|
ATMOSPHERE AND CLIMATE STUDIES OF MARS USING THE MARS OBSERVER PRESSURE MODULATOR INFRARED RADIOMETER
D. J. McCleese, R. D. Haskins, J. T. Schofield, and R. W. Zurek
Earth and Space Sciences Division Jet Propulsion Laboratory California Institute of Technology Pasadena, California
C. B. Leovy
Department of Atmospheric Sciences University of Washington Seattle, Washington
D. A. Paige
Dept. of Earth and Space Sciences University of California Los Angeles, CA
F. W. Taylor
Dept. of Atmospheric, Oceanic and Planetary Physics Oxford University Oxford, England
February 16, 1992
ABSTRACT
Studies of the climate and atmosphere of Mars are limited at present by the lack of meteorological data sets having systematic global coverage with good horizontal and vertical resolution. The Mars Observer spacecraft, in its low, nearly circular, polar orbit will provide an excellent platform for acquiring the data needed to advance significantly our understanding of the Martian atmosphere and its remarkable variability. The Mars Observer Pressure Modulator Infrared Radiometer (PMIRR) is a nine-channel limb and nadir scanning atmospheric sounder which will observe the atmosphere of Mars globally from 0 to 80 km for a full Martian year. PMIRR employs narrow-band radiometric channels and two pressure modulation cells to measure atmospheric and surface emission in the thermal infrared; a visible channel (0.39 - 4.7 µm) is used to measure solar radiation reflected from the atmosphere and surface. Vertical profiles of atmospheric temperature, the infrared extinction of dust suspended in the atmosphere, atmospheric water vapor, and condensate hazes will be retrieved from infrared measurements having a vertical resolution of 5 km, which is half an atmospheric scale height. PMIRR infrared and visible measurements will be combined to determine the radiative balance of the polar regions, where a sizeable fraction of the global atmospheric mass annually condenses onto and sublimes from the surface. Derived meteorological fields, including the diabatic heating and cooling of the atmosphere and the vertical variation of its winds, will be computed from the globally mapped fields retrieved from PMIRR data. Analyses of these observed and derived fields will address many key questions regarding the atmosphere and climate of Mars.
1. INTRODUCTION
The study of the atmosphere and climate of Mars requires a comprehensive investigation of atmospheric and surface phenomena linked by the physical mechanisms of atmospheric transport, surface-atmosphere interaction, radiative heating and cooling, and latent heat exchange. To date, such studies have been severely limited by the lack of simultaneous and systematic measurements of the distributions of dust, volatiles, and key meteorological fields. Previous missions, beginning with the fly-by of Mariner 4 in 1965 and continuing through the more recent Viking and Phobos missions, were constrained by their limited lifetimes, the instrument technology available at the time, or the lack of basic data (e.g., atmospheric composition and mass) which these missions have now provided.
The Mars Observer mission, with the spacecraft in a polar near-circular orbit, is ideally suited to global synoptic mapping of atmospheric constituents and fields. Two of its primary mission goals are to:
¥ Determine the structure of the atmosphere and aspects of its circulation. ¥ Determine the time and space distribution, abundance, sources, and sinks of volatile material and dust over a seasonal cycle.
Within the Mars Observer (MO) instrument payload, the Pressure Modulator Infrared Radiometer (PMIRR) is unique in that it is dedicated to observing atmospheric fields globally and nearly continuously in the lower and middle atmosphere of Mars. PMIRR is a nine-channel limb and nadir scanning atmospheric sounder designed to provide global coverage of atmospheric temperature, dust, water vapor, and condensate clouds and of surface radiative balance. Specifically, its measurement objectives are to:
¥ Map the three-dimensional and time varying thermal structure of the atmosphere from the surface to 80 km altitude, as a function of atmospheric pressure. ¥ Map the atmospheric dust loading and its global, vertical and temporal variation. ¥ Map the seasonal and spatial variation of the vertical distribution of atmospheric water vapor in the lower atmosphere ( ² 30 km). ¥ Distinguish atmospheric condensates of H2O and CO2 from each other and from suspended dust and map their spatial and temporal variation. ¥ Monitor the seasonal and spatial variation of atmospheric surface pressure. ¥ Monitor the radiative balance of the polar regions. ¥ Derive surface temperatures, albedo and thermal inertia to understanding the atmosphere's interaction with the ground at large scales.
The combination of systematic viewing of the atmosphere, permitted by the Mars Observer orbit and PMIRR's primary observation strategy, the relatively high vertical resolution provided by limb sounding, the simultaneous, co-located measurement of temperature, dust and water vapor, and the synergisms with other Mars Observer instruments will enable the PMIRR investigation to extend significantly our present understanding of the atmosphere and climate of Mars.
Section 2 addresses further the specific scientific objectives of the PMIRR investigation. The existing state of knowledge and the outstanding scientific questions related to the atmosphere and climate of Mars are discussed, together with statements of key measurements to be made by PMIRR and of quantities to be derived from those measurements. Section 3 outlines the instrument observational strategy required to obtain the most representative climatology possible and the instrument itself, including calibration procedures. Section 4 outlines the retrieval approach, as it applies to PMIRR, and presents some numerical simulations of the ability of PMIRR to achieve its measurement objectives.
2. SCIENTIFIC OBJECTIVES
2.1 ATMOSPHERIC THERMAL STRUCTURE AND GENERAL CIRCULATION
Previous measurements, analyses and modeling studies have identified several of the major components of the general circulation of the atmosphere of Mars [see the reviews by Leovy (1985) and Zurek et al. (1992)]. These include: the meridional flow due to the seasonal condensation and sublimation of CO2 in the polar regions (Hess et al., 1980; James and North, 1982), the seasonally dependent zonally symmetric mass transport in the meridional plane (Leovy and Mintz, 1969; Leovy et al., 1973; Haberle et al., 1982), traveling and stationary planetary waves in the northern hemisphere (Pollack et al., 1981; Barnes, 1980,1981, 1990a), atmospheric thermal tides (Pirraglia and Conrath, 1974; Zurek, 1976; Zurek and Leovy, 1981), internal gravity waves (Pirraglia, 1976; Barnes, 1990b), regional slope winds, dust devils (Ryan, 1964; Ryan and Lucich, 1983), local dust storms (Gifford, 1964; Martin and Baum, 1969; Peterfreund, 1985) and episodically, regional and even planetary-scale dust storms (Gierasch, 1974; Martin, 1974a,b, 1976; Pollack et al., 1979; Zurek, 1982).
On Mars, dust suspended in the atmosphere can modify the atmospheric thermal structure and various components of the atmospheric circulation to a degree that is quite remarkable by comparison with Earth (Gierasch and Goody, 1972; Hanel et al., 1972; Conrath et al., 1973; Moriyama, 1976; Haberle et al., 1982; Zurek et al., 1992). The nonlinear feedbacks involving the transport of dust and the diabatic heating by airborne dust underlie much of the complexity apparent in seasonal and interannual variations of the atmosphere of Mars. Other features somewhat unique to Mars and important to its general circulation are the condensation and sublimation of a sizeable fraction of its total atmospheric mass and the presence of its substantial topographic relief. Furthermore, while Mars does not have continents in the sense of a land-sea distribution, there are vast coherent regions of the planet with high surface albedo and low apparent thermal inertia (Palluconi and Kieffer, 1981). These "thermal continents" may have significant effects on the atmospheric general circulation (Zurek and Christensen, 1990).
Existing data lack the spatial and temporal coverage of the Martian atmosphere needed to define the full range of its variability and to relate quantitatively possible radiative and mechanical forcing to observed features. The key objectives of the PMIRR investigation in the study of the general circulation of Mars are: first, to describe the time-space structure of the circulation; second, to understand quantitatively the relationship between general circulation features and radiative forcing, as influenced by such factors as atmospheric mass, surface thermal properties, and topography; and third, to constrain quantitatively wherever possible, the role of the general circulation in atmospheric transports and the evolution of climate.
PMIRR will characterize large-scale atmospheric variability by repeated scanning of the atmospheric limb of Mars on each of 12-13 orbits per day to provide global, daily views of atmospheric temperatures as a function of pressure and the distribution of atmospheric dust, both from 0 - 80 km. Although winds cannot not be observed directly by PMIRR, the vertical variation of the horizontal wind components (i.e., the wind shear) can be derived from systematic global mapping of temperature as a function of pressure.
Deriving Winds from Temperatures The only direct measurements of atmospheric winds on Mars are those made by the two Viking Landers at a single height of 1.6 m. Presently, there are no plans to fly instruments that could measure Martian winds directly from orbit, although such instruments are now observing Earth's upper atmosphere (Reber, 1990). Fortunately, the vertical variation of the horizontal wind components (i.e., the wind shears) can be computed from PMIRR's observations of global temperature fields, using the geostrophic or higher-order approximations (Zurek et al., 1992). Determination of the winds themselves depends upon a boundary condition, such as the precise measurement of surface pressure or the knowledge of the wind at some known height. The first is difficult to achieve by passive remote sensing from orbit. Cloud tracking by the Mars Observer Camera (MOC) can provide the required reference-level winds, on those occasions where there are identifiable cloud features which can be located in altitude and tracked from one orbit to the next. This is most likely to occur at high latitudes.
The magnitude of the horizontal components of the wind at high altitude (say above 20 km) is determined largely by the wind shear, since winds near the surface are expected to be small by comparison. For these altitudes it is adequate to assume that u Å v Å 0 near the surface. Figure 1 shows a meridional cross-section of temperature constructed from Mariner 9 IRIS data, together with the computed geostrophic zonal wind. Similar cross-sections will be computed for each orbit of Mars Observer using PMIRR's observations of temperature, but with corresponding maps of dust opacity, haze distribution and water vapor.
Zonally Symmetric Components of the Circulation The variations of the zonally averaged temperature and zonal-mean wind with pressure, latitude and season together comprise two of the most important characteristics of a planet's general circulation (e.g., Lorenz, 1967). Among other things, they are first-order dynamical properties against which dynamical models, including GCMs, must be validated. PMIRR will provide zonally averaged temperatures at altitudes below 80 km for a full Mars year. Previous latitude-height cross-sections of the zonal-mean temperatures have been confined to altitudes below 45 km and largely to northern spring (Leovy, 1982).
The region above 45 km is of particular interest in that it is virtually unexplored. There is some evidence (Deming et al., 1986) that, just as is the case in the Earth's mesosphere, the latitudinal variation of temperature may be far from the radiative equilibrium distribution. Observation of such departures over seasonal time scales would provide clear evidence that the breaking of vertically propagating gravity waves and atmospheric tides is playing a significant role in the momentum balance of the middle atmospheres of both Earth and Mars. Estimates of the effective "drag" exerted on the zonal-mean Martian circulation by these breaking waves can be derived from temperature observations, just as on Earth (Smith and Lyjak, 1985).
The zonal-mean component of the zonal wind can be derived in a rather straightforward way from the observed temperature field, as discussed above [also, Fig. 1]. The zonally averaged meridional wind cannot be derived in this fashion since the zonal average of the longitudinal pressure gradient is zero, and thus the zonal-mean component of the meridional wind results from a complex interplay of pressure forces and wave-induced fluxes of heat and momentum. Yet, it may be a thermally driven Hadley-type circulation, with its zonal-mean meridional wind, which is most effective in transporting dust and volatiles across the planet (e.g., Haberle et al., 1982; Zurek et al., 1992). Furthermore, the intensity of the cross-equatorial Hadley circulation and its meridional extent are closely tied to the opacity and the vertical and horizontal extent of dust hazes in the atmosphere; the intensification and extent of this circulation is likely to be greatest during the episodic planet-encircling dust storms (e.g., Haberle et al., 1982; Zurek et al., 1992).
One signature of the Mars Hadley circulation during a great dust storm is the reversal aloft of the poleward temperature gradient due to strong descent and adiabatic warming in the down-welling branch of the circulation. The position of this relatively warm air will move with latitude as the amount of dust and the area covered by dust changes with time. PMIRR has the spatial resolution and coverage to identify this change in the upper layers of the dust haze.
By combining the thermodynamic energy and mass continuity equations, the zonal-mean meridional and vertical winds can each be represented as the sum of two components, one related to the diabatic heating of the atmosphere and the other to the meridional flux of heat due to atmospheric waves. This second, wave induced, component is difficult to estimate from Mars Observer data, in part because of the reference level problem discussed above. However, theory and terrestrial experience indicate that the zonally averaged transports due to advection by that first, diabatically induced, component of the zonal-mean meridional and vertical winds is a good approximation to the net zonal-mean meridional transport. Since this net transport is the difference between advection by the zonally averaged winds and the zonal-mean meridional fluxes induced by planetary waves, the diabatically related components of the zonal-mean meridional and vertical velocities are together referred to as the residual zonal-mean circulation (Andrews et al., 1987).
Derivation of residual zonal-mean circulations has proved useful for the study of the transports of ozone, water vapor and other trace species in the Earth's stratosphere (Garcia and Solomon, 1983; Gille et al., 1987). For Mars, the residual-mean circulation can be estimated from PMIRR data, using observed temperatures and a computed diabatic heating which depends primarily on the distribution of atmospheric dust.
Long-Period Waves: Traveling Planetary Waves. The two Viking Lander meteorology stations detected traveling wave disturbances during winter at both the 23¡N and 48¡N sites, with the wave activity more prominent and more frequent at the more poleward site. With the aid of the geostrophic approximation, these waves were inferred to be typically of zonal wavenumber 3 or 4 with eastward phase speeds in the range 10-20 m/s. On this basis they have been tentatively attributed to baroclinic instability (Barnes, 1980, 1981). The location and character of these waves were observed to vary with atmospheric dust loading. Occasionally, the waves themselves appeared to trigger local dust storms (Briggs and Leovy, 1974; Ryan et al., 1981; Arvidson et al., 1983).
Planetary-scale waves were also detected in the Mariner 9 IRIS observations, but the orbital geometry precluded the unambiguous determination of zonal wavenumber and period (Conrath, 1981). It is not known if wintertime wave disturbances occur in the mid-latitudes of the southern hemisphere. Simulations using general circulation models (GCM) of the atmosphere Mars suggest that the occurrence of these waves in the southern mid-latitudes is highly sensitive to the uncertain features of the large-scale topography there (Barnes, Haberle and Pollack, private communication).
Stationary Waves: Inferences based on Mariner 9 IRIS data (Conrath, 1981) and simulations using a variety of atmospheric circulation models indicate that the large-amplitude planetary-scale orography and variation in surface properties (Webster, 1977; Moriyama and Iwashima, 1980; Pollack et al., 1981) will drive a strong atmospheric response. The long-period traveling component of this response was discussed above. Model simulations suggest that the stationary, or quasi-stationary (periods > 10 days or so), response may be even larger. The nature of this response is thought to take the form of vertically propagating Kelvin waves at low latitudes and of vertically trapped Rossby waves at higher latitudes (Webster, 1977). The theory of these disturbances is well-developed and abundant observations of these phenomena exist for Earth's atmosphere. Amplification of orographically forced waves during planetary-scale dust storms can provide a very effective means of transporting heat and dust into the polar regions (Barnes and Hollingsworth, 1987).
PMIRR observations have the spatial resolution to reveal unambiguously the characteristic zonal wavenumbers, phase speeds, vertical scales and meridional extent of stationary and of most traveling planetary-scale waves (see Section 3). Its observations of the wintertime southern hemisphere will establish whether or not traveling waves occur in the south, as they are known to do in the north, even as the Mars Observer Laser Altimeter (MOLA) is defining the planetary-scale terrain height. This will provide a critical test of hypotheses about the influence of topography on these traveling waves. PMIRR observations of the variations of the dust distribution and of temperature will also link quantitatively the effects of net radiative forcing due to suspended dust and the modification or generation of atmospheric waves.
Short-Period Planetary-Scale Waves Atmospheric tides. These are planetary-scale oscillations of the meteorological fields driven principally by the daily varying solar heating of the atmosphere and surface (Zurek, 1976). Observations of surface pressure and winds by the Viking Landers (Leovy, 1981), together with atmospheric temperature observations by the Mariner 9 IRIS (Pirraglia and Conrath, 1974), the Viking Lander entry science package (Seiff and Kirk, 1977), and the Viking IRTM (Martin, 1981), show that atmospheric tides are a significant component of the general circulation, particularly during planetary-scale dust storms (Zurek and Leovy, 1981). The largest components of the diurnal (i.e., once daily) and semidiurnal (i.e., twice daily) tidal fields are fixed in local time, but the large-amplitude planetary-scale orography of Mars and the presence of its "thermal continents" may induce components which are not locked to the sun's apparent motion. One eastward traveling tidal component of particular interest is an apparent diurnal Kelvin mode (Zurek, 1976; Conrath, 1976), which may be greatly amplified during the early stages of a great dust storm and thereby provide a positive feedback leading to the full onset of the storm (Zurek and Leovy, 1981; Tillman, 1988).
In its nominal observational mode PMIRR will characterize diurnal atmospheric tides using the day-night (2 p.m.-2 a.m.) differences in the temperatures observed on constant pressure surfaces. These observations will be supplemented on occasion by viewing local time differences of about three hours using the ability of PMIRR to look in the cross-track direction at the atmosphere at both side limbs (Sec. 3.5).
Some tidal components will grow in amplitude with height, especially at low latitudes, and these will be readily observed by PMIRR in the middle atmosphere. The structure of the waves in the vertical above 30 km will also permit the identification of the heights of tidal breaking, a process which may drive mechanically the circulation of the Mars middle atmosphere through turbulent mixing and momentum dissipation (Zurek, 1986).
Free Modes. In contrast to the thermally driven atmospheric tides discussed above, atmospheric free modes rise above the meteorological noise because of the resonant properties of the global wave-guide provided by the background temperature structure. The Viking Lander surface pressure record suggests that these free modes are present, and it has been suggested that they may play a role in "triggering" large-scale dust storms (Tillman, 1988). On Mars the frequencies of these resonant modes are close to the diurnal and semidiurnal harmonic frequencies of the thermally forced tides (Hamilton and Garcia, 1986; Zurek, 1976, 1988). PMIRR data will be used to isolate modes of nearly diurnal frequency, while providing simultaneous information on the background temperatures which determine the frequencies of the resonant, or free-mode, response through the formation of the atmospheric waveguide.
2.2 DUST IN THE ATMOSPHERE
Radiative Effects of Airborne Dust Dust suspended in the atmosphere of Mars directly absorbs and scatters the incoming solar radiation, while it absorbs and re-emits thermal radiation emitted by the surface. In so doing, suspended dust greatly affects atmospheric heating (Gierasch and Goody, 1972; Moriyama, 1976; Zurek, 1978; Pollack et al., 1979; Zurek et al., 1992) and modifies net radiation at the surface (Davies, 1979; Ryan and Henry, 1979). The key parameters determining the radiative effects of dust suspended in the atmosphere are: first, the spatial distribution of the dust and its variation with time, and second, the optical properties of the dust, including the visible and infrared opacities and the effective single scattering albedo and phase function of the dust particles.
Infrared cooling by CO2 as a function of height can be computed from PMIRR observations of temperature and aerosol opacity. The computation of atmospheric heating and cooling due to airborne dust are constrained by the vertical variations of infrared opacities and temperatures retrieved from PMIRR observations. If the composition, shape and size distribution of the suspended dust particles were known, including their variation with space and time, then the optical parameters of the dust could be computed for all the wavelengths of interest (Kahn et al., 1992). While the variation of opacity with wavelength, as observed by PMIRR's mid-IR broadband radiometric channels, provides some constraint on the dust particle size distribution, more refined estimates of microphysical properties can be derived in combination with the higher spectral resolution data of the Thermal Emission Spectrometer (TES). This information can be used to refine computation of the visible opacities needed to determine solar heating due to suspended dust. Estimates of visible opacity are constrained by the limb and on-planet views of PMIRR's wideband solar channel, by planned sequences in which the TES wideband solar channel will view the same surface area at different zenith angles, and by MOC images of the limb and surface.
The Dust Cycle The most dramatic component of the seasonal and interannual variation of atmospheric dust is the episodic occurrence of planetary-scale dust storms (Gierasch, 1974; Zurek, 1982; Martin, 1984; Kahn et al., 1992). Earth-based observations (Martin, 1974a,b; 1976) suggest that such storms originate when one or more localized centers of dust-raising expand and coalesce until most of a zonal corridor in the southern hemisphere is obscured. Once the storm encircles the planet, it may expand to other latitudes, including the northern hemisphere. This onset and expansion phase lasts several days and is followed in just a few days by a more leisurely clearing over several weeks and months of the airborne dust (Conrath, 1975; Pollack et al., 1979).
Major questions regarding these great dust storms remain unanswered. What are the conditions that lead to the occurrence of one or more such dust storms in some years, yet none in others (Leovy et al., 1985; Haberle, 1986; Tillman, 1988)? The fact that these storms occur most frequently during southern spring and summer, when Mars is closest to the sun and the rate of solar heating is greatest, provides a clue that remains to be deciphered (Zurek, 1982). What causes just a few of the many local dust storms to expand to regional scale? What are the mechanisms controlling the rate of clearing, which seems remarkably uniform (Conrath, 1975; Pollack et al., 1979)? Model simulations (Murphy et al., 1990) indicate that particle size and shape may play roles, while major influences are exerted by atmospheric mixing and large-scale circulation.
With regard to the dust cycle as a whole, major questions remain as to the relative roles of the circulations driven by the diabatic heating due to dust itself (Leovy et al., 1973; Haberle et al., 1982; Kahn, 1983; Zurek et al., 1992) and of other components of the general circulation, like the traveling planetary waves (Briggs and Leovy, 1974; Zurek et al., 1992). Where are the surface sources and sinks of the airborne dust? Once suspended in the atmosphere, dust may be transported into the north polar regions and incorporated into the polar layered terrains (Pollack et al., 1979); dust may also be redistributed into those regions of the northern subtropics having high albedo and low thermal inertia (Christensen, 1986).
PMIRR addresses these questions in two ways. First, the large-scale thermal structure of the atmosphere will be observed prior to, during and after the onset of a great dust storm, should one occur during the Mars Observer mission. In addition, the vertical and meridional spreading and subsequent clearing of the dust hazes will be mapped with good spatial and temporal resolution. This will permit the estimation of the radiative forcing and the monitoring of changes in: 1) the background atmospheric static stability; 2) in the residual-mean circulations; and 3) in large-scale stationary and traveling waves, including atmospheric tides. In this manner, the PMIRR data should illuminate the radiative-dynamical mechanisms involved in the generation and dissipation of the planetary-scale dust storms on Mars.
2.3 THE WATER CYCLE
Much of the past and present interest in Mars and its climate stems from speculation on existence and behavior of water on its surface and in its atmosphere. Water in the current epoch is characterized by a few to several tens of precipitable microns (pr µm) of vapor in the atmosphere and by a permanent water-ice polar cap in the north which is the only known surface reservoir. Other possible reservoirs are, in likely order of importance to the atmosphere, water vapor adsorbed in the regolith, near-surface ground ice and the south polar residual CO2 cap.
Ground-based and spacecraft data, most notably those from the Mariner 9 and Viking missions, have established clear signatures of seasonal variability in the abundance and distribution of atmospheric water vapor (Hanel et al., 1972; Farmer et al., 1977; Farmer and Doms, 1979; Jakosky and Farmer, 1982; Jakosky, 1985; Haberle and Jakosky, 1990; Jakosky and Haberle, 1991). Data from the Mars Atmospheric Water Detector (MAWD) onboard Viking orbiter has contributed most to our understanding of the water cycle. The instrument measured column amounts of water over the sunlit portion of the planet for more than a full seasonal cycle (Fig. 2). Earth-based observations by Barker et al (1970), combined with more recent measurements by ground-based infrared and microwave techniques, suggest substantial interannual variability in the amount of atmospheric water vapor, particularly in the southern hemisphere (Jakosky and Barker, 1984). Such variability may be related to the annual persistence or disappearance of the south polar residual CO2 cap (Kieffer, 1979; Haberle and Jakosky, 1992).
[Deleted / Combined Paragraphs]
Attempts to use the MAWD, IRTM, and Earth-based observations to estimate quantitatively surface and near-surface sources and sinks of water vapor globally have thus far met with only limited success. Studies using numerical models of surface-atmosphere exchange and atmospheric circulation (Davies, 1981; Jakosky, 1983a, b; James, 1985; Haberle and Jakosky, 1990) have failed to reproduce the full seasonal variation of the distribution and abundance of atmospheric water that is apparent in the data. One reason for this difficulty is the remaining uncertainty as to whether or not the regolith plays a significant role as a source of water for the atmosphere.
A related problem is the apparent difficulty in transporting enough water from high to low latitudes to duplicate the MAWD observations (Haberle and Jakosky, 1990). Such studies suggest local sources of water at tropical and middle latitudes. However, uncertainties in sublimation and desorption rates, and the currently limited knowledge of atmospheric relative humidity and both local and global transports of water severely constrain our ability to describe quantitatively the Mars water cycle.
The PMIRR investigation addresses these uncertainties principally by simultaneously observing surface temperatures, the vertical profiles of temperature and of water vapor abundance and by detecting the presence and location of water-ice hazes in the atmosphere. Mapping the vertical variations of water vapor with good resolution can potentially constrain atmospheric transports by cross-equatorial Hadley circulations and by atmospheric waves, if only by establishing upper limits. Determining the vertical distributions of relative humidity, water-ice and dust hazes can constrain models of the vertical redistribution of water vapor in the atmosphere and its loss to the surface by condensation and snow (Kahn, 1990). The simultaneous, co-located observation of temperature and water vapor and the location of dust and ice hazes may also provide useful constraints on the horizontal transport of water ice, including effects of the nucleation of water ice on suspended dust particles.
2.4 CO2 CYCLE
Carbon dioxide is the dominant constituent of the Martian atmosphere, and the most volatile and dynamic component of the Mars climate system. A full understanding of the Martian climate requires determination of the sizes of surface and subsurface reservoirs of condensed CO2 and their behavior over seasonal and climatic time scales, as well as understanding the interactions between the CO2, dust and water cycles, both seasonally and long-term. Existing spacecraft and ground-based observations have provided us with a good first-order picture of the present seasonal CO2 cycle (Paige and Ingersoll, 1985; James et al., 1992), but to extend this understanding to interannual timescales and beyond will require a considerably more detailed understanding of the properties and processes that are responsible for the presently observable behavior.
Reservoirs of CO2 identified in earlier observations include: 1) the atmosphere with Å14 g/cm2 of CO2 averaged over the globe, 2) the seasonal polar caps which contain 0-3 g/cm2 of CO2 averaged over the globe, and 3) a residual CO2 frost deposit near the south pole (Kieffer, 1979; Paige et al., 1990), which has been estimated to contain less than 14 g/cm2 of CO2 averaged over the globe (Fanale, 1976) and which may disappear completely in some years. Reservoirs for which we have fragmentary evidence include: 1) adsorbed CO2 in the Martian regolith (Fanale and Cannon, 1979) and carbonates (Pollack, 1979; Kahn, 1985; Pollack et al., 1987). If present, either of these potential unobserved reservoirs may contain significantly more volatile carbon than is presently contained in the observed polar cap-atmosphere sub-system.
Models for the behavior of Martian CO2 over seasonal and climatic timescales indicate that, in equilibrium, the partitioning of exchangeable CO2 between the gas and the condensed phases is governed primarily by temperature. When a permanent CO2 cap is present, the mass of CO2 gas in the Martian atmosphere is determined by the vapor pressures of the permanent frost deposits, as determined by their annual averaged temperatures (Leighton and Murray, 1966). When a permanent CO2 cap is not present, the mass of the Martian atmosphere could be determined by the vapor pressures of adsorbed CO2 deposits, which are strongly temperature dependent (Fanale and Cannon, 1979), or by the formation of carbonates, which are formed most rapidly when atmospheric temperatures permit the presence of liquid water, possibly in disequilibrium (Kahn, 1985).
The recession and growth of the Martian seasonal polar caps is governed largely by their heat balance. The Viking IRTM observations provided the first direct measurements of the most important heat budget quantities over an annual cycle for the north and south polar residual caps (Paige and Ingersoll, 1985). PMIRR has a wideband solar and several radiometric filter channels in the thermal infrared to measure reflected solar and thermal radiation emerging from the top of the Martian atmosphere. The PMIRR observations of the polar heat balance will go beyond the Viking data by: 1) providing a more systematic and accurate set of heat balance measurements over wider geographic regions at high northern and southern latitudes; and 2) making simultaneous measurements of polar surface temperatures and of atmospheric fields necessary to constrain quantitatively individual terms in the heat balance of the polar surface and atmosphere.
The major term in the heat balance is the net radiative gain of the atmosphere/surface system and is computed using the measured broadband visible and the several infrared PMIRR radiances. Smaller terms include: 1) rates of subsurface heat conduction, determined through measurements of bare-ground and water-ice surface temperature variations; 2) atmospheric heat storage, determined by measurements of polar atmospheric temperature profiles; and 3) net atmospheric heat transport, constrained through measurements of global temperature and aerosol fields combined with computed estimates of atmospheric winds or with dynamical models. The computation of these terms, together with observations by PMIRR of dust and ice hazes at high latitudes and of surface and atmospheric fields at lower latitudes will form a basis for a better understanding of the present seasonal CO2 cycle on Mars and the factors controlling it, including its interaction with the cycles of water and dust.
2.5 SURFACE SCIENCE
Characterizing the physical and chemical properties of the surface of Mars is one of the principal scientific goals of the Mars Observer mission. Existing images of the surface of Mars show evidence for a wide range of surface modification processes, some undoubtedly atmospheric in origin, that have operated over Mars' history and have shaped the Martian terrain.
PMIRR measurements will provide an important boundary condition for considerations of the physical and chemical properties of the surface and subsurface. Mars surface temperatures are affected by a wide range of parameters, including the partitioning of solar heating and infrared cooling between the surface and atmosphere, subsurface thermal conduction and latent heat effects due to the presence of volatiles. Bulk thermal properties of near-surface and sub-surface soil layers can be estimated from diurnal and seasonal surface temperatures derived from PMIRR observations by modeling the surface heat budget, including the net effects of the Martian atmosphere on surface temperature variations as constrained by PMIRR observations of atmospheric temperature and aerosol opacity.
3. MEASUREMENT APPROACH AND INSTRUMENT DESCRIPTION
Confidence in the ability of PMIRR to achieve the necessary measurement objectives is based in part on numerical simulations [see Sec. 4] and in part on its heritage. The observational study of Earth's stratosphere over the past 25 years serves as a paradigm for the investigation of the atmosphere of Mars, a regime of comparable pressures, temperatures, and water vapor abundances. Detailed observations of the Earth's middle atmosphere have been and are now largely performed by limb viewing experiments, because of the need to achieve vertical resolutions better than one atmospheric scale height.
The need to achieve such resolution has been demonstrated for Earth by the scientific impact of data from the Limb Infrared Monitor of the Stratosphere (LIMS) and the Stratospheric and Mesospheric Sounder (SAMS) onboard the Nimbus 7 satellite (Gille and Russell, 1984; Barnett et al., 1985). In the fall of 1992, a new generation of limb sounders deployed on the Upper Atmosphere Research Satellite (UARS; see Reber, 1990) began to map the global distributions of temperature and minor constituent concentrations in Earth's middle atmosphere. These instruments, several employing filter and gas correlation spectro-radiometry, achieve vertical resolutions of half an atmospheric scale height or better.
Characteristic thermal structures in the Mars atmosphere also have vertical scales of one-to-two atmospheric scales (e.g., Seiff and Kirk, 1977; Leovy, 1985; Barnes, 1986). In addition, Mars presents the difficulty that dust may be suspended in its atmosphere much of the time (e.g., Pollack et al., 1979) and so its contribution must be separated from that due to thermal emission by CO2 or H2O gas. Thus, the design of PMIRR and its observational strategy is driven by three needs: 1) to retrieve atmospheric fields with adequate vertical resolution, 2) to retrieve vertical profiles of temperature and water vapor in the presence of aerosols and to retrieve simultaneously the vertical profile of aerosol extinction; and 3) to quantify the radiative fluxes of the polar regions. To achieve these goals, PMIRR has adapted the techniques of limb sounding and filter and pressure modulation radiometry proven for study of Earth's stratosphere to the global mapping of Martian atmospheric properties.
3.1 INSTRUMENT SPECIFICATIONS
A detailed description of the PMIRR instrument design was published by McCleese et al. (1986). The PMIRR instrument itself is shown in Fig. 3. Table 1 summarizes the bandpass, type and measurement function of the PMIRR spectral channels, while the chief physical and operational parameters of the instrument are given in Table 2. Briefly, PMIRR observes in a broadband visible channel and in eight spectral intervals in the 6-50 µm range. Spectral discrimination within this infrared range is achieved using spectral filters and pressure modulator units. Adequate signal-to-noise in the temperature and water vapor sounding channels 1-4 is ensured through the placement of their detectors on a cold focal plane assembly cooled to 80K by a passive radiative cooler. Vertical profiles of atmospheric properties are constructed from observations in three fields-of-view (FOV) scanned across the limb and onto the planet using a two-axis scan mirror situated in the optical chain ahead of the primary telescope.
Optics The PMIRR optical layout is shown schematically in Figure 4. All channels share a common scan mirror and telescope primary, and view Mars and space alternately at 800 Hz via a rotating double-sided mirror chopper. The CO2 and H2O PMCs are located in the optical paths of channels 1 through 4 and impress a 50 Hz amplitude modulation on to the chopped signal.
Optical Mechanisms The most important PMIRR optical mechanisms are the pressure modulator units (PMUs), the scan mirror assembly, and the chopper assembly. The PMUs mechanically modulate the pressure of gas within a sealed optical cell, using dynamically balanced pistons driven at resonance (Taylor, 1983). The scan mirror assembly, consisting of elevation and azimuth drive actuators mounted orthogonally in a yoke arrangement, allow the mirror orientation to be stepped over a large angular range in both axes. The chopper assembly is a rotating, toothed disc which provides continuous and full (100%) signal modulation at 800 Hz. Detectors The nine PMIRR channels employ single element detectors. Channels 1-5 use cooled HgCdTe photon detectors, whereas channels 6-9 use ambient temperature deuterated tri-glycine sulfate (DTGS) pyroelectric detectors. Channels 1 and 4 make use of photovoltaic detector technology while the remaining cooled detectors are photoconductive. Channels 6 and 7 each have two detectors and hence two FOVs; all others view through just one of the three FOVs.
Cold Focal Plane Assembly and Passive Radiative Cooler The PMIRR cold focal plane assembly (CFPA) consists of channel 1-5 detectors, condensing optics, bandpass filters, and their mounting structures cooled by the cold stage of the PMIRR passive radiative cooler and surrounded by concentric housings cooled in turn by the radiator intermediate and warm stages. The mechanical and optical design of the CFPA are driven by the need to minimize the radiative energy from the 300 K instrument cavity falling on the 80 K focal plane whilst maximizing the signal throughput in the bandpass of each channel. PMIRR uses a three-stage passive radiator to satisfy its cooling requirements. The radiator has a reclosable door which acts as a planet shield when open and protects cold surfaces from contamination when closed.
Electronics and Signal Processing Digital logic control and data processing electronics interface with the spacecraft payload data system (PDS), receiving and executing commands and processing instrument science and engineering data. A microprocessor is used for executive control, for example, to implement an automatic PMIRR scan mirror cycle. The primary instrument programs reside in read-only-memory (ROM) and are copied into random-access-memory (RAM), where they may be modified through links to changes made in RAM using small-scale memory loads.
While all PMIRR channels produce an 800 Hz wideband signal, the PMUs also produce a 50 Hz sideband signal which appears as an amplitude modulation on the wideband signal. The function of the analog signal processing electronics is to convert these signals into digital voltage levels. A wideband signal chain amplifies and filters each 800 Hz signal. These signals are then demodulated, filtered, and fed into a voltage to frequency converter which is accurately gated into a counter for the signal integration period. The sideband signal chain is very similar in function, except that its demodulator reference is taken from the PMU drive waveform rather than the chopper, and its input is the demodulated wideband signal.
Expected Instrument Performance Basic figures of merit for a PMIRR channel can be defined as follows:
NER = NEP ¥ ¹ / ( 4/¹ A½ W F (2t)1/2) W ¥ cm-2 ¥ sr-1/cm-1, S/N = B(n0,T0) / NER, NEÆT = NER / (dB(n0,T0)/dT) K, Tmin = B-1(n0, NER) K. where NER is the noise equivalent radiance, or the radiative flux incident on the instrument required to produce an r.m.s signal referenced to the detector equal to the detector noise equivalent power (NEP, W ¥ Hz-1/2). NER depends only on detector NEP, instrument etendue (AW, cm2 ¥ sr-1), equivalent channel spectral width (W, cm-1), channel optical transmission F, and signal integration time (t, seconds). S/N is the channel signal-to-noise ratio for a view of a blackbody at wavelength n0 and temperature T0, where B(n0,T0) is the Planck function. NEÆT is the noise equivalent temperature change, or the blackbody target temperature change that produces a signal change equal to the noise. Finally Tmin is the minimum detectable temperature, or the blackbody target temperature that gives a signal equal to the noise, where B-1(n0,NER) is the inverse Planck function .
The PMIRR far-IR and visible broadband channels have a detector-noise limited performance. When observing a blackbody target at a typical Martian temperature of 225K, channels 6-9 have a S/N Å 100 and an NEÆT Å 1K. The wideband signals for channels 1-5, with their cooled detectors, are digitizer-limited. At 225K, their (digitized) S/N performance ranges from 3600 for channel 4 (at 6.9 µm) to 13000 for channel 3 (at 14.5 µm, closer to the peak blackbody intensity). Corresponding NEÆTs are 0.006 and 0.003K. The broadband visible channel 9 has a S/N Å 3500 when viewing a normally illuminated, unit albedo surface at the mean Mars distance. [Should we include Tminxxxx]
3.2 PRESSURE MODULATION RADIOMETRY
Gas correlation radiometry combines the high spectral discrimination available to advanced Fourier transform spectrometers with the simplicity, high energy grasp, and low data rate of broadband filter radiometers. Pressure modulation radiometry is a flight-proven technique developed for temperature and species profiling in the terrestrial upper atmosphere (Taylor, 1983). Its principal advantage is that it is species specific. For Mars, this means that radiation originating from the ubiquitous airborne dust can be distinguished unambiguously from gaseous emission by atmospheric carbon dioxide and water vapor. This is especially important when limb sounding, given the heightened sensitivity to suspended dust (see Sec. 3.4).
A PMIRR pressure modulator radiometer (PMR) channel is shown schematically in Fig. 5. A PMR detects radiation from the emission lines of a specific atmospheric gas by directing atmospheric emission through an optical cell containing the same gas. In PMIRR incident radiation is chopped mechanically at 800 Hz and the pressure of the gas in the cell is modulated at approximately 50 Hz. Chopped signals at 800 Hz and pressure modulated signals, which appear as a 50 Hz amplitude modulation on the 800 Hz signal, are selected by electronic processing. Because the transmission of the pressure modulator cell (PMC) varies at the frequency of modulation only near gaseous absorption lines, the pressure modulated signal quantifies atmospheric emission in spectral regions in and near the absorption lines in the PMC, which match the emission lines of the same gas in the atmosphere. The chopped, or wideband, signal provides a measure of uncorrelated scene radiation while largely rejecting atmospheric emission from the gas of interest. Figure 6 illustrates the spectral discrimination of a PMR for a single CO2 line and a limb view of the Martian atmosphere. The curves labelled Tp1 and Tp2 show the transmission of the PMC at the two extremes of the modulator pressure cycle, and the curve labelled Tp1-Tp2 is the effective transmission at the modulation frequency. The spectral resolution of a PMR is determined primarily by cell length and the mean pressure of the gas in the cell, and can be designed to match atmospheric emission line profiles (Fig. 6). Because a PMR channel is defined by a broad bandpass filter, 50-100 emission lines can be observed simultaneously giving the PMR a multiplex advantage. As compared with an interferometer of comparable resolution, instrument etendue is not limited by the high resolution element, so that the PMR is able to collect a great deal more energy (Houghton and Smith, 1970).
3.3 INSTRUMENT CALIBRATION
Calibration is the process that allows the relationship between the radiance incident on the instrument in each spectral channel and the digital data output by that channel to be determined. PMIRR radiance measurements are made in spectral channels that cover a wide spectral range and are used to determine a wide range of physical parameters. This is only possible if the radiometric, spectral, and FOV characteristics of each channel are determined accurately. PMIRR calibration measurements fall into five basic categories, these are: a. Radiometric Calibration b. Linearity Calibration c. Field-of-View Calibration d. Wavelength Calibration e. PMC Spectral Response Calibration
These categories are addressed through two activities: in-flight and laboratory calibration. In-flight calibration is used to correct for small, time-varying changes in the properties of the instrument, whereas laboratory calibration is required to relate in-flight calibration sources to absolute standards and to measure instrument calibration parameters that are not expected to change during the mission.
In-Flight Calibration Radiometric calibration is the only calibration activity performed during flight, and provisions for accurate radiometric calibration have played a major role in the PMIRR instrument design. The calibration goals for PMIRR are an absolute radiometric accuracy ± 0.5% for the thermal infrared channels and ± 3% for the solar channel, and a radiometric precision of ± 0.1% for all channels.
Absolute calibration is provided for nearly the full dynamic range of each channel by measurements which establish radiometric offset and gain. The offset calibration is supplied by a view of space, which is an excellent radiometric zero for every channel. Gain calibration is achieved using an internal 300 K blackbody for the thermal channels and an external diffusely scattering target for the solar channel. Radiometric precision is obtained by minimizing signal drifts. This is achieved by chopping all signals against space at 800 Hz, and by accurately controlling the PMIRR scan mirror and optical bench temperatures.
During flight the PMIRR thermal channels are calibrated at 10-20 minute intervals by views of space and the internal black body, which is a small convoluted cone that can be introduced into the optical path of the instrument by an actuator. The accuracy of the gain calibration depends entirely on the accuracy with which emission from the target is known. In order to meet the calibration goals, its effective emissivity must exceed 0.995 from 6.5 to 53 µm and its temperature must be monitored to an absolute accuracy of < 0.25 K. A high effective emissivity is achieved by treating the surface of the target with a high emissivity coating, and by ensuring that the target and the instrument optical cavity that surrounds it are at the same temperature.
The solar channel is calibrated using views of space and the external, diffusely scattering target illuminated by the sun. The target is an electric arc-sprayed aluminum disc which fills the instrument aperture and FOV. It is viewed via the scan mirror up to once per orbit as the spacecraft approaches the terminator over the northern hemisphere of Mars. The gain calibration depends entirely on the accuracy with which the target (absolute) reflectivity and bi-directional reflectance function can be determined. These are measured in the laboratory, and the target is subsequently protected against contamination by a reclosable cover throughout the launch, cruise, and orbit insertion phases of the mission.
During the mapping orbit, a novel calorimetric method is used to measure changes in target reflectivity, should they occur. The target disc is thermally isolated from the instrument and its temperature is monitored continuously. Any long-term changes in maximum temperature not consistent with the dominant heat balance terms of solar absorption, thermal emission, and conductance to the instrument are assigned to changes in target reflectivity.
Laboratory Calibration The laboratory radiometric, linearity, FOV and wavelength calibration of the PMIRR were performed in a thermal vacuum chamber under conditions designed to approximate, as closely as possible, the instrument environment during a Mars Observer mapping orbit. The sources and targets viewed during calibration were also mounted within the chamber.
3.4 STANDARD DATA PRODUCTS
The continuous data taken by PMIRR is processed to five standard levels defined as follows:
Level 0: PMIRR data numbers.
Level 1: PMIRR calibrated radiances, calibration files, housekeeping records and geometry files.
Level 2: Geophysical fields: Surface parameters in the on-planet FOVs and vertical profiles of atmospheric fields associated with individual limb scans.
Level 3: Globally mapped fields.
Level 4: Highly derived fields (e.g., diabatic heating, winds).
The PMIRR data are packetized by the instrument and held by the spacecraft Payload Data System (PDS) until transmitted to Earth. There the PMIRR packets are separated from the PDS data stream and, together with instrument health parameters stripped from the spacecraft engineering data, are relayed as Level 0 data to the PMIRR Analysis and Data Archive Center. The raw Level 0 data are checked, assigned a quality flag, and converted into calibrated radiances. These calibrated radiances, calibration files, housekeeping records and associated geometry fields form the Level 1 data set. Geometry consists of absolute pointing predicts or knowledge as supplied by the Mars Observer project SPICE toolkit.
Calibrated radiances are separated into space, limb, partially on-planet, on-planet off-nadir, and on-planet nadir segments. On-planet views associated with the physical space traversed by the limb scans are identified. Atmospheric profiles of temperature, dust extinction, water vapor, haze distribution, and visible reflectance are then retrieved as a function of pressure, and thus relative height. These profiles, together with surface fields, form the Level 2 data set. The Level 3 data set contains individual atmospheric profiles and surface fields are combined into global maps, accumulated for various time periods: daily, mapping cycle or monthly, annually. These are the meteorological, surface properties, and radiation budget maps. Two techniques for estimating synoptic maps each day from the asynoptic data acquired from the polar orbiting Mars Observer are currently being investigated: sequential estimation (Rodgers, xxxx) and asynoptic Fourier transformation (Salby, xxxx). More highly derived products are then computed from these mapped fields [e.g., wind shears, atmospheric heating], and these form the Level 4 data set.
3.5 OBSERVATIONAL APPROACH
Mars Observer provides a three-axis stabilized platform in a 371 km, sun-synchronous, near polar orbit about Mars. PMIRR employs its two-axis elevation/azimuth scan mirror to obtain both in-track and cross-track limb, nadir, off-nadir and space views for all its channels. Measurements are accumulated continuously and repetitively throughout the mapping mission in order to assemble the necessary climatological data set. Functionally, the PMIRR instrument has a single operational mode. Measurements are integrated simultaneously in each channel and are obtained every two seconds, regardless of the position of the scan mirror. This yields a constant data rate of 156 bits per second (bps).
Observational flexibility is provided by the scan mirror. Because of the repetitive nature of PMIRR's observations, the mirror actuators are driven by simple tables in memory on board the instrument. When PMIRR arrives at Mars it will contain in memory a nominal set of all tables required to meet its basic measurement objectives. The pre-loaded tables are read from ROM to RAM, where they can be altered by uplink commands to change the mix or duration of the observational sequence modes. These basic observational modes are reviewed below, following a brief discussion of the limb sounding approach.
Limb Sounding The effective limiting vertical resolution of a passive remote sounding measurement is determined by viewing geometry, the pressure dependence of atmospheric opacity in the wavelength interval selected, and the instrument signal-to-noise. In practice, the vertical resolution of a limb measurement is determined by the instrument field-of-view, and current flight systems achieve better than one-half scale height resolution. By contrast, the vertical resolution of a nadir measurement typically approaches an atmospheric scale height. In limb sounding, the effective averaging distance along the line-of-sight also depends on the instrument FOV and the pressure dependence of atmospheric opacity in the selected wavelength interval. For PMIRR, the approximate line-of-sight averaging is 240 km, centered about the tangent point.
Limb sounding enhances sensitivity to, and thereby extends vertical coverage for, atmospheric trace species by increasing the mass-weighted geometrical optical path observed relative to nadir sounding [by a factor of 45 in the case of Mars]. Of course, this increased path also enhances the line-of-sight opacity of airborne dust. Although these dust opacities are smaller at the infrared wavelengths observed by PMIRR than in the visible, the dust contribution must still be taken into account [see Sections 3.2, 4].
The species sensitivity of a limb sounder is not influenced significantly by the vertical atmospheric temperature gradient. For nadir sounding, in contrast, errors in derived profiles and abundances increase rapidly as the surface and atmosphere become more nearly isothermal. Such conditions are far from rare in either the terrestrial or Martian atmospheres.
Nominal Observational Sequence In its primary observation mode, PMIRR looks beneath the spacecraft to view the atmosphere at the aft (in-track) limb, scanning down onto the planet; it then slews further onto the planet to obtain an on-planet, off-nadir view at Å60¡ emission angle, followed by a nadir view (Figs 7,8). The mirror then slews up to obtain a space view for calibration, and the sequence is repeated. Over the polar regions, the repetitive in-track scanning described above is interleaved with 'buckshot' sequences which consist of nadir and off-nadir measurements made over a wide range of elevation and azimuth angles using the two-axis scan mirror. Views of space and the internal blackbody are performed several times per orbit and provide complete gain and offset calibrations for the thermal channels.
For limb scanners the tangent point recedes as the instrument views lower in the atmosphere. When PMIRR observes the aft limb, this recession of the tangent point is partly compensated by the forward motion of the Mars Observer spacecraft during the two-second integration time at a given altitude (Fig. 8). Thus, downward scanning of the aft limb produces a more compact profile in latitude. The effective horizontal resolution achieved when PMIRR views the aft in-track limb is Å4¡ in latitude, determined by the line-of-sight averaging, and Å28¡ in longitude at the equator, determined by the 12-13 orbits of Mars Observer each day (Fig. 9). Equivalently, longitudinal structures having zonal wavenumbers (i.e., the number of maxima or minima around a latitude circle) ² 6 can be characterized.
This combination of repetitive aft-limb scanning and polar 'buckshot' sequences comprises the nominal observational mode, whose daily coverage is illustrated in Fig 9. The nominal limb scanning sequence provides contiguous vertical coverage from 90 to -20 km on the limb in 5 km steps (Fig. 8). This extended scanning allows each channel to observe the necessary height range whilst accommodating uncertainties in spacecraft attitude. In the time taken to complete this sequence the spacecraft moves 115 km. Given a line-of-sight averaging of 240 km, it is possible to interleave other observation sequences with the in-track limb and on-planet scans without seriously compromising latitudinal coverage. Alternatively, successive scans can be averaged together to increase signal-to-noise when required.
The polar spatial coverage provided by a single day of 'buckshot' sequences is shown in Fig. 10; this pattern also provides the angular coverage required for polar radiative balance studies. Since the aft-limb tangent point is approximately 25¡ of latitude behind the nadir view location, the aft-limb scans become less frequent in middle and high latitudes as the 'buckshot' mode is executed at polar latitudes; also note that the planetary rotation slightly offsets the on-planet views from the points beneath the aft-limb tangent points (Fig. 9).
Special-Purpose Sequences Buckshot and specialized surface sounding sequences will be employed occasionally at low latitudes to supplement the nadir observations included in the nominal in-track limb and nadir scan. Other special-purpose sequences include off-track and cross-track limb scans to give increased local-time of day coverage and greater longitudinal or latitudinal resolutions. Figure 11 shows the local times of day at the tangent points of in-track and cross-track limb views. The FOV rotates slightly when projected onto the side limb, producing lower vertical resolution (Å 7 km), while achieving much finer latitudinal resolution (Å 0.5¡). In this mode of scanning line-of-sight averaging (now more zonal) remains 240 km.
The number of steps in the limb scans and the dwell time at a given position are programmable. Each FOV can be stepped by half its vertical dimension (Å 2.5 km at the aft limb) to explore for finer scale vertical structure in the atmosphere or on the surface. Also, the mix of nadir, on-planet and limb views can be changed as conditions require. However, the need to obtain a long record with systematic coverage of the meteorological fields means that the special purpose sequences will necessarily be interleaved with the nominal observational sequence.
Coordinated Observations PMIRR will carry at launch specialized sequences for coordinated measurements with other experiments. Coordinated measurement with the Radio Science (RS) investigation will provide the best means of validating PMIRR's temperature retrieval algorithms. The RS temperatures are particularly valuable as they are derived from measurements of physically independent parameters (i.e. refractive index as opposed to thermal emission) and have very high vertical resolution (Å1 km). In the southern hemisphere, off-track limb scans allow PMIRR to make nearly simultaneous measurements of the same parcel of atmosphere sampled by radio occultation. Since it cannot view the forward limb, PMIRR views near the occultation tangent point seen by RS in the northern hemisphere some 10 to 20 minutes after the occultation has occurred.
TES views the forward in-track limb on a regular, though infrequent, basis. Thus, both PMIRR and TES look at the in-track limb, but in opposite directions. PMIRR cross-track limb scanning, interleaved with in-track measurements, will be used to make limb measurements coincident with those obtained by the MOC wide-angle camera. MOC observations of the limb can reveal with 2-km vertical resolution tenuous hazes known to exist from Mariner and Viking limb imaging (Jaquin et al., 1986). When superimposed on the 5-km resolution PMIRR profiles, these data will yield additional information on the vertical structure and saturation state of the atmosphere.
4. RETRIEVAL APPROACH AND SIMULATED RESULTS
This section shows how we intend to retrieve atmospheric profiles and surface properties from PMIRR data, given the instrument characteristics and observational strategy discussed above. One means of assessing the anticipated impact of PMIRR data is through retrieval simulations using synthetic radiance data. Such simulations have been carried out for PMIRR, and representative samples are shown here.
4.1 RETRIEVAL TECHNIQUES
The retrieval of atmospheric variables from remotely sensed infrared radiances requires the solution of a nonlinear radiative transfer equation, namely a Fredholm integral equation of the first kind. The technique used by PMIRR is the iterative relaxation method (Chahine, 1970). This method starts with initial estimates of surface temperature and pressure and of the atmospheric profiles of temperature and dust extinction. Instrument radiances are then calculated using these fields as input to a forward radiative transfer model. Initial estimates are updated using perturbations based on the difference between the calculated radiances and the actual radiance measurements.
These perturbations alter the input profile at levels corresponding to the peaks of the weighting functions, and the updated profile is constrained by linear interpolation between these levels. This approximation relies on the mean-value theorem and works best for narrow, spatially independent weighting functions. At one-half scale height vertical resolution, PMIRR provides a near optimum set of largely spatially independent samples of the Martian atmosphere, and tests show that within the range covered by these samples, the retrieved profile is insensitive to the initial estimate. After the input fields are perturbed, new instrument radiances are calculated and compared with the observed radiances. This process is repeated iteratively until the measured and calculated radiances converge to within the noise of the measurement.
Iterative retrieval methods of this kind require no climatological information (a distinct advantage for Mars), work well for non-linear problems, and can be adapted easily to apply ad hoc physical constraints to retrieved profiles. Their chief disadvantages are the greater use of computer time, the lack of formal error statistics associated with the retrieved profile, and their limited ability to make full use of a priori information, when available. Considerable effort is being devoted to making the PMIRR retrieval scheme computationally fast, so that retrieval can keep pace with the rate at which observations are made during the mission.
Temperature Figure 12 shows the vertical response or weighting functions for the three PMIRR temperature sounding channels and a single limb scan. Field-of-view (FOV) averaging has been included. PMIRR sounds temperature using a novel approach in which three pressure modulator radiometer channels observe different spectral emissions features of CO2 by viewing through a single PMC containing a known mixture of CO2 isotopes. This cell, 3.7 cm long, has a total pressure of 80 mbar.
In the lower atmosphere, where the amounts of CO2 along the limb path are extremely large, measurements are made near 7 µm in the weak fundamental n1 band of the isotope 12C16O18O. The concentration of this isotope is enhanced in the PMC to increase the pressure modulated signal component. Models derived from Mariner 9 suggest that 7 µm is also a region of minimum dust extinction in the Martian atmosphere (Fig. 13; Toon et al., 1977). In the middle atmosphere, where CO2 amounts in the limb path are smaller, measurements are made using the wing and, at higher altitudes, the R-branch of the fundamental n2 band of the most abundant isotope. At these levels, dust extinction is generally small.
Pressure All atmospheric profiles derived from PMIRR measurements are referenced to constant pressure levels, independent of uncertainties in the knowledge of the spacecraft attitude. For a given limb view, atmospheric pressure at the tangent altitude can be derived from the pressure modulated and wideband signal components of the PMIRR temperature sounding channels, provided the opacity of the path is dominated by CO2. The ratio of these signal components is insensitive to atmospheric temperature but is a strong function of pressure at the tangent point of the limb path. Pressure and temperature are therefore retrieved simultaneously with a precision limited only by instrument signal-to-noise performance.
For PMIRR the best results are obtained from the 14.6 µm channel in the 30 to 50 km region because of good signal-to-noise performance and generally negligible (infrared) dust opacity. Given pressure at these altitudes, pressure at other levels are derived from the hydrostatic relation using the retrieved temperature profile and the accurate relative altitude scale provided by the PMIRR limb scan steps. Retrieved temperature, dust, and water vapor profiles are then referenced to simultaneously obtained, co-located pressures.
The pressure retrieval capabilities of PMIRR are similar to those of earth orbiting limb sounders such as the SAMS on Nimbus 7. This instrument retrieved pressures to approximately 1.5% using temperature sounding CO2 pressure modulator channels (Barnett et al., 1985); PMIRR should do as well at altitudes above 40 km on Mars in the absence of significant dust or condensate haze at those altitudes. Estimation of surface pressure is limited primarily by the difficulty of locating the surface at the limb within PMIRR's finite field of view.
By combining nadir and limb sounding observations, it is estimated that PMIRR will detect changes of 5 - 7% in surface pressure with an absolute accuracy of 15 - 20%. This absolute inaccuracy is due mainly to systematic errors in the nadir measurements produced by uncertainties in surface emissivities and the spectroscopy of CO2 over the inhomogeneous nadir path. These systematics should be reduced by intercomparison with TES, to better determine surface characteristics, and with the Radio Science (RS) experiment, which derives temperatures from spacecraft occultation data. PMIRR utilizes its 2-axis scanning mirror to observe the atmosphere near the tangent points of the occultation path; for Mars Observer the RS occultations occur at high latitudes.
Non-LTE Radiative transfer models of the Martian atmosphere indicate that local thermodynamic equilibrium (LTE) in the fundamental n2 band of CO2 will hold to 70 km and perhaps higher. Under non-LTE conditions atmospheric emission is represented not by the the Planck function, but by a band source function which is determined by a complex balance between radiative, vibrational, and translational excitation and relaxation processes for that band. The effect of non-LTE on PMIRR temperature retrievals is negligible below 70 km. At greater heights non-LTE effects can be taken into account using models of the n2 band source function. The day-night contrasts measured by PMIRR at high altitudes should yield information on non-LTE radiative transfer processes in the upper atmosphere of Mars, since solar photons absorbed in the near-IR bands of CO2 can be exchanged as vibrational quanta which are re-emitted in the 15-µm band.
Dust and Condensate Extinction Profiling PMIRR measures aerosol extinction profiles in eight spectral channels covering the range 6.5 to 50 µm. These profiles are derived from limb emission measurements performed simultaneously and with the same vertical and horizontal resolution as the temperature measurements. Observations from the Mariner 9 and Viking spacecraft have shown that the infrared opacity of airborne dust is strongly dependent on wavelength (Toon et al., 1977; Martin et al., 1979), and have allowed opacities at infrared and visible wavelengths to be related (Zurek, 1982; Martin, 1986). Models derived from this data have been used to guide the spectral placement of PMIRR channels and to relate opacities in different channels. For altitudes with relatively low dust opacity, estimates of the dust optical depth can be made independent of a priori knowledge of dust spectral characteristics in those channels having both a pressure-modulated and a wide-band signal.
Four PMIRR spectral channels are designed primarily for dust and condensate profiling (Fig. 13, Table 1). Channels at 11.8 and 20.5 µm are chosen for their contrasting sensitivity to small amounts of water ice and dust. A channel at 32.6 µm is located in a spectral interval where gaseous absorption is negligible and serves as a reference for total line-of-sight opacity due to airborne particulates. The 20.5 and 32.6 µm channels each has two detectors to cover the full height range (Fig. 8, Table 1). Water-ice hazes are covered throughout this range by the combination of the 11.8 and 48.3 µm channels. The latter is particularly useful when dust opacities are high. The 32.6 and 48.3 µm channels are relatively insensitive to dust and condensates having the size distributions and number densities expected in the Martian atmosphere but, again, have contrasting sensitivities to water vapor, water ice and dust. Figure 13 compares the spectral location of all the PMIRR infrared channels with extinction spectra for dust, H2O ice, and CO2 ice, computed using particle sizes suggested by Mariner 9 and Viking orbiter observations (Toon et al., 1977; Curran et al., 1973).
The independent retrieval of temperature in the presence of particulates is essential for reliable dust and condensate profiling, and is made possible by the two signal components of the pressure modulator temperature sounding channels. Because the pressure modulated and wideband signal components share fields-of-view and spectral bandpasses, they have identical responses to emission from particulates. The large contrast in response to CO2 emission for the two signal components therefore allows particulate and CO2 emission to be differentiated by each temperature sounding channel, once pressure is determined in that field of view as described earlier.
Given the vertical profile of temperature as a function of pressure, profiles of dust and condensate extinction coefficients are derived from limb and nadir emission measurements made in each PMIRR infrared channel. Two processes contribute to particulate emission: thermal emission and scattering. As particle single scattering albedos are small at infrared wavelengths, the primary scattering of upwelling radiation from the surface dominates the scattering source function and multiple scattering can be ignored. Under these conditions multi-spectral extinction profiles can be retrieved and used to distinguish between dust and condensates and to investigate their microphysical properties.
PMIRR will also measure on-planet and limb radiances in its broad-band solar channel (Table 1). The limb-viewing solar channel will observe solar radiation scattered to the instrument by atmospheric hazes; the measured radiances can be used to locate the altitudes of the hazes and to constrain their visible opacities (e.g., Jaquin et al., 1986). A direct measurement of the visible broad-band solar extinction can be made over the north polar region on those occasions when PMIRR continuously views its solar calibration target as Mars Observer approaches the terminator from the night side. At those times the solar calibration target, which faces the forward limb, is illuminated by the rising sun occulted by the Martian atmosphere. Finally, the combination of on-planet views of the broad-band solar channel which have different reflection and incidence angles can be used to constrain broad-band visible opacities and scattering properties of suspended dust and the reflectance of the surface itself (e.g., Thorpe, 1977). This enables PMIRR to monitor the advance and retreat of the seasonal polar caps (e.g., Christensen and Zurek, 1984), and the redistribution of surface aeolian materials (e.g., Christensen, 1988).
Water Vapor Profiling PMIRR derives water vapor profiles from emission measurements in channels centered at 6.9 and 48.3 µm in the spectral features of H2O. These measurements are coincident with, and have the same spatial resolution as, those of temperature and particulates.
The 6.9 µm channel is centered in the near wing of the H2O n2 vibration-rotation band, close to a minimum in the dust opacity spectrum, and uses pressure modulation radiometry to achieve high sensitivity to water. The pressure modulator cell is 10 cm long and contains H2O at a pressure of 17 mbar. Water vapor concentration and 6.9 µm dust extinction are retrieved independently from the pressure modulated and wideband components of the 6.9 µm signal, given the temperature profile. The 48.3 µm channel is located in the water vapor rotation band and employs simple passband filter radiometry. At this wavelength, the channel is relatively insensitive to dust opacity and has less sensitivity to a given temperature error, as determined by the Planck function, and so it can be used to augment water vapor profile retrievals derived from measurements with the 6.9 µm channel.
Given the retrieved temperature profile, profiles of water vapor concentration will be derived from limb and nadir emission measurements made in the 6.9 and 48.3 µm channels. For low water abundances (² 10 pr. µm) and background dust opacities, the 6.9 µm channel yields the best results, whereas the 48.3 µm channel is best suited to dustier conditions and greater abundances of water.
Polar Radiative Balance The PMIRR solar and infrared channels directly measure the solar and infrared radiation emerging from the top of the Martian atmosphere. The nominal spectral response of the PMIRR solar channel covers 97% of the incident solar flux. Observations in this channel, obtained over a wide range of viewing geometries, will be used to determine the bi-directional reflectance of the surface-atmosphere system. Multi-spectral measurements in the eight PMIRR infrared channels will be used to synthesize bolometric emission spectra as a function of emission angle.
Figure 13 shows the spectral location of the PMIRR infrared channels relative to the cumulative Planck functions for 150 and 300 K blackbodies. Errors in net emission produced by the necessary interpolation and extrapolation of emission spectra are small because the bolometric integral is well sampled by PMIRR. Errors are reduced further by modeling the upward flux at the top of the atmosphere as a function of wavelength, using atmospheric properties derived from simultaneous PMIRR observations. Studies using Viking observations have shown that the net flux of solar and IR radiation at the top of the atmosphere dominates the polar heat balance during most seasons (Paige and Ingersoll, 1985); this makes it possible to determine condensation and sublimation rates of CO2 on the surface and in the atmosphere from remote observations.
Surface Temperatures PMIRR will measure on-planet radiances in several FOVs at eight infrared wavelengths as part of its nominal observational mode. At nadir each smeared FOV is Å 7 km along-track and 5 km cross-track; channels 6 and 7 (at 20.5 and 32.6 µm) have two FOVs each and all other surface sensing channels are superimposed on one of these two. In the 33 and 48 µm channels, the radiances are dominated by thermal emission from the surface, and they do also not exhibit the non-linear response to non-isothermal surface fields-of-view which complicates the interpretation of data obtained at shorter wavelengths. An integral part of the PMIRR data reduction plan is to use these radiances to retrieve the temperature and spectral emissivity of the surface. Surface temperatures can be used to map at relatively coarse resolution (e.g., 2¡ x 2¡) the bulk thermal properties of the Martian surface and to locate surface frost and ice deposits. CO2 frost deposits can be identified by temperatures close to the CO2 solid-vapor equilibrium temperature of Å148 K at 6 mbar CO2 partial pressure.
4.2 ANTICIPATED RESULTS
The ability to retrieve atmospheric profiles from PMIRR radiance measurements cannot be simulated using existing data alone, since the data now in hand lack global coverage or adequate vertical resolution and coverage. There are, for instance, very few meridional cross-sections of temperature, such as the ones retrieved from Mariner 9 (Fig. 1). Furthermore, the region 40 - 80 km in altitude is virtually unexplored, and only column values of dust extinction or water vapor amounts are available for much of the planet (e.g., Fig. 2).
For these reasons, the retrieval simulations shown here have been extended using atmospheric fields computed by models representing the range of atmospheric conditions now anticipated at Mars. From these fields, whether observed or computed, synthetic PMIRR radiances are generated to simulate limb and on-planet observations. These synthetic radiances include the appropriate geometrical factors and known instrumental noise. Using the current retrieval algorithms, vertical profiles are retrieved from these synthetic radiances and can then be compared with the original input fields.
Temperature, Dust Extinction, and Pressure Measurements Figure 14 shows a synthetic climatology used to assess the ability to retrieve vertical profiles of temperature and dust as a function of pressure from PMIRR measurements. The one-orbit cross-section was constructed using vertical profiles of temperature taken from history tape generated as part of a southern summer solstice simulation by the Ames Research Center Mars General Circulation Model or GCM (Pollack et al., 1990). The Mars GCM currently has a longitude-latitude resolution of 9¡ x 7.5¡ and extends to Å 50 km in altitude. Dayside profiles were taken from the longitude whose local time was closest to 2 p.m. and nightside profiles were appended from the longitude whose local time was closest to 2 a.m.. The resulting cross-section (Fig. 14) thus extends from 90¡N through the south pole (dayside, descending portion of the orbit) and back to 90¡N (nightside, ascending portion of the orbit). Above 40 km the MGCM temperature profiles were smoothly merged with profiles obtained from the Mars Global Reference Atmospheric Model (MARS-GRAM) engineering model (Justus, 1990). Note that PMIRR obtains the coverage shown in Fig. 14 for each orbit and with a latitudinal resolution of Å 4¡.
Figures 15-16 show the differences between the retrieved temperatures and dust extinction values and the original values input from the model fields, as a function of latitude and height. The dust opacity in the MGCM run was specified to have a visible optical depth tv = 2, and the dust was assumed to be uniformly mixed with height. In order to examine the effects of different dust concentrations on the combined retrieval of temperature and dust extinction, the synthetic radiances were generated assuming that tv from 0.5 at the poles to 2 at the equator. This spans a range representative of the background dust loading above the Viking Lander sites to that characteristic of a planetary-scale dust storm (Pollack et al., 1979). As a further challenge to the retrieval algorithm, the vertical distribution of suspended dust was randomly perturbed with height; these variations are evident in Fig. 17, where several individual profiles from the simulation run are shown.
Below 80 km, temperature errors in the retrieved profiles are typically 2 K or less (Fig. 15) and the percent error in retrieved dust extinction (opacity) is typically 10% or less (Fig. 16). The largest errors occurred when the opacity was greatest at the region of transition between channels 1 and 2 (Fig. 8). [The large relative errors in Fig. 16 above 50 km in the polar regions are due to the very cold, nearly isothermal atmosphere and to the very small amounts of dust assumed to be present at those altitudes.] Errors in retrieved pressures (not shown) are typically 1-2% from profile to profile. Errors in estimated surface pressure are less than 3%, but do not include uncertainties in locating the surface.
While Figs. 15-16 are excellent representations of the relative error characteristics of retrieved profiles, not all systematic errors have been included. Additional systematic errors in temperature and pressure retrievals are expected due to uncertainties in the calculation of CO2 transmittances in the Martian atmosphere. Also, it was assumed that the limb FOVs were either entirely on or entirely off the ground. Sources of systematic errors not accounted for in the simulated dust retrievals include the scattering of surface emission into the limb field-of-view and the assumed ratio of dust opacity between the various IR wideband (WB) channels. These ratios are needed when the dust loading is so large that the ratio of PMC-to-WB signals essentially saturates and cannot provide an independent estimate of the dust opacity; the variation of dust opacity with wavelength (Fig. 8) was computed for the particle size distribution and compositions suggested by the analysis of Mariner 9 and Viking data (Toon et al., 1977; Hunt, 1979; Zurek, 1982; Kahn et al., 1992).
Water Vapor and Condensates Atmospheric condensates have been identified in Mariner 9 and Viking Orbiter visible images of Mars (Anderson and Leovy, 1978; Jaquin et al., 1986). They are thought to be composed of CO2 and H2O ices, occur in distinct layers as well as broadly distributed hazes, and may frequently be mixed with airborne dust. PMIRR must therefore identify the various compositional components of the Martian aerosols and map their distribution. Figure 18 shows a simulated retrieval of H2O ice and dust mass loading profiles for dust and ice mixed in the same atmospheric column. For this test case dust and condensate mass loadings are determined with a precision of Å 20% by extinction measurements in PMIRR channels 5-8. The broad band solar channel can also be used to estimate the optical properties and so constrain the microphysical characteristics of condensate hazes, but this has not yet been included in the operational retrieval algorithm.
Given the retrieved temperature profile, water vapor profiles are derived from limb radiance measurements in channels 4 and 8. Figure 19 shows simulated water vapor profile retrievals for two very different temperature profiles. A background column dust visible optical depth of 0.4 is assumed, and the water model is based on the altitude dependence of the saturated vapor pressure of water. Water vapor mixing ratios are retrieved with a precision of 10-20% in the lower atmosphere. The upper limit of PMIRR vertical coverage is determined largely by the atmospheric holding capacity (i.e., saturation vapor pressure for water), which decreases rapidly with decreasing temperature.
Polar Heat Balance PMIRR will acquire the spatial, temporal and angular coverage necessary to determine the reflected and emitted radiation fields in the polar regions by scanning a pre-selected set of 31 azimuth and elevation angles. Figure 13 showed the spatial coverage that can be obtained over the north pole by PMIRR in just 13 orbits during one day. PMIRR observations span most of the range of emission and solar azimuth angles needed to construct the bi-directional reflectance. The main gap in angular coverage is due to PMIRR's inability to look very far toward the forward limb. However, this relatively small, but important, gap is well sampled by TES which observes in-track toward the forward limb with its own solar channel.
CO2 Clouds Atmospheric temperature profiles provide the most important clue to the existence of CO2 clouds because, whatever their scattering and emission properties, the temperatures at the tops of these clouds must be at or below the local CO2 saturation temperature. The pressure-levels of the effective top of the clouds can be determined using the PMIRR wide-band and PMC radiances in Channels 1, 2 and 3. Below the tops of the clouds, the PMIRR limb radiances will depend in detail on the optical properties and opacity of the cloud particles. PMIRR's ability to observe at nadir, off-nadir and limb geometries should resolve, once and for all, whether the anomalously low brightness temperature regions observed by Viking (Kieffer et al., 1976) are due to surface or atmospheric phenomena. Vertical profiles of temperature and aerosol extinction would provide useful constraints on the condensation of volatiles in the winter polar atmosphere and the incorporation of dust and water into the seasonal polar caps.
Surface Science All standard PMIRR observational sequences include regular observations of solar reflectance and infrared emission. Retrieving surface temperatures from these observations will be an integral part of the PMIRR standard data reduction scheme. During a mapping cycle, the surface coverage that will be obtained by PMIRR will be sufficient to map globally surface temperatures at 2 a.m. and 2 p.m. and top-of-the atmosphere solar reflectances to spatial resolutions of 2¡ x 2¡. At high latitudes, the increased temporal and spatial sampling that results from the spacecraft orbit and PMIRR's frequent on-planet viewing for polar radiative balance measurements, will enable PMIRR to obtain maps of surface temperatures and top-of-the-atmosphere solar reflectances to spatial scales of 0.5¡ x 0.5¡ every 5 days.
PMIRR's thermal and reflectance maps will be used for a wide range of purposes, including determinations of: 1) the apparent thermal inertia of the surface of Mars from diurnal surface temperature variations (Palluconi and Kieffer, 1981); 2) the boundaries of the seasonal polar caps; 3) the actual surface albedo and surface thermal inertia using one-dimensional radiative-convective models in conjunction with PMIRR-derived atmospheric temperature and opacity fields; and 4) high-latitude subsurface temperatures and thermal properties using the observed seasonal cycles of surface temperatures.�
5.0 SUMMARY
The exploration of Mars will enter a new phase with Mars Observer. The low, circular and nearly polar orbit of the spacecraft will provide an ideal platform for the systematic remote investigation of the atmosphere and seasonal climate of Mars. PMIRR is designed to take full advantage of that platform to map the global fields of temperature, dust distribution, water vapor, surface ice and atmospheric condensates, and the reflected solar and emitted thermal emission at the top of the atmosphere. From these fields, maps can be derived of horizontal wind (shear), diabatic forcing of the atmosphere, zonal-mean transport (i.e., residual-mean) velocities, surface radiation components, and key parameters of the surface energy budget, such as the change of CO2, the surface albedo and the apparent thermal inertia.
The climatologies generated for Mars by PMIRR will be unprecedented in their systematic global coverage and good vertical and horizontal resolution. Just as for the case of Earth, such climatologies provide a basis for seeking an improved understanding of the physical processes underlying the present structure and future change of the atmosphere and climate of Mars (Zurek and McCleese, 1989). We may gain insight into the past climates of Mars and perhaps, through comparative study, of Earth's as well. Each of the past spacecraft investigations of Mars has yielded new, and often dramatically changed, perspectives of that planet. Mars Observer is unlikely to be an exception.
ACKNOWLEDGMENTS
We gratefully acknowledge the contributions of the following to preparation of this manuscript: S. Calcutt and T. Kingsmill-Vellicott of Oxford University; A. Ono of UCLA; S. Gaiser, D. Rider, H. Stone and R. West of JPL/Caltech. A heartfelt thanks is extended to all those members of the PMIRR Hardware and Pre-Launch Science Teams whose contributions to design, fabrication and calibration of the instrument are making this investigation possible. We also thank J. Pollack, R. Haberle and J. Schaeffer of NASA Ames Research Center for providing the Mars GCM history tape, as well as L. Chen and J. Knighton of JPL for assisting in the creation of the combined atmospheric model for testing the retrieval schemes. This work was funded by the Mars Observer Project and was carried out at the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration.
References
Anderson, E., and Leovy, C.B., Mariner 9 television limb observations of dust and of ice hazes on Mars, J. Atmos. Sci., 35, 723-734, 1978.
Andrews, D. G., J. R. Holton, and C. B. Leovy, Middle Atmosphere Dynamics, Academic Press, 490pp., 1987.
Arvidson, R. E., Guiness, E. A., Moore, H. J., Tillman, J. E., and Wall, S. D., Three Mars years: Viking Lander 1 imaging observations, Science, 222, 463-468, 1983.
Barker, E. S., R. A. Schorn, A. Woszcsyk, R. B. Tull, and S. J. Little, Mars: Detection of atmospheric water vapor during the southern hemisphere spring and summer season, Science, 170, 1308-1310, 1970.
Barnes, J.R., Time spectral analysis of midlatitude disturbances in the Martian atmosphere, J. Atmos. Sci., 37, 2002-2015, 1980.
Barnes, J. R., Midlatitude disturbances in the Martian atmosphere: A second Mars year, J. Atmos. Sci., 38, 225-234, 1981.
Barnes, J. R., Transport of dust to high northern latitudes in a Martian polar warming, J. Geophys. Res., 95 , 1381-1400, 1990a.
Barnes, J. R., Possible effects of breaking gravity waves on the circulation of the middle atmosphere of Mars, J. Geophys. Res., 95 , 1401-1422, 1990b.
Barnes, J. R., and Hollingsworth, J. L., Dynamical modeling of a planetary wave mechanism for a Martian polar warming, Icarus, 71, 313-334, 1987.
Barnett, J. J., et al., Global and seasonal variability of the temperature and composition of the middle atmosphere, Nature, 313, 439- , 1985.
Chahine, M. T., Inverse problems in radiative transfer: Determination of atmospheric parameters, J. Atmos. Sci., 27, 960- , 1970.
Christensen, P. R., Regional dust deposits on Mars: Physical properties, age, and history, J. Geophys. Res., 91, 7611-7624, 1986.
Christensen, P. R., Global albedo variations on Mars: Implications for active aeolian transport, deposition and erosion, J. Geophys. Res., 93, 7611-7624, 1988.
Christensen, P. R., and Zurek, R. W., Martian north polar hazes and surface ice: Results from the Viking Survey/Completion Mission, J. Geophys. Res., 89, 4587-4596, 1984.
Conrath, B. J., Thermal structure of the Martian atmosphere during the dissipation of the dust storm of 1971, Icarus, 24, 36-46, 1975.
Conrath, B. J., Influence of planetary-scale topography on the diurnal thermal tide during the 1971 Martian dust storm, J. Atmos. Sci., 33, 2430-2439, 1976.
Conrath, B. J., Planetary-scale wave structure in the Martian atmosphere, Icarus, 48, 246-255, 1981.
Conrath, B., R. Curran, R. Hanel, V. Kunde, W. Maguire, J. Pearl, J. Pirraglia, J. Welker, and T. Burke, Atmospheric and surface properties of Mars obtained by infrared spectroscopy on Mariner 9, J. Geophys. Res., 78, 4267-4278, 1973.
Curran, R. J., Conrath, B. J., Hanel, R. A., Kunde, V. G., and Pearl, J. C., Mars: Mariner 9 spectroscopic evidence for H2O ice clouds, Science, 182, 381-383, 1973.
Davies, D. W., Effects of dust on the heating of Mars' surface and atmosphere, J. Geophys. Res., 84, 8289-8293, 1979.
Davies, D. W., The Mars water cycle, Icarus, 45, 398-414, 1981.
Deming, D., M. J. Mumma, F. Espenak, T. Kostiuk, and D. Zipoy, Polar warming in the middle atmosphere of Mars, Icarus, 66, 366-368, 1986.
Fanale, F. P., Martian volatiles: Their degassing history and geochemical fate, Icarus, 28, 179-202, 1976.
Fanale, F. P., and W. A. Cannon, Mars: CO2 adsorption and capillary condensation on clays¾Significance for volatile storage and atmospheric history, J. Geophys. Res., 84, 8404-8414, 1979.
Farmer, C. B., D. W. Davies, A. L. Holland, D. D. LaPorte, and P. E. Doms, Mars: Water vapor observations from the Viking orbiters, J. Geophys. Res., 82, 4225-4248, 1977.
Farmer, C. B., and P. E. Doms, Global seasonal variation of water vapor on Mars and the implications for permafrost, J. Geophys. Res., 84, 2881-2888, 1979.
Garcia, R. R., and S. Solomon, A numerical model of zonally averaged dynamical and chemical structure of the middle atmosphere, J. Geophys. Res., 88, 1379-1400, 1983.
Gierasch, P. J., Martian dust storms, Revs. Geophys. Space Phys., 12, 730-734, 1974.
Gierasch, P. J., and R. M. Goody, The effect of dust on the temperature of the Martian atmosphere, J. Atmos. Sci., 29, 400-402, 1972.
Gifford, F. A., A study of Martian yellow clouds that display movement, Mon. Wea. Rev., 92, 435-440, 1964.
Gille, J. C., and J. M. Russell III, The Limb Infrared Monitor of the Stratosphere: Experiment description, performance and results, J. Geophys. Res., 89, 5125, 1984.
Gille, J. C., L. V. Lyjak, and A. Smith, The global residual mean circulation in the middle atmosphere for the northern winter period, J. Atmos. Sci., 44, 1437-1452, 1987.
Haberle, R. M., Interannual variability of global dust storms on Mars, Science, 234, 459-461, 1986.
Haberle, R. M., C. B. Leovy, and J. B. Pollack, Some effects of global dust storms on the atmospheric circulation of Mars, Icarus, 50, 322-367, 1982.
Haberle, R. M., and Jakosky, B. M., Sublimation and transport of water from the north residual polar cap on Mars, J. Geophys. Res., 95, 1423-1437, 1990.
Hamilton, K., and Garcia, R. R., Theory and observations of the short-period normal mode oscillations of the atmosphere, J. Geophys. Res., 91, 11867-11875, 1986.
Hanel, R., B. Conrath, W. Hovis, V. Kunde, P. Lowman, W. Maguire, J. Pearl, J. Pirraglia, C. Prabhakara, B. Schlachman, G. Levin, P. Straat, and T. Burke, Investigation of the Martian environment by infrared spectroscopy on Mariner 9, Icarus, 17, 423-442, 1972.
Hess, S. L., J. A. Ryan, J. E. Tillman, R. M. Henry, and C. B. Leovy, The annual cycle of pressure on Mars measured by Viking Landers 1 and 2, Geophys. Res. Let., 7, 197-200, 1980.
Houghton, J. T. and S. D. Smith, Remote sounding of atmospheric temperature from satellites, I. Introduction, Proc. Roy. Soc. London Ser., A 320, 23-, 1970.
Hunt, G. E., Thermal infrared properties of the Martian atmosphere. 4. Predictions of the presence of dust and ice clouds from Viking IRTM spectral measurements, J. Geophys. Res., 84, 2865-2874, 1979.
Jakosky, B. M., The role of seasonal reservoirs in the Mars water cycle. I. Seasonal exchange of water with the regolith, Icarus, 55, 1-18, 1983a.
Jakosky, B. M., The role of seasonal reservoirs in the Mars water cycle. II. Coupled models of the regolith, the polar caps, and atmospheric transport, Icarus, 55, 19-39, 1983b.
Jakosky, B. M., The seasonal cycle of water on Mars, Space Sci. Rev., 41, 131-200, 1985. Jakosky, B. M., and E. S. Barker, Comparison of groundbased and Viking Orbiter measurements of Martian water vapor: Variability of the seasonal cycle, Icarus, 57, 322-334, 1984.
Jakosky, B. M., and C. B. Farmer, The seasonal and global behavior of water vapor in the Mars atmosphere: Complete global results of the Viking atmospheric water detector experiment, J. Geophys. Res., 87, 2999-3019, 1982.
Jakosky, B. M., and R. M. Haberle, The seasonal behavior of water on Mars, The MARS Book, U. Arizona Press, in revision, 1992.
James, P. B., The Martian hydrological cycle: Effects of CO2 mass flux on global water distribution, Icarus, 64, 249-264, 1985.
James, P. B., and G. R. North, The seasonal CO2 cycle on Mars: An application of an energy balance climate model, J. Geophys. Res., 87, 10271-10283, 1982.
Jaquin, F., Gierasch, P., and Kahn, R., The vertical structure of limb hazes in the Martian atmosphere, Icarus, 68, 442-461, 1986.
Justus, C. G., A Mars global reference atmospheric model (MARS-GRAM) for mission planning and analysis. Submitted.
Kahn, R., Some observational constraints on the global-scale wind systems of Mars, J. Geophys. Res., 88, 10189-10209, 1983.
Kahn, R., The evolution of CO2 on Mars, Icarus, 62, 175-190, 1985.
Kahn, R., S. Lee, T. Z. Martin and R. Zurek, The Martian dust cycle, The MARS Book, U. Arizona Press, in revision, 1992.
Kieffer, H. H., Mars south polar spring and summer temperatures: A residual CO2 frost, J. Geophys. Res., 84, 8263-8288, 1979.
Leighton, R. B., and B. C. Murray, Behavior of carbon dioxide and other volatiles on Mars, Science, 153, 136-144, 1966.
Leovy, C. B., Observations of Martian tides over two annual cycles, J. Atmos. Sci., 38, 30-39, 1981.
Leovy, C. B., Martian meteorological variability, Adv. Space Res., 2, 19-44, 1982.
Leovy, C. B., The general circulation of Mars: Models and observations, Adv. Geophys., 28a, 327-346, 1985.
Leovy, C. B., and Y. Mintz, Numerical simulation of the atmospheric circulation and climate of Mars, J. Atmos. Sci., 26, 1167-1190, 1969.
Leovy, C. B., Zurek, R. W., and Pollack, J. B., Mechanisms for Mars dust storms, J. Atmos. Sci., 30, 749-762, 1973.
Leovy, C. B., Tillman, J. E., Guest, W. R., and Barnes, J., Interannual variability of Martian weather, Recent Advances in Planetary Meteorology, ed. G. Hunt, Cambridge University Press, 69-84, 1985.
Lindal, G. F., Hotz, H. B., Sweetnam, D. N., Shippony, Z., Brenkle, J. P., Hartsell, G. V., Spear, R. T., and Michael, Jr., W. H., Viking radio occultation measurements of the atmosphere and topography of Mars: Data acquired during 1 Martian year of tracking, J. Geophys. Res., 84, 8443-8456, 1979.
Lorenz, E. N., The Nature and Theory of the General Circulation of the Atmosphere, World Meteor. Org., 162pp, 1967.
Martin, L. J., The major Martian yellow storm of 1971, Icarus, 22, 175-188, 1974a.
Martin, L. J., The major Martian dust storms of 1971 and 1973, Icarus, 23, 108-115, 1974b.
Martin, L. J., 1973 dust storm on Mars: Maps from hourly photographs, Icarus, 29, 363-380, 1976.
Martin, L. J., Clearing the Martian air: The troubled history of dust storms, Icarus, 57, 317-321, 1984.
Martin, L. J., and W. A. Baum, A study of could motions of Mars, Final Report B, JPL Contract 951547, Lowell Observatory, Flagstaff, Ariz, 1969.
Martin, T. Z., Mean thermal and albedo behavior of the Mars surface and atmosphere over a Martian year, Icarus, 45, 427-446, 1981.
Martin, T. Z., Thermal infrared opacity of the Mars atmosphere, Icarus, 66, 2-21, 1986.
Martin, T. Z., A. R. Peterfreund, E. D. Miner, H. H. Kieffer, and G. E. Hunt, Thermal infrared properties of the Martian atmosphere. 1. Global behavior at 7, 9, 11 and 20 µm, J. Geophys. Res., 84, 2830-2842, 1979.
McCleese, D. J., J. T. Schofield, R. W. Zurek, J. V. Martonchik, R. D. Haskins, D. A. Paige, R. A. West, D. J. Diner, J. R. Locke, M. P. Chrisp, W. Willis, C. B. Leovy, and F. W. Taylor, Remote sensing of the atmosphere of Mars using infrared pressure modulation and filter radiometry, Appl. Optics, 25, 4232-4345, 1986.
Moriyama, S., Effects of dust on radiation transfer in the Martian atmosphere. III. Numerical experiments of radiative-convective equilibrium of the Martian atmosphere including the radiative effects due to dust, J. Meteor. Soc. Japan, 54, 52-57, 1976.
Moriyama, S., and T. Iwashima, A spectral model of the atmospheric general circulation of Mars: A numerical experiment including the effects of the suspended dust and the topography, J. Geophys. Res., 85, 2847-2860, 1980.
Murphy, J. R., Toon, O. B., Haberle, R. M., and Pollack, J. B., Numerical simulations of the decay of Martian global dust storms, J. Geophys. Res. , 95 , 14,629-14,648, 1990.
Paige, D. A., The Annual Heat balance of the Martian Polar Caps: Viking Observations, Ph.D. Thesis, California Institute of Technology, 207pp., 1985.
Paige, D. A., and A. P. Ingersoll, Annual heat budget of martian polar caps: Viking Observations, Science, 228, 1160-1168, 1985.
Paige, D. A., K. E. Herkenhoff, and B. C. Murray, Mariner 9 observations of the south polar cap of Mars: Evidence for residual CO2 frost, J. Geophys. Res., 95, 1319-1335, 1990.
Palluconi, F. D., and H. H. Kieffer, Thermal inertia mapping of Mars from 60¡S to 60¡N, Icarus, 45, 415-426, 1981.
Peterfreund, A. R., Contemporary aeolian processes on Mars: Local dust storms, Ph. D. Thesis, Arizona State University, 247 pp., 1985.
Pirraglia, J., Martian atmospheric lee waves, Icarus, 7, 517-530, 1976.
Pirraglia, J., and B. Conrath, Martian tidal pressure and wind fields obtained from the Mariner 9 infrared spectroscopy experiment, J. Atmos. Sci., 31, 318-329, 1974.
Pollack, J. B., Climatic change on the terrestrial planets, Icarus, 37, 479-553, 1979.
Pollack, J. B., Colburn, D. S., Flasar, F. M., Kahn, R., Carlston, C. E., and Pidek, D., Properties and effects of dust particles suspended in the Martian atmosphere, J. Geophys. Res., 84, 2929-2945, 1979.
Pollack, J. B., C. B. Leovy, P. W. Greiman, and Y. Mintz, A Martian general circulation experiment with large topography, J. Atmos. Sci., 38, 3-29, 1981.
Pollack, J. B., J. F. Kasting, S. M. Richardson, and K. Poliakoff, The case for a wet, warm climate on early Mars, Icarus, 71, 203-224, 1987.
Pollack, J. B., Haberle, R. M. , Schaeffer, J., and Lee,H., Simulations of the general circulation of the Martian atmosphere, 1, Polar processes, J. Geophys. Res., 95, 1447-1474, 1990.
Reber, C. A., The Upper Atmosphere Research Satellite, Eos, 71, 1867-1878, 1990.
Ryan, J. A., Notes on the Martian yellow clouds, J. Geophys. Res., 69, 3759-3770, 1964.
Ryan, J. A., and R. M. Henry, Mars Atmospheric phenomena during major dust storms, as measured at the surface, J. Geophys. Res., 84, 2929-2945, 1979.
Ryan, J. A., and R. D. Lucich, Possible dust devils, vortices on Mars, J. Geophys. Res., 88, 11005-11011, 1983.
Seiff, A., and D. B. Kirk, Structure of the atmosphere of Mars in summer at mid-latitudes, J. Geophys. Res., 82, 4364-4378, 1977.
Smith, A. K., and L. V. Lyjak, An observational estimate of gravity wave drag from the momentum balance in the middle atmosphere, J. Geophys. Res., 90, 2233-2241, 1985.
Taylor, F. W., Pressure modulator radiometry, Spectrom. Tech., 4, 137- , 1983.
Thorpe, T. E., Viking orbiter observations of atmospheric opacity during July-November 1976, J. Geophys. Res. , 82, 4151-4159, 1977.
Tillman, J. E., Mars global atmospheric oscillations: Annually synchronized, transient normal mode oscillations and the triggering of global dust storms, J. Geophys. Res., 93, 9433-9451, 1988.
Toon, O. B., J. B. Pollack, and C. Sagan, Physical properties of the particles composing the Martian dust storm of 1971-1972, Icarus, 30, 663-696, 1977.
Webster, P. J., The low-latitude circulation of Mars, Icarus, 30, 626-649, 1977.
Zurek, R. W., Diurnal tide in the Martian atmosphere, J. Atmos. Sci., 33, 321-337, 1976.
Zurek, R. W., Solar Heating of the Martian dusty atmosphere, Icarus, 35, 196-208, 1978.
Zurek, R. W., Martian great dust storms: An update, Icarus , 50, 288-310, 1982.
Zurek, R. W., Atmospheric tidal forcing of the zonal-mean circulation: The Martian dusty atmosphere. J. Atmos. Sci., 43, 652-670, 1986.
Zurek, R. W., Free and forced modes in the Martian atmosphere, J. Geophys. Res., 93, 9452-9462, 1988.
Zurek, R. W., and P. R. Christensen, Dust redistribution by east-west circulations on Mars, Bull. Amer. Astron. Soc., 22, 1990.
Zurek, R. W., and C. B. Leovy, Thermal tides in the dusty Martian atmosphere: A verification of theory, Science, 213, 437-439, 1981.
Zurek, R. W., and D. J. McCleese, The role of climate studies in the future exploration of Mars. The Case for Mars III: Strategies for Exploration - General Interest and Overview, Ed. Carol Stoker, Science and Technology Series, 74, 277-285, 1989.
Zurek, R. W., J. R. Barnes, R. M. Haberle, J. B. Pollack, J. E. Tillman, and C. B. Leovy, Dynamics of the Atmosphere of Mars, The MARS Book, U. Arizona Press, in revision, 1992.
FIGURE LEGENDS
Figure 1: Latitude-pressure cross-section of temperatures [K, solid lines] retrieved from Mariner 9 IRIS data for the middle of northern spring (areocentric longitude Ls = 43 - 54¡). Zonal winds [m s-1, broken lines] were computed assuming geostrophic balance and no wind at the surface. The shaded region with the undefined equatorial boundary indicates a region of easterlies (i.e., westward) wind; elsewhere westerlies prevail. (Leovy, 1982, from data provided by B. Conrath.)
Figure 2: The seasonal cycle of the zonally averaged vertical column abundance of atmospheric water vapor, as derived from Viking MAWD measurements. Seasonal date is given in terms of Ls, the areocentric longitude of the sun, with the corresponding seasons for the northern and southern hemispheres given on the top and the bottom axes, respectively. Units are precipitable microns. Shaded regions indicate areas of no data, principally in the polar nights. The two horizontal arrows indicate the periods of onset and dissipation of the two planetary-scale dust storms that occurred that year. (Jakosky and Farmer, 1982).
Figure 3: The PMIRR instrument. The scan mirror assembly is in the foreground, with the dark instrument aperture next to the circular solar calibration target, shown with its deployable cover. The scan mirror has been rotated into the white hemispherical housing, where it is stored to avoid damage or contamination. A separate box housing electronics is seen to the right of the covered solar target. The passive radiative cooler is on the right, with its cover extended; once in orbit this placement provides the radiator a view of space and shields it from emission from Mars ("upward" in this perspective).
Figure 4: PMIRR schematic optical layout.
Figure 5: Schematic of a PMIRR Pressure Modulator Radiometer channel emphasizing mechanics of the pressure modulator unit. Pistons are set in motion at a resonant frequency (Å 50 Hz here), producing pressure/density variations of CO2 or H2O in the optical path through modulation of pressure in the gas cell. Mean pressure is determined using a temperature-controlled molecular sieve containing the gas of interest.
Figure 6: The spectral discrimination of a PMR for the CO2 line at 698.436 cm-1. The pressure modulator cell is 3.7 cm long and contains CO2 at a mean temperature of 300K modulated between pressures of p1 = 60 mbar and p2 = 120 mbar. Atmospheric emission (dashed line) is calculated for a limb view of the Martian atmosphere at 40 km tangent height, using the COSPAR standard model temperature profile. Dotted lines show cell transmission profiles at the two pressure extremes. The solid line is the difference between these profiles and represents the spectral response of the PMR.
Figure 7: Observing geometry for the nominal observation mode of PMIRR. The instrument field-of-view is stepped down the aft limb onto the surface and then to on-planet fields-of-view at local zenith angles of 60¡ and 0¡ (nadir). This sequence is then repeated. The projected field-of-view at the in-track, aft limb is 5 km high by 25 km cross-track; at nadir the unsmeared field-of-view is 1 km x 5 km cross-track. The line-of-sight averaging (LOS) is Å 240 km, or 4¡ along a great circle.
Figure 8: Locations of the projected centers (tangent points) of the three PMIRR detector fields-of-view when scanning down the aft limb of Mars as part of the nominal observational sequence. Numbers at the beginning and end of each limb segment indicate elapsed time in seconds from an arbitrary starting point. Note that the three middle segments shown are simultaneous, though displaced both vertically and horizontally from one another. The line-of-sight averaging distance for limb views is indicated at the top and the height of a single FOV at the aft limb is also shown. Calibration, space views and on-planet segments of the nominal observational sequence are not shown. Figure 9: Daily coverage of Mars achieved during the nominal observational sequence of PMIRR. Each point represents an on-planet view, while the heavy dashes indicate the horizontal range of tangent points during an aft limb scan. The latter run together at low latitudes, but are fewer in number (and so separate) as the spacecraft moves over the polar regions and the 'buckshot' mode commences. The nadir points taken during the aft limb scan are acquired more than eight minutes earlier than the limb scan at the same latitude, and so their longitudes are separated by the rotation of Mars.
Figure 10: Daily coverage of the Martian north pole achieved by continuous buckshot observations. Each point represents a PMIRR sounding in 9 spectral channels. Line segments indicate the horizontal trace of tangent points during a limb scan. During the nominal observational sequence, these traces would be nearly continuous on the equatorward (descending from the north pole) portions of the orbits. Note the nearly random coverage of the on-planet views.
Figure 11: Local times of day accessible to PMIRR. The spacecraft track and thus PMIRR's nominal (in-track) observations are near local times of 2 p.m. and 2 a.m., as shown by the heavy line, except very near the poles. Local times as much as 1.7 hours earlier and later are accessible by viewing cross-track; i.e., toward the side limbs orthogonal to the spacecraft track. Above 60¡ latitude, nearly all local times can be viewed. The positions shown indicate local times that could be viewed during a day in which PMIRR systematically scanned out to both side limbs, as well as to the aft limb. Dots and dashes indicate on-planet measurements; heavy dashes indicate (horizontal) range of tangent points in various vertical limb scans.
Figure 12: A typical suite of PMIRR wideband weighting functions chosen for limb temperature sounding from the total set of such functions provided by observations in three spectral intervals. The weighting function is defined here by K(z) = (dT(z,°)/dz), where T(z,°) is the effective transmission from level z to space along the line-of-sight of the instrument. The response has been calculated for a representative temperature profile at 1-km intervals in height and then convolved with the 5-km vertical FOV of the instrument.
Figure 13: PMIRR spectral channel locations compared with cumulative Planck functions and with the wavelength variation of atmospheric extinction due to suspended dust, water ice or carbon dioxide ice, as computed for representative particle size distributions (Curran et al., 1973; Toon et al., 1977). Absorption by CO2 gas is indicated by the Mariner 9 IRIS spectrum (Hanel et al., 1972).
Figure 14: Temperatures (K) used to simulate one orbit of PMIRR radiances. Values were taken from a GCM history tape generated by the NASA Ames Research Center Mars GCM (Pollack et al., 1990) in a simulation of a global dust storm at southern summer solstice; above 40 km these values were merged with temperatures from the MARS GRAM engineering model for the same season. Note the cold north polar atmosphere and strong temperature gradients at high latitudes.
Figure 15: Difference (model - retrieved, in Kelvins) between retrieved temperatures and model temperatures used to synthesize PMIRR radiances.
Figure 16: Relative difference in per cent of dust extinction; i.e., 100 x {ln [retrieved] - ln [model]}. The assumed column dust amount varies (as a cosine function in latitude) from a Channel 1 opacity of t7.6µm = 0.05 at the poles to a maximum of t7.6µm = 0.2 at the equator. Visible opacities tv (at Å 0.6 µm) are estimated to be 10 times larger (Fig. 8, Zurek, 1982); i.e., tv = 0.5 - 2.
Figure 17: Individual profiles comparing retrieved (dotted lines) versus model atmosphere (solid). Largest errors are encountered above the winter pole because of the low values of thermal emission from the cold polar atmosphere.
Figure 18: Simulating the simultaneous retrieval of water ice and dust mass loading.
Figure 19: Simulated retrieval (dashed lines) of water vapor profiles corresponding to: a) the Viking Lander 2 entry temperature profile (Seiff and Kirk, 1977), and b) a radio occultation temperature profile (Lindal et al., 1979). The vertical distributions of "atmospheric" vapor mixing ratio (solid lines) were constructed using partial pressures needed to give a constant relative humidity.
|