Localization and physical properties experiments conducted by Spirit at Gusev Crater.

Abstract

The precise location and relative elevation of Spirit during its traverses from the Columbia Memorial station to Bonneville crater were determined with bundle-adjusted retrievals from rover wheel turns, suspension and tilt angles, and overlapping images. Physical properties experiments show a decrease of 0.2% per Mars solar day in solar cell output resulting from deposition of airborne dust, cohesive soil-like deposits in plains and hollows, bright and dark rock coatings, and relatively weak volcanic rocks of basaltic composition. Volcanic, impact, aeolian, and water-related processes produced the encountered landforms and materials.

Full text

Page 1

DOI: 10.1126/science.1099922
, 821 (2004);
305
Science
et al. R. E. Arvidson,
Conducted by Spirit at Gusev Crater
Localization and Physical Properties Experiments

www.sciencemag.org (this information is current as of June 11, 2007 ):
The following resources related to this article are available online at
http://www.sciencemag.org/cgi/content/full/305/5685/821
version of this article at:
including high-resolution figures, can be found in the online
Updated information and services,
found at:
http://www.sciencemag.org/cgi/content/full/305/5685/821#related-content
, 7 of which can be accessed for free:
cites 10 articles
This article
can be
related to this article
A list of selected additional articles on the Science Web sites

http://www.sciencemag.org/cgi/content/full/305/5685/821#otherarticles
41 article(s) on the ISI Web of Science.
cited by
This article has been
http://www.sciencemag.org/cgi/content/full/305/5685/821#otherarticles
11 articles hosted by HighWire Press; see:
cited by
This article has been
http://www.sciencemag.org/cgi/collection/planet_sci
Planetary Science
:
subject collections
This article appears in the following
http://www.sciencemag.org/about/permissions.dtl
in whole or in part can be found at:
this article
permission to reproduce
of this article or about obtaining
reprints
Information about obtaining
registered trademark of AAAS.
c 2004 by the American Association for the Advancement of Science; all rights reserved. The title SCIENCE is a
CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005.
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
on June 11, 2007
www.sciencemag.org
Downloaded from

Page 2

dark parts of the two-toned rocks, the height
of the largest of the bright (light-toned) rocks,
and the perched rocks would suggest local
deflation of 5 to 60 cm. Thus, there must have
been previous deposition on this order.
References and Notes
1. Reviewed by R. Greeley, N. Lancaster, S. Lee, P. Thom-
as, in Mars, H. H. Kieffer, B. Jakosky, Eds. (Univ. of
Arizona Press, Tucson, AZ, 1992), pp. 730–766.
2. S. W. Squyres et al., Science 305, 794 (2004).
3. J. A. Crisp et al., J. Geophys. Res. 108, 10.1029/
2002JE002038 (2003).
4. M. P. Golombek et al., J. Geophys. Res. 108, 10.1029/
2003JE002074 (2003).
5. M. P. Golombek et al., Lunar Planet. Sci. XXXV, abstr.
2185 (2004).
6. At 0.19 (36), the landing site has the lowest albedo as
determined from orbit of the entire landing ellipse.
From analysis of broadband (0.4 to 1.0 ?m) images
from the MER Pancam; the average albedo measured
from the surface is 0.25 ? 0.05 (10).
7. A martian solar day has a mean period of 24 hours 39
min 35.244 s and is referred to as a sol to distinguish
this from a roughly 3% shorter solar day on Earth. A
martian sidereal day, as measured with respect to the
fixed stars, is 24 hours 37 min 22.663 s, as compared
with 23 hours 56 min 04.0905 s for Earth. See www.
giss.nasa.gov/tools/mars24/ for more information.
8. “Dust” is an imprecise term but is commonly used
for material on Earth that is smaller than about 40
?m and that is transported primarily in suspension.
Estimates for dust on Mars suggest diameters of a
few micrometers, although these particles are
thought to stick together in some cases, possibly
by electrostatic charges, to form aggregates of
larger sizes (37).
9. R. E. Arvidson et al., Science 305, 821 (2004).
10. Names have been assigned to areographic features
by the Mars Exploration Rover (MER) team for plan-
ning and operations purposes. The names are not
formally recognized by the International Astronomi-
cal Union.
11. J. F. Bell III et al., Science 305, 800 (2004).
12. K. E. Herkenhoff et al., J. Geophys. Res. 108, 10.1029/
2003JE002076 (2003).
13. J. B. Pollack et al., J. Geophys. Res. 82, 4479 (1977).
14. P. H. Smith et al., Science 278, 1758 (1997).
15. M. G. Tomasko, L. R. Doose, M. Lemmon, P. H. Smith,
E. Wegryn, J. Geophys. Res. 104, 8987 (1999).
16. P. R. Christensen et al., Science 305, 837 (2004).
17. H. Y. McSween et al., Science 305, 842 (2004).
18. K. E. Herkenhoff et al., Science 305, 824 (2004).
19. J. F. Bell et al., Icarus 158, 56 (2002).
20. J. L. Bandfield, J. Geophys. Res. 107, 10.1029/
2001JE001510 (2002).
21. R. P. Sharp, J. Geol. 71, 617 (1963).
22. Ripples are bedforms composed of sand and granules
that are moved by surface creep induced by the
impact of saltating sands, whereas dunes are larger
bedforms typically composed of finer sands and not
formed directly by saltation processes.
23. R. Greeley et al., J. Geophys. Res. 104, 8573 (1999).
24. M. P. Golombek, N. T. Bridges, J. Geophys. Res. 105,
1841 (2000).
25. J. A. Grant et al., Science 305, 807 (2004).
26. R. Greeley, J. D. Iversen, Wind as a Geological Process:
Earth, Mars, Venus, and Titan (Cambridge Univ. Press,
Cambridge, 1985).
27. R. Greeley, J. D. Iversen, Geophys. Res. Lett. 14, 925
(1987).
28. M. R. Raupach, Boundary-Layer Meterol. 60, 375 (1992).
29. M. R. Raupach, D. A. Gillette, J. F. Leyes, J. Geophys.
Res. 98, 3023 (1993).
30. N. T. Bridges et al., J. Geophys. Res. 104, 8595 (1999).
31. R. Greeley et al., J. Geophys. Res. 87, 10009 (1992).
32. N. T. Bridges et al., Planet. Space Sci. 52, 199 (2004).
33. S. P. Gorevan et al., J. Geophys. Res. 108, 10.1029/
2003JE002061 (2003).
34. S. R. C. Rafkin, T. I. Michaels, J. Geophys. Res. 108,
10.1029/2002JE002027 (2003).
35. D. Toigo, M. I. Richardson, J. Geophys. Res. 108,
10.1029/2003JE002064 (2003).
36. R. Greeley et al., J. Geophys. Res. 108, 10.1029/
2002JE002006 (2003).
37. M. T. Mellon, B. M. Jakosky, H. H. Kieffer, P. R.
Christensen, Icarus 148, 437 (2000).
38. R. Greeley, J. Geophys. Res. 84, 6248 (1979).
39. Our work was supported by NASA by contracts
through the Jet Propulsion Laboratory.
10 May 2004; accepted 23 June 2004
R E P O R T
Localization and Physical Properties Experiments
Conducted by Spirit at Gusev Crater
R. E. Arvidson,1R. C. Anderson,2P. Bartlett,3J. F. Bell III,4D. Blaney,2P. R. Christensen,5P. Chu,3L. Crumpler,6
K. Davis,3B. L. Ehlmann,1R. Fergason,5M. P. Golombek,2S. Gorevan,3J. A. Grant,7R. Greeley,5E. A. Guinness,1
A. F. C. Haldemann,2K. Herkenhoff,8J. Johnson,8G. Landis,9R. Li,10R. Lindemann,2H. McSween,11D. W. Ming,12
T. Myrick,3L. Richter,13F. P. Seelos IV,1S. W. Squyres,4R. J. Sullivan,4A. Wang,1J. Wilson3
The precise location and relative elevation of Spirit during its traverses from the
Columbia Memorial station to Bonneville crater were determined with bundle-
adjusted retrievals from rover wheel turns, suspension and tilt angles, and overlap-
ping images. Physical properties experiments show a decrease of 0.2% per Mars solar
day in solar cell output resulting from deposition of airborne dust, cohesive soil-like
deposits in plains and hollows, bright and dark rock coatings, and relatively weak
volcanic rocks of basaltic composition. Volcanic, impact, aeolian, and water-related
processes produced the encountered landforms and materials.
During the first few Mars solar days (sols) (1)
of operations, we determined the landed loca-
tion in inertial coordinates by analyzing Spirit-
to-Earth two-way X-band Doppler transmis-
sions and two passes of ultrahigh-frequency
two-way Doppler between Spirit and the Mars
Odyssey orbiter. The equivalent location in the
International Astronomical Union (IAU) 2000
body-centered reference frame is 14.571892°S,
175.47848°E. The location with respect to sur-
face features was derived by the correlation of
hills and craters observed in images taken by
the Pancam, the Entry Descent and Landing
(EDL) Camera, and the Mars Orbital Camera.
On the basis of these analyses, the landing site
is located at 14.5692°S, 175.4729°E in IAU
2000 coordinates, ?300 m north-northwest of
the radiometric solution. This offset is consis-
tent with the map tie errors between inertially
derived coordinate systems and those derived
from image-based coverage of the planet.
Localization experiments during traverses
focused on systematic acquisition of forward-
and backward-looking overlapping images, on-
board inertial measurement unit (IMU) obser-
vations to derive rover tilt, and tracking the
number of wheel turns to provide wheel-based
odometry. These observations were employed
inaleast-squaresbundleadjustmenttosolvefor
the position and orientation of Spirit in local
Cartesian coordinates at discrete locations dur-
ing traverses (Fig. 1 and Plate 14). In addition,
measurements of differential rocker and bogie
angles in the suspension system, together with
IMU data, were used to reconstruct the eleva-
tion of each wheel at a 2- to 8-Hz sample rate
relative to the start of each traverse (Fig. 1).
Localization results were extracted for 33
traverse segments from the Columbia Memori-
1Department of Earth and Planetary Sciences, Washington
University, St. Louis, MO 63130, USA.2Jet Propulsion Lab-
oratory, California Institute of Technology, Pasadena, CA
91109, USA.3Honeybee Robotics, 204 Elizabeth Street,
New York, NY 10012, USA.4Department of Astronomy,
Space Sciences Building, Cornell University, Ithaca, NY
14853, USA.5Department of Geological Sciences, Arizona
State University, Tempe, AZ 85287, USA.6New Mexico
MuseumofNaturalHistoryandScience,Albuquerque,NM
87104, USA.7Center for Earth and Planetary Studies, Na-
tional Air and Space Museum, Smithsonian Institution,
Washington,DC20560,USA.8U.S.GeologicalSurvey,Flag-
staff, AZ 86001, USA.9National Aeronautics and Space
Administration (NASA) Glenn Research Center, Cleveland,
OH 44135, USA.10Department of Civil and Environmental
Engineering and Geodetic Science, Ohio State University,
Columbus, OH 43210, USA.11Department of Earth and
Planetary Sciences, University of Tennessee, Knoxville, TN
37996, USA.12NASA Johnson Space Center, Houston, TX
77058, USA.13Deutsche Luft und Raumfahrt Institut fu ¨r
Raumsimulation, Linder Hoehe, Koln, DJ-51170, Germany.
S P I R I TA T G U S E V C R A T E R
www.sciencemag.org SCIENCEVOL 3056 AUGUST 2004
821
SPECIAL SECTION
on June 11, 2007
www.sciencemag.org
Downloaded from

Page 3

al station to the rock Route 66 (2) (Fig. 1). The
terrain from the landing site to the location of
therockHumphreyinMiddleGroundhollowis
flat and level, with a relative-elevation decrease
of only 1 m over the horizontal distance of
197 m between localization stations. From
Humphrey to the crater rim, the relative eleva-
tion increased 7.4 m over a distance of 135 m
between stations, consistent with the increased
slope expected as the rim of Bonneville crater
wasapproached.Fromthecraterrimtotherock
Mazatzal,therelativeelevationdecreased0.6m
over a distance of 80 m, consistent with the
traverse azimuths that kept the vehicle near the
rim. Finally, the distance from Mazatzal to
Route 66 is 95 m, with a decrease in relative
elevation of 3 m, consistent with travel away
from the rim. Based on wheel odometry, Spirit
traversed 637 m from Columbia Memorial sta-
tion to the rock Route 66 and the summed
distance between the 33 stations is 507 m.
Notably, the bundle adjustments increased po-
sitional accuracy by 2% relative to the use of
wheel odometry alone.
Rocks litter the landscape at the landing site
and are larger and more abundant near the
Bonneville crater rim (3). Soil (4) deposits
dominated by particles ?1 mm in diameter
cover the plains and occasionally are found as
aeolian drifts (3, 5, 6). Soil deposits not within
hollows or on hollow rims are covered with an
evenly spaced set of angular to smooth rocks
ranging in size from granules (2 to 4 mm) to
pebbles (4 to 16 mm). Furthermore, soils are
covered with a thin (?1-mm) layer of bright
red dust. These dust deposits are easily dis-
turbed, as shown by airbag bounce marks gen-
erated during landing and observations that
wheel motions associated with traversing re-
moved the dust covers and exposed darker,
underlying deposits (Fig. 1). The short-circuit
current monitor solar cell showed a decrease in
current of 0.2% per sol (corrected for seasonal
variations in Mars-Sun distance and solar ele-
vation angle) over the first 93 sols of operation,
showing airborne dust accumulation on the rov-
er solar panels comparable to that observed
during the Pathfinder mission (7). The ubiqui-
tous nature of the dust cover on soils, combined
with discernable dust accumulation rates,
serves as evidence that the Gusev site has ac-
cumulated dust deposits for a number of years.
Soil deposits typically display surface crusts
a few millimeters thick beneath the thin dust
covers. For example, airbag retraction scars
include thin crustal plates a few centimeters in
width that were laterally displaced as the deflat-
ed airbags were retracted after landing. Surface
crusts also have been observed in wheel track
disturbances, especially where rocks wobbled
by wheels created moats as small amounts of
adjacent crust were displaced. Imaging of
wheel tracks shows well-defined soil casts in-
dicative of materials with a range of grain sizes,
fromcoarsesand(0.5to1mm)todiameterstoo
fine to be discerned with the 30 ?m/pixel spa-
tial resolution of the Microscopic Imager (MI)
(Fig. 2A). Imaging data also show that smooth
indentations were produced by the Mo ¨ssbauer
Spectrometer contact plate as it pushed into
soils before the contact switch initiated, stop-
ping movement at ?1 N of applied force (Fig.
2A). The finer grains were molded into inter-
stices of the larger grains when displaced
by the wheels and faceplate, producing
well-defined soil casts. Unfortunately, no
in situ measurements have been done yet on
cloddy soils that are dust free. Thus, it is
impossible to determine whether the cloddy
nature is due to enrichment of cementing
agents such as iron oxides or sulfates.
To explore the degree to which soil proper-
tieschangewithdepth,a6-to7-cm-deeptrench
was excavated with the right front wheel in
Laguna hollow on sol 47. Trench walls in La-
guna hollow have slopes of up to 65°, values
steeper than the angle of repose for most gran-
ular materials (?30°). This result indicates the
presence of slightly cohesive soils throughout
the upper 6 to 7 cm beneath the surface, al-
though the exact degree of cohesion is difficult
to estimate with certainty. An additional exper-
imentexcavatedintothedriftSerpent(Fig.2B).
The surface layer of the drift consists of closely
packed, very coarse sand-sized particles (?1 to
2 mm in diameter) overlying finer grained soils.
The very coarse sand armors the surface and
protects the drift from aeolian erosion and
transport of the underlying particles. The armor
is covered with the thin dust deposit that is
typical of the soil surfaces examined during
Spirit’s traverses and observations. The slightly
cohesive nature of the upper portion of the drift
soil is evident based on the presence of clods
produced by the excavations (Fig. 2).
Fig. 1. (A) MER EDL Camera image acquired from 1430 m above the
surface shows Spirit’s localization-derived traverses and positions for
physical property experiments during the first 93 sols in Gusev crater.
Tick marks along traverse path show localization stations. Some
stations that are close together are not plotted to avoid clutter.
Image identification is 2E126462405EDN0000F0006N0M1. (B) Nav-
cam image (2N130812149EFF1000P1901L0M1) of Laguna hollow
with tracks that are about 10 cm wide. The inset shows a vertical
topographic profile across the hollow as derived from rover tilt and
wheel suspension data.
S P I R I TA T G U S E V C R A T E R
6 AUGUST 2004VOL 305SCIENCEwww.sciencemag.org
822
SPECIAL SECTION
on June 11, 2007
www.sciencemag.org
Downloaded from

Page 4

With regard to standard terramechanical
analyses, estimates of soil-bearing strength of
?5 kPa, cohesive strength of ?1 kPa, and an
angle of internal friction of ?20° were derived
from wheel sinkage depths in a hollow adjacent
to the lander traversed on sol 15 (8). More
resistantdeposits(withlesssinkage),suchasthose
foundintheegressareatraversedonsol12andin
theregioninfrontofthedriftArenacrossedonsol
42, provided values of ?200 kPa, 15 kPa, and
25°, respectively, for the same parameters. These
values are comparable to the range of soil values
found from analyses of Viking Lander and Path-
finder physical properties data (9).
Rocks ranging up to boulder sizes were
observed during Spirit’s traverses (Fig. 1).
Rocks tend to be massive in appearance, al-
though a few rocks, including Mimi (Fig. 2A),
have a layered appearance. The layered rocks
are interpreted to be weathered in place by
mechanical spalling along lines of weakness.
Low-lying rocks (less than ?25 cm tall) of the
more typical massive variety in many cases
appear to have been faceted by wind action and
tend to be light-toned relative to surrounding
soils, e.g., Mazatzal rock on the rim of Bonne-
ville crater (Fig. 3). Additionally, these rocks
are commonly embedded in soil deposits. All
rocks show strong variations in brightness with
lighting and viewing geometries in imaging
data. At small phase angles, flat surfaces of indi-
vidual rocks tend to be brighter and redder in
visiblewavelengthsrelativetocornersoredgesor
tops,consistentwiththepresenceofbrightercoat-
ings that are optically thicker on the facets as
opposed to the corners or edges. These observa-
tions imply that rocks are coated with bright, red
materialspreferentiallyonsmooth,flatfacets.Ac-
cumulation of airborne dust is a plausible mecha-
nism for formation of these bright coatings.
Rock Abrasion Tool (RAT) deployments
were made on Adirondack (?20-cm-high light-
toned rock), Humphrey (?50-cm-high rock,
variable brightness), and Mazatzal (?20-cm-
high, faceted, light-toned rock) (Figs. 1 and 3).
After initial brushing experiments designed to
remove loose deposits, the RAT was used to
grind into the rock facets, extending 2.7 and 2.1
mm (based on RAT motor drive data) into
Adirondack and Humphrey, respectively. For
Mazatzal, two rattings were conducted with a
Fig. 2. (A) Pancam
image(2P130088011-
EFF0514P2538L2M1)
of Mimi rock and rov-
er front wheel tracks,
along with an MI (2M-
130169106EFF0514-
P2953M2M1) inset
showing the tracks in
more detail. The MI
inset shows the im-
print of the MB con-
tact plate (arrows)
into moldable soil.
The inset
about 3 cm across. (B)
Pancam image (2P13-
2756681EFF1957P23-
52L2M1) of Serpent
drift after scuffing by
the rover front wheel,
with MI (2M132842-
058EFF2000P2977M-
2M1) inset showing
dust-coated armor
of very coarse sand
grains over sand-sized and finer-grained particles. Soil clods are evident just to the upper right of the inset location. The MI inset covers about 1.5 cm
in width on the drift.
covers
Fig. 3. (A) Navcam
image (2N13400813-
9EFF2238P1959L0M1)
showing the brushed
(Brush) and abraded
(Grind) areas on the
2.3-m-wide Mazatzal
rock. (B) Mosaic (2M-
P085IOF22ORT32P2-
959L456F1_qn)
four MI images ac-
quired of the Mazat-
zal RAT hole after
the second grinding
operation, showing
that only a small
amount of dark coat-
ing remains. The RAT
hole is about 4.5 cm
across. Black and
whiteversion
merged color Pan-
cam and MI mosaic
is shown.
of
of
S P I R I TA T G U S E V C R A T E R
www.sciencemag.orgSCIENCE VOL 305 6 AUGUST 2004
823
SPECIAL SECTION
on June 11, 2007
www.sciencemag.org
Downloaded from

Page 5

3.8-mm penetration on the first attempt and a
4.1-mm penetration on the second attempt. The
brushing and grinding operations on Mazatzal
showed the presence of a dark, smooth, indu-
ratedcoatingbeneaththelight-tonedloosecoat-
ing (Fig. 3). This dark coating was largely
removed by the second grinding, revealing the
underlying rock surface (Fig. 3B). There is also
someevidencethatasecondbrightcoatingmay
underlie the dark coating (10, 11).
Grind motor currents, the depths achieved,
and grinding areas provided estimates of the
amount of energy consumed by the RAT while
removing a unit volume of material. Because
grind energy density is a nontraditional means
of quantifying rock mechanical properties,
three terrestrial rocks were abraded in the lab-
oratory for calibration, with the use of a flight-
like RAT: a fresh, nonvesicular basalt sample
from Ash Fork, Arizona; a fine-grained dolos-
tone sample collected from the Soda Mountains
north of Silver Lake, California; and Cleveland
Member (fissle shale) of the Ohio Formation.
These experiments yielded energy densities of
?166, 83, and 11 J/mm3, respectively. For
comparison, Humphrey required 83 J/mm3,
whereas Adirondack and Mazatzal required 51
and 65 J/mm3, respectively. Thus, rocks abrad-
ed by Spirit required less energy per volume
than the particular basalt sample ground in
the tests, and comparable grind energy den-
sities to the two terrestrial sedimentary rock
samples, even though composition and min-
eralogy data from Spirit demonstrate that the
rocks encountered are basalts (10, 12–14).
The physical properties experiments con-
ducted by Spirit at Gusev crater show that
surface soils are cloddy, rock coatings are ubiq-
uitous, and rocks are easily abraded and thus
mechanically weaker than the fresh, nonvesicu-
lar basaltic sample used in the laboratory tests.
Furthermore, the abraded surfaces of the rocks
at Gusev exposed vugs and cracks filled with
bright material suggestive of aqueous mineral-
ization (Fig. 3) (10, 11). There is also a sugges-
tion of a vertical weathering profile for Hum-
phrey rock, where the grinding direction was
accomplished at a slight angle from the surface
normal, thereby exposing shallow to deep
surfaces (10). The presence of liquid water,
even for brief periods of time, is one way to
cement surface soils, form rock coatings,
deposit minerals in vugs and cracks, and
weather the surfaces of rocks. Liquid water
might occur for brief periods when the spin
axis obliquity and atmospheric relative hu-
midity are high and precipitation occurs as
snow or frost (15). A modest temperature
enhancement as a result of absorption of
solar radiation by underlying regolith and
rocks, with “greenhouse”-related absorp-
tion of outgoing thermal radiation by the
ice and snow, could generate thin films of
liquid water that would mobilize soluble
species and produce the features observed
by Spirit. Other models are also being ex-
plored to place the physical properties ex-
periments in an environmental context, in
addition to further measurements designed
to test hypotheses.
References and Notes
1. A martian solar day has a mean period of 24 hours 39
min 35.244 s and is referred to as a sol to distinguish
this from a roughly 3% shorter solar day on Earth. A
martian sidereal day, as measured with respect to the
fixed stars, is 24 hours 37 min 22.663 s, as compared
with 23 hours 56 min 04.0905 s for Earth. See
www.giss.nasa.gov/tools/mars24 for more informa-
tion.
2. Names have been assigned to areographic features
by the Mars Exploration Rover (MER) team for plan-
ning and operations purposes. The names are not
formally recognized by the IAU.
3. J. A. Grant et al., Science 305, 807 (2004).
4. The term martian soil is used here to denote any
loose unconsolidated materials that can be distin-
guished from rocks, bedrock, or strongly cohesive
sediments. No implication of the presence or ab-
sence of organic materials or living matter is in-
tended.
5. P. R. Christensen et al., Science 305, 837 (2004).
6. R. Greeley et al., Science 305, 810 (2004).
7. G. Landis, P. Jenkins, J. Geophys. Res. 105, 1855
(2000).
8. Based on wheel track sinkage values for terrains
covered by one to three MER wheels described in L.
Richter and P. Hamacher, paper presented at the 13th
Conference of the International Society for Terrain-
Vehicle Systems, Munich, Germany, 14 to 17 Sep-
tember 1999.
9. H. Moore et al., J. Geophys. Res. 104, 8729 (1999).
10. K. E. Herkenhoff et al., Science 305, 824 (2004).
11. H. Y. McSween et al., Science 305, 842 (2004).
12. S. W. Squyres et al., Science 305, 794 (2004).
13. J. F. Bell III et al., Science 305, 800 (2004).
14. P. R. Christensen et al., Science 305, 837 (2004).
15. M. A. Mischna et al., J. Geophys. Res. 108, 5062
(2003).
16. Work funded by NASA through the Mars Exploration
Rover Project. We thank the MER team of scientists
and engineers, who made the landing, traverses, and
science observations a reality.
4 May 2004; accepted 2 July 2004
R E P O R T
Textures of the Soils and Rocks at Gusev Crater
from Spirit’s Microscopic Imager
K. E. Herkenhoff,1* S. W. Squyres,2R. Arvidson,3D. S. Bass,4J. F. Bell III,2P. Bertelsen,5N. A. Cabrol,6
L. Gaddis,1A. G. Hayes,2S. F. Hviid,7J. R. Johnson,1K. M. Kinch,8M. B. Madsen,5J. N. Maki,4
S. M. McLennan,9H. Y. McSween,10J. W. Rice Jr.,11M. Sims,12P. H. Smith,13L. A. Soderblom,1
N. Spanovich,13R. Sullivan,2A. Wang14
The Microscopic Imager on the Spirit rover analyzed the textures of the soil and rocks
at Gusev crater on Mars at a resolution of 100 micrometers. Weakly bound agglom-
erates of dust are present in the soil near the Columbia Memorial Station. Some of
the brushed or abraded rock surfaces show igneous textures and evidence for
alteration rinds, coatings, and veins consistent with secondary mineralization. The
rock textures are consistent with a volcanic origin and subsequent alteration and/or
weathering by impact events, wind, and possibly water.
The Microscopic Imager (MI) is a fixed-focus
camera mounted on a robotic arm (1, 2). The MI
was designed to function like a geologist’s hand
lens, acquiring images at aspatial resolution of 31
?m/pixel (picture element) over a broad spectral
range (400 to 700 nm). The MI uses the same
electronics design as the other Mars Exploration
Rover(MER)cameras(3,4),butitsopticsyielda
field of view of 32 by 32 mm across a 1024- by
1024-pixel charge-coupled device image. The MI
acquires images with only solar or skylight illu-
mination of the target surface. A contact sensor is
used to place the MI slightly closer to the target
surface than its best-focus distance of about 66
mm,whichallowsconcavesurfacestobeimaged
in good focus. The depth of field of the MI is ?3
mm; coarse focusing (?2-mm precision) is
achieved by moving the arm away from a rock
target after contact is sensed. The MI optics are
protected from the martian environment by a re-
tractabledustcover.ThiscoverincludesaKapton
(polyimide film, DuPont, Wilmington, Delaware)
windowthatistintedorangetorestrictthespectral
bandpass to 500 to 700 nm, which allows crude
color information to be obtained by acquiring
images with the cover open and closed.
S P I R I TA T G U S E V C R A T E R
6 AUGUST 2004VOL 305SCIENCE www.sciencemag.org
824
SPECIAL SECTION
on June 11, 2007
www.sciencemag.org
Downloaded from