The primary objectives of this investigation are to: 1) observe martian atmospheric processes at global scale and synoptically, 2) study details of the interaction of the atmosphere with the surface at a variety of scales in both space and time, and 3) to examine surface features characteristic of the evolution of the martian climate over time. These objectives will be met through the acquisition of images over a range of time- and spatial scales.

2.1. Atmospheric Studies

Atmospheric investigations will include studies of: a) the distribution of dust, condensates, and ozone in the atmosphere, b) the structure of condensate clouds and their relationship to topography and to circulation patterns predicted by global circulation models (GCMs), c) dust storms ranging in magnitude from localized clouds to planet encircling storms, and d) polar phenomena.

Among the major scientific questions to be addressed are:

1. What is the seasonal and geographical variability in dust, ozone, and condensates in the martian atmosphere?

2. What is the general circulation of Mars? How reliable are the predictions of GCMs, and how much interannual variability exists?

3. What role do clouds in general, and the polar hoods in particular, play in the global water cycle on Mars? Do they control the net interhemispheric transfer of water?

4. How do local, regional, and global dust storms evolve? How are they related to the local, meso-, and synoptic-scale circulations?

5. How are the seasonal and residual caps involved in the global transport of water and dust on Mars? What is the direction of the net annual water transport on the planet?

2.1.1. Clouds

Clouds probe many important properties of the martian atmosphere and may be used to test dynamical atmospheric models. The distribution of clouds depends in part on the global distribution of atmospheric water (both absolute content and relative saturation), and can be used to extract information about the water distribution when independent constraints can be obtained from atmospheric temperatures. The distribution of clouds also depends on the regional and local dynamics of atmospheric circulation, in turn dependent on heat transport processes in the atmosphere that respond to the transport of dust and condensates. Thus, clouds can also be used to probe these transport mechanisms. Global, synoptic observations are the best way to study the geographical and temporal distribution of martian clouds in a manner that allows them to be used to constrain circulation models. In addition, multispectral observations, when combined with modeling of the physical and chemical attributes of atmospheric constituents that affect color, can address questions of cloud optical depths and water content.

2.1.2. General Circulation

The general circulation of the martian atmosphere determines the transport of dust and water around the planet which can significantly affect the rate at which CO₂ condenses at the poles (Pollack et al., 1990; 1993). Studies of the general circulation on Mars are relevant to terrestrial GCMs because of the similar diurnal and seasonal cycles but different compositions and masses for the two planets. Unfortunately, observations pertinent to the general circulation on Mars lack the spatial and temporal density needed to describe it. Thus, a primary goal of this investigation is to help describe the structure and variability of the general circulation.

Winds can be determined from imaging data by tracking cloud systems and dust storms, and can also be inferred from topographically generated gravity wave trains whose wavelengths are diagnostic of the local wind field. Of course, such observations are unable to describe the full three-dimensional structure of the general circulation, since they depend on aerosols and condensates as "tracers" to "see" the wind field. These materials actively participate in creating the conditions that generate the field in the first place. However, both at global and regional scale the effects of these materials can be relatively small. Previous experience on Mars suggests that they can be used to follow atmospheric motion.

One very important feature of the general circulation that will be addressed is the nature of traveling baroclinic wave systems. GCMs suggest that they may be much less vigorous in the southern hemisphere during winter than in the northern hemisphere, owing to hemispheric differences in topography (Barnes et al., 1993). It is necessary to characterize the dominant zonal wave numbers, phase speeds, and meridional structure of these systems in both hemispheres if the general circulation models are to be validated.

Models also predict that, at least in the northern hemisphere, baroclinic disturbances should grow and decay in preferred longitudinal regions called "storm zones" (Hollingsworth et al., 1995; Barnes et al., 1995). In these regions, the poleward fluxes of heat and momentum are maximized and dust and water transport should be as well. Topography is mainly responsible for the stationary eddies and should produce inter-hemispheric differences in storm zone patterns. Atmospheric opacity and cloud frequency and optical properties should reveal these zones as persistently dustier and cloudier regions. Models also predict that the martian Hadley cell is much more longitudinally variable than previously thought. In particular, the meridional flow in the lower branch of the Hadley cell should be channeled into narrow currents along the eastward flanks of major topographic rises such as Tharsis. Imaging data can be used to test this model prediction.

2.1.3. Interannual Climate Variations

Ground-based temperature profiling and Hubble Space Telescope (HST) imaging since 1988 show much colder atmospheric temperatures and lower dust loading than during the Viking mission (Clancy et al., 1990; James et al., 1994). This distinction appears most significant for the northern Spring/Summer season of Mars (around Mars aphelion); the typical behavior of the Mars aphelion atmosphere may include low dust loading, reduced atmospheric temperatures, low altitudes water vapor saturation, and a 10° S to 30° N global belt of moderate opacity clouds (Clancy et al., 1995a). Low altitude saturation may impede transport of water into the southern hemisphere during this season and thereby have a significant impact on the global water distribution. The Mars Surveyor '98 Orbiter mission permits an imaging experiment to make observations relevant to several important aspects of this purported "new climate". Not only can the direct seasonal variation of clouds be observed, but also indirectly related properties such as limb measurements of atmospheric ozone concentration which, by providing a proxy for water saturation altitudes, can be used to address the variability issue.

2.1.4. Dust Storms

Global maps are important for monitoring all but the smallest scale dust activity on Mars. Viking observations showed that local sand or dust storms are fairly common in certain locales and during certain seasons on Mars, e.g. at the edge of the south polar cap in Spring, in Hellas and Argyre Basins, in the Solis Planum region, etc. On rare occasions during the Viking mission, single storms were monitored with sufficient temporal sampling density to determine the evolution of these events. However, major questions remain about the diurnal development of local storms, especially in the polar regions.

The expansion of global storms has been observed from Earth as well as by Viking using sequences of images (Martin, 1976), and the path of a large regional storm in the northern hemisphere was deduced from Viking sequences obtained on three consecutive days in 1978 (James, 1985). Observation of Mars for one year at a scale of better than 10 km should be sufficient to define the global dust cycle for that year and help to identify sources and sinks for dust. Coupled with a successful MGS MOC experiment, this would provide complete records of two consecutive seasonal dust cycles that could then be applied to addressing the circumstances that lead to planet encircling storms and why the dust cycle shows such interannual variability.

There also remain significant issues regarding dust processes at smaller scales. Little is known about dust devils of the type observed by Thomas and Gierasch (1985) and, indirectly, by Ryan and Lucich (1983) using Viking Lander meteorology. Regions where local dust storm activity is known to occur, such as preferred longitudes at the edge of the receding spring caps, are likely sites for dust devil activity. Inspection of these areas might shed light on how local storms are started, and how topography, surface, local sources of dust, and other physical factors participate in their generation.

2.1.5. Polar Processes

The annual cycles of the martian polar caps are the most obvious seasonal signpost on the planet, having been recognized as such by Herschel (1784). A good data set at 1 km/pixel resolution was obtained by Viking Orbiter 2 for the south cap (James et al. 1979), while the best spacecraft views of a north polar recession were obtained by Mariner 9 and HST. Neither spacecraft nor Earth-based imaging have provided much information on the deposition of either cap because most condensation takes place during polar night. However, this does not mean visible monitoring is unimportant. Global monitoring using a multispectral, wide angle camera for an entire martian polar cycle would provide important clues to polar processes by showing the relationships between the polar hood and cap edges, even if the waxing pole were not illuminated.

The observational situation is somewhat better for the residual, late summer phases of the polar caps. Untangling the physics of the residual caps is one of the most interesting problems in Mars science because of the dichotomy in composition between the water ice north cap and the (at least partially) dry ice south cap. Involved in this puzzle are clues to the water cycle on Mars; net annual transfer of water from one hemisphere to another may be revealed by interannual changes in the residual caps. Such changes were seen in a comparison of the very good Mariner 9 and Viking data sets, and two additional data sets from MGS and the Mars Surveyor '98 orbiter could provide not only two additional reference points for identification of possible monotonic or cyclic variations with periods greater than one Mars year, but also such reference points at considerably finer temporal and spatial resolution. The residual polar caps should be monitored at medium resolution (tens of meters per pixel) in order to study the fractional coverage of the surface and to estimate the thickness of the frost deposits.

Observing the reflectivity of surface frost at several wavelengths and at a variety of photometric angles is pertinent to deducing the nature of the ice surface, and changes in the frost albedo as a function of wavelength can be compared to the atmospheric opacity to attempt to correlate changes in the frost surface with dust deposition and other seasonal and geographic variables. Various regions near the edge of the seasonal caps should be observed on consecutive days at medium resolution to high spatial resolution in order to observe the sublimation (and condensation) processes as the edge of the cap passes those particular points. In addition, regions in the cap's interior should be observed to test the hypothesis that albedo variations within the cap are due primarily to fractional coverage of the surface. Specific locations within the cap such as the Mountains of Mitchel, craters with frost streaks, and craters with dunes within them are excellent candidate locations, as they have been observed on previous missions to lend credibility to the aforementioned hypothesis, although not with the time and space resolution afforded by the Mars Surveyor '98 opportunity.

2.1.6. Atmospheric Opacity

A particularly exciting capability that is proposed here is mapping the entire planet, daily and at a nadir scale of better than 7 km/pixel, at ultraviolet wavelengths. UV imaging has the advantage for atmospheric purposes of minimizing the contribution to the reflectance from the surface of Mars because of its very low UV albedo and lack of variation in same. Most of the reflectance at these wavelengths results from scattering and absorption in the atmosphere. Inasmuch as the Rayleigh contribution can be calculated exactly, contributions from aerosol scattering can then be isolated to the extent that the phase functions for those processes are known or can be approximated. In addition, comparison of reflectances at, for example 250 nm (deep within the Hartley band) and 330 nm (which is unaffected by ozone absorption), can be used to estimate atmospheric ozone, which also constrains the amount of water vapor through photochemistry. This method has been successfully used in analyses of the HST Mars data set (James et al., 1994)

2.2. Studies of Surface/Atmosphere Interactions

Given the current lack of fluvial activity on Mars, processes involving surface/atmosphere interactions (SAI) are the primary agents of surficial change at present. Manifestations of SAI include regional and global-scale dust storm activity, the variability of surface albedo features related to eolian transport of dust and sand, and the condensation/sublimation of volatiles at the surface. The accumulated historical evidence (ground-based, spacecraft, and HST observations) indicates that a great deal of seasonal and interannual variability exists in eolian activity on Mars. Returning data for more than one Martian year will significantly expand the temporal observations that are crucial to understanding the seasonal, interannual, and long-term efficacy of SAI. In addition, multi-wavelength capabilities can yield observations not possible from previous spacecraft. The combination of regular, repeatable global mapping (using a wide angle camera) and selected detailed observations (with a medium resolution camera) allows the temporal and spatial resolution, and the multispectral observations, necessary to significantly expand understanding of many of the details of SAI on Mars.

2.2.1. Dust Storm Activity

Dust storms are the primary source of short-term changes in the appearance of Mars at present, providing the driver for the numerous surface features attributed to eolian processes. Understanding eolian processes on the planet requires knowledge of the location, genesis, timing, and frequency of dust storms. The historical record indicates that local dust storms can occur at any season, while great dust storms of global extent are most common near Mars' perihelion (Slipher, 1982; Zurek, 1982; Martin, 1984; Peterfreund, 1985); it is also evident that a great deal of interannual variability exists in the timing, location, and extent of dust storms.

Regularly repeated global mapping by the low resolution, wide angle camera is the primary method of monitoring dust storm activity, being of sufficient spatial and temporal resolution to allow detailed mapping of dust storm timing, location, and evolution. Coupled with contemporaneous information on dust loading obtained by PMIRR, camera observations from the Mars Surveyor '98 orbiter will provide unprecedented insight into the processes involved in Martian dust storms. Where dust storms are detected, observations at medium resolution can be targeted for detailed investigation of the surface where storms arise. A multispectral capability is important for distinguishing between dust and condensate clouds, and for investigating mixtures of the two. Equally important will be the expansion of the historical record of dust storm monitoring, adding more than one Martian year of semi-continuous monitoring. If dust storms are observed by MGS during the preceding Mars' year, Mars Surveyor '98 will be in the position to provide detailed characterization of the source regions and give insight into the repeatability of dust storm activity.

2.2.2. Variable Features Resulting From Dust Transport

There is abundant evidence that, at present, eolian processes are active over most of the surface of Mars (cf. Veverka et al., 1977; Thomas et al., 1981). A wide variety of surface features, such as wind streaks, sand dunes, and regional albedo features, are attributed to eolian deposition and erosion. Previous studies have demonstrated that variations in regional albedo and wind streak patterns are indicative of sediment transport through a region (e.g., Figure 2(Lee et al., 1982, 1994; Kahn et al., 1992), while thermal inertia data (derived from Viking IRTM and to be derived at high spatial resolution by the TES on MGS) are indicative of the degree of surface mantling by dust deposits (Kieffer et al., 1977; Christensen, 1982, 1986a,b, 1988; Jakosky, 1986). Experimental studies (Wells et al., 1984) show that small amounts of dust deposited/eroded from the surface can explain the observed surface albedo changes (modifying surface dust cover by less than a micron of dust can alter the albedo by several tens of percent). The colors of martian albedo features can also be related to variable amounts of dust cover (Soderblom et al., 1978; Singer and McCord, 1979; McCord et al., 1982a,b). Visual data are therefore diagnostic of net erosion or deposition of dust-storm fallout that is taking place currently; combined with available thermal data, inferences can be made as to whether such processes have been active in a region over the long term.

GIF = 78 KBytes
Figure 2: Variations in a Major Albedo Feature on Mars (Cerberus)

While the basic distribution, season of formation, effective wind directions, and perhaps meteorological conditions involved in the formation of variable features have been documented with previous observations, the actual conditions of dust entrainment and deposition related to these features have remained elusive. The orbit, and proposed payload, of Mars Surveyor '98 is particularly well suited to provide the detailed temporal, spatial, and multi-wavelength studies needed to significantly advance understanding of these processes; the scientific yield will be further enhanced by coupling with observations of atmospheric dust loading to be made by PMIRR. Regularly repeated global mapping by a low resolution camera is ideal for monitoring regional albedo features and for detecting areas undergoing albedo variations. Observations by a medium resolution camera can then be targeted for detailed examination of currently active variable features. Such observations are pertinent at a variety of scales, allowing comparisons to general circulation models (global scale wind patterns) as well as investigation of individual features (local winds, cf. Magalhaes and Young, 1995).

2.2.3. Variable Features Resulting From Sand Transport

Sand dunes are markers of a specific physical process, saltation, and hence are powerful indicators of both short-term and accumulated surface wind stress. Their wide spread, but not ubiquitous, distribution on Mars allows a wide sampling of eolian transport conditions (Thomas and Weitz, 1989; Thomas and Gierasch, 1995; Greeley et al., 1992). Multispectral observations by the medium resolution camera can address several key questions related to Martian dunes, such as interannual variability (based on comparisons with Mariner 9, Viking, and MGS observations of specific dune fields), components of dunes and their relationship to sediments in the source deposits, detailed morphology of dune fields and the relationship to local, regional, and global atmospheric circulation (i.e. using dunes as wind, paleowind, and climate indicators).

2.3. Surface Studies

The Mars Surveyor Orbiter permits imaging studies of surface features and processes that were not possible with the Viking Orbiter or MGS cameras. Of most interest to surface studies is the ability to acquire moderate resolution data through multiple filters over sufficient area to provide both morphologic and surface unit information.

2.3.1. Morphology and Pattern

2.3.1.1. Eolian Sediments and Morphology

Eolian deposits on Mars range from micron thick dust coatings to multi-km thick polar deposits. Eolian transport has been important in moving material for much of Mars' history and the transport has been global in scale. In the previous sections, eolian materials were discussed as measures of past and present atmospheric phenomena and tracers of surface/atmosphere interactions. Here they are considered in their own right, as deposition and erosion of eolian materials are geologic processes as well, and provide great insight into the geological history of Mars. Deciphering the materials and stratigraphy of the wide variety of eolian deposits on Mars is crucial to understanding the geological influences of climate on Mars and the relationship of current processes to past ones. Appropriately selected pixel scales, image formats, and spectral wavelengths will allow investigation of eolian materials and stratigraphy not possible with previous Mars missions, including MGS.

2.3.1.2. Polar Deposits

The polar deposits, thought to be climatically sensitive because of their association with frost deposition, include at least 2 non-volatile components as indicated by Viking Orbiter color data (Herkenhoff and Murray, 1990; Thomas and Weitz, 1989). The layered deposits appear to reflect cycles of deposition and erosion on a variety of time scales, and their overall extent has been reduced from previously larger deposits (Thomas et al. 1992). Both the makeup of the layers, the relationship of unconformities to the various layers and other sedimentary units, and the erosional forms at several scales around the deposits will be incompletely explored before Mars Surveyor '98. Coverage of selected areas of the layered deposits, chasmata, deposits marginal to the layered deposits, and polar dunes with four or five filters between 450 and 900 nm can provide far more discrimination of surface units than was possible using Viking images. Even with two broad band filters centered at 450 and 590 nm, Viking was able to discriminate several stages of mixing of at least three end members in many regions of Mars. With filters giving some discrimination between 800 and 900 nm (where iron oxidation state greatly varies reflectances), it is possible to anticipate at least an order of magnitude improvement in spectrally mapping exposed units, which the Viking, and even HST data (Bell et al. 1995) suggest are present in considerable complexity and discernible because of patchiness in dust covering. Spatial resolution of tens of meters permits local morphology to be associated with color units. This is particularly useful in relating materials exposed on dunes and wind streaks to current wind regimes, and relating morphologies to source areas as can be done in terrestrial dune fields that have multiple sources (Blount et al. 1990). The primary questions to be addressed include: How many different materials make up the polar deposits and on what scales are they segregated? Are the layers distinct materials or distinct surface textures? Are the peripheral sediments sources of, or derived from, the polar layered deposits? How do the sediment transport patterns shown by older deposits compare to present day winds? Is there a correlation of size of features (reconstitution time?) and orientation?

2.3.1.3. Mid- and Low Latitude Deposits

Transport of dust and sand at mid- and low latitudes is presently dominated by Hadley circulation in southern Summer (Greeley et al. 1993; Thomas and Gierasch, 1995). Many dune deposits, wind streaks, and other deposits of likely eolian (or possibly lacustrine) origin occur in these latitudes. As with polar deposits, the number of materials involved, and their scales of interlayering, are unknown. A primary goal of any study of such deposits is the search for indications of relict forms with local or global asymmetry different from that of the present dominant flow. Such asymmetry may be expected because the present winds are very hemispherically asymmetric, and are expected to switch with the 51,000 year cycle of perihelion precession. Multicolor observations at moderate to high resolution are needed to map the morphologies and color units in key areas identified from earlier Viking Orbiter images. Study of mid-to-low latitude deposits is required at all scales that may reflect regional transport of materials or that may improve knowledge of the eolian transport regime on Mars. The significance of these older deposits is necessarily closely tied to present day variable features (See Section 2.2).

2.3.1.4. Non-eolian Morphology

Although present material transport on Mars is dominated by wind, the surface shows effects of volcanic eruptions of lavas and pyroclasts, standing and flowing water, faulting, mass wasting, thermokarst, and possibly glacial action. Volumetrically, the morphology of Mars' surface is dominated by volcanoes, canyons, channels, and impact craters, not eolian deposits. The global distribution, context, and general morphology of the non-eolian forms has been well established by Mariner 9 and Viking, and will be examined in extremely limited locations at very high resolution by MGS. There remains, however, a gap in resolution coverage, as the global/regional resolution of Viking data are not generally much better than MGS wide angle observations (100's of meters). Much to be accomplished by regional mapping at 10's meters resolution over sufficient areas to supply context for individual forms, many of which at present are assigned interpretations that range from glacial to structural. As with the eolian studies, multicolor images will be important in discriminating differences in units in which these landforms have developed.

2.3.2. Composition

The measurement of normal albedo and relative reflectivity spectra of areas on Mars can provide qualitative and sometimes quantitative information on the composition and relative maturity of the terrain. The materials that control the reflectivity of the martian surface are primarily Fe-bearing minerals in various oxidation and hydration states Therefore, spectrophotometry in the near-UV, visible, and near-IR can give compositional information complementary to the mineralogical and petrological observations to be made by the MGS TES.

The reflection spectrum of Mars in most areas increases monotonically with wavelength, with more or less the same shape, throughout the visible portion of the spectrum. To first order, the spectra of bright areas and dark areas are remarkably similar, varying in magnitude but not relative shape. Such spectra differ, however, in the details, with broad absorption bands of low contrast (in general, these absorptions have relative magnitudes measured in the few percent). It is these differences that permit different materials to be discriminated. Only a few absorption bands are distinguishable in the visible spectrum of Mars; a few filters are sufficient to delineate their positions. Thus, it is possible to distinguish compositional differences regardless of the specific or quantitative interpretation attached to these spectral features.

It is important to note that it is not the primary objective of this imaging experiment to provide a detailed understanding of the composition of the surface of Mars. The technical and cost challenges associated with developing hyperspectral imaging systems for flight are far beyond the scope of this effort. Rather, it is the intent here to acquire multispectral (or multicolor) observations to supplement the MGS TES observations in the thermal infrared with higher spatial resolution "unit" maps. These maps reflect the mineralogical composition, but are more or less independent of the ability to devine percentages of specific minerals at high spatial resolution from specific absorption features or sets of features. Observations from 400 to 1000 nm are generally sensitive to variations in the crystallinity and oxidation state of iron-bearing minerals, and as such will represent primarily the distribution of weathering products, and secondarily the occurrence of bedrock.

The martian regolith can be considered to consist of three types of materials (e.g., Singer et al., 1979): (1) crystalline rock fragments, at a variety of sizes, consisting of basaltic materials rich in clino- and orthopyroxenes (ferro-augite, pigeonite, etc.; enstatite, hypersthene, etc.) (e.g., Singer et al., 1979; Bell et al., 1990; Mustard et al., 1993) and probably other Fe-bearing minerals, including, for example, olivine in accessory amounts (Bell et al., 1993; Geissler et al., 1993; Mustard and Sunshine, 1995); (2) weathering products, in particular including ferric oxides/oxyhydroxides such as goethite and hematite; and (3) as yet poorly characterized salts and other non-silicate minerals of undetermined origin, composition, and abundance.

Table 1: Some Absorption Features in the Visible and Near-Infrared

   Center                       Origin of Absorption
 Wavelength
    (nm)
                                                            
     430     "Blue"; Center of Fe³⁺ electronic transition band in ferric
	     oxides/oxyhydroxides

     500     Edge of deep near-UV absorption in goethite and lepidocrocite

     550     "Green"; Edge of deep near-UV absorption in hematite

     650     Center of Fe³⁺ ET band in ferric oxides/oxyhydroxides; also near
             strong Fe²⁺ -> Fe³⁺ charge transfer (CT) band in many
             iron-bearing silicates

     750     "Red"; Local maximum in ferric oxide/oxyhydroxide spectra

     860     Center of Fe³⁺ electronic transition (ET) band in hematite

     900     Center of Fe³⁺ ET band in goethite; also center of "1-µ m" Fe²⁺
             band in orthopyroxenes (e.g., enstatite, hypersthene)

     930     Center of "1-µ m" Fe²⁺ band complex in low-Ca clinopyroxenes
             (e.g., pigeonite)

     950     Center of "1-µ m" Fe²⁺ band complex in moderate-Ca  clinopyroxenes
             (e.g., augite); also center of  Fe³⁺  ET band in maghemite,
             lepidocrocite

    1000     Center of "1-µ m" Fe²⁺ band in high-Ca clinopyroxenes  (e.g.,
             diopside); also near center of Fe²⁺ feature in  Mg-olivine
             (forsterite)

The spectra of martian soil can be interpreted primarily in terms of absorption by Fe in various forms in the surface materials (Table 1). Three zones of absorption-related features are superimposed on the shape of the curve in the visible and near-infrared: on the blue end is the edge of near-UV absorptions of Fe oxides, which shifts blue-ward with hydration; in the center are various absorptions associated with Fe³⁺ electronic transitions in ferric oxides/oxyhydroxides; and at the red end (at and beyond 900 nm) is the famous "1 µ m" Fe²⁺ absorption found in many primary Fe-bearing silicate minerals. Characterization of the martian surface with respect to these broad compositional types, at tens of meters resolution, would greatly contribute to understanding of martian geology.

2. SCIENTIFIC OBJECTIVES