Rocky 7 prototype Mars rover field geology experiments: 1 ... - Wiley

JOURNAL OF GEOPHYSICAL

RESEARCH, VOL. 103, NO. El0, PAGES 22,671-22,688, SEPTEMBER 25, 1998

Rocky 7 prototype Mars rover field geologyexperiments 1. Lavic Lake and Sunshine Volcanic Field, California R. E. Arvidson,• C. Acton,2 D. Blaney,2 J. Bowman,1 S. Kim,2 G. Klingelh6fer,3 J. Marshall, 4 C. Niebur, • J. Plescia,s R. S. Saunders,2 and C. T. Ulmer 6 Abstract. Experimentswith the Rocky 7 rover were performed in the Mojave Desert to better understandhow to conductrover-based,long-distance(kilometers)geological traverseson Mars. The rover was equippedwith stereo imagingsystemsfor remote

sensing science andhazardavoidance and57FeM6ssbauer andnuclearmagnetic resonancespectrometersfor in situ determinationof mineralogyof unpreparedrock and soil surfaces.Laboratorydata were also obtainedusingthe spectrometersand an X ray diffraction(XRD)/XRF instrumentfor unpreparedsamplescollectedfrom the rover sites. Simulated orbital and descentimage data assembledfor the test siteswere found to be critical for assessing the geologicsetting,formulatinghypothesesto be testedwith rover observations,planningtraverses,locatingthe rover, and providinga regional contextfor interpretationof rover-basedobservations. Analysesof remote sensingand in situ observationsacquiredby the rover confirmedinferencesmade from orbital and simulated descentimagesthat the SunshineVolcanic Field is composedof basalt flows.Rover data confirmedthe idea that Lavic Lake is a rechargeplaya and that an alluvial fan composed of sedimentswith felsiccompositionshas progradedonto the playa. Rover-based discoveries include

the inference

that the basalt flows are mantled

with aeolian

sediment

and coveredwith a densepavementof varnishedbasaltcobbles.Resultsdemonstratethat the combinationof rover remote sensingand in situ analyticalobservationswill significantlyincreaseour understandingof Mars and provide key connectinglinks between orbital and descentdata and analysesof returned samples. challengesthat mustbuild on the highlysuccessful Pathfinder missionwith the Sojournerrover [e.g., Golombeket al., 1997; The Mars SurveyorProgram is embarkingon an ambitious Smithet al., 1997].A major differencebetweenthe Pathfinder suite of missions,with two launchesduring each energetically and future missions will be the need for the rover to leave the favorableopportunityover the next decade.The overall focus vicinity of the landing site and thus to operate without the is to understandcurrent and past climates,whether Mars de- benefit of the Lander imagingsystem. veloped prebiotic compounds,and whether or not life develTo help identify and solve key challengesassociatedwith oped and evolvedon or beneath the planet'ssurface.In par- long-distanceroving, field testsare being conductedwith the ticular, long-distance(severalto 10 km) roverswill be a key Jet PropulsionLaboratory'sprototype Mars rovers in joint elementof the program.For example,currentplansincludea efforts involving small scienceteams. The experimentsare roverwith the Athena sciencepayload,whichwill includehigh focusedin the Mojave Desert to testtraversecapabilities,measpatial resolutionstereo imaging and emissionspectroscopic surementsystems,samplingprocedures,and how to best use observationsfor remote sensingand an arm with analytical prototype Mars roversfor field-basedscience.The first of these instrumentation for in situ determination of soil and rock texexperimentswas conductedwith the Rocky 7 rover, a derivature, chemistry,and mineralogy[Squyres et al., 1998;Arvidson tive of the Sojournerrover capableof long-distancetraverses et al., 1998].Key soil samplesand rock coreswill be collected and cachedas part of the rover missions,and sampleswill be (Figure 1). The experimentwasconductedfrom May 23 to 20, returned either as part of those missionsor afterward. These 1997, in the Lavic Lake playa and SunshineVolcanic Field rover missionspresent major technologicaland operational areas on the 29 Palms Marine Base, California. During the teststhe rover was equippedwith instrumentationfor charac•Department of EarthandPlanetary Sciences, Washington Univer- terization of the sites,navigationand hazard avoidance,and determinationof rock and soilmineralogy(Table 1). Approxsity, St. Louis, Missouri. 2jetPropulsion Laboratory, CaliforniaInstituteof Technology, Pas- imately 500 m of traversingwas accomplished,and numerous 1.

Introduction

adena.

3Institute for NuclearPhysics, Darmstadt University of Technology, Darmstadt, Germany.

4Search for Extraterrestrial Intelligence Institute,MountainView, California.

SU.S.Geological Survey, Flagstaff, Arizona. 6UlmerSystems, Inc.,Pasadena, California. Copyright1998 by the American GeophysicalUnion. Paper number 98JE01768. 0148-0227/98/98JE-01768509.00

observations

were made on basalt flows with aeolian accretion

mantles and desert pavements,mud-crackedplaya surfaces with impact (bomb) craters,and an alluvialfan composedof sediments

derived

from andesitic

and dacitic source rocks.

The first purposeof this paper is to presentresultsfrom the firstfield tests.A secondpurposeis to uselessonslearnedfrom theseactivities,the Pathfindermission[Golombeket al., 1997], and Marsokhodrover field experiments[Christianet al., 1997] to developa strategyfor conductingrover-basedoperationson 22,671

22,672

ARVIDSON

ET AL.' PROTOTYPE MARS ROVER FIELD EXPERIMENTS

Figure 1. Rocky7 on the SunshineField flowswith mast-based nuclearmagneticresonance(NMR) spectrometer againstboulder conductingmeasurements.

Mars that maximize scientificreturns.The Marsokhod experiments are particularlyrelevant becausethis rover has also been used to test how to conduct long-distance science traverses,although the instrumentationand telescienceapproachare quite differentfrom thoseusedduringthe Rocky7 tests and from those that will be used on Mars. A prime conclusionof this paper is the critical importanceof orbital and descentdata setsfor selectinglandingsitesthat enablethe directtestingof geologicalhypotheses from rover observations (Figure 2). The paperbeginswith an overviewof the geologyof the field site. The importanceof orbital and descentdata is then illustrated, usingsurrogatedata collectedover the traversesites.

Hypothesesto be testedfrom roverobservations are developed from analysesof thesesimulatedorbital and descentdata sets. Sections5-7 documenthow rover remote sensingand in situ observationsallowedtestingand updatingof thesehypotheses for each of the major geologicunits traversed,togetherwith discoveriesthat would probablyhave come only from roverbased measurements.

Section 8 includes lessons learned

and a

synopsisof the strategyfor conductingrover operationson Mars. Becauseof the importanceof the knowledgeof rover locationrelative to other data setsand body-centeredcoordinate systems,the Appendix included illustrates the use of spacecraft,planet,instrument,C matrix,and events(SPICE) concepts[Acton,1996] appliedto rover observations.

Table 1. Summaryof Instrumentation Instrument Stereopanoramic imaging system

Key Parameters 5.7-mm focal length, f/2 optics,1/3inch CCD, 45ø field of view, and

Mass and Power 170 g, 1.2 W per camera

Field Use integratedonto mast and used to define terrain digital elevation modelfor waypointdefinitions terrain hazard avoidancesystem

Universe Kogaku lens model PT-51 Navigationimagingsystem 2-mm focal length,f/2 optics,1/3-inch 20 g excludingelectronicsin CCD, and 124ø field of view rover body, 1.2 W per camera located on front and back of rover 0.45 kg, 1.5 W integratedonto mast, connected M6ssbauerspectrometer S7Fespectrometer in backscatter via RS232 interface, and mode, Co/Rh source, Si-PIN diode detectors,and field of view -1.5 cm2 measurementsovernight 2-kg breadboard,0.3-kg mock up placed on mast and X ray diffractometer/X ray 9-W Fe X ray source(30/kV measurements done in deployabledevice fluorescencespectrometer operation)and Si-PIN diode detectors laboratory integratedonto mast, connected Nuclear magnetic scansradio frequenciesbetween43-61 0.10 kg, 0.2 W resonance

and 61-73 MHz to detect S7Fe-

to PC, and scansover 5-15,

spectrometer

bearingminerals

min

Si-PIN, silicon-positive intrinsicnegative.

ARVIDSON

ET AL.: PROTOTYPE

MARS ROVER

FIELD

EXPERIMENTS

22,673

Address overarching questions: *Warm

and wet?

*Prebioticcompounds? *Life?

Rover

operations and

analyses Orbital and

other data •iteselection I

i"""I•;•;'•'•'"'I• ....... i• Traverse planning

Imaging and •_• Remote Sensing Science

point spectra

............................................................. i I Intensive science, Insitu

i'::.......• sampling and caching

observations and !.

sampling i

'I

"'] Respond

!•s;;vWeries

'

........................ I........

ifunctions !:' •& ... Ireturn Sample Figure 2. Flow diagram for implementationof Mars rover missionthat beginswith overarchingscience questionsand proceedswith use of orbital, descent,and rover-basedobservationsto addressthe questions.

Alsoincluded is thenecessity of responding to newdiscoveries..Analysis of returned samples is alsoshown.

surfacestrap aeolian materials at the same time that cobbles and bouldersare erodedfrom local high areassuchaspressure The studysite is centeredon a playain the Mojave Desert of ridges.The resultis a mixture of aeoliansilt and basaltcobbles CaliforniacalledLavicLake (Figure 3) that waspart of a suite and boulders.The flow surfacebecomessmootherby accumuof lacustrinesystemsprevalent in the area during Pleistocene glaciationepochs[Norrisand Webb,1990].The playahas also lation of aeolian materialsin low areasand downslopetransbeen occasionallycovered with shallow lakes over the past port of bouldersand cobblesfrom high areas.Aeolian deposeveralthousandyears. Rainfall and runoff infiltrate into the sitionratesdeclineasthe surfaceroughnessdecreasesand thus subsurfaceof Lavic Lake, thereby placingthis playa into the trappingof aeolian materialsbecomeslessefficient.New barechargeclass[Rosen,1994].As a consequence, the playased- saltbouldersand cobblesare still erodedfrom local high areas iments are dominatedby fine-grainedclasticsas opposedto and are depositedonto the surfaceof the mantle. Further, the evaporitedeposits.Mud-cracked surfacesare common,with basalt boulders and cobblesexposedon the surface spall by cracksrangingfrom centimetersto approximatelya meter in physicalweatheringto produce a fairly uniform size distribuwidth.The playamaterialsconsistof claysand silt-sizedquartz, tion, with a mean size of approximatelyseveralcentimeters. feldspar,carbonate,and lithic fragments.Finally,the playaand Wetting and drying of the silt and clay minerals within the other areasused during the rover testsare locatedwithin the accretionmantle movesthe surfacecobblesboth laterally and air-groundwarfare training facility of the 29 Palms Marine vertically,allowingmantle sedimentto accumulatebeneaththe Base. Thus the playa has numerous craters and associated cobbles.At the same time that pavementsare forming, dark, ejecta fields of varying ages.Craters are typically