Each Mars Exploration Rover carries an integrated suite of scientific instruments and tools called the Athena science payload. The primary objective of the Athena science investigation is to explore two sites on the Martian surface where water may once have been present, and to assess past environmental conditions at those sites and their suitability for life. The remote sensing portion of the payload uses a mast called the Pancam Mast Assembly (PMA) that provides pointing for two instruments: the Panoramic Camera (Pancam), and the Miniature Thermal Emission Spectrometer (Mini-TES). Pancam provides high-resolution, color, stereo imaging, while Mini-TES provides spectral cubes at mid-infrared wavelengths. For in-situ study, a five degree-of-freedom arm called the Instrument Deployment Device (IDD) carries four more tools: a Microscopic Imager (MI) for close-up imaging, an Alpha Particle X-Ray Spectrometer (APXS) for elemental chemistry, a Mössbauer Spectrometer (MB) for the mineralogy of Fe-bearing materials, and a Rock Abrasion Tool (RAT) for removing dusty and weathered surfaces and exposing fresh rock underneath. The payload also includes magnets that allow the instruments to study the composition of magnetic Martian materials. All of the Athena instruments have undergone extensive calibration, both individually and using a set of geologic reference materials that are being measured with all the instruments. Using a MER-like rover and payload in a number of field settings, we have devised operations processes that will enable us to use the MER rovers to formulate and test scientific hypotheses concerning past environmental conditions and habitability at the landing sites.
 The main objective of the Mars Exploration Rover mission is to explore two sites on the Martian surface where water may once have been present, and to assess past environmental conditions at those sites and their suitability for life. Each MER rover functions as a remotely operated robotic field geologist, reading the record that is contained in the rocks and soils at its landing site. Like a human field geologist, each rover is equipped with the capabilities and tools that it needs to carry out its tasks. The suite of scientific instruments and tools carried by each rover is called the Athena Science Payload.
 The Athena payload was first developed in 1995 for a Mars rover mission that was proposed to NASA's Discovery Program. It was subsequently selected by NASA as the science payload for a rover mission planned for launch in 2001. While NASA's Mars exploration program has gone through some significant changes since that time, the fundamental scientific objectives and structure of the Athena payload have remained unchanged. When NASA selected the Mars Exploration Rover mission for flight in 2000, they chose to send two copies of the Athena payload, each on a large and capable roving vehicle, to two sites on the Martian surface.
 The science objectives of the Mars Exploration Program are focused on understanding the current and past habitability of the red planet on a global scale. This focus is not simply one of conducting direct life detection experiments either in situ or with returned samples. Rather, the intent is to develop a deep intellectual understanding of the spatial and temporal patterns and interactions associated with global-scale geologic and climatic processes that have operated on and within Mars, including how these processes have modulated geochemical cycles of biological importance [Greeley et al., 2001]. Understanding the nature, extent, and timing of interactions between surface and subsurface water and crustal materials is of paramount importance because water is central to the development, evolution, and vigor of life.
 Our understanding of the evolution of Mars and its hydrologic cycle is incomplete and can best be cast as a suite of multiple working hypotheses that are undergoing rapid revision as new Mars Global Surveyor and Mars Odyssey orbiter data are becoming available for analyses [Albee et al., 2001; Saunders et al., 2003]. One view that has had an extensive following in the science community is that the climate was once warm enough to support liquid water on the surface. The rapid introduction of greenhouse gases, such as carbon dioxide, into the atmosphere during periods of enhanced volcanism or during giant impact events would have provided the enhanced warming of the lower atmosphere needed to keep surface temperatures above the triple point temperature for water for some period of time [e.g., Phillips et al., 2001]. Removal of these gases by geochemical reactions with surface materials and/or by stripping by solar wind would have led to global cooling and collapse of the warm, wet environments [Jakosky and Jones, 1997]. Valley networks may have formed during this early warm, wet climate. The northern lowlands may have been the site of a “northern ocean” into which a massive amount of sediment eroded from the uplands was deposited, forming the plains that now dominate the landscape [Head et al., 1999]. In addition there may have been many inter-crater lakes and associated sediment deposition in the cratered uplands [Malin and Edgett, 2000].
 An alternate view is that Mars may never have been warm and wet. At best it may have been cold and wet at the surface in the past [e.g., Squyres and Kasting, 1994]. Any water bodies would have been ice covered and precipitation may have been dominated by snow. As this cold, wet climate decayed by removal of greenhouse gases the system became even colder and evolved to current conditions in which shallow ice deposits are only stable at higher latitudes [Feldman et al., 2002]. In fact, this view of Mars is consistent with mounting evidence from analyses of reflectance and emission spectra that indicate that the surface is covered with mafic volcanic materials, without extensive exposures of carbonates, clays, or other minerals that would have formed during wet, warm surface conditions. In particular, pyroxenes and feldspars, together with volcanic glasses and palagonitic material, seem to be ubiquitous on the surface [e.g., Bell, 1999; Bandfield, 2002; Bandfield et al., 2003]. In addition, olivine has been mapped in several locations using both TES and Odyssey THEMIS data [Christensen et al., 2003b]. Although it has been argued that TES data for the northern plains of Mars are consistent with the presence of weathered basalts [Wyatt and McSween, 2002], the lack of clay signatures in visible and infrared reflectance data for these regions argues against the presence of large exposures of weathering products. A cold, wet Mars could have inhibited extensive surface weathering while still allowing ice-covered lakes, seas, and riverine systems to produce the lacustrine and glacio-fluvial landforms and features evident today.
 In either of the scenarios discussed above, enhanced heat flow due to impacts or magmatic activity in regions with groundwater or subsurface ice deposits would have generated hydrothermal systems, particularly early in Martian history when the frequency of impacts was higher and magmatic activity occurred at enhanced rates relative to the present [e.g., Clifford, 1993]. These warm water systems would have been corrosive and altered host rocks, forming assemblages of alteration minerals such as silica deposits, sulfates, and oxides whose nature and extent would have been dependent on local conditions. Surface exposures of these deposits would be limited to locations where the circulating fluids reached the surface and to regions where subsurface waters altered host rocks and the materials were then exposed by erosion.
 The Mars Exploration Rovers will explore two key sites and acquire key data sets to allow development and testing of the types of hypotheses presented in the previous paragraphs, with an emphasis on understanding the nature and extent of interaction of water and crustal materials. For example, the Meridiani Planum landing site is hypothesized to expose hematite in association with basalt, produced either in an oxygenated lacustrine environment that prevailed during an earlier warmer, wetter period, or via anhydrous or hydrothermal alteration of volcaniclastic deposits [Christensen et al., 2001; Arvidson et al., 2003a; Golombek et al., 2003]. The Gusev Crater floor site may expose sedimentary rocks of lacustrine origin produced when water or ice-covered water was abundant at the surface, or as volcanic units from the Apollinaris Patera volcano located to the north [Cabrol et al., 2003; Golombek et al., 2003]. Even if the surface deposits within Gusev Crater are mostly of volcanic origin, there may evidence in the rock record for the nature and extent of alteration via circulating groundwater and hydrothermal fluids.
 During the operational lifetimes of the two rovers on Mars, Pancam and Mini-TES observation will be used to map the morphology and mineralogy at a number of locations at each site. Analyses of the data will focus on evidence for the nature and origin of the materials investigated, including aqueous phases that would provide information on ambient conditions that pertained during emplacement and/or modification of the units. Key rock and soil targets will be selected on the basis of Pancam and Mini-TES data, and will be approached for close-up imaging and acquisition of Mössbauer and Alpha Particle X-Ray Spectra. Natural rock surfaces will be examined with Athena instrumentation and then re-examined after removing any surface coatings using the Rock Abrasion Tool. Particular emphasis will be placed on using the rover's mobility to reach exposed bedrock, although our ability to do so may be limited by rover range (expected to be hundreds of meters over the course of the mission) and by the availability of bedrock outcrops at our landing sites. If bedrock is absent, the geologic inferences that we can draw will have to take into account the appropriate uncertainties concerning the provenance of the materials investigated. All of the data will be used to test and refine hypotheses concerning the geologic and climatic evolution of the sites and the role of water in producing and modifying crustal materials. Further, results will be used to better understand the global evolution of Mars and its habitability, particularly the role of water and its effects on geochemical cycles of possible biological importance.
3. Payload Description
 The Athena payload (Figure 1) is a suite of scientific instruments and tools for geologic exploration of the Martian surface. In order to address the primary science objectives discussed above, the payload is designed to: (1) provide color stereo imaging and remotely sensed mineralogical information for Martian surface materials, (2) determine the elemental and mineralogical composition of Martian surface materials, including soils, rock surfaces, and rock interiors, (3) determine the fine-scale textural properties of these materials. Two identical copies of the Athena payload are carried by each of the two Mars Exploration Rovers.
 The major components of the Athena payload are (1) for remote sensing, Panoramic Camera (Pancam), a high-resolution stereo color panoramic imager; Miniature Thermal Emission Spectrometer (Mini-TES), a mid-infrared spectrometer for remote investigation of mineralogy of rocks and soils, Pancam Mast Assembly (PMA), an articulated mast supporting both Pancam and Mini-TES; and (2) for in situ sensing, Microscopic Imager (MI), for close-up imaging of rock and soil surfaces; Alpha Particle X-Ray Spectrometer (APXS), for in-situ elemental analysis; Mössbauer Spectrometer (MB), for in-situ determination of the mineralogy of Fe-bearing rocks and soils; Rock Abrasion Tool (RAT), for removing weathered rock surfaces and exposing fresh material for characterization; Instrument Deployment Device (IDD), a robotic manipulator for positioning the above four in-situ payload elements.
 The payload also includes a Magnetic Properties Experiment that can separate magnetic soil particles from non-magnetic ones, enabling the composition of the magnetic materials to be studied by the instruments.
 In addition to these scientific instruments, the rover also carries six engineering cameras [Maki et al., 2003]: Navigation Cameras (Navcams), two wide-angle monochromatic cameras, also mounted on the PMA and Hazard Avoidance Cameras (Hazcams), four fisheye monochromatic stereo cameras, mounted in two stereo pairs on the rover body, viewing forward and backward. While not formally part of the science payload, these engineering cameras also play important roles in science operations.
3.1. Remote Sensing Payload Elements
 A primary objective of color, stereo, panoramic imaging is to provide high spatial resolution information on the morphology of the landing sites, on the lithology, texture, distribution, and shape of nearby rocks, and on local geologic features that may be present. This information is critical for understanding what geologic processes have affected the sites, particularly when merged with compositional data from other instruments on the payload. Pancam imaging will also be the primary tool used for establishing the nature of stratigraphy, cross-cutting, and other geologic relationships.
 Multispectral panoramic imaging will provide information on the mineralogy of materials to supplement and complement data obtained by other instruments. Spectra of Mars in the 0.4–1.1 μm range covered by Pancam are dominated by iron oxides and oxyhydroxides with varying degrees of crystallinity. These oxides are presumably mostly weathering products. Multispectral imaging will help determine the oxidation state of iron, identify the secondary iron minerals and their crystallinity, and identify primary mafic minerals. Pancam filters have been chosen to detect absorptions from hydrated minerals and other secondary minerals like Fe-bearing carbonates and sulfates. Multispectral imaging also will help determine if there are weathering rinds on rocks, and if so, how they compare compositionally with the regolith and, by inference, how weathering processes have changed with time.
 Pancam uses 1024 × 2048-pixel CCD detectors. The CCDs are operated in frame transfer mode, with one 1024 × 1024-pixel region constituting the active imaging area and the another adjacent 1024 × 1024 region serving as a frame transfer buffer. The frame transfer buffer has an opaque cover that prevents >99% of light at all wavelengths from 400 to 1100 nm from being detected by this region of the CCD. The pixel pitch is 12 μm in both directions. The arrays are capable of exposure times from 0 msec to 335 s in 5.12 msec steps. Under expected operating conditions, the arrays have at least 150,000 electrons of full-well depth, and a read noise of less than 50 electrons. Dark current is low and varies with temperature; it is negligible at −55°C and is <100 electrons/s at 0°C. Analog-to-digital converters provide a digital output with 12-bit encoding, and signal-to-noise ratio (SNR) > 200 at all signal levels above 50% of full scale. The detector response has a linearity >99% for signals between 10% to 90% of full well.
 Each array is combined with optics and a small filter wheel to form one “eye” of a multispectral, stereoscopic imaging system. The optics for both cameras consist of identical 3-element symmetrical lenses with an effective focal length of 43 mm and a focal ratio of f/20, yielding an instantaneous field of view (IFOV) of 0.28 mrad/pixel and a square FOV of 16.1° × 16.1° per eye. The optics and filters are protected from direct exposure to the Martian environment by a sapphire window at the front of the optics barrel. The optical design allows Pancam to maintain optimal focus from infinity to within about 1.5 meters of the cameras.
 Each filter wheel has eight positions, allowing multispectral sky imaging and surface mineralogic studies in the 400–1100 nm wavelength region. The left wheel contains one “clear” (empty) position. The remaining filter wheel positions are filled with narrowband interference filters that are circular with a 10 mm diameter clear aperture. One filter on each eye has an ND5.0 coating to allow direct imaging of the Sun at two wavelengths.
 Radiometric calibration of Pancam has been performed with an absolute accuracy of 7% or better and a relative precision (pixel-to-pixel) of 1% or better. Calibration at Mars will be achieved using a combination of preflight calibration data and inflight images of a Pancam calibration target carried by each rover [Bell et al., 2003].
 The two Pancam eyes are mounted on a mast on the rover deck. The mast is referred to as the Pancam Mast Assembly (PMA, see section 3.1.3), and also includes several key components used by Mini-TES. The cameras are located on a “camera bar” with a boresight 180° from the Mini-TES boresight. The rover navigation cameras (Navcams, see Maki et al. ) are also located on this same camera bar, and point in the same direction as Pancam. The boresight of the Pancam cameras is approximately 1.52 m above the Martian surface with the PMA in the deployed position. The cameras are moved together from +90° to −97° in elevation using a geared brush motor on the camera bar. The entire PMA head, including the cameras, can be rotated 370° in azimuth by a geared brush motor assembly.
 The two Pancam eyes are separated by 30 cm horizontally and have a 1° toe-in. The separation and toe-in provide an adequate convergence distance for scientifically useful stereo topographic and ranging solutions to be obtained from the near-field (5–10 m) to approximately 100 m from the rover.
 Pancam will be commanded by and will return digital data directly to the rover computer. The computer provides the capability to perform a limited set of image processing tasks on Pancam data prior to transmission. These tasks include (1) electronic shutter effect (frame transfer) correction, (2) bad pixel replacement, (3) rudimentary correction of flat field artifacts, (4) automatic exposure control capability to maximize the SNR of downlinked data while preventing data saturation, (5) relative scaling of exposure times through different filters using a stored onboard table, (6) image downsampling and subframing, and (7) image compression using bit scaling and/or the high-performance JPL-developed compression algorithms ICER (wavelet-based) or LOCO.
 The primary objective of Mini-TES is to obtain mineralogical information for rocks and soils surrounding each rover. These data provide fundamental scientific information about Mars and, like Pancam images, will be used to select materials to be investigated in more detail by the APXS, the Mössbauer Spectrometer, and the Microscopic Imager.
 Mini-TES acquires spectra in the 5–29 μm range, which for rocks and soils are dominated by molecular vibrations. The vibrational energies (wavelengths) are controlled by the anion compositions, coordination numbers, and bond lengths. Mini-TES measurements therefore provide a direct means of identifying crystal structure, and hence mineralogy, of geologic materials including silicates, carbonates, sulfates, phosphates, oxides, and hydroxides. In silicates, for example, the vibrational motions associated with the Si-O stretching modes occur between 8 and 12 μm. The Si-O absorption band decreases from 11 to 9 μm in a uniform succession for minerals with chain, sheet, and framework structure, and so provides a means of discriminating minerals with these structures. Additional bands occur throughout the 12–25 μm region associated with a variety of Si, O and Al stretching and bending motions. Carbonates have strong absorption features associated with CO3 internal vibrations in the 6–8 μm region that are easily distinguished from silicate bands. Hydroxide-bearing minerals like clays have spectral features due to fundamental bending modes of OH attached to various metal ions. Salts like phosphates, sulfates, nitrites, and chlorides all have characteristic bands. Feldspar and pyroxene compositions can be determined within their respective solid solution. And, compared to past IR observations of Mars, the power of Mini-TES to discriminate among minerals should be aided substantially by the ability to resolve individual rocks and the negligible path length of atmospheric gas between the instrument and most targets.
 Mini-TES is a Michelson interferometer that provides a spectral resolution of 10 cm−1 over the wavelength range from 5–29 μm (2000–345 cm−1). The instrument is mounted inside the rover, and views the terrain around the rover by using the PMA as a periscope. A scan mirror assembly atop the PMA reflects radiation down through the PMA and into the telescope and interferometer. The scan mirror assembly allows Mini-TES to provide spectral image cubes over a 360° range in azimuth and from −50° to +30° in elevation. The scan mirror assembly also provides views of internal and external full-aperture calibration targets. The elevation mirror can be slewed to a stowed position in which a cover blocks the Mini-TES aperture in the PMA, protecting the optics from dust accumulation.
 Mini-TES has two spatial resolution modes. A solenoid-activated field stop can be removed from the optical path to provide an IFOV of 20 mrad, or inserted to provide an IFOV of 8 mrad. During data acquisition, the PMA's elevation mirror and azimuth actuator are sequenced to generate a raster image of the scene. The scan mirror assembly can also be commanded to allow Mini-TES to view the internal and external calibration targets regularly in order to maintain instrument calibration during an image acquisition.
 The Mini-TES telescope at the base of the PMA uses a reflecting Cassegrain configuration with a mirror diameter of 6.35 cm, a focal ratio of f/12, and an intermediate field stop that feeds an approximately collimated beam into the Mini-TES interferometer. The 6.35-cm telescope diameter defines the minimum size of the Mini-TES beam; the beam diverges further at an angle of either 8 or 20 mrad, depending on the resolution mode chosen. The optical design provides for more than 85% of the encircled energy to be contained in an area equal to a single IFOV, 98% within an area equal to 2 × 2 IFOV, and 99.8% within an area equal to 3 × 3 IFOV. Focus is maintained from 2 meters to infinity, with a blur of no more than 15% of an IFOV at infinity focus.
 The Mini-TES Michelson interferometer uses the same linear motor mechanism and drive electronics as the Mars Observer (MO)/Mars Global Surveyor (MGS) TES instruments [Christensen et al., 1992]. Double-sided interferograms at a spectral resolution of 10 cm−1 are obtained with a mirror travel distance of 0.55 mm in 1.8 s
 Mini-TES uses a single uncooled deuterated triglycine sulfate pyroelectric detector sized to define the instrument's 20-mrad IFOV. The IFOV, dwell time, and interferometer scan rate have been selected to produce frequencies in the range of 15 to 120 Hz which is the range over which minimum noise equivalent spectral radiance (NESR) can be achieved. The detector provides the necessary performance over a temperature range from −10 to +20°C and with reduced performance from −40 to +35°C.
 The NESR of the Mini-TES for a single spectral accumulation interval at 10 μm observing a scene at 270 K and 20 mrad will be <1.25 × 10−8 W cm−2 s−1 sr−1, corresponding to a SNR of at least 450 for co-addition of two observations. Radiometric calibration of Mini-TES over its full spectral range has been performed with an absolute accuracy of 5% or better and a relative precision (pixel-to-pixel) of 2% or better, viewing a 270 K blackbody. The internal calibration target is located inside the head of the PMA, and the external target is located on the deck of the rover. Both targets have V-grooved surfaces and are coated with high emissivity paint. Temperature sensors affixed to both targets have an absolute accuracy of ±0.2°C and a precision of ±0.1°C.
 Mini-TES acquires data in a cyclic fashion in two-second increments, with each two-second period corresponding to the Michelson mirror scan followed by its retrace. Spectral integration is coordinated with the PMA elevation and azimuth drive mechanisms using the rover computer. Once the data have been transferred to the rover memory, flight software performs a Fourier transform on the interferogram in order to generate a spectrum. It then performs data aggregation in order to reduce the total volume of data to be downlinked. Data volume is further reduced via lossless compression using a Rice algorithm.
3.1.3. Pancam Mast Assembly
 Pointing of both Pancam and Mini-TES is provided by the Pancam Mast Assembly (PMA), shown in Figure 2. The Pancam cameras (and also the Navcam cameras) are mounted on a movable camera bar at the top of the PMA. Mini-TES is mounted within the rover body at the base of the PMA, and views the terrain around the rover by using the PMA as a periscope. The hollow cylindrical mast serves as an optical pathway for Mini-TES, and a scan mirror assembly at the top of the mast provides the Mini-TES pointing capability. Having Pancam, Navcam, and Mini-TES all view the scene around the rover from atop the same mast provides for optimal viewing of terrain, and also allows for relatively straightforward coregistration of data acquired from all of those sensors.
 The PMA uses three independent actuators for instrument pointing. One near the base of the PMA provides 360° of azimuth actuation for all instruments, plus 5° of overtravel in each direction to allow data overlap. A second one near the top of the mast moves the camera bar in the elevation direction, providing +90/−97° of elevation pointing capability for Pancam and Navcam, from zenith to just beyond nadir. The cameras are pointed down whenever they are not in use, minimizing exposure of the optics to dust settling from the atmosphere. A third actuator, also near the top of the mast, provides Mini-TES elevation pointing.
 Mini-TES pointing is accomplished using two mirrors (Figure 3). One is a fixed 45° fold mirror at the top of the mast, directly above the Mini-TES telescope. The other is a rotating 45° Fold mirror that is rotated about a horizontal axis by the Mini-TES elevation actuator. Use of two mirrors in this fashion enables elevation pointing to be performed with reflection angles that are always 90°, eliminating radiometric calibration problems that could be caused by reflections at variable angles.
 Both of the Mini-TES pointing mirrors are enclosed within a cylindrical “can” at the top of the PMA. There are actually two cans, one nested within the other. The outer can is fixed, and the inner one rotates with the elevation mirror. Both have openings that allow-passage of the Mini-TES beam. The opening in the inner can is just large enough to admit the beam, while the opening in the outer can is large enough to permit Mini-TES pointing over elevation angles ranging from 30° above the horizon to 50° below it.
 The elevation mirror and the inner can that surrounds it can be rotated over an elevation range of about 210°. From +30° to −50°, the instrument views the scene around the rover. Beyond −50°, the surface of the inner can blocks the opening in the outer can, closing off the PMA head and protecting the mirrors from dust. The head is kept closed whenever Mini-TES is not in use. Rotating the actuator to an elevation angle of −180° allows the instrument to view a full-aperture calibration target [Christensen et al., 2003a] that is set into the inner surface of the outer can. Figure 4 summarizes the pointing capabilities of the PMA.
 During launch, cruise, and landing, the PMA is stowed, lying parallel to the rover deck. It is restrained to the deck in several positions at its head, allowing it to sustain launch and landing loads. Shortly after landing, pyrotechnic devices are fired to release these restraints. Then, a fourth actuator called the Mast Deployment Drive (MDD) is used to erect the mast to the vertical position. The MDD uses an unusual design in which the whole mast is cut through near its base at a 45° angle. The portion of the PMA below this cut is vertical and fixed to the rover deck, while the portion above it is movable and lies horizontal when the PMA is stowed. A large circular ring gear that is coplanar with the cut is rotated by a motor after the mast head has been released. This motion rotates the upper portion of the PMA about an axis perpendicular to the ring gear, sweeping it through half a cone to reach the vertical position. Once the upper portion of the PMA is vertical, it is latched in place. After deploying the PMA, the MDD does not move again for the remainder of the mission.
 The portion of the PMA tube that lies below the azimuth actuator and includes the MDD is made of titanium. Above the azimuth actuator the primary structural elements, including the inner and outer cans around the Mini-TES mirrors, are made of graphite-epoxy composite. Roundwire cabling that wraps around the MDD is used below the azimuth actuator to carry power and data lines for the actuators and cameras. Above the azimuth actuator, most cabling is flexprint.
 The MDD, camera bar actuator, and azimuth actuator all use Maxon RE020 brush motors, while the Mini-TES elevation actuator uses the slightly larger RE025. Each actuator uses a custom gearbox tailored to its particular performance requirements, and each uses an encoder to provide fine position knowledge. The camera bar, azimuth, and Mini-TES elevation actuators all have hardstops at both ends of their range of travel. Initialization against these hardstops can provide unambiguous position knowledge whenever the position knowledge provided by the encoders is lost. Each actuator is equipped with a heater that allows it to be warmed to its minimum operating temperature of −55°C under cold environmental conditions. This low-temperature capability is particularly important for performing Mini-TES observations of the Martian atmosphere at night (see section 4.1 below).
 Because the PMA serves as the fore-optics of the Mini-TES, we devoted considerable attention to its optical properties in the thermal infrared. The Mini-TES mirrors are both made of polished beryllium, providing reflectance of ≥97.5% across the full spectral range of the instrument. Honeycomb-like ribs on the back of each mirror provide the necessary combination of stiffness and low mass. The combined wave front distortion introduced by the mirrors is less than two waves (peak to valley) at >8 cycles per aperture.
 Interior baffles are used along the full length of the PMA to prevent single-bounce stray energy from outside Mini-TES' 20-mrad field of view from reaching the instrument. It is also important to maximize the thermal emissivity of the inner surfaces of the PMA. The inner surface of the titanium structure that forms the lower portion of the PMA is painted with Aeroglaze Z306 high-emissivity paint. The inner surface of the upper portion of the PMA is bare composite, with an emissivity of approximately 0.8. The entire outer surface of the PMA is painted white to minimize its temperature (and hence its radiation into the instrument) under solar illumination.
 The PMA provides absolute pointing knowledge errors for all instruments, relative to the elevation and azimuth hardstops, of roughly 0.1° (∼2 mrad). The stepping capability of the Mini-TES elevation actuator was designed with the internal timing of the Mini-TES instrument in mind. As noted above, Mini-TES acquires data in 1.8 second integration periods, separated by 200-msec interferometer retrace periods. The Mini-TES elevation mirror is able to move in steps of up to 20 mrad in size within one 200-msec retrace period. This capability allows a Mini-TES raster to be constructed using 20-mrad elevation steps but requiring no integration periods to be wasted during mirror moves. Azimuth steps can take as long as 1 second, so one Mini-TES integration period typically is discarded during an azimuth move. We therefore normally construct Mini-TES rasters using a “snake” pattern in which we scan down one column, take a small step in azimuth, scan up the next column, and so forth. This approach minimizes the ratio of (slow) azimuth steps to (rapid) elevation steps.
3.2. In Situ Payload Elements
3.2.1. Microscopic Imager
 Much information can be obtained by studying rocks and soils with a close-up imager that has a resolution sufficient to enable detailed characterization of coatings, weathering rinds, individual mineral grains, clasts, or other particles. For sedimentary rocks, the size, angularity, shape, and sorting of grains can reveal much about conditions of transport and deposition. For lavas, vesicularity gives an indication of volatile content. Grain size and texture of igneous rocks provide information on crystallinity of the magma when emplaced and how quickly it cooled. Optical properties of mineral grains allow improved mineral identification. Microscopic imaging can also be used to identify small-scale veins of precipitated minerals.
 The Athena Microscopic Imager (MI) is a high-resolution imaging system mounted on the Instrument Deployment Device (IDD). The camera body is identical to the ones used by Pancam, so the field of view is 1024 × 1024 pixels in size and the instrument has the same basic radiometric performance characteristics as Pancam. There is a single broad-band filter, so imaging with the Microscopic Imager is monochromatic.
 The MI optics use a simple, fixed focus design at f/15 that provides ±3 mm depth-of-field at 30 μm/pixel sampling. The field of view is therefore 31 × 31 mm at the working distance. The focal length is 20 mm, and the working distance is 66 mm from the front of the lens barrel to the object plane. The object-to-image distance is 100 mm. Focus adjustment is provided by using the IDD to move the MI forward or backward along its optical axis.
 The spectral bandpass of the MI optical system is 400–680 nm. At best focus, the modulation transfer function of the optics is at least 0.35 at 30 lp/mm over this bandpass. Radiometric calibration of the MI has been performed with a relative (pixel-to-pixel) accuracy of ≤5%, and an absolute accuracy of ≤20% over the instrument's full spectral bandpass. The MI signal to noise ratio is at least 100 for exposures of ≥20% full well over the spectral bandpass and within the calibrated operating temperature range (−55 to +5°C).
 No onboard radiometric calibration target is provided for inflight calibration of the MI. Flat fields can be obtained by imaging the Martian sky. The MI is also able to view the Compositional Calibration Target (see section 3.2.3 below), and this target includes fiducial marks that can be used to perform a focus check.
 Whenever the MI is not in use, the MI optics are protected from contamination by a transparent cover. The cover is opened only for MI imaging sequences. A contact sensor attached to the MI will be used to detect rock and other hard surfaces, to help ensure accurate positioning and protect the MI from accidental damage.
 The MI acquires images using only solar or skylight illumination of the target surface. Stereoscopic observations and mosaics can be obtained by moving the MI between successive frames. Stereo images and images taken at various distances from the target will be used to derive the 3-dimensional character of the target surface. Optical sections will also be combined to produce an image of the target that is well-focused across the entire frame.
 In-focus images obtained by the Microscopic Imager will have a spatial resolution of 30 μm per pixel. This resolution meets or exceeds the effective resolution provided by hand lenses typically used by field geologists. The MI will be used in a similar fashion to reveal the fine-scale petrographic texture of rocks, as an aid in interpretation of their composition and history. MI images of fresh rock surfaces exposed by the RAT may be particularly useful in this regard.
3.2.2. Alpha Particle X-Ray Spectrometer
 The primary objective of the APXS is to determine the elemental chemistry of rocks and soils, complementing the textural and mineralogical analyses of the other Athena instruments.
 A key APXS objective is examination of the products of water-induced erosion, sedimentation, solution, and evaporation. Elemental analyses of rocks or clasts that resulted from these processes will constrain mineralogical analyses. Moreover, such analyses can be compared with analyses of local soils, showing whether the conditions under which weathering occurred early in the planet's history differed substantially from conditions later. Comparisons of concentrations of elements with differing mobilities during weathering, e.g., Al, Si, and Fe, can be used for this purpose, as can concentrations of S and halogens that are indicators of salts.
 The APXS can also reveal the chemistry of primary crustal rocks. Such measurements are essential for understanding under what environmental conditions Martian crustal materials formed, and how they formed. For example, highland rocks are presumably a mixture of primordial crust, ancient volcanic rocks, and ancient sediments, all stirred by impact. Differences are likely from rock to rock, and even among clasts within the same rock. Chemical analyses may therefore shed light on a variety of processes that were important during early Martian history.
 Other APXS objectives deal with the elemental chemistry of soils. Minerals produced by weathering in particular tend to be complex, and permit many substitutions, like halogens for water, or Al for Fe. The environmental conditions under which weathering took place on Mars, as well as the composition of the source rocks, are largely unknown, so it is important that mineralogical determinations of soils and weathering products provided by Mini-TES and the Mössbauer Spectrometer be constrained by chemistry.
 The APXS works by exposing Martian materials to energetic alpha particles and x rays from a radioactive 244Cm source, and then measuring the energy spectra of backscattered alphas and emitted x rays. The instrument is conceptually similar to the APXS instrument that flew on the Mars Pathfinder mission. However, there are several differences that improve the instrument's reliability and performance. Unlike the Pathfinder APXS, the Athena APXS does not have a proton mode. The proton mode has been dropped because recent increases in the spectral resolution and sensitivity of the x-ray mode have made it superfluous. Significant modifications have also been made to the instrument to reduce the CO2-induced background that was observed on Pathfinder, to improve x-ray spectral resolution, and to decrease susceptibility to electromagnetic interference. The Athena APXS has undergone extensive preflight calibration under Mars-ambient conditions, and has an onboard reference target for post-landing calibration on Mars.
 The x-ray mode is sensitive to major elements, such as Mg, Al, Si, K, Ca, and Fe, and to minor elements, including Na, P, S, Cl, Ti, Cr, and Mn. The alpha mode is sensitive to lighter elements, particularly C and O. The depth of analysis varies with atomic number, ranging from approximately 10 to 20 micrometers for sodium, to approximately 50 to 100 micrometers for iron. The detection limit is typically 0.5 to 1 weight percent, depending on the element. The APXS is insensitive to small variations of the geometry of the sample surface because all major and minor elements are determined, and can be summed to 100 weight percent.
 The APXS instrument consists of a sensor head mounted on the rover's Instrument Deployment Device, and electronics mounted in the rover's warm electronics box.
 The sensor head contains six 244Cm alpha radioactive sources with a total source strength of about 30 mCi. The sources are each covered with 3-μm titanium foils that reduce the energy of emitted alpha particles from the initial value of 5.8 MeV to about 5.2 MeV. At this energy, the alpha particle scattering cross section of carbon is significantly reduced. The reduction is accompanied by a slight degradation of the alpha spectral resolution caused by broadening of the excitation spectrum, but the net result is a significant suppression of atmospheric background in the alpha spectra. Collimators in front of the sources define the instrument's field of view, which is about 38 mm in diameter at the nominal working distance of 29 mm.
 Surrounding the sources are six alpha detectors. The FWHM for the alpha mode of a 244Cm peak at 5.8 MeV is less than 100 keV. Interior to the ring of sources is a single high-resolution silicon drift x-ray detector with a 5-μm beryllium entrance window. The FWHM of this detector at 6.4 keV is about 160 eV, compared to 260 eV for the Pathfinder APXS. The noise level in the x-ray mode is less than 600 eV at temperatures below −30°C. The overall sensitivity of the x-ray mode is roughly 20 times greater than it was for the Mars Pathfinder instrument, enabling measurement durations for x-ray spectra that are considerably shorter than was required on the Pathfinder mission.
 The entrance to the detector head is normally protected from Martian dust and other potential contaminants by a pair of doors. These doors swing inward and lock open when the sensor head is pressed against a target or other hard surface. They can be closed again by actuation of a release mechanism. The inner surfaces of the doors provide the calibration reference surface for the instrument.
 Proper preflight calibration is essential to analysis of APXS data, so the Athena APXS instruments have undergone an extensive calibration program. Most calibration measurements were made in a chamber filled with a mixture of gases that closely matches the composition of the Martian atmosphere, at the appropriate atmospheric density. Calibration measurements included spectral “library” measurements of pure elements and oxides, geochemical standards that span a wide range of plausible Martian surface compositions, standard targets under a range of atmospheric densities and measurement geometries, standard targets in both natural and powdered form, to investigate texture effects, the APXS flight calibration target, the magnets of the magnet array, and several blind certified geochemical reference standards, for independent assessment of the accuracy with which compositions can be measured. All of these measurements were made using the flight radiation sources.
3.2.3. Mössbauer Spectrometer
 The primary objective of the Mössbauer Spectrometer is to reveal the valence state, molecular structure, and magnetic properties of iron-bearing minerals in rocks and soils. A 57Fe Mössbauer spectrometer uses the resonance absorption of recoil-free emitted gamma-rays by 57Fe nuclei in a solid to investigate the splitting of nuclear levels due to the interaction of the Fe atom with its surrounding electronic environment. This hyperfine interaction is different for Fe nuclei with different electronic environments, so each Fe-bearing mineral has its own characteristic Mössbauer spectrum.
 Detailed objectives of the Mössbauer spectrometer are to
 1. Determine the oxidation state of iron: The Fe2+/Fe3+ ratio provides information on the oxidation state of the soils and rocks. Comparison of these oxidation states can indicate the extent to which the oxidation state was enhanced during weathering, and hence can give insights into the processes involved, the nature of surface-atmosphere interactions, and likelihood of the preservation of organics against the oxidation process.
 2. Identify the iron oxides and the magnetic phase in the Martian soil: Individual iron oxide and oxyhydroxide minerals have different chemical pathways of formation. For instance, iron oxides or hydroxides formed via precipitation in abundant liquid water will be different from the oxidation products formed via solid-gas reactions. Identification of ferric phases in the soil can therefore contribute to the understanding of the history of Martian water.
 3. Identify iron-bearing minerals in rocks: What igneous rocks are present? By Mössbauer spectroscopy, Fe-bearing silicate minerals like pyroxene and olivine, as well as ilmenite and other Fe oxides, can be identified.
 4. Search for Fe-sulfates, Fe-nitrates and Fe-carbonates: These could be important irreversible volatile reservoirs, and their identification would aid in understanding of Martian volatile evolution.
 The Athena Mössbauer spectrometer uses a vibrationally modulated 57Co source to illuminate target materials. Backscattered gamma signals are binned according to the source velocity, revealing hyperfine splitting of 57Fe nuclear levels that provides mineralogical information about the target. The main parts of the instrument are the Mössbauer drive that moves the 57Co source with a well-known velocity, the γ- and x-ray detectors that detect the backscattered radiation, the microcontroller unit, the 57Co/Rh Mössbauer source, and the radiation collimator and shielding.
 Like the APXS, the Mössbauer spectrometer is split into the sensor head-on the IDD and the electronics in the rover's warm electronics box. The sensor head carries the Mössbauer drive with the analog part of the drive control unit, the 57Co/Rh Mössbauer source, the radiation collimator and shielding, the four PIN-diode detector channels including pulse amplifiers, and one reference detector channel to monitor the velocity of the drive using a weak 57Co source and a well known Mössbauer reference absorber in transmission geometry. The electronics in the rover body include an internal microcontroller, so that the instrument can collect data independently of the rover computer. The analog signals of the five detector channels are analyzed by discriminators for 14.4 keV and 6.4 keV peaks. Mössbauer spectra for the two different energies of 6.4 keV and 14.41 keV are sampled separately.
 Measurements are made by placing the instrument directly against a rock or soil sample. Physical contact is used to provide an optimal measurement distance and to minimize possible microphonics noise on the velocity-modulated energy of the emitted γ rays. The mechanical construction of the IDD and the interface limit vibration-induced velocity noise at the sensor head to less than 0.1 mm/s. A contact plate is mounted at the front part of the sensor head, assuring an optimal distance from the sensor head to the sample of about 9 to 10 mm. A heavy metal collimator in front of the source provides an irradiated spot of nominally 15 mm (up to 20 mm, depending on actual sample distance and shape) in diameter on the surface of the sample. The average depth of sampling by Mössbauer data is about 200 to 300 μm.
 Mössbauer parameters are temperature dependent. Especially for small particles exhibiting superparamagnetic behavior (e.g., nanophase Fe oxides), the Mössbauer spectrum may change significantly with temperature. The observation of such changes will help in determining the nature of the iron-bearing phases. Therefore Mössbauer measurements will be performed over a range of diurnal temperatures spanning both the daytime maxima and the nighttime minima.
 One Mössbauer measurement takes approximately 12 hours, depending on the phases present in the sample and the total iron content. The temperature variation for one spectral accumulation interval will not be larger than about ±10°C. When larger variations occur, spectra for different temperature ranges are stored separately, resulting in an increase in the total data volume (depending on the number of temperature intervals required), and a decrease of statistical quality for the individual subspectra.
 In parallel with the measurements of samples, calibration spectra will be taken using the reference channel implemented in the instrument. A Compositional Calibration Target containing a thin slab of magnetite-rich rock is also included on the rover where it can be viewed directly by the instrument immediately after landing, as well as later in the mission if necessary.
3.2.4. Rock Abrasion Tool
 The primary scientific objectives of the Athena investigation involve using the rover and its payload to read the geologic record at the landing sites and to assess past environmental conditions and possible former habitability of the site. Because weathering processes can alter the texture and composition of rock surfaces, they could act to alter or destroy evidence of the conditions under which rocks formed. We therefore have included in the payload a tool that can penetrate through the possibly weathered outer regions of rocks, exposing fresh rock underneath. This device, called the Rock Abrasion Tool (RAT), is a diamond-tipped grinding tool capable of removing a cylindrical area 4.5 cm in diameter and at least 0.5 cm deep from the outer surface of a rock. The exposed area is large enough to admit the APXS and Mössbauer sensor heads, and to fill a Microscopic Imager frame almost completely. All of the in-situ instruments can therefore examine a surface that has been exposed by the RAT. RAT operation takes a few hours for penetration into dense basalt.
 The RAT has a total of three actuators. One causes each of two grinding wheels to rotate at high speeds. One wheel has two diamond teeth, which cut out a circular area associated with each grinding head as the head rotates. The other wheel has a brush that helps to remove rock cuttings from the freshly ground surface. A second actuator causes the two grinding wheels to revolve around one another at a much slower rate, sweeping the two circular cutting areas around the full 4.5-cm diameter cutting region. Another brush associated with this revolve axis pushes rock cuttings to the periphery of the freshly ground region. Finally, a third “z axis” actuator translates the entire cutting head toward the rock, causing it to penetrate to the commanded depth.
 In order to grind a rock, the IDD places the RAT directly against it. Contact is made on two small knurled balls external to the grinding heads, and a ring surrounding the heads can adjust in two orthogonal axes to the orientation of the rock surface. Once pressed firmly against the rock by the IDD, all further actuations take place within the RAT itself. Rotation and revolution of the grinding wheels is initiated, and they are slowly translated toward the rock surface by the z axis actuator until contact is made. Encoders monitor penetration progress, and allow closed-loop control of the grinding process.
 The RAT is designed to preserve petrographic textures of the prepared rock surfaces as fully as possible, so that they can be viewed effectively using the MI. The grinding process is slow enough that no measurable modification of rock chemistry or mineralogy by frictional heating is anticipated. The grinding wheels are designed so that contamination of the exposed surface by cuttings from the rock (and previous rocks) is minimized. Grinding wheel materials have been selected so that there should be no detectable contamination of rock surfaces due to wear of the grinding heads themselves. If necessary, the grinding wheels can be cleaned by actuating the RAT against a stiff wire brush that is mounted on the forearm link of the IDD.
3.2.5. Magnetic Properties Experiment
 The objective of the Athena magnetic properties experiment is to attract Martian magnetic materials, and to hold them in a way that is optimized for investigation by means of the payload instruments. The magnets are similar in some respects to the magnet arrays carried on the Viking and Mars Pathfinder missions [Hargraves et al., 1977; Madsen et al., 1999]. However, the experiment is a significant scientific step beyond its predecessors, primarily because of the unique mineralogical capabilities of the Mössbauer Spectrometer that is part of the Athena payload.
 The nature of the magnetic phase that Viking and Mars Pathfinder detected in the Martian soil is uncertain. There is some evidence from reflectance spectroscopy that superparamagnetic particles (nanophase iron oxides with diameters of less than about 50 nm) are present. The shape of the Mössbauer spectrum, especially of such small particles, depends strongly on temperature and particle size. By measuring the spectrum at different temperatures one may obtain semi-quantitative information on the crystallite size and whether superparamagnetic particles are indeed present. Whether the iron oxides are poorly crystalline (e.g., nanophase or superparamagnetic) or well crystalline also has implications for the environmental conditions at the time they formed.
 The Athena magnets include (1) “filter” and “capture” magnets mounted on the rover's Magnet Array, (2) a “sweep magnet” mounted on the rover deck, and (3) RAT magnets, mounted within the Rock Abrasion Tool.
 The filter and capture magnets are mounted on a Magnet Array on each rover that is accessible to the Mössbauer Spectrometer and the APXS. The stronger capture magnet is designed to attract all ferro/ferrimagnetic dust, while the weaker filter magnet is designed to attract only the most magnetic dust.
 The capture magnet is designed to accumulate a homogenous layer of airborne Martian dust as efficiently as possible, and to provide a relatively constant magnetic field at the position of the dust layer. The magnetic field strength is approximately 280 mT at the active surface. On the basis of experience from the magnetic properties experiment on Mars Pathfinder, it is expected that the capture magnet will collect sufficient material for Mössbauer analysis in about 15 sols.
 The filter magnet also will collect airborne dust particles carried to the magnet by the atmosphere. The filter magnet is designed to accumulate a homogenous layer of strongly magnetic dust, and to attract weakly magnetic dust as little as possible. Thus this magnet will separate out and to keep attached to its surface a magnetic subset of the Martian dust particles from the bulk material (if the properties of the dust allow this). The magnetic field strength at the active surface is approximately 140 mT. On the basis of experience from the magnetic properties experiment on Mars Pathfinder, it is expected that the filter magnet will collect sufficient material for Mössbauer analysis in about 30 sols.
 The filter and capture magnets are each contained within an aluminum disk 45 mm in diameter. Each is positioned near the base of the PMA, mounted such that their surface normals are oriented 45° above horizontal. Both may be viewed by Pancam, though not by Mini-TES.
 An unresolved question from the Viking and Mars Pathfinder missions is whether magnets are culling a population of more strongly magnetic particles from the airborne dust, or whether all dust particles have similar magnetic properties [e.g., Hargraves et al., 1977; Madsen et al., 1999]. The sweep magnet is designed to answer this question. This magnet consists of a thin-walled magnetic tube magnetized along its symmetry axis. With this configuration it is possible to make a strong magnet (∼350 mT at the surface) capable of deflecting the paths of wind-transported, magnetic particles arriving at the surface of the magnet. Magnetic particles will accumulate on a narrow ring corresponding to the magnetic tube. The central surface inside the ring magnet will only collect non-magnetic dust particles. At greater radial distances from the ring magnet, both magnetic and non-magnetic particles will accumulate. Pancam images of the sweep magnet will provide spectral information on the dust collected, and thus provide information on the magnetic versus nonmagnetic particles in the Martian dust. The sweep magnet is mounted on the rover deck immediately adjacent to the Pancam calibration target, so that it will be observed during each Pancam calibration sequence.
 Because the scientific objectives to be addressed using the filter, capture, and sweep magnets all require just a single measurement once adequate dust buildup has occurred, we have not included any mechanism for removing dust from the magnets.
 Use of the RAT will produce small particles of abraded Martian rock. Four 7 mm diameter × 9 mm thick magnets are mounted within the RAT to sample and concentrate the magnetic portion of the abraded rock. After a RAT operation, Pancam can be used to examine any cuttings that may have adhered to the magnets. The RAT magnets have different strengths, providing a range of conditions for magnetic particles to be attracted and held.
3.2.6. Instrument Deployment Device
 The four in-situ payload elements (MI, APXS, MB, RAT) are mounted on a five-degree-of-freedom (DOF) manipulator called the Instrument Deployment Device (IDD), shown in Figure 5.
 The IDD is a robotic arm, with a shoulder, an elbow, and a wrist. Its dimensions are similar to those of a human arm. Two joints at the shoulder provide actuation in azimuth and elevation. A single joint at the elbow provides additional elevation actuation. The four payload elements are mounted pointing radially outward at roughly 90° to one another in a “turret” arrangement (Figure 6); actuators at the wrist provide pitch and rotation actuation for the turret.
 The combined ranges of motion of these five actuators define a five-dimensional IDD “work volume” at the front of the rover. Any target surface within the work volume can be defined by five parameters: its x, y, and z position, plus two angles that describe the orientation of the surface normal vector. Use of five DOFs means in principle that we can orient each payload element normal to any surface in the work volume, but without a sixth DOF we have no control of the rotational or “twist” angle of the payload element around the surface normal.
 The original design goal for the IDD work volume was a cylindrical volume in front of the rover directly along the rover's centerline, 50 cm in diameter and is 70 cm high. The actual work volume approximates this goal, but is significantly more complex in shape. Joint motion restrictions and potential collisions with rover structure define keep-out zones within this five-dimensional space. There are also secondary work volumes on the rover body that allow the APXS and Mössbauer Spectrometer to be placed against the Compositional Calibration Target, and also against the capture and filter magnets.
 During launch, cruise, and landing, the IDD is stowed beneath the rover body, and is restrained so that it can sustain launch and landing loads. Pyrotechnic devices are fired after landing to release these restraints, allowing the IDD to be deployed from its launch locks once the rover has successfully traversed off the lander.
 When not in use during surface operations, the IDD is stowed in a position similar to the launch position. Stereo images obtained using the front Hazcams show the IDD work volume and any science targets that may be present in it. Hazcam images are used on the ground to identify target surfaces within the work volume, and to determine their position and orientation relative to the rover. This position and orientation information is then used to generate command sequences that direct the IDD to position payload elements against target surfaces. This commanding approach is simple and robust, but it imposes important restrictions. No closed-loop positioning using Hazcam images onboard is performed, and no deployment of the IDD without analysis of Hazcam images on the ground is permitted.
 The upper arm and forearm links of the IDD are built with titanium tubes and end fittings. Flexprint cables running the length of the IDD carry power and data signals for the instruments and the IDD actuators. All five actuators use identical Maxon RE020 brushed motors. Each has a custom gearbox with a design specific to the requirements of that joint. Heaters are provided on all five actuators, allowing them to be brought to their minimum operating temperature of −55°C even under cold nighttime environmental conditions. This capability will be used routinely to switch from use of one in-situ instrument to another during the Martian night.
 Potentiometers on each joint provide coarse position information, and encoders provide fine position information. The encoder information is used by the IDD control software that runs on the rover CPU to position the payload elements. The software permits all five joints to be operated simultaneously, allowing considerable flexibility in how the IDD can be operated. Each IDD-mounted payload element also carries two redundant contact sensors that can be used to determine when it has been brought into contact with a target surface. This contact information can be used in a closed-loop fashion to terminate IDD movement.
 Motion of the IDD and its instruments can be accomplished in joint space or Cartesian space. Joint space moves are typically used to move the IDD in and out of the stowed-for-driving location, move the RAT to the RAT brushing station located on the IDD forearm, and rotate the instruments on the turret. Cartesian space moves are typically used to place an instrument above a desired target (both rover-mounted and rock and soil targets), place the instrument on a target, and retract the instrument off of a target. Joint space moves can be specified as relative or absolute joint motions while Cartesian space moves can be specified as absolute or relative moves with respect to the rover coordinate frame or tool frame moves relative to an instrument's pointing vector with respect to the coordinate frame attached to the turret.
 Uncertainties in the kinematics of the IDD (knowledge of link lengths, joint offsets, joint positions relative to hardstops, etc.) and inaccuracies in the target positions and orientations as determined from front Hazcam images contribute errors to the positioning capability of the IDD. Hazcam images generally allow determination of surface positions to within 5 mm, and surface orientations to within 5°. When combined with the capabilities of the IDD hardware itself, these uncertainties mean that the IDD can position each payload element within 10 mm of a science target that has not previously been contacted by another in-situ instrument, and to within 10° of the target's surface normal. Once contact has been made, the IDD is capable of repeatably positioning instruments to ±2.5 mm in position and ±3° in orientation.
 For the APXS and the Mössbauer Spectrometer, the IDD's job is simply to place these instruments into physical contact with their targets so that measurement integrations can take place. The other two in-situ payload elements, however, impose other special requirements. The RAT must be placed against rock targets with a preload, and the grinding capability of the RAT increases with increasing preload. The minimum preload requirement for the RAT is 10 N. The IDD can provide ≥20 N over its full work volume, ≥50 N over 60% of the work volume, and ≥80 N over 10% of the work volume. For the Microscopic Imager, which has no focus mechanism, the IDD is used to change the instrument's focus position. The instrument's depth of field is ±3 mm. The IDD provides a minimum controllable motion along a science target's surface normal vector of 2 ± 1 mm RMS, allowing it to image a rough surface in a sequence of images. After placing the MI in position for imaging, the motion of the IDD damps down to an amplitude of less than 30 microns (i.e., less than one MI pixel) within 15 seconds.
4. Additional Science Objectives
4.1. Atmospheric Science
 The primary scientific focus of the Athena payload is the composition of Martian rocks and soils, and what they indicate about past environmental conditions and habitability at the two landing sites. However, Pancam and Mini-TES also have the ability to view the Martian sky, and hence to provide information about the present Martian climate.
 Mini-TES can view upward to provide high-resolution temperature profiles in the Martian atmospheric boundary layer. Temperatures are retrieved from the wings of the 15-μm CO2 band.
 The Martian boundary layer has large temperature contrasts across it, varying from extremely stable at night to extremely unstable during the day. Over sloping terrain the cold nighttime boundary layer may produce unusual drainage winds. Other mesoscale circulation phenomena are also expected. With temperature profiles obtained through diurnal and annual cycles, models of winds and surface stresses can be constrained. Improved understanding of surface stresses could lead in turn to improved understanding of dust storm generation.
 Water abundance may be estimated by upward viewing of rotational H2O lines. Separate measurements of water near the ground might be obtained by viewing distant surface obstacles. The broad librational water ice feature near 800 cm−1 may allow monitoring of ground ice hazes. Together, these measurements may help illuminate the behavior of water in lower atmosphere and of water transport between the atmosphere and surface. Observing the sky, the spectrum of high ice clouds will appear in emission.
 Atmospheric dust abundance may be obtained using the redundant temperature information in the two sides of the 15-μm CO2 band, together with differential absorption across the dust band in that region. Combining dust and temperature data, profiles of atmospheric heating rates can be calculated that will help constrain general circulation models.
 Pancam will also be used to conduct atmospheric observations. The instrument can image the full Martian sky, including direct imaging of the sun. A time series of atmospheric dust opacity in the visible, complementary to the one in the IR from the Mini-TES, can therefore be obtained by imaging the sun through neutral density filters and applying Beer's Law. Systematic diurnal variations in atmospheric opacity, probably due to variations in water ice clouds, have been observed by Viking and Mars Pathfinder [e.g., Pollack et al., 1979; Tomasko et al., 1999]. Pancam sky images and direct determination of atmospheric opacity over two wavelength regions can provide more detailed data on diurnal opacity changes, and hence, in concert with Mini-TES observations, on diurnal variation of water ice or dust aerosols.
 Aerosol properties like mean size, single scattering phase function, and single scattering albedo are important for atmospheric modeling, and can also be obtained from sky imaging [Tomasko et al., 1999]. Variation of sky brightness near the sun defines the diffraction peak of the aerosols, and from that allows approximate determination of mean particle size. Particle phase function can be determined from variation of sky brightness over a large range of angles from the sun. Once particle size and phase function are known, single scattering albedo may be determined from absolute sky brightness [e.g., Ockert-Bell et al., 1997].
4.2. Soil/Rock Physical Properties
 The Athena payload and the MER rovers can also be used to study a variety of physical properties of Martian rocks and soil [Arvidson et al., 2003b].
 Mini-TES observations of rocks and soils will reveal these materials' temperatures. Data obtained over diurnal cycles can be used to determine thermophysical properties (primarily thermal inertia) of Martian materials. Layering in soils, including crusts up to 5 cm deep like those seen in local patches by Viking, should be detectable. These observations can provide information on formation of crusts and deposition of unconsolidated sediments. They should provide useful ground truth for orbital thermophysical observations that average over larger spatial scales.
 Viking Lander and Pathfinder observations showed that the soils at these three Mars landing sites are diverse, with surface deposits of aeolian dust, drifts and dunes, and an underlying indurated deposit that has been called duricrust. Viking Lander observations [Clark et al., 1982] showed that duricrust contains more sulfur and chlorine than loose surface deposits. Many researchers believe that duricrust formed when thin films of water migrated from the subsurface to the surface, evaporated, and left behind salts. Others believe that the duricrust formed as sulfur and chlorine rich volcanic aerosols settled onto the surface under relatively moist conditions that would allow cementation to occur. The origin of duricrust remains uncertain, although the correlation with volatile species and probable association with water remain key elements of extant hypotheses.
 The relevance to the MER mission is that the rovers are capable of excavating in soils to a depth of about half a wheel diameter (approximately 10 cm), well beneath the loose deposits and into the duricrust found at all three Mars landing sites and therefore presumably widespread on Mars. Exposure of fresh subsurface deposits by wheel excavation experiments will help us to characterize the texture, mineralogy, and chemistry of soils at the landing site and along the traverses, in particular, deposits hypothesized to have been modified by aqueous activity. Preflight wheel calibrations experiments will be conducted to be able to understand the relationships among motor voltage, motor current, and wheel torque as a function of temperature.
 A prime subsurface soil experiment will be to lock all but one wheel, spin that wheel, and let it dig into the soil up to a depth of half the wheel diameter. During the experiments, the rover would first characterize the undisturbed surface, conduct subsurface soil excavation experiments, image the wheel, back up, and then survey results with imaging systems and, as appropriate, with Mini-TES and the in-situ instrument suite. The engineering and science instrument data will be used to infer physical properties, chemistry, and mineralogy as a function of depth. Of particular interest is determining whether or not there are stratigraphic horizons in the soil, and whether or not duricrust is a ubiquitous substrate, along with inferring the properties of all units encountered.
 The RAT can also be used to investigate the physical properties of rocks. During operation of the RAT, the rover will monitor currents, temperatures, and encoder readouts for all three RAT actuators. These data can be used to infer information about the strength properties of the rocks that have undergone grinding. A test program is underway with an engineering model of the RAT to establish some of the relationships among these parameters and rock strength.
5. Payload Calibration
 In May of 2001, before calibration of the Athena instruments began, the MER Project convened a Payload Calibration Peer Review Board. This group was drawn from the planetary science community at large, and it conducted a review of calibration procedures for all of the scientific instruments on the Athena payload. The review objectives were (1) to document for users of MER mission data that the Athena instruments are properly calibrated and (2) to assist the Principal Investigator and Payload Element Leads (PELs) in establishing calibration procedures and priorities within budget and time constraints. Separate reviews were held for each instrument, except that the Pancam and MI were combined into one review. The board was chaired by S. Baloga. Members of the board for Pancam and MI were S. Baloga, P. Lucey, K. Klaasen and R. West; for Mini-TES, S. Baloga, W. Calvin, and P. Lucey; for the Mössbauer spectrometer, S. Baloga, B. Fultz, and R. Housley; and for APXS, S. Baloga, B. Clark, and J. Crisp. PELs made oral presentations and provided written documentation of calibration plans. The review board sessions were also attended by the Athena PI, the MER Project Scientist, instrument development engineers, and members of the Athena science team.
 The Payload Calibration Peer Review Board produced a written report to the MER Project, which contained a number of recommendations. One of the key recommendations involved measurement of a well-characterized set of geologic samples by all of the Athena instruments, including a “round-robin” sample exchange program with a subset of samples for a blind test. Such measurements (1) validate, for the Athena science team and the general scientific community, the primary instrument calibration, (2) document the quality of measurement and data reduction procedures on complex geologic samples, (3) provide cross-calibration among Athena instruments and thus enhance the ability of the scientific community to correlate interpretations of Mars data from one instrument to another, and (4) aid in characterizing measurement errors and identifying anomalies in data returned from Mars. This recommendation from the calibration review board has been followed in full.
 To validate the calibration of the Athena instruments, we measured a number of well-characterized rock slabs using actual flight instruments or engineering models. We chose to use rock slabs instead of powders because they are relatively clean, do not pose a dust contamination problem, and present reproducible surfaces to instruments. Rock slabs used during system-level thermal vacuum tests as Mini-TES and Pancam targets were required to be mechanically strong and, because of bake-out requirements for use in a vacuum chamber with the fully assembled MER rovers, mechanically and mineralogically stable at 100°C for 50 hours.
 The rock slabs, sawed from large specimens, were approximately square with side dimensions ranging from ∼3 to 15 cm and thicknesses ranging from ∼0.5 to 1.5 cm. Most slabs had two nearly parallel flat faces, one polished with 60 grit paper and the other polished with 600 grit paper. Some slabs had natural or broken surfaces with the opposite face polished with 60 grit paper. The 60 grit polish approximates the surface finish that will be obtained with the RAT [Gorevan et al., 2003]. The bulk elemental and mineralogical compositions of the rocks were determined by laboratory counterparts of Athena instruments (major and selected minor element concentrations and Fe2+/Fe3+ by x-ray fluorescence and wet chemistry, transmission Mössbauer spectroscopy, thermal emission spectroscopy, and visible, near-IR spectroscopy) and by X-ray diffraction and magnetic properties methods using procedures described previously [e.g., Ruff et al., 1997; Morris et al., 2000, 2001]. In addition, we obtained color digital images of polished rock faces at a resolution of ∼42 μm/pixel using a flat-bed scanner. This resolution approximates the resolution of the MI (∼30 μm/pixel [Herkenhoff et al., 2003]), and the resolution can be degraded to Pancam image resolution. These rock images also were useful in testing of image compression algorithms.
 Approximately 200 rock slabs were fabricated. Their mineralogical compositions ranged from nearly monomineralogic rocks (e.g., olivine, pyroxene, feldspar, amphibole, oxides, sulfates, carbonates, phosphates) to complex assemblages (e.g., basalt and gabbro, andesite and diorite, iron formations, and breccias).
 For the three in-situ instruments (MI, Mössbauer Spectrometer, and APXS), individual rock slabs were of course analyzed one at a time. To date, MI measurements have been completed during instrument-level tests with flight instruments, Mössbauer measurements are underway with an engineering model instrument, and APXS measurements are scheduled to begin in the near future with an engineering model instrument.
 For Pancam and Mini-TES, two targets with multiple rock slabs were built. The rock target used during Pancam instrument-level thermal vacuum tests is shown in Figure 7a. The target consists of 45 individual rock slabs, white and gray scale standards, and red, green, blue, and yellow color standards fixed on an optical breadboard with a flat-black finish. The target was used at room temperature outside the thermal-vacuum chamber and was viewed by the Pancam at a distance of 2.3 m through the optical window of the chamber using a front surface mirror. The second target (Figure 7b) was used during system-level thermal vacuum tests inside the thermal vacuum chamber with both MER rovers. Individual rocks were glued to the aluminum plate, which was painted with high-emissivity, flat-black paint. Strip heaters were glued to the back of the plate to produce a temperature contrast between the rock surfaces and the chamber walls and rover. (Such heaters are necessary to obtain good Mini-TES data in an enclosed vacuum chamber, since the whole chamber forms a large blackbody cavity.) The large rock slabs on the target were sized so that they completely enclosed the Mini-TES beam at the target distance.
 The MER rovers are robotic tools for conducting field geology on the surface of another planet. Field geology is an iterative process of scientific hypothesis formulation and testing, performed in a field setting. Most geologists are used to doing field work on foot, using simple tools and working alone or with a partner. For the MER mission, we have had to develop new processes for dealing with the unique challenges of performing field geology on a distant planet with a robotic vehicle.
 The MER rovers have many shortcomings compared to the capabilities that field geologists are used to having. Their mobility is restricted to tens of meters per sol, even in rather benign terrain. The total traverse distance over the full duration of the mission for each rover is not expected to substantially exceed half a kilometer. Navigation errors and long response times can stretch out to several sols the period of time that is needed to approach a rock that has been observed in remote sensing data. Data bandwidth is many orders of magnitude lower than the equivalent of the human eye-brain combination. And the ability of the rover to manipulate and interact with its environment is extremely limited.
 The challenge of developing an operations approach for these rovers has been to find ways to use the unique strengths of the rovers and their payloads to overcome the unavoidable weaknesses. Pancam has the equivalent of 20/20 visual acuity, but it also provides a visible and near-IR radiometric imaging that can be used to provide remote assessment of some aspects of rock variability and composition that would not be apparent to the naked eye. Mini-TES has excellent mineralogical capabilities, allowing the basic compositional characteristics of geologic materials to be identified and compared from a distance, without having to drive the rover to them. In situ APX and Mössbauer spectroscopy provide compositional data that field geologists typically only obtain by returning hand samples to the laboratory. And the slow pace of operations, combined with the significant budget available to a space flight project, allow the knowledge of many experienced geologists to be brought to bear simultaneously on the scientific problems that are encountered.
 When we first began work on the problem of rover operations, we knew very little about how to conduct remote geologic field exploration with a vehicle that has MER-like capabilities. Our primary learning tool has been JPL's FIDO rover (Figure 8). This vehicle is MER-like in many respects, and in using it over a series of four field deployments we have been able to devise and refine our approach to doing robotic field science.
 The first FIDO field deployment involved a great deal of trial and error. But as we broke our activities down into the equivalent of one-sol segments, we realized that there were some operations that were performed repeatedly, in a consistent way. This led us to a concept of distinct “sol types” that can form the building blocks of a complex multisol science operations sequence. After the first field deployment, and through three subsequent ones, we were able to develop and then refine a simple flowchart that is based on these building blocks and that describes the process of MER science operations at a high level (Figure 9).
 There are five basic sol types, represented by the rectangular boxes in the flowchart (1) Panorama Sol: Acquire remote sensing data of the scene around the rover; (2) Drive Sol: Move tens of meters in the direction of some selected target or area of interest; (3) Approach Sol: Attempt to move close enough to a target to place it within the work volume of the IDD; (4) Spectroscopy Sol: Perform detailed in-situ analyses on a target; and (5) “Scratch and Sniff” Sol: Use the RAT to expose a rock target surface and perform some in-situ analyses on it.
 Most of the MER surface mission will be a sequential ordering of these sol types - in essence, a 90-sol traversal of this flowchart. The five basic sol types are therefore the fundamental building blocks from which most MER science operations are developed. The typical content of each is described in detail in section 6.2 below.
 Operational decisions must be made on both a sol-to-sol “tactical” timescale and a longer multisol “strategic” timescale. The tactical decisions are often simple ones; for example, choosing between an approach sol and a spectroscopy sol based on whether or not some intended target is actually found to be within the work volume of the IDD. The main sol-by-sol tactical decisions faced by the science team are represented by the diamond-shaped boxes on the flowchart, and will be made based on the data products indicated next to each box.
 Decision-making on a strategic timescale requires us to look a number of sols into the future, and to consider how best to test the multiple scientific hypotheses that may be in work at any given time. Our primary tools for strategic planning are “sol trees” that project multiple possible outcomes of mission activities for a number of sols into the future (Figure 10). The sol trees are formulated and refined on a daily basis, considering our progress against mission objectives, all the data obtained to date, and the key scientific questions that we face. Old branches of the sol tree are “pruned” and new ones are added as some scientific questions are resolved and new ones are generated.
 There is significant operational flexibility within each of the five basic sol types. The activities listed on the flowchart for each sol type form a menu from which the specific activities to be executed can be selected. Not every one will necessarily be executed on each sol, and in fact resource constraints (e.g., available time or energy) often prevent all of them from being executed.
 There are activities as part of each sol type that differ from and supplement the main objectives of that sol. A primary example of these are what we call “target of opportunity” activities performed with instruments on the IDD. On panorama, approach, and drive sols, work with the IDD is not the primary objective of the sol. However, front Hazcam images obtained from the rover's start-of-sol position - obtained, for example, at the end of the previous sol - can document the appearance of the work volume of the IDD. There will always be some Martian surface materials within reach of the IDD; these constitute targets of opportunity. If the rover's configuration permits an IDD deployment, one or more targets of opportunity can be investigated with IDD instruments in addition to the main objectives of the sol.
 Some important scientific activities are not included in Figure 9. These include calibration measurements for the Mössbauer Spectrometer and APXS, observations of the Magnet Array by the Mössbauer Spectrometer and the APXS, Mini-TES observations of rocks and soils at multiple times of day to determine thermal inertia, trenching activities that use the rover wheels to expose subsurface materials, and observations of the Martian sky with Mini-TES and Pancam.
 These additional activities will be worked into the main operations flow as the scientific circumstances and the availability of time, power, and downlink bandwidth permit. Mössbauer and APXS calibration measurements will be performed as soon as possible after landing, and again subsequently if necessary. The magnet array will be observed with the APXS and the Mössbauer when Pancam images show that an adequate buildup of dust has taken place. Physical properties experiments like thermal inertia and trenching can be performed at any point, but will require one or more sols of highly focused and specialized operations. Atmospheric observations are mostly independent of the rover's position, and are inserted into each sol as necessary to support our atmospheric science objectives. Any sol type can, in principle, be used to conduct atmospheric observations.
6.2. Sol Types
6.2.1. Panorama Sol
 A panorama sol is used when detailed remote sensing information is needed to establish geologic context or to select candidate targets for subsequent in-situ analysis.
 The main events of a panorama sol are collection of large Pancam and/or Mini-TES panoramas. There is considerable flexibility in how observing time and data volume can be distributed between the two instruments, and in how the observations can be laid out. In some instances it may be beneficial to perform two successive panorama sols from the same location, to provide a more comprehensive view of the scene.
 Although its focus is on remote sensing, a panorama sol can also include target-of-opportunity science with IDD instruments, observing any target that is within the IDD's work volume. The rover does not move on a panorama sol, and the rover's design allows IDD instruments to be used at the same time that Pancam or Mini-TES data are being collected. Panorama sols therefore permit all three of the IDD instruments to be used on targets of opportunity, including long APXS and/or MB integrations if desired. The RAT is not normally used for target-of-opportunity science, since RAT operations require several hours and cannot take place simultaneously with other payload activities.
6.2.2. Drive Sol
 The primary objective of a drive sol is to move the rover from one place to another. In particular, drive sols are used when the distance to be traversed is more than ∼10 meters. Because the rover's navigational errors can be as large as 10%, drive sols typically end with a significant positional uncertainty.
 The first events of a drive sol can include collection of Pancam and/or Mini-TES data on areas of interest around the rover, and also may include target-of-opportunity in-situ observations within the work volume of the IDD. Unlike target-of-opportunity activities on a panorama sol, these must be brief. They therefore are typically limited to MI images and/or short APXS integrations that are aimed at obtaining x-ray data only.
 Once the IDD has been stowed, the rover drives to locations specified in lander-centered coordinates, using a combination of low-level motion commands and higher-level “waypoint” commands in which specific points along the traverse can be specified by the operator. When executing waypoint commands, the rover will autonomously detect and avoid obstacles (e.g., rocks and steep slopes) using its front Hazcams and navigation software.
 In order to help reconstruct the events of the traverse after it has happened, Navcam panoramas can be obtained at points along the traverse. Hazcam images are also acquired at frequent intervals for onboard processing. Some or all of these images can be downlinked at the end of the sol to aid traverse reconstruction, although data bandwidth restrictions are significant. At the end of the traverse, front/rear Hazcam images and a full or partial Navcam panorama are acquired to document the rover's new location and the contents of the IDD work volume. Front Hazcam images that were acquired ∼50 cm before the end of the drive also are downlinked, showing the configuration of the terrain that lies underneath the rover at the end of the drive. Any remaining time and power for that sol can then used to acquire Pancam and/or Mini-TES data.
6.2.3. Approach Sol
 Approach sols are used to attempt to place the rover close enough to a science target that it can be investigated on the next sol with the in situ elements of the payload. Depending on their geometry and their initial distance from the rover, some targets may require two or more successive approach sols before the target is found to be within the work volume of the IDD. Because this work volume is roughly 20 cm in radius (and because rover navigation errors can be as large as 10%), the distance traveled during the final approach sol in an approach sequence often will be no more than ∼2 meters.
 An approach sol can begin with Pancam and/or Mini-TES investigation of areas of interest, and/or target-of-opportunity MI and/or APXS observations on some area within the arm's work volume. The rover is then commanded to drive to a location what will place the target that has been selected for detailed in situ analysis within the work volume of the IDD. Again, front Hazcam images acquired ∼50 cm before the end of the show the configuration of the terrain under the rover. The sol concludes with Navcam and Hazcam images to document the new position, and Pancam and/or Mini-TES data, usually in the direction of the intended target. The front Hazcam images acquired at the completion of the drive are particularly important, since they reveal whether or not the target is actually within the IDD's work volume.
6.2.4. Spectroscopy Sol
 The purpose of a spectroscopy sol is to perform detailed in-situ investigation of a rock or soil target.
 A spectroscopy sol can begin with Pancam and/or Mini-TES observations of areas of interest around the rover. If resources allow, brief target-of-opportunity science with the MI and/or the APXS can be performed on some target other than the main spectroscopy target. The main events of the sol are observations of the main target with the Microscopic Imager, the Mössbauer Spectrometer, and the APXS. The effort required to get a selected target within the IDD work volume is substantial, so all three in situ instruments normally are used. MI imaging takes place first, followed by long (∼10-hour) integrations with the Mössbauer and the APXS. APXS measurements are normally performed last, taking advantage of the cold late-night/early-morning temperatures to maximize the performance of the x-ray mode. The switch from Mössbauer to APXS takes place during the night, requiring a rover wakeup and heating of the IDD actuators to above their minimum operating temperature. All IDD activities on a spectroscopy sol are documented with front Hazcam images that show the instrument placement. Because the Mössbauer-to-APXS switch takes place in the dark, the APXS placement is documented with Hazcam images on the morning of the next sol, before the IDD is stowed.
 After the final sol of in situ investigation of a target has been completed, a crucial event on the next sol, before leaving the target for good, is to acquire high-resolution Pancam and Mini-TES data for the target. These data provide a complete suite of observations with all five Athena instruments, at the highest resolution possible. For many targets this requires that the rover drive backward a few tens of cm so that the target is placed within the fields of view of both Pancam and Mini-TES. Whenever such observations are made, the risk of leaving the target without verification on the ground that the data were obtained successfully must be weighed against the time that would be lost in waiting for such verification.
6.2.5. “Scratch and Sniff” Sol
 “Scratch and sniff” sols are used to expose a fresh rock surface with the RAT, and then to conduct in-situ analyses on that surface.
 A scratch and sniff sol can begin with Pancam and/or Mini-TES investigation of any areas of interest around the rover, and/or a Microscopic Imager and/or APXS sequence on some area within the arm's work volume. The target typically would be the same spot to be exposed by the RAT, if it has not already been investigated on a prior sol. The main event of the day is the use of the RAT to grind away rock over an area 4.5 cm in diameter. The nominal depth of grinding is 5 mm. In some geometries grinding to slightly greater depths may be possible. Shallower depths may also be used if circumstances warrant it. Once grinding has been finished, the remaining time and power for the sol can be devoted to some combination of Microscopic Imager, Mössbauer Spectrometer, and APXS data collection on the exposed surface. As is the case for a spectroscopy sol, all IDD activities are documented as fully as possible with front Hazcam images.
 Note that the rover operations flow depicted in Figure 9 allows for considerable flexibility in RAT use. For example, multiple cycles of rock abrasion - grinding a short distance followed by in situ observations and then more grinding - are possible if the scientific circumstances warrant it. Such activities take significant time, however, and we will have to balance the scientific return against the intense time demands imposed by the short MER mission duration.
6.3. Other Operational Issues
6.3.1. Impact Through Egress
 The period of time at the beginning of the landed mission, from lander impact through rover egress from the lander, is different from other mission phases. At landing, the rover is folded up inside the lander, with the PMA and the IDD stowed. Of course, the primary focus during impact-through-egress is on the engineering activities needed to support deploying the rover and getting it off the lander. However, some important science activities are conducted at this time. The PMA is deployed on Sol 1 (the landing sol), making remote sensing possible very early. Pancam and Mini-TES each undergo basic health checks, and each view their respective calibration targets. Pancam will probably be used to acquire some images of the scene around the lander even before the rover “stands up” - i.e., before the mobility system is deployed and the rover is at its full height.
 After stand-up, three key scientific observations will be made. Pancam will be used to acquire a high resolution color stereo panorama of the full scene around the lander, and Mini-TES will be used to acquire a full panorama as well. These two data products document the terrain around the lander, and provide a framework of observations within which the first sols of deployed surface operations will occur. Third, Mini-TES will be used to observe the Martian sky, providing the first upward sounding the Martian atmospheric boundary layer. Because the Mini-TES data in particular are of such high science priority, they will be downlinked in full as a risk mitigation measure before egress occurs.
 The IDD cannot be used until after egress has taken place. However, pre-egress health checks of the IDD-mounted instruments are possible, including Mössbauer spectra of the internal reference target and APXS spectra of the Martian atmosphere (the APXS is launched with its doors latched open). Also during this timeframe, Pancam will be used to provide baseline images of the filter, capture, and sweep magnets, as well as the rovers' solar panels, before dust accumulation occurs.
 As soon as possible after egress the IDD will be deployed and the APXS and Mössbauer Spectrometer will make measurements on their respective calibration targets.
6.3.2. Support Imaging
 Navcam and Hazcam are not formally part of the Athena payload, but support imaging with these cameras is an essential part of MER science operations. Extensive support imaging is crucial for determination of rover position relative to both prospective science targets and terrain obstacles. Whenever possible, a complete set of Navcam (360° panorama) and Hazcam (front and rear) images is acquired at the end of every rover move. Determination of rover position after a traverse, as well as analysis of events that occurred during a traverse, is aided significantly by searching for and identifying rover tracks in end-of-traverse images. Hazcam images of the IDD work volume must be acquired after every rover move and analyzed on the ground before the IDD can be deployed safely from the new rover position.
 Navcam panoramas can be particularly useful for efficient planning of Mini-TES and Pancam activities. If Navcam data exist from a new rover location, one of the most effective ways to use Mini-TES on a sol when other activities are planned is to start the sol with small Mini-TES rasters or linear scans on targets that have been selected in Navcam data. If the targets are chosen on the basis of their appearance to test specific scientific hypotheses, such Mini-TES observations can provide as much science value as a large Mini-TES panorama, but in much less time. Similar time savings can be afforded by using Navcam panoramas to target high-resolution multispectral Pancam coverage.
6.3.3. Pre-uplink Activities
 One of the most important characteristics of the MER rovers is that each sol's uplink is not received onboard and ready for execution until an hour or more after the rover wakes up. This means that “overnight” operations planning must in fact include both the post-uplink activities for the coming sol and the pre-uplink activities for the sol after that. Other key characteristics of the rovers are that front Hazcam images of the IDD work volume must be acquired before any IDD deployment, and that Navcam images can be necessary for planning accurate pointing of some Pancam and Mini-TES observations. Because the Hazcam and Navcam images that are needed to plan some pre-uplink activities of sol N + 2 may not be available on sol N if the rover is going to move on sol N + 1, some special considerations apply.
 1. If sol N + 2 follows a panorama sol, a spectroscopy sol, or a scratch and sniff sol (none of which involve moving the rover), there are no restrictions; all the necessary support images to support IDD operations or PMA pointing should be in hand.
 2. If it has been decided that sol N + 2 is going to be a panorama sol, which does not involve rover motion, then panoramic data acquisition on sol N + 2 can begin immediately upon rover wake-up. Note, however, that any target-of-opportunity in situ science on that sol cannot begin until after the uplink has been received.
 3. If sol N + 2 follows an approach sol, rover position knowledge will be imperfect. Pointing of pre-uplink Pancam and Mini-TES observations therefore must be done based primarily on rover attitude knowledge, plus images acquired from the rover's location before the approach. Such observations must be designed with likely position errors, and hence pointing errors, in mind. In general, study of distant targets is favored under such circumstances, since the azimuth to a distant target in Mars surface coordinates will have changed relatively little due to the rover motion.
 4. If sol N + 2 follows a drive sol, knowledge of the scene around the rover may be poor. Pre-uplink use of Pancam and Mini-TES on such a sol is best performed in a panoramic fashion, surveying a large portion of the scene rather than trying to point at small targets whose positions may be poorly known or unknown.
6.3.4. Low-Power Operations Near Mission End
 Near the end of the mission when output from each rover's solar arrays is diminishing, operations will be changed to conserve electrical power and extend mission duration. Pixel summing can be used to reduce the acquisition time for image panoramas. Even summed at 3 × 3 pixels - reducing data volume by almost an order of magnitude - Pancam images will still have resolution comparable to the highest resolution images returned from past Mars surface missions. Mini-TES panoramas can be acquired using fewer 1.8-second integration periods per pixel, diminishing signal-to-noise ratio, or they can be acquired with larger azimuth and elevation step sizes between pixels, diminishing spatial resolution. APXS and Mössbauer data can be acquired with shorter integration times, decreasing signal-to-noise ratio for these instruments. Note, however, that good x-ray data, providing the abundances of nearly all the major rock-forming elements, can still be obtained with very short integration times. And by limiting operations to the warmest parts of the Martian day, requirements for heating of actuators and other payload elements can be diminished, further lengthening the mission.
 The MER rovers, each equipped with the Athena payload, are robotic tools for field geologic exploration. Their primary scientific objective is to explore two sites on the Martian surface where water is likely to have been present in the past, and to read the geologic record there. Companion papers in this volume describe each of the scientific elements of the payload in more detail. We have devised science operations processes for the rovers that should allow us to formulate and test scientific hypotheses at each of the MER landing sites, and to assess past environmental conditions at those sites and their suitability for life.
 The Athena payload is the work of literally hundreds of dedicated engineers and scientists, all of whom have our heartfelt thanks for their efforts. Sam Dallas of the Jet Propulsion Laboratory managed the original Athena mission proposal in 1995–96, and provided invaluable engineering insights during the earliest stages of payload development. Barry Goldstein of JPL held the same role on the Athena payload proposal written in 1997, and provided superb engineering leadership for the development of the payload up until the start of the MER Project in 2000. Joel Rademacher managed the development of the Mini-TES, APXS, and Mössbauer payload elements, Mark Schwochert led the engineering teams for Pancam and the Microscopic Imager, and Steve Kondos and Mark Johnson managed the development of the Rock Abrasion Tool. Innumerable other scientists and engineers contributed their time and talents as well, and the on-time delivery of two fully tested and calibrated science payloads is testimony to their efforts.