Vikingera diurnal waterice clouds - Wiley Online Library

3 downloads 0 Views 898KB Size Report
[1] Water-ice clouds during the Viking era are mapped as a function of season, ... Citation: Tamppari, L. K., R. W. Zurek, and D. A. Paige, Viking-era diurnal ...
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 108, NO. E7, 5073, doi:10.1029/2002JE001911, 2003

Viking-era diurnal water-ice clouds Leslie K. Tamppari and R. W. Zurek Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California, USA

D. A. Paige Department of Earth and Space Sciences, University of California, Los Angeles, Los Angeles, California, USA Received 9 April 2002; revised 22 November 2002; accepted 27 January 2003; published 15 July 2003.

[1] Water-ice clouds during the Viking era are mapped as a function of season, latitude,

longitude, and time of day. In the springtime hemisphere, where data are available, clouds are observed to decrease from morning to midday. They are observed to increase in extent from midday to afternoon, possibly due to increased atmospheric temperatures causing uplifting of dust which then acts as cloud condensation nuclei. During the summer hemisphere clouds are observed to decrease in extent throughout the day, indicating sufficient warming to prevent water reaching the condensation level. The Hadley cell upwelling branch is seen throughout the year via cloud belts in both the northern (aphelion cloud belt) and southern hemispheres in their respective springs and summers. Clouds extending equatorward of 60N in late northern summer and early northern autumn are seen primarily over the Acidalia Planitia longitudinal sector and may indicate the southernmost extent of the north polar hood or clouds associated with storms. Water-ice clouds associated with topographic features are seen, including Olympus Mons, the Tharsis volcanoes, and the northern rim of Hellas basin. There is little evidence for interannual variability in the prominent cloud features (e.g., the aphelion cloud belt) since they are present in both Martian years examined, though the detailed structure does INDEX TERMS: 5464 Planetology: Solid Surface Planets: Remote sensing; 5494 Planetology: change. Solid Surface Planets: Instruments and techniques; 6225 Planetology: Solar System Objects: Mars; 5445 Planetology: Solid Surface Planets: Meteorology (3346); 5409 Planetology: Solid Surface Planets: Atmospheres—structure and dynamics; KEYWORDS: Hellas, Tharsis, equatorial, polar, morning, afternoon Citation: Tamppari, L. K., R. W. Zurek, and D. A. Paige, Viking-era diurnal water-ice clouds, J. Geophys. Res., 108(E7), 5073, doi:10.1029/2002JE001911, 2003.

1. Introduction [2] Water-ice clouds are known to form in the Martian atmosphere. Historically, ‘‘yellow,’’ ‘‘white,’’ and ‘‘bluish’’ clouds were recorded, the latter two subsequently being inferred to be water-ice clouds [e.g., Capen and Martin, 1971]. Slipher [1962] discussed telescopic observations of clouds on Mars starting in 1907. Many of these clouds were identified as condensate clouds over the volcanoes or other uplift regions, shown by Mariner 9 to be topographic highs [Hartmann, 1978]. However, the first positive identification of water-ice in the Martian atmosphere came from the Mariner 9 Infrared Interferometer Spectrometer [Curran et al., 1973]. [3] With the dawn of the spacecraft age, observations of Mars have dramatically increased. We now have in hand observations of water-ice clouds from the Mariner 9 [Briggs and Leovy, 1974; Curran et al., 1973], Viking [Briggs et al., 1979; Christensen and Zurek, 1984], Phobos [Petrova, 1993], Mars Pathfinder [Smith and Lemmon, 1999], and Mars Global Surveyor [Clancy et al., 1999; Pearl et al., Copyright 2003 by the American Geophysical Union. 0148-0227/03/2002JE001911

2001] spacecraft, from the Hubble Space Telescope [James et al., 1994, 1996], as well as from continuing ground-based observations [Parker et al., 1999]. [4] The analyses of these data have recently allowed the community to recognize that water-ice clouds are an important part of the Martian climate system. In particular, water-ice clouds form in a low-latitude belt during the aphelion season [James et al., 1996; Pearl et al., 2001], which is currently northern spring and summer. This prominent feature appears to be consistent over the last 25 years of observing [Tamppari et al., 2000] and this season is now often referred to as the cloudy season of Mars, analogous to the southern summer dusty season. Initially, it was thought that this cloud belt was strictly a more recent phenomenon due to potential climatic change on Mars causing cooler temperatures since the Viking era (mid-1970s) [Clancy et al., 1996]. However, Tamppari et al. [2000] showed that water-ice clouds were indeed present during the Viking era, and in fact, were omnipresent on Mars, occurring in at least some locations in every seasonal bin examined. In that paper, however, only a subset of the Viking Infrared Thermal Mapper (IRTM) data were studied: data taken during midday (hours 10 –14 local time (LT)). This time frame was chosen for several reasons. One, it provided a

9-1

9-2

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS

manageable data set for initial examination. Two, it was the time of day in which the surface to atmosphere temperature contrast was greatest, allowing easier identification of the water-ice clouds. In this paper, two additional times of day (morning: 9 –10 LT and afternoon: 14– 17 LT), have been examined and are presented. The goals for examining different times of day were to map the distribution of water-ice clouds during the Viking era as completely as possible, to provide a comprehensive data set for use by those examining Mars Global Circulation Model (MGCM) output, and to provide a comprehensive data set that can be used for interannual comparisons. While data were not available for all locations at all times, the maps presented show the water-ice clouds for 360 longitude between ±60 latitude. The resulting set of maps, presented below, span 1.25 Martian years and three local time bins. [5] One reason to map water-ice clouds over nearly the entire daylight time period is to aid in understanding the Martian atmosphere and its interaction with the surface. Clouds cap the water vapor column [Davies, 1979; Jakosky, 1985] and may also cap the dust column [Clancy et al., 1996]. This capping could then lead to an increased surfaceatmosphere exchange [Kahn, 1990], which could play a role in the transport of dust [Clancy et al., 1996]. Additionally, this mechanism for allowing the regolith to become a water reservoir, via adsorbed water or surface ice, may explain the increase in northern springtime atmospheric water vapor prior to the exposure of the residual water-ice cap [Jakosky and Haberle, 1992]. Changes in the diurnal signature of water-ice clouds could be caused by ground fogs which form when the near-surface atmosphere is colder and dissipate as the atmosphere warms due to surface heating. [6] One method for understanding the Martian atmosphere and surface-atmosphere interactions is through the use of Mars Global Circulation Models (MGCMs). While these models have evolved from validated Earth GCMs, and much of the physics employed in these models is well understood, there is still much validation of the MGCMs to be done. In particular, in this era of Martian research, MGCMs are just beginning to incorporate cloud microphysics. At this time it is important to support these efforts with data products to which the MGCM output can be compared. There are a number of research topics being pursued currently, for which this data set may lend insight, especially when combined with MGCM modeling efforts. For example, these data may teach us about water transport, since water-ice clouds are our best tracers for upward air motion on Mars. In particular, this paper shows evidence for topographic clouds, polar hood clouds, and widespread hazes associated with the Hadley cell circulation. These cloud and haze features were seen in the work of Tamppari et al. [2000] within the midday water-ice cloud maps and are shown here in the morning and afternoon times as well. The afternoon maps show the north to south transition of the Hadley cell, occurring near Ls = 140, which was not seen examining only the midday maps. Finally, these water-ice clouds may help to characterize the weather on Mars through examination of possible storm systems. Clouds are shown to exist in locations that models have shown tend to generate storm systems. [7] Another reason to examine the cloudiness on Mars as extensively as possible, during the Viking era, is for

comparison to other time periods. Initial debates on whether or not the climate of Mars has changed over the past 25 years [Clancy et al., 1996], involved HST images of a northern springtime low-latitude cloud belt, which had not been seen through Viking data analysis prior to Tamppari et al. [2000]. Clancy et al. [1996, 2000] also compared Martian atmospheric measurements derived using microwave data to to Viking IRTM 15-mm channel derived atmospheric temperatures and to TES derived temperatures. Initially, the microwave measurements were found to be 15 K colder than the Viking measurements [Clancy et al., 1996]. However, if correction for a potential surface contribution is taken into consideration [Wilson and Richardson, 2000], the Viking-derived temperatures match the microwave temperatures to within 5 K [Clancy et al., 2000]. Possible increased cloudiness plus decreased temperatures indicated that the climate on Mars may have shifted to a colder and more water-ice cloudy climate within the last 20 years or so. This idea intrigued the Martian research community and generated significant interest in the intercomparison of data sets from these different time periods. As shown by Tamppari et al. [2000], it is shown here that there were in fact clouds during the Viking era. Interest in understanding the Martian climate remains, and as the first long term orbiting mission, Viking provided the means for establishing a climate baseline. The water-ice cloud maps presented here allow understanding an important component of that climate. The data set will be available for comparison to ongoing and future missions as well, such as the Mars Global Surveyor (MGS), Mars Odyssey, and the Mars Reconnaissance Orbiter. As of this writing, however, the Viking data set provides the longest baseline diurnal coverage of any Martian data. This is a consequence of the fact that MGS is in a mapping orbit which provides data at only 2 LT and 14 LT (i.e., 2 AM/2 PM). [8] Section 2 reviews the Viking IRTM data set used and the technical approach to retrieving the water-ice clouds. Section 3 presents and discusses the water-ice clouds for the three times of day examined. Specifically, the additional times of day confirm the Hadley cell circulation and add the northern- to southern-cell transition, they confirm polar hood and storm zone detections of Tamppari et al. [2000] and add new detections between Ls = 335– 20, they indicate seasonal trends such as the spring season increase in water-ice clouds as a function of time of day, the summer season decrease as a function of time of day, and an equatorward shift in the upwelling branch of the Hadley cell in the afternoon, and they indicate thicker or colder clouds over the volcanoes in the morning. Finally, section 4 discusses the conclusions.

2. Approach 2.1. Viking Infrared Thermal Mapper Data [9] A complete discussion of the Viking Infrared Thermal Mapper (IRTM) data set used for this study has been given previously by Tamppari et al. [2000]. A summary of the data coverage and additional data used for this paper are given below. [10] Between Viking Orbiter 1 (VO1) and Viking Orbiter 2 (VO2), a little over two Martian years-worth of data were obtained, spanning the time period from June 1976 to July

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS

9-3

Figure 1. Shown in color are the water-ice clouds present during Ls = 80– 170 Year 1, late northern spring to late northern summer. The season progresses down the page in 15 Ls increments (i.e., the top set of panels is for Ls = 80– 95 and the bottom panel is for Ls = 155– 170. The left panels show the morning time frame (9 – 10 LT also called H9 – 10), the middle panels show the midday time frame (10 – 14 LT), and the bottom panels show the afternoon time frame (14 –17 LT). Ls = 80– 95 is the season in which the residual (water-ice) north polar cap becomes exposed allowing more water vapor into the atmosphere [Jakosky and Haberle, 1992]. Near Ls = 150 is the season during which the northern hemisphere low-latitude water vapor was shown to be at a maximum [Jakosky and Haberle, 1992].

1980. This covered portions of three Martian calendar years defined by the Ls range 84 in the first Martian year to 142 in the third [Snyder, 1977; Snyder and Moroz, 1992]. Ls is the areocentric longitude of the sun, a seasonal parameter for which Ls = 0 is the start of northern spring and which progresses through 360 over one Martian year. The data presented here concentrates on the period from Ls = 84 in the first Martian year to Ls = 170 in the second Martian year, due to the more extensive coverage during this period. [11] In the work of Tamppari et al. [2000], the maps used only data taken during the midday: 10– 14 LT. The Martian day (or sol) is divided, arbitrarily, into 24 equal parts. In this paper, two additional times of day have been examined. The morning time period which extends only from 9 – 10 LT and the afternoon time period which extends from 14– 17 LT. These times of day were chosen to be during the daylight hours and to be far enough from the terminator to avoid temperature inversions. In other words, the times chosen

were such that the surface temperature was likely to be several degrees warmer than water-ice clouds forming above the atmospheric boundary layer (>5 km). The technique used here for identifying water-ice clouds is the same as that used by Tamppari et al. [2000] (see Figure 2) and depends on the cloud-surface temperature contrast. A brief summary is given below. [12] The cloud detection technique uses the difference between the IRTM 11- and 20-mm channels. Since the 11-mm channel is within the water-ice absorption band [e.g., Curran et al., 1973], a cooler cloud over a warmer surface causes the 11-mm channel to have a colder brightness temperature (T11) than the 20-mm channel (T20). However, the surface has wavelength-dependent, non-unit surface emissivities [Christensen, 1998]. To account for that we model the wavelength-dependent surface temperature via a surface thermal model [Paige et al., 1994] with the surface emissivities applied [Christensen, 1998]. Therefore

9-4

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS

Figure 2. Similar to Figure 1, but for the seasonal range Ls = 170 –260. the T11 – T20 difference caused by the surface alone can be compared to the actual T11 – T20 difference observed. If the absolute value of the observed difference is greater than the modeled difference, a water-ice cloud is detected.

3. Map Presentation and Discussion [13] The sections below discuss, in detail, the water-ice clouds seen in the maps presented. First, the maps provide evidence for the Hadley cell circulation, for both the southern-dominated and northern-dominated cells, as well as the transitions from the one to the other. The afternoon maps provide the evidence for the northern- to southerndominated cell transition. Second, polar hood clouds and clouds occurring in zones of possible storm development are discussed. Third, time of day trends are presented and fourth, topographically forced clouds and their time of day trends are discussed. Much of the time, water-ice clouds form in the same spatial regions at the different times of day. During the northern summer, the season for which there are two Martian years of data to examine, the behavior of the clouds is similar in spatial location and extent. [14] The water-ice cloud maps are presented in Figures 1 through 5. Each figure has three columns. The left column is the morning (9 – 10 LT), the middle column is the midday (10 – 14 LT), and the right column is the afternoon (14 – 17 LT). The rows are ordered, top to bottom, according to seasonal start date. Each map spans 15 of Ls, which corresponds to about one-half of a Martian month. The maps are presented in west longitude (0 at the right-hand

side of the map) and latitude. The latitude axis extends from pole to pole, but the data examined only span from 60 to +60. The water-ice cloud mapping was restricted to ±60 latitude, as done by Tamppari et al. [2000] and consistent with the original mapping of IRTM observed surface albedo and derived thermal inertia used in the Paige et al. [1994] thermal model. On each map, black represents locations where either no data existed for examination or data were not included in the examination (from 60 poleward). White represents locations where data existed, but no water-ice clouds were present, and blue represents locations containing water-ice clouds, with darker blues representing thicker or colder clouds. 3.1. Hadley Cell Circulation [15] Models indicate that the rising branch of the Hadley cell circulation should be in the spring/summer hemisphere and the descending branch in the fall/winter hemisphere. Near equinox there may be two overturning cells, with rising air near the equator and sinking air a few tens of degrees poleward in both hemispheres or there may also be a rapid transition from one hemisphere’s summer pattern to the opposite hemisphere’s spring/summer pattern. This section will show water-ice cloud evidence for the Hadley cell. In particular, the transition between the northern- to southern-hemisphere dominated cell, not previously seen in the work of Tamppari et al. [2000], is seen with the addition of the afternoon maps. [16] The Hadley cell is best identified through the transition between a late southern summer zone (after Ls = 335;

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS

9-5

Figure 3. Similar to Figure 1, but for the seasonal range Ls = 260 – 350. This seasonal period included the start of the second planet-encircling dust storm on Mars (1977b), which extended from Ls = 275 – 320. As such, the water-ice cloud detections, shown in color, are only possible water-ice clouds. The water-ice cloud detections in the northern hemisphere are more likely real, but further work would need to be done to confirm this. Figure 3) in southern mid-latitudes (centered near 45S) through a period with less frequent clouds occurring near the equator in early-mid northern spring (Ls = 20 –50; Figure 4) to a cloudy zone in the northern tropics. This suggests the annual reversal of the Hadley circulation occurred in early northern spring. The northern zone is quite prominent after mid-spring and persists through solstice until northern mid-summer (Figures 4 –5). The pattern of equatorial water-ice clouds during Year 2 in the late northern spring and northern summer is consistent with the clouds observed during Year 1 (see Figures 1 and 5), even though the coverage is not as extensive. The latitudes of the rising branch are qualitatively consistent with modeling [e.g., Houben et al., 1997], though the extent and location of the rising branch depends on the amount of dust suspended in the atmosphere as well [Haberle and Leovy, 1982]. With the addition of the morning and afternoon maps, particularly the latter, the northern cloud belt was seen to dissipate by L s = 140 Year 1 (Figure 1). The coverage is limited for Ls = 155 Year 1, however, no equatorial clouds are seen. For this season in Year 2 (Figure 5) the diminishing of the northern cloud belt was

also suggested. This is consistent with [Pearl et al., 2001], who observed the cloud belt vanish between Ls = 140 150. Starting at Ls = 170 Year 1 and continuing through Ls = 275 Year 1, Figures 1 – 3 show water-ice clouds consistently forming in the southern hemisphere zone (45S) in the afternoon maps. Many of these clouds are detached from the southernmost extent of the data, while some extend fully to 60S. This transition from a northern low-latitude cloud zone in mid-northern summer and the subsequent formation of a southern cloud zone by late northern summer indicates the overturning of the Hadley cell from a northern-dominated cell to a southern-dominated cell. This is the first cloud observation of the northern to southern transition of the Hadley cell, although it has been previously predicted by models [e.g., Houben et al., 1997]. This northto-south transition (Ls = 140 –170) occurs slightly earlier than predicted as well [Haberle and Jakosky, 1990]. [17] In summary, the Hadley cell circulation is seen throughout the entire Mars year. The northern spring/summer show a clear water-ice cloud band in the low-latitudes (between 10S and 30N) and the southern spring (before the planet-encircling dust storms) show a water-ice cloud

9-6

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS

Figure 4. Similar to Figure 1, but for the seasonal range Ls = 350 – 80. band in the southern mid-latitudes (30S to 45S). In particular, the addition of the afternoon maps presented here allowed detection of the northern- to southernhemisphere transition. The transitions between the northernand southern-dominated cells both occur closer to northern summer solstice than models would predict [Haberle and Jakosky, 1990]. This may occur for one of several reasons. One, it may be that since perihelion occurs during southern summer, the southern summer Hadley cell re-asserts itself earlier and persists longer due to the greater insolation at that time of year. Two, it may also be that since the southern summer is dustier, particularly during the Viking period, that the Hadley cell may begin earlier and be maintained longer during this season. In any case, it seems not to be a reflection of the seasonal water vapor content [Jakosky and Farmer, 1982], which has its peak in the northern hemisphere during northern summer, since a short transition season is seen followed immediately by the onset of a water-ice cloud band in the opposite hemisphere. 3.2. Polar Hood Clouds: Storm Zones [18] Beginning at Ls = 155 Year 1 (Figure 1), the first indications are present of possible polar hood water-ice clouds [Leovy et al., 1972; Briggs and Leovy, 1974; James et al., 1994; Pearl et al., 2001] or clouds associated with zones in which storms preferentially develop [Hollings-

worth et al., 1996, 1997]. These phenomena have been discussed fully by Tamppari et al. [2000]. The coverage in the morning was not very good during late northern summer and early northern autumn (Ls = 155 – 200), during which the northern polar hood clouds were seen in the work of Tamppari et al. [2000]. In contrast, the afternoon coverage is often comparable and sometimes better than the midday coverage. This section shows that the addition of the morning and afternoon data confirm the polar hood clouds and storm zone clouds seen in the work of Tamppari et al. [2000] and extend the detection of these features in the Ls = 335 – 20 seasonal bins. [19] The water-ice clouds that extend southward of 60N at Ls = 155 Year 1 (Figure 1) are seen not only in the midday map, but also in the afternoon map. The longitude extent of the coverage is very limited in this season, so the full extent of the clouds is not known. However, this map can be compared to Ls = 155 Year 2 (Figure 5), in which the longitudinal coverage is nearly global. This latter map contains clouds for similar longitudinal regions (hereafter ‘‘longitudes’’ for simplicity) which suggests that the cloudy region is preferentially between about 120 – 200 W. These longitudes correspond roughly to one of three preferred storm regions as modeled by Hollingsworth et al. [1996]. [20] In the Ls = 170 Year 1 and Ls = 185 Year 1 bins (Figures 1 and 2), the morning coverage is very sparse and

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS

9-7

Figure 5. Similar to Figure 1, but for the seasonal range Ls = 80 – 185. This is the season during which the residual (water-ice) north polar cap becomes exposed (Ls = 80), allowing water vapor subliming off the cap into the atmosphere, reaching a maximum near 110 [Jakosky and Haberle, 1992]. the afternoon coverage is better than the midday coverage. In these two seasonal bins, it appears that clouds are extending southward from 60N at nearly all longitudes. The afternoon maps confirm the hints of cloud extent given in the midday maps. The best example is the afternoon map at Ls = 170 (Figure 1). [21] As noted by Tamppari et al. [2000], the most convincing storm zone clouds are those near Acidalia Planitia (30– 60N/0– 60W), especially in the Ls = 245– 275 bins (Figures 2 and 3). Acidalia Planitia is a region in the northern hemisphere modeled by Hollingsworth et al. [1996, 1997] to be a zone where storms preferentially occur. It is also characterized by low albedo, high thermal inertia, and low elevation. During this southern spring season, the coverage in the northern hemisphere during the afternoon was very sparse. However, the morning maps confirm waterice clouds extending southward of 60N during Ls = 245 (Figure 2) and detached from the polar hood clouds in the Ls = 260– 275 bins (see Figure 3). [22] Maps with starting seasonal parameters of Ls = 275 –320 (Figure 3) are during the 1977(b) planet-encircling dust storm [Martin and Richardson, 1993; Zurek and Martin, 1993; Martin and Zurek, 1993]. As noted by Tamppari et al. [2000], for the midday maps, the cloud

detections during these dusty seasons were not considered robust and therefore these seasonal bins were not examined. However, the dust was typically higher opacity in the southern hemisphere than in the northern, so the water-ice cloud signature over Acidalia Planitia (45N, 40W) at Ls = 275 and Ls = 290 in Figure 3 may in fact be real. Without further consideration of this time period to ensure proper identification of water-ice clouds during high dust-loading conditions, it remains uncertain. [23] Probable north polar hood clouds can also be seen, following a gap in northern hemisphere coverage, in the afternoon map of the Ls = 335 bin (Figure 3). The water-ice cloud signature is strongest over the Acidalia Planitia region (45W). The afternoon maps in the following two seasonal bins (Ls = 350 – 20; Figure 4) have better coverage at these high latitudes than the midday maps and show that there are water-ice clouds at some longitudes. In general, however, their southward extent at west longitudes >90, outside the Acidalia Planitia region, is not as great as during the northern late summer (e.g., Ls = 170; Figure 1). By Ls = 20, Figure 4, the only southward extent of these possible north polar hood clouds is in the Acidalia Planitia region. [24] The north polar hood was identified in MGS TES 2 p.m. data by Pearl et al. [2001]. They noted a dramatic

9-8

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS

Figure 6. The zonal percent cloud coverage (solid line) is shown for each latitude and seasonal bin for 9 – 10 LT. The zonal percent IRTM coverage (dotted line) is shown as well, since full coverage was rarely obtained. Ls = 200 – 245 and Ls = 275– 335 have been annotated with 1977a and 1977b indicating the first and second major dust storms, respectively, during the Viking mission. During the 4 seasonal bins of the 1977b dust storm, the dust opacity was high enough that they have been over-corrected in our surface emissivity removal process and therefore the clouds shown should not be trusted. increase in zonally averaged water-ice cloud opacities between Ls = 150– 175, after which time a slight decrease followed. Between 40 – 60N, they observed a smaller increase in opacities. Since their data are zonally averaged, no distinction was made between polar hood clouds and possible storm zone clouds. However, the timing in these increases in opacity is consistent with the increase in cloudiness at higher latitudes presented here. [25] In summary, the extended diurnal coverage presented here showed similar seasonal trends to Tamppari et al. [2000] for the polar hood and storm zone cloud detections. However, the addition of the afternoon maps allowed these clouds to be seen at more longitudes and in more seasons as the coverage was better at the high northern latitudes. 3.3. Water-Ice Cloud Changes as a Function of Time of Day [26] Examination of the water-ice clouds over three times of day allowed detections of certain features, such as the

upwelling branch of the southern hemisphere-dominated Hadley cell, discussed above. Figures 1 through 5 show the entire water-ice cloud map set, which reveals time of day trends. The afternoon maps during the northern- to southerndominated Hadley cell transition show an interesting shift toward the equator in cloud location. For use in the discussion of these trends below, the data coverage and percent cloud detection are shown for each time of day in Figures 6 (9– 10 LT), 7 (10 – 14 LT), and 8 (14 – 17 LT). Note that the clouds, particularly in the southern hemisphere, derived during the 1977b dust storm between Ls = 275 – 335, may be artifacts of the cloud detection scheme [see Tamppari et al., 2000]. Figures 9 and 10 compare the cloud coverage, as a percentage of available data, between the three times of day. However, the coverage varied between times of day even for the same season, which causes some artifacts. These data are also summarized in Tables 1 through 5. [27] During the southern springtime (Ls = 180– 270), water-ice clouds in the southern hemisphere increased in

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS

9-9

Figure 7. Same as Figure 6 except for 10– 14 LT. extent as the day progressed from midday to afternoon (see Figures 2, 9, and 10 and Table 2). During this season the coverage of the morning time frame and of the northern hemisphere was not extensive. [28] A similar pattern was present throughout the entire northern springtime (see Figures 9 and 10 and Table 4); cloud extent increased from the midday to afternoon in the equatorial region and the northern latitudes. Although coverage is limited, especially in the southern hemisphere and equatorial region in mid-late northern spring, this trend is still apparent in the low-latitude cloud belt. Water-ice clouds showed a stronger signature, particularly over volcanoes, in the afternoon than in the midday, as well as the increased extent (see Ls = 5– 65 Year 2; Figures 9 and 10 and Table 4). [29] The enhancement of clouds in the afternoon is unexpected since the atmospheric temperatures are warming. Emission angles as a function of time of day for two seasonal bins were examined (Ls = 35 and Ls = 65) and found to be similar for all times of day. Thus this enhancement of clouds was not due to higher emission angle data during the afternoon. One possible explanation is that as the atmosphere heats, convection is enhanced, allowing water vapor to be uplifted to altitudes at which it will form ice clouds.

[30] Contrary to the springtime case, in the two southern summer seasonal bins in which water-ice clouds could be confidently identified (i.e., outside of the 1977(b) planetencircling dust storm), cloud extent decreased as the day progressed from morning through afternoon (see Figures 9 and 10). In the Ls = 335 –350 bins (Figure 3), this trend is very apparent. See also Table 3. [31] A similar pattern was seen in the northern latitudes during early northern summer; clouds appeared to diminish in extent from morning to midday (see Tables 1 and 5 and Figures 9 and 10). There was only adequate coverage in Year 2 to identify this trend. This enhancement in morning cloudiness can be seen in seasonal bins Ls = 80– 110 (Figure 5). This is the season in which the water vapor starts rapidly increasing due to the exposure of the residual north polar water ice cap [Jakosky and Haberle, 1992]. In seasonal bins Ls = 125 –155 (Figure 5), the water-ice cloud extent was not as great as the previous three seasonal bins, but the clouds were still more widespread in the morning than during midday. Unfortunately, during this season, there was minimal coverage of the northern hemisphere during the afternoon. Atmospheric temperatures increased during the Viking years as the northern summer season progressed (Figure 7) [Zurek et al., 1992]. Thus, in

9 - 10

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS

Figure 8. Same as Figure 6 except for 14– 17 LT. Table 1. Time of Day Changes for Northern Summer Year 1a N. Latitudes Ls 80 – 95 95 – 110 110 – 125 125 – 140 140 – 155 155 – 170 170 – 185

AM MID X X X X X X X

X X X X X C C

PM X X C Clear C C C

Equatorial AM MID C C C C C X X

PM

S. Latitudes AM

MID

PM

C X X X X C X Clear Clear X C C C C+ X C C C C C C C C C C Clear Clear X Clear Clear C C X C C+

a Ls = 80 – 185. The table shows a qualitative picture of how clouds change as a function of time of day for different regions on Mars. Northern latitudes were those north of 30N, the equatorial region was between ±30, and the southern latitudes were those south of 30S. AM refers to the morning time frame (9 – 10 LT), MID refers to the midday time frame (10 – 14 LT), and PM refers to the afternoon time frame (14 – 17 LT). ‘‘X’’ denotes times and seasons in which there was either not enough data to make an assessment of the cloud coverage or there weren’t any data at all. ‘‘Clear’’ denotes times and seasons which contained very few or no waterice clouds: a very clear period. ‘‘C’’ denotes water-ice clouds present. ‘‘C+’’ and ‘‘C ’’ denote an increase or a decrease in the apparent cloudiness relative to the first local time period with clouds present within the same season and latitude region. For example, in the Ls = 155 – 175 season and the northern latitude region, the apparent cloudiness decreases in the afternoon compared to the midday, and there is no data present in the morning. ‘‘C 2’’ denotes an even greater decrease in clouds with respect to the previous bin.

contrast to the spring season, if the temperatures heat enough, perhaps even increased convection is not sufficient to lift water vapor to the condensation level. This trend is analogous to what was seen in the southern hemisphere in the southern summer, although these atmospheric temperatures appear to decrease as the season progresses (see Ls = 335 and Ls = 350 in Figures 3, 4, and 7 [Zurek et al., 1992]). [32] In the early northern springtime, (Ls = 5 – 35) water-ice clouds appeared to form in the southern hemisphere in a southern mid-latitude belt during the morning and midday (see Figures 9 and 10 and Table 4). By the afternoon time, clouds appeared to be forming closer to the

Table 2. Time of Day Changes for Northern Fall Year 1a N. Latitudes Ls

MID

PM

Equatorial

S. Latitudes

AM

MID

PM

185 – 200 X C C X 200 – 215 X C X X 215 – 230 X Clear Clear X 230 – 245 X C X X 245 – 260 C C X C 260 – 275 Clear C X Clear

Clear Clear Clear Clear Clear Clear

C Clear C Clear Clear Clear

a

AM

Ls = 185 – 275. The table is similar to Table 1.

AM MID PM C X C X C C

C Clear C C C C

C+ C C+ C+ C+ C+

9 - 11

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS Table 3. Time of Day Changes for Northern Winter Year 1a N. Latitudes

Equatorial

Ls

AM

MID

PM

AM

335 – 350 350 – 5

X X

X C

C C

C C

MID C C

Table 5. Time of Day Changes for Northern Summer Year 2a

S. Latitudes PM

AM

Clear C

C C

AM

MID

PM

AM

MID

PM

AM

MID

PM

C C

C 2 C 2

95 – 110 110 – 125 125 – 140 140 – 155 155 – 170 170 – 185

C C C C C C

C C C C C C

C X X X X X

C C C C C C

C C C+ C C ? C

C C X X X X

X X X X X X

X X X X X X

X X X X X X

3.4. Topographic Water-Ice Cloud Features [33] Topographic water-ice clouds are seen, particularly those forming over Olympus Mons, the Tharsis volcanoes, and Elysium Mons. Olympus Mons is the largest volcano in the solar system, peaking at about 25 km elevation. It and the three Tharsis volcanoes, Pavonis, Arsia, and Ascraeus, are all within 40 longitude and 30 latitude. Elysium Mons is about 90 west of Olympus Mons and peaks near 16 km elevation. The topographic clouds were discussed in detail by Tamppari et al. [2000] and here only time of day effects are noted. The topographically formed clouds over Olympus Mons and the Tharsis volcanoes are best seen in Figure 5 (Ls = 80 – 110), during the second Mars year of data taking, as darker blue. The locations are Olympus Mons (17N/133W), Ascraeus Mons (11N/104W), Pavonis Mons (0N/113W), and Arsia Mons (9S/ 121W). Coverage of Olympus Mons and the Tharsis volcanoes is fairly good in both the morning and midday times (Ls = 95 and Ls = 110 in Figure 5) and it is clear that the clouds have a stronger signature in the morning than in Table 4. Time of Day Changes for Northern Spring Year 2a Ls

AM

MID

PM

AM

5 – 20 X C C Clear 20 – 35 Clear Clear C C 35 – 50 C C C C 50 – 65 C C C C 65 – 80 Clear Clear X C 80 – 95 C Clear Clear C a

S. Latitudes

MID PM

AM

C C C C C C

C C C C C X Clear Clear X Clear Clear X X X X X X X

Ls = 5 – 95. The table is similar to Table 1.

C+ C+ C+ C+ C+ C

S. Latitudes

Ls

equator (e.g., Ls = 5 Year 2, Figure 4). A similar enhancement of clouds near the equator is evident in the Ls = 20 Year 2 bin, although the lack of coverage in the southern mid-latitudes in this season makes it impossible to tell whether or not clouds were also forming there. Christensen [1998] suggests that clouds intensify from midday to afternoon near 10 S/45 W. This was also seen here between Ls = 5 – 35 (Figure 4). The seasonal timing of this equatorward shift of clouds as the day progresses occurs just before and overlaps the beginning of the transition from a southern- to a northern-dominated Hadley cell (Ls = 20 – 50). During the transition period, models predict that the Hadley circulation will change from a single cell structure to a double cell structure, with the upwelling branches closer to the equator [Haberle and Jakosky, 1990]. It is possible that the equatorial clouds in the afternoon time period are evidence for this double cell structure, or in other words, represent the Martian Intertropical Convergence Zone (ITCZ). The ITCZ on Earth is characterized by a narrow band of clouds near the equator.

Equatorial

Equatorial

PM

a Ls = 335 – 5. The time period of the second planet-encircling dust storm, Ls = 275 – 335, has not been included. The table is similar to Table 1.

N. Latitudes

N. Latitudes

MID

MID

PM

a Ls = 95 – 185. The table is similar to Table 1, except here, a ‘‘?’’ denotes uncertainty due to lack of coverage.

the midday. This is in contrast to Slipher [1962] who stated that clouds occurring over the Tharsis volcanoes (the ‘‘W’’ cloud) were strictly an afternoon feature. In our technique, this stronger signature is associated with higher opacity or higher altitude (colder) clouds. Pearl et al. [2001] see the highest opacities over volcanic features indicating that our stronger signature may also imply higher opacity. For example, tracks over Arsia Mons [Pearl et al., 2001] achieve an opacity of t12mm = 0.6 during early northern fall.

4. Conclusions [ 34] The conclusions of this paper are as follows: (1) water-ice clouds are able to be mapped with confidence during the Viking era and are found to be present to some extent in most seasons and times of day examined, (2) the first cloud evidence for the Hadley cell circulation throughout the entire Martian year is seen, particularly the northernto southern-hemisphere transition as seen in the afternoon maps, (3) the southernmost extent of the polar hood clouds are seen as well as clouds over a location modeled by Hollingsworth et al. [1996, 1997] to have frequent storms, (4) the lack of evidence for substantial interannual variability between the two Viking years of data examined here, (5) the evidence for topographically forced clouds, and (6) that there are cloudiness changes as a function of timeof-day and that there are seasonal time-of-day trends exhibited. The most significant conclusions among these are (1), (2) and (6). These are all discussed in greater detail below. [35] The techniques developed by Tamppari et al. [2000] and employed here allowed water-ice clouds to be identified in the Viking IRTM data set with greater confidence than in previous work. This paper expands upon the work done by Tamppari et al. [2000] by additionally providing the morning and afternoon water-ice cloud maps. Thus the maps presented here provide not only longitudinal, latitudinal and seasonal variation of the occurrence of water-ice clouds for the Viking period, but also time of day occurrence. The Mars Global Surveyor (MGS) spacecraft mapping data (Mars arrival: September 11, 1997 and ongoing at this writing) have much more uniform and complete spatial coverage, but are limited to just two times of day. While even better space-time coverage may be provided by future missions, the clouds identified in the Viking data set provide a global reference of water-ice clouds during the Viking era for use in the study of the Mars water cycle and its interannual variation. In particular, the cloud maps derived here indicate that there were extensive water-ice clouds throughout much of the day during the Viking period,

9 - 12

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS

Figure 9. Percent cloud coverage (of available data) as a function of latitude and season for northern spring (both Mars years) and northern summer year 1. Morning (9 – 10 LT) is shown in green, midday (10 – 14 LT) is shown in blue, and afternoon (14 – 17 LT) is shown in red. Note that the coverage varied between times of day even for the same season. This causes some artifacts; for example, the afternoon coverage was very poor in the southern hemisphere in the Ls = 50 Year 2 bin, yet there were clouds in much of the data. Refer to Figures 6, 7, and 8 to see the coverage. comparable to that suggested by more recent HST monitoring of Mars, even given the presence of a significant amount of atmospheric dust during the period of Viking observations. [36] The cloud maps derived by Tamppari et al. [2000] and here show evidence for a cloudy zone due to the rising branch of the Mars Hadley circulation. Specifically, the southern-northern transition is seen via the development of a cloud band centered near 45S in late northern winter (Ls = 335), followed by the disappearance of low latitude clouds and cloud bands in early spring (Ls = 20), and the emergence of a cloudy zone in northern low latitudes during northern mid-spring. This cloud band is present during the morning and afternoon maps, as well as the previously derived midday maps, although the cloud extent diminishes as the day progresses. However, a significant result of the derived morning and afternoon cloud maps presented here, is that a southern cloud belt was also seen during the southern springtime (Ls = 185), prior to the onset of the second planet-encircling dust storm (1977(b)). During this season, the afternoon time period typically had better global

coverage. In addition, the cloudiness increased later in the daytime, facilitating the detection of this belt. Prior to and with the onset of this belt (Ls  170), the water-ice clouds in the equatorial region diminish between Ls = 140– 170, indicating a possible transition from a northern- to a southern-hemisphere dominated Hadley cell. [37] Polar hood and storm system clouds are not well resolved with the IRTM temporal coverage and our restriction to non-polar latitudes. As stated by Tamppari et al. [2000], there was some evidence for extension of the north polar hood into northern middle latitudes and for the presence of northern storm zones, the latter occurring most frequently and appearing to penetrate farthest toward the equator in the Acidalia Planitia longitudinal sector. Similar excursions of the south polar hood were not observed probably due to the very dusty spring and summer observed by Viking and due to poor southern high-latitude coverage in southern autumn. [38] As in the work of Tamppari et al. [2000], there was little evidence for substantial interannual variability between the two successive northern summers observed by the Viking

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS

9 - 13

Figure 10. Percent cloud coverage (of available data) as a function of latitude and season for northern summer year 2, and northern autumn and winter year 1. Morning (9 – 10 LT) is shown in green, midday (10 – 14 LT) is shown in blue, and afternoon (14 – 17 LT) is shown in red. Note that the coverage varied between times of day even for the same season. This causes some artifacts; for example, the afternoon coverage was very poor in the Ls = 155 – 170 bins, yet there were clouds in much of the data. Refer to Figures 6, 7, and 8 to see the coverage. Orbiters. The additional maps of the morning and afternoon were consistent with this finding. In addition, the low-latitude aphelion cloud belt seen in northern spring and summer are similar in extent and seasonal occurrence to cloud observations by HST [James et al., 1996], ground observations in years since the Viking mission ended, and the recent TES observations [Pearl et al., 2001; Smith et al., 2001]. [39] Water-ice clouds over many topographic highs were seen at all times of day, including the largest volcanoes, Olympus Mons, Pavonis Mons, Arsia Mons, Ascraeus Mons, Elysium Mons, and a local topographic high, Alba Patera. Discrete clouds were often identified in the data even when superimposed on a background cloud band as in James et al. [1996]. A strong water ice cloud signature over Hellas basin was seen at Ls = 110 Year 1 in the morning and midday maps (this location was not covered in the afternoon), and at Ls = 125 Year 1 in the morning and midday maps. However, by the afternoon of Ls = 125 Year 1, this cloud had dissipated. [40] The examination of the water-ice clouds as a function of time of day uncovered two interesting seasonal

trends. First, in the springtime hemisphere, (i.e., southern hemisphere in southern spring or northern hemisphere in northern spring) water-ice clouds tended to increase in extent from midday to afternoon. This may imply that dust is being uplifted to higher altitudes as the atmospheric temperatures increase, acting as cloud condensation nuclei and allowing cloud formation to become more efficient. In the northern hemisphere during northern spring, sufficient morning data exist to observe that clouds decreased from morning to midday during this season as well. This implies that the midday is the least cloudy. A similar statement cannot be made for the southern hemisphere in southern spring due to the lack of morning coverage. Second, in the summertime hemisphere, clouds tended to decrease in extent from morning to midday. This may be evidence for clouds re-vaporizing as the atmospheric temperatures warm during the day. In addition, during late-southern summer in the southern hemisphere, the clouds extent continued to decrease into the afternoon. Data were not extensive enough in the afternoon in the northern hemisphere during northern summer to make a similar statement in that season. The

9 - 14

TAMPPARI ET AL.: MARTIAN WATER-ICE CLOUDS

difference in late afternoon behavior between spring and summer in the southern hemisphere may be due to the atmosphere heating sufficiently in summer to prevent water from being lifted to the condensation level, even if convection is increased. [41] The cloud maps derived here provide a valuable basis for comparison with both global circulation models and current and future data sets. In particular, comparison can be made to observations by the Mars Global Surveyor TES [Pearl et al., 2001] and other future instruments. A major advantage of the MGS observations is the daily global coverage achieved, albeit at just two times of day. Other advantages over Viking IRTM come from the ability of TES to simultaneously retrieve temperature and aerosol extinction. Limb views by TES are less sensitive to the surface emissivity effects dealt with here and will directly provide some information on cloud altitudes. With TES data, an analysis similar to the one used here can be performed, allowing a direct comparison between these two data sets spanning 12 Martian years. In particular, the TES spectra can be convolved to the IRTM bandpasses and the cloud detection technique used here can be applied. This work is in progress and will be the subject of a future paper. [42] Acknowledgments. The authors thank Phil Christensen for his invaluable surface emissivity data set, without which thermal IR water-ice clouds could not be properly derived from the Viking IRTM data set. The authors would also like to thank two anonymous reviewers for thorough and constructive comments. The research described in this paper was carried out at the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration.

References Briggs, G. A., and C. B. Leovy, Mariner 9 observations of the Mars north polar hood, Bull. Am. Meteorol. Soc., 55, 278 – 296, 1974. Briggs, G. A., W. A. Baum, and J. Barnes, Viking orbiter imaging observations of dust in the Martian atmosphere, J. Geophys. Res., 84, 2795 – 2820, 1979. Capen, C. F., and L. J. Martin, The developing stages of the Martian yellow storm of 1971, Lowell Obs. Bull., 7, 211 – 216, 1971. Christensen, P. R., Variations in Martian surface composition and cloud occurrence determined from thermal infrared spectroscopy: Analysis of Viking and Mariner 9 data, J. Geophys. Res., 103, 1733 – 1746, 1998. Christensen, P. R., and R. W. Zurek, Martian north polar hazes and surface ice: Results from the Viking survey completion mission, J. Geophys. Res., 89, 4587 – 4596, 1984. Clancy, R. T., A. W. Grossman, M. J. Wolff, P. B. James, D. J. Rudy, Y. N. Billawalla, B. J. Sandor, S. W. Lee, and D. O. Muhleman, Water vapor saturation at low altitudes around Mars aphelion: A key to Mars climate?, Icarus, 122, 36 – 62, 1996. Clancy, R. T., M. Wolff, M. Smith, and J. Pearl, The 1999 aphelion cloud belt: MGS TES measurements of cloud particle properties, Eos Trans. AGU, 80(46), Fall Meet. Suppl., F608, 1999. Clancy, R. T., B. J. Sandor, M. J. Wolff, P. R. Christensen, M. D. Smith, J. Pearl, B. J. Conrath, and R. J. Wilson, An intercomparison of groundbased millimeter, MGS TES, and Viking atmospheric temperature measurements: Seasonal and interannual variability of temperatures and dust loading in the global Mars atmosphere, J. Geophys. Res., 105, 9553 – 9571, 2000. Curran, R. J., B. J. Conrath, R. A. Hanel, V. G. Kunde, and J. C. Pearl, Mars: Mariner 9 spectroscopic evidence for H20 ice clouds, Science, 183, 381 – 383, 1973. Davies, D. W., The vertical distribution of Mars water vapour, J. Geophys. Res., 84, 2875 – 2879, 1979. Haberle, R. M., and B. M. Jakosky, Sublimation and transport of water from the north residual polar cap on Mars, J. Geophys. Res., 95, 1423 – 1437, 1990. Haberle, R. M., and C. B. Leovy, Some effects of global dust storms on the atmospheric circulation of Mars, Icarus, 50, 322 – 367, 1982.

Hartmann, W. K., Mars: Topographic control of clouds, 1907 – 1973, Icarus, 33, 380 – 387, 1978. Hollingsworth, J. L., R. M. Haberle, J. R. Barnes, A. F. C. Bridger, J. B. Pollack, H. Lee, and J. Schaeffer, Orographic control of storm zones on Mars, Nature, 380, 413 – 416, 1996. Hollingsworth, J. L., R. M. Haberle, and J. Schaeffer, Seasonal variations of storm zones on Mars, Adv. Space Res., 19, 1237 – 1240, 1997. Houben, H., R. M. Haberle, R. E. Young, and A. P. Zent, Modeling the Martian seasonal water cycle, J. Geophys. Res., 102, 9069 – 9083, 1997. Jakosky, B. M., The seasonal cycle of water on Mars, Space Sci. Rev., 41, 131 – 200, 1985. 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, in Mars, edited by H. H. Kieffer et al., pp. 969 – 1016, Univ. of Ariz. Press, Tucson, 1992. James, P. B., R. T. Clancy, S. W. Lee, L. Martin, R. Kahn, R. Zurek, R. Singer, and E. Smith, Monitoring Mars with the Hubble Space Telescope: 1990 – 1991 observations, Icarus, 109, 79 – 101, 1994. James, P. B., J. F. Bell III, R. T. Clancy, S. W. Lee, L. J. Martin, and M. J. Wolff, Global imaging of Mars by Hubble Space Telescope during the 1995 opposition, J. Geophys. Res., 101, 18,883 – 18,890, 1996. Kahn, R., Ice haze, snow, and the Mars water cycle, J. Geophys. Res., 95, 14,677 – 14,693, 1990. Leovy, C. B., G. A. Briggs, A. T. Young, B. A. Smith, J. B. Pollack, E. N. Shipley, and R. L. Wildey, The Martian atmosphere: Mariner 9 television experiment progress report, Icarus, 17, 272 – 292, 1972. Martin, L. J., and R. W. Zurek, An analysis of the history of dust activity on Mars, J. Geophys. Res., 98, 3221 – 3246, 1993. Martin, T. Z., and M. I. Richardson, New dust opacity mapping from Viking infrared thermal mapper data, J. Geophys. Res., 98, 10,941 – 10,949, 1993. Paige, D. A., J. E. Bachman, and K. D. Keegan, Thermal and albedo mapping of the polar regions of Mars using Viking thermal mapper observations: 1. North polar region, J. Geophys. Res., 99, 25,959 – 25,991, 1994. Parker, D. C., J. D. Beish, D. M. Troiani, D. P. Joyce, and C. E. Hernandez, Telescopic observations of Mars, 996 – 997: Results of the Marswatch program, Icarus, 138, 3 – 9, 1999. Pearl, J. C., M. D. Smith, B. J. Conrath, J. L. Bandfield, and P. R. Christensen, Observations of Martian ice clouds by the Mars Global Surveyor Thermal Emission Spectrometer: The first Martian year, J. Geophys. Res., 106, 12,325 – 12,338, 2001. Petrova, E. V., Ice clouds over Martian volcanoes: KRFM experimental data (Phobos project), Sol. Syst. Res., 27, 14 – 19, 1993. Slipher, E. C., Mars: The Photographic Story, Northland, Flagstaff, Ariz., 1962. Smith, M. D., J. C. Pearl, B. J. Conrath, and P. R. Christensen, Thermal Emission Spectrometer results: Mars atmopsheric thermal structure and aerosol distribution, J. Geophys. Res., 106, 23,929 – 23,945, 2001. Smith, P. H., and M. Lemmon, Opacity of the Martian atmosphere measured by the Imager for Mars Pathfinder, J. Geophys. Res., 104, 8975 – 8985, 1999. Snyder, C. W., The missions of the Viking orbiters, J. Geophys. Res., 82, 3971 – 3982, 1977. Snyder, C. W., and V. I. Moroz, Spacecraft exploration of Mars, in Mars, edited by H. H. Kieffer et al., pp. 71 – 119, Univ. of Ariz. Press, Tucson, 1992. Tamppari, L. K., R. W. Zurek, and D. A. Paige, Viking era water ice clouds, J. Geophys. Res., 105, 4087 – 4107, 2000. Wilson, R. J., and M. I. Richardson, The Martian atmosphere during the Viking mission, 1: Infrared measurements of atmospheric temperatures revisited, Icarus, 145, 555 – 579, 2000. Zurek, R. W., and L. J. Martin, Interannual variability of planet-encircling dust storms on Mars, J. Geophys. Res., 98, 3247 – 3259, 1993. 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, in Mars, edited by H. H. Kieffer et al., pp. 835 – 933, Univ. of Ariz. Press, Tucson, 1992.

D. A. Paige, Department of Earth and Space Sciences, University of California, Los Angeles, Los Angeles, CA 90024, USA. L. K. Tamppari and R. W. Zurek, Jet Propulsion Laboratory, California Institute of Technology, MS 301-422, 4800 Oak Grove Drive, Pasadena, CA 91109, USA. ([email protected])