunar Pla-
net. Sci. Conf., 37, 1254.
Keller, J. M., W. V. Boynton, R. M. S. Williams, S. Karunatillake, and GRS
Science Team (2006), Analysis of layering at Mars near-surface using
attenuation of Chlorine gamma rays, (abstract), Lunar Planet. Sci. Conf.,
37, 1184.
Lanagan, P. B., and A. S. McEwen (2003), Cerberus Plains volcanism:
Constraints on temporal emplacement of the youngest flood lavas on
Mars, paper presented at 6th International Conference on Mars, Lunar
and Planet. Inst., Houston, Tex.
Malin, M. C. (1999), Oceans or seas in the Martian northern lowlands;
high-resolution imaging tests of proposed coastlines, Geophys. Res. Lett.,
26(19), 3049?3052, doi:10.1029/1999GL002342.
Malin, M. C., and K. S. Edgett (2000), Sedimentary rocks of early Mars,
Science, 290(5498), 1927?1937, doi:10.1126/science.290.5498.1927.
Malin, M. C., et al. (1998), Early views of the Martian surface from the
Mars Orbiter Camera of Mars Global Surveyor, Science, 279(5357),
1681?1685, doi:10.1126/science.279.5357.1681.
McCauley, J. F. (1973), Mariner 9 evidence for wind erosion in the equatorial
and mid-latitude regions of Mars, J. Geophys. Res., 78, 4123?4137,
doi:10.1029/JB078i020p04123.
McCauley, J. F., M. J. Grolier, and C. S. Breed (1977a), Yardangs of Peru
and Other Desert Regions (Interagency Report: Astrogeology), 177 pp.,
U.S. Geol. Surv., Reston, Va.
McCauley, J. F., C. S. Breed, and M. J. Grolier (1977b), Yardangs, in Geo-
morphology in Arid Regions, edited by D. O. Doehring, pp. 233?269,
Allen and Unwin, Boston.
McColley, S. M., J. W. Head, and G. Neukum (2005), The Medusae Fossae
Formation: Geological characteristics and topographic and stratigraphic
relationships of the lower member along southeastern Elysium Planitia
(abstract), Proc. Lunar Planet. Sci. Conf., 36, 1184.
Mouginis-Mark, P. (1993), The influence of oceans on Martian volcanism
(abstract), Proc. Lunar Planet. Sci. Conf., 24, 1021?1022.
Nussbaumer, J. (2005), Extent and further characteristics of former gla-
ciated terrain in Elysium Planitia, Mars (abstract), Proc. Lunar Planet.
Sci. Conf., 36, 1949.
Nussbaumer, J., R. Jaumann, and E. Hauber (2003), Evidence for a surging
ice sheet in Elysium Planitia, Mars, paper presented at 6th International
Conference on Mars, Lunar and Planet. Inst., Houston, Tex.
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
14 of 15
E12011
Parker, T. J. (1991), A comparison of the Martian Medusae Fossae Forma-
tion with terrestrial carbonate platforms (abstract), Proc. Lunar Planet.
Sci. Conf., 22, 1029?1030.
People?s Republic of China (1982), 30 Years? Review of China?s Science
and Technology, 1949?79, World Sci., Hackensack, N. J.
Roever, W. P. D. E. (1966), Dacitic ignimbrites with upward increasing
compactness near Sibolangit (NE Sumatra, Indonesia) and their peculiar
hydrology, Bull. Volcanol., 29, 105?122, doi:10.1007/BF02597147.
Ross, C. S., and R. L. Smith (1961), Ash-flow tuffs: Their origin,
geologic relations, and identification, U.S. Geol. Surv. Prof. Pap.,
366, 1?81.
Sakimoto, S. E. H., H. V. Frey, J. B. Garvin, and J. H. Roark (1999),
Topography, roughness, layering, and slope properties of the Medusae
Fossae Formation from Mars Orbiter Laser Altimeter (MOLA) and
Mars Orbiter Camera (MOC) data, J. Geophys. Res., 104(E10),
24,141?24,154.
Schultz, P. H., and A. B. Lutz (1988), Polar wandering on Mars, Icarus, 73,
91?141, doi:10.1016/0019-1035(88)90087-5.
Scott, D. H., and K. L. Tanaka (1982), Ignimbrites of Amazonis Planitia
region of Mars, J. Geophys. Res., 87, 1179 ? 1190, doi:10.1029/
JB087iB02p01179.
Scott, D. H., and K. L. Tanaka (1986), Geologic map of the western equa-
torial region of Mars, U.S. Geol. Surv. Misc. Invest. Map, I-1802-A.
Selby, M. J., P. Augustinus, V. G. Moon, and R. J. Stevenson (1988), Slopes
on strong rock masses: Modeling and influences of stress distribution and
geomechanical properties, in Modelling Geomorphological Systems,
edited by M. G. Anderson, pp. 341?374, John Wiley, Hoboken, N. J.
Smith, D. E., et al. (1999), The global topography of Mars and implications
for surface evolution, Science, 284, 1495?1503, doi:10.1126/science.
284.5419.1495.
Takagi, M., and J. R. Zimbelman (2001), Geomorphic mapping and ana-
lysis of the eastern Medusae Fossae Formation, Mars (abstract), Proc.
Lunar Planet. Sci. Conf., 32, 1579.
Tanaka, K. L. (2000), Dust and ice deposition in the Martian geologic
record, Icarus, 144, 254?266, doi:10.1006/icar.1999.6297.
Tanaka, K. L., and G. J. Leonard (1998), Martian paleopolar deposits:
Evidence for a stable pole and periods of climate change, paper presented
at 1st International Conference on Mars Polar Science and Exploration,
Lunar and Planet. Inst., Camp Allen, Tex.
Tramp, K. L., G. S. Soreghan, and R. D. Elmore (2004), Paleoclimatic
inferences from paleopedology and magnetism of the Permian Maroon
Formation loessite, Colorado, USA, Geol. Soc. Am. Bull., 116, 671?686,
doi:10.1130/B25354.1.
Ward, A. W. (1979), Yardangs on Mars: Evidence of recent wind erosion,
J. Geophys. Res., 84, 8147?8166, doi:10.1029/JB084iB14p08147.
Ward, A. W., and R. Greeley (1984), Evolution of yardangs at Roger Lake,
California, Geol. Soc. Am. Bull., 95, 829 ? 837, doi:10.1130/0016-
7606(1984)95<829:EOTYAR>2.0.CO;2.
Watters, T. R., et al. (2007), Radar sounding of the Medusae Fossae For-
mation, Mars: Equatorial ice, or dry, low-density deposits, Science, 318,
1125?1128.
Wilson, L., and J. W. Head (1994), Mars: Review and analysis of volcanic
eruption theory and relationships to observed landforms, Rev. Geophys.,
32, 221?264.
Wyrick, D., D. A. Ferrill, A. P. Morris, S. L. Colton, and D. W. Sims
(2004), Distribution, morphology, and origins of Marian pit crater chains,
J. Geophys. Res., 109, E06005, doi:10.1029/2004JE002240.
Zimbelman, J. R. (2003), Evidence supporting an ignimbrite origin for the
Medusae Fossae Formation, Mars, Geol. Soc. Am. Abstr. Programs, 35,
265.
Zimbelman, J. R., and S. D. Patel (1998), Photogeologic constraints on the
emplacement of the Medusae Fossae Formation, Mars (abstract), Lunar
Planet. Sci. Conf., 29, 1085.
Zimbelman, J. R., D. Crown, and D. Jenson (1996), Initial investigation of
the enigmatic massive deposits in Amazonis Planitia, Mars (abstract),
Proc. Lunar Planet. Sci. Conf., 27, 1495.
Zimbelman, J. R., D. A. Crown, J. A. Grant, and D. M. Hooper (1997a),
The Medusae Fossae Formation, Amazonis Planitia, Mars: Evaluation of
proposed hypotheses of origin (abstract), Proc. Lunar Planet. Sci. Conf.,
28, 1482.
Zimbelman, J. R., A. K. Johnston, P. S. Russell, and C. G. Lovett (1997b),
Regional geologic setting of the Medusae Fossae Formation on Mars, Eos
Trans. AGU, 78(46), 411.
Zimbelman, J. R., S. E. H. Sakimoto, and H. Frey (2000), Evidence for a
fluvial contribution to the complex story of the Medusae Fossae Forma-
tion on Mars, Geol. Soc. Am. Abstr. Programs, 32, 303.
Zimbelman, J. R., K. C. Bender, and J. C. Harris (2003), Geologic mapping
applications using THEMIS data for the Medusae Fossae Formation,
Mars (abstract), Proc. Lunar Planet. Sci. Conf., 34, 1390.























D. A. Crown, Planetary Science Institute, 1700 East Fort Lowell Road,
Suite 106, Tucson, AZ 85719, USA.
S. L. de Silva, Department of Geosciences, Oregon State University, 104
Wilkinson Hall, Corvallis, OR 97331, USA.
K. E. Mandt, Space Science and Engineering Division, Southwest
Research Institute, 6220 Culebra Road, P.O. Drawer 28510, San Antonio,
TX 78228, USA. (kathymandt@yahoo.com)
J. R. Zimbelman, Center for Earth and Planetary Sciences, National Air
and Space Museum, Smithsonian Institution, P.O. Box 37012, MRC 315,
Washington, DC 20013, USA.
E12011 MANDT ET AL.: MFF ORIGIN?A SYNOPTIC APPROACH
15 of 15
E12011