Thermal Infrared Properties of the Martian Atmosphere

VOL. 84, NO. B6 JOURNAL OF GEOPHYSICAL RESEARCH JUNE 10, 1979 Thermal Infrared Properties of the Martian Atmosphere 3. Local Dust Clouds ALAN R. PETERFREUND AND HUGH H. KIEFFER ' Department of Earth and Space Sciences, University of California at Los Angeles Los Angeles, California 90024 Prior to and during the Martian global dust storms in 1977, local dust clouds were identified from thermal observations. Local dust clouds appear up to 40 K cooler than the surrounding region in the daytime and appear warm at night. In contrast to their well-developed and consistenthermal signatures, local dust clouds exhibited little albedo contrast. Multiple observations of the same areas show that local dust clouds form and dissipate in a few days or less, and clouds were observed to move with velocities of from 14 to at least 32 m s-. Local dust clouds are common along the receding edge of the south polar cap (observed primarily for orbiter imaging) and in the restricted area - 10 ø to -30 ø, 80 ø to 125øW (observed primarily by the infrared thermal mapper); several of the latter occurred over a ridge in Claritas Fossae, indicating strong topographic control. In their active growing stage, local dust clouds are probably about 8 km high and have relatively abrupt tops. The accumulation of dust from local clouds reached similar levels immediately prior to both global storms, supporting the hypothesis that such dust accumulations are responsible for triggering the global dust storms. INTRODUCTION Dust storms, ranging in size from local (104 km:) to global, have been observed by Viking spacecraft during the course of their missions. Earth-based observers have long noted the presence of 'yellow' dust clouds; these occur in all seasons and in locations spanning the entire planet [Capen and Martin, 1971]. Major storms, usually global in nature, have been documented by Capen [1974], Capen and Martin [1972], and L. J. Martin [1974] and have been found to begin near Mars' perihelion, when solar insolation is maximum in the southern midlatitudes. Mariner 9 arrived during such a global storm in 1971 [Leovy et al., 1972] and was able to observe the waning stages of the storm as well as numerous local dust clouds. (For an extended discussion of historical observations and review of Now at Department of Geology, Arizona State University, Tempe, Arizona 8528 I. 'Now at U.S. Geological Survey, Flagstaff, Arizona 86001. Copyright 1979 by the American Geophysical Union. Paper number 8BI058. 0148-0227 / 79/008 B- 1058 $01.00 An increase in global atmospheric opacity prior to the onset of each of the global storms was observed from both orbiter IRTM observations [T. Z. Martin et al., 1979] and lander imaging [Pollack et al., 1977, 1979]. The occurrence of numerous local dust clouds prior to the global storms thus appears to be an important mechanism for injecting considerable quantities of dust into the atmosphere. In this paper we describe the thermal and visual characteristics of the local dust clouds as detected by the Viking IRTM. Attention is given to documenting the occurrence of the local clouds in time and location on the planet, their thermal and visual properties, and the association of the clouds with local and regional surface features. IRTM DATA DESCRIPTION dust storm dynamical theories, see the work of Briggs et al. [1979].) Thus the Viking observations of storms came as no The data presented in this paper were collected between surprise. What they do provide is high-resolution imagery of June 1976 and October 1977 by the infrared thermal mappers dust activity by the Viking imaging system (VIS) [Briggs et al., aboard the two Viking orbiters. The I RTM is an infrared 1979], in situ measurements [Pollack et al., 1979; Ryan and radiometer which collects data in six spectral bands: five ther- Henry, 1979; Tillman et al., 1979], and visual and thermal data mal channels, centered at 20, 15, 11, 9, and 7 #m, and one from the infrared thermal mapper (IRTM), which detected visual channel which covers from 0.3 to 3.0 #m (see the work and mapped storms with and without the aid of visual imagof T. Z. Martin et al. [1979] for an extended discussion of the ing. Also, the Viking experiments have extended observations IRTM characteristics as they relate to atmospheric studies). to portions of the Martian year previously undocumented by Brightness temperatures derived from the thermal channels are spacecraft instrumentation. used extensively in this paper, the channel reference being Two major dust storms and numerous local dust clouds given as Tx, where x equals the wavelength in microns correwere observed from June 1976 (Ls85) to October 1977 (Ls342), sponding to the center of that channel (for example, a brightthe greatest concentration of activity occurring during the ness temperature derived from the 7-#m channel will be re- Martian southern spring (January 1977 to June 1977, L, 180- ferred to as T7). A comprehensive description of the IRTM is 270). The two global storms were observed near perihelion, given by Chase et al. [1978]. although the first storm (L,207) appeared earlier than expected The Viking mission has produced an enormous quantity of from the historical record; the occurrence of two major dust I RTM data, an average of more than 1 hour of observation storms in a single season, however, is not uncommon (Mars per spacecraft per day. Some selectivity among data sets was Scientific Model [Michaux and Newburn, 1971]). Although therefore necessary to search for local storm signatures. The observations of local dust clouds were made by Viking search has concentrated on the time between the two global throughout the year, the majority of the storms occurred durdust storms. Although considerable earlier IRTM data have been examined without detection of a dust storm, this examiing spring in the southern hemisphere. nation neither was exhaustive nor did it include dust storms 2853 as an objective. There are additional constraints on IRTM coverage of local and regional dust storms placed by Viking orbital parameters and viewing opportunities. A profile of the Viking orbiter

- 2854 PETERFREUND AND KIEFFER: MARS VOLATILES primary mission and IRTM operations have been described by Snyder [1977] and Kieffer et al. [1977], respectively. In the northern hemisphere, orbiter coverage during the period of the two dust storms was such that synoptic observations were not feasible. Synoptic coverage of the southern hemisphere was possible, and thus the detection of local dust clouds is biased toward this hemisphere. The observations are not uniformly spaced in time, but an average of about one sequence per day obtained between the two global storms was examined. This study covers all longitudes reasonably uniformly, with special examination of areas where local dust storms were known to occur. The recognition of a dust cloud from the thermal data was based on two criteria: the presence of an unexpected area of thermal contrast in a region where such a contrast was not observed in previous mapping and a large variation in brightness temperature with wavelength. DUST STORM OBSERVATIONS The majority of local dust clouds and the two major global storms originated in the southern hemisphere between Ls170 and 290. The major global events corresponding to this time period are the rapid sublimation of the seasonal south polar cap, which produces a large mass flux of COo. into the atmosphere, Mars perihelion (Ls250), and peak solar insolation at southern mid-latitudes corresponding to the southern solstice (L,270). Brightness temperatures in excess of 300 K were measured during this period for some areas of low thermal inertia. The local dust clouds observed in the southern hemi- sphere during this time by the orbiter imaging and/or the IRTM are represented in Figure 1 as a function of time and latitude. While a considerable number of local dust clouds have been identified, complete and continuous sampling at all longitudes, latitudes, and times was not possible; thus some occurrences have certainly not been recorded. The local dust clouds are not uniformly distributed in latitude and season and can be conveniently grouped for purposes of discussion. In the southern hemisphere the clouds can be divided into two groups: those which occurred along the receding edge of the south polar cap prior to each global storm and those which occurred in a narrow tropical band in the period following the beginning of the first global storm. Only two local dust clouds have been observed in the northern hemisphere at L,201 and L,340 at +42 ø and +22 ø, respectively, the occurrence of the second one being over the Viking Lander 1 site. The small number of northern storms observed may be due to the lack of synoptic coverage of the northern hemisphere during this period. Numerous local dust clouds were observed by the VIS along the receding edge-of the south polar cap throughout the southern spring [Briggs et al., 1979]. At L8176, when two yellow dust clouds occurred in the Argyre basin (-47 ø, 39øW and -48 ø, 31øW), To.0 ranged from 150 K over the COo. frost at -55 ø to 210 K at the latitude of the local clouds. The thermal characteristics of these clouds were difficult to ascertain because of the limited IRTM spatial resolution of these observations, the large temperature contrasts associated with the edges of frosted regions, and the similarity of surface and atmospheric temperatures. The locations of the centers of the tropical dust clouds are shown in Figure 2. During the period between the two global storms, orbiter observations covering all longitudes were checked for evidence of local dust activity, but only in the region between 80 ø and 130 ø longitude was such evidence detected. This is the same longitude band in which the global dust storms were first detected. The fact that local dust clouds were not detected elsewhere in the southern hemisphere, except near the south polar cap edges as described above, was surprising because of the many historical observations of dust clouds in the Hellespontus (-40 ø, 315øW)and Noachis (-40 ø, 345øW) regions during the southern spring [L. J. Martin and W.,4. Baum, 1969]. In 1977, local dust clouds formed repeatedly over particular locations: Solis Planurn (-20 ø, 90øW) and C!aritas Fossae (-20 ø, l l0øw). At L,221.4 a small dust cloud was observed by the VIS in Solis Planum (Figure 3a), at a time when the atmosphere was still dust laden from the first global storm, which had begun only I month earlier. At Ls227.7 a larger, more sharply developed cloud was imaged over the same location (Figure 3b). Both images were taken at -- 11.6 hours local time with violet 1976 J 1977 UTC DAY OF YEAR ß = VIS = IRTM = BOTH 0 30 60 90 120 150 150! I ' -10-20 - o -70 - I I I 170 180 190 200 210 220 230 240 250 260 270 250 290 i i EQUINOX PERIHELION SOLSTICE AREOcENTRIC LONGITLDE OF SUN, L s Fig. 1. Time line of the occurrences of local dust clouds in the southern hemisphere. Time is given in both universal time and solar longitude. The receding edge of the south polar cap is shown as two lines at the approximate time that the cap becomes asymmetrical [see James et al., 1979].

PETERFREUND AND KIEFFER: MARS VOLATILES 2855 0 0 0 m 0 0 0 ' 0. 0 0 0 0 0

2856 PETERFREUND AND KIEFFER: MARS VOLATILES

filters and similar viewing geometry. The dust cloud at Ls227.7 has an area of 3 X 105 km: and shadowsuggesting heights of 10 km. The IRTM observed a dust cloud in the same location on the preceding day (Ls227) in a set of four observations which spanned a midday period of 5.1 hours. The spectral signature and behavior in time of this cloud were studied in detail. The spectral signature of the Martian atmosphere as a function of season is discussed by T. Z. Martin et al. [1979]. In general, dust absorption has the greatest effect on T9, somewhat less and similar on T and T:0, and least on T7. This was shown to agree with the Mariner 9 infrared interferometer spectrometer spectra, which noted broad absorption bands due to the dust near 480 cm -, within the 20-#m band, and between 900 and 1200 cm -, which is centered at 9 #m and extends toward the 11-#m band [Hand et al., 1972]. The local dust clouds exhibit the same relative pattern, T7 being the warmest and T9 the coolest of the IRTM 'surface' temperatures at midday. Figure 4 shows these spectral differences for the L 227 dust cloud at 12.8 local hour. The shape of the cloud and the location of the thermal minimums are essen- tially the same for all of these four spectral bands. Because of the strong gradients of brightness temperature across the dust cloud the spectral changes can best be established by differencing the simultaneous individual measurements. Since T and T,are measured by two offset triples of detectors in the PETERFREUND AND KIEFFER: MARS VOLATILES 2857 same telescope, they cannot be directly differenced but can be compared to either T or T:o, which each have a full complement of seven detectors; T is used here. In :a 150 X 300 km area centered over the dust storm (selected as the area of the lowest temperatures), T,, T,, and T:o were + 16.0, -3.4, and -1.5 K relative to Tn, respectively, and 500 km to the southeast, off the edge of the dust cloud thermal depression, the cor- responding spectral contrasts were +9.8, -4.0, and -3.8 K. T, - T, however, was as low as -7 K near the strongest temperature gradients on the east and south boundaries of the dust cloud. Over the next 4-hour period of general cooling the spectral contrast between the current center of the cloud and the region 500 km southeast, (T, - rll)clouci- (T, - T,,)reg,on was essentially constant at 6.2 K for T - T, decreased from 0.6 to -2.5 K for T,- T, and decreased from 2.3 to 1.5 K for T 0 - T. The difference between T o and T was less over.the cloud than in the surrounding region at all times observed. There was in general a remarkable lack of spectral effect, apart from T,, considering that the cloud had a 35 K thermal amplitude. The other local clouds observed in the southern tropical latitudes had similar spectral clifferences...' 'able I presents the locations, observational ge0:rfietry, minimum temperatures, and estimated velocities for these clouds. The absolute values 290 J/.15 ø -20 ø --265-25 ø -25 ø _30 ø 105 ø 100 ø 95 ø 90 ø 85 ø 80 ø 17 Fig. 4a.30 ø 105 ø 100 ø 95 ø 90 ø 85 ø 80 ø T 9 Fig. 4b -15 ø.15 ø.25 o -30 ø 105 ø 100 ø 95 ø 90 ø 85 ø 80 ø 105 ø 100 ø 95 ø 90 ø 85 ø 80 ø Tll T20 Fig. 4c Fig. 4d Fig. 4. Brightness temperatures in 7, 9, 11, and 20 #m of the Ls227 dust cloud at 12.8 local hour. The contour interval in each map is 5 ø. The spacing between samples is somewhat less than the resolution, which was 227 km for these observations.

2858 PETERFREUND AND KIEFFER: MARS VOLATILES TABLE 1. IRTM Observations of Local Dust Clouds Revolu- Local Incidence Emission Resolu- fix, Velocity, S/C tion L, Hour Angle Angle tion Latitude Longitude T:0 T, Tg T7 T,, Tmoae km m/s 2 210 227 11.8 5 43 217-19 89 241 244 242 258 202 291 2 210 227 13.2 15 44 227-22 88 239 241 237 255 209 292 96 19 2 210 227 14.9 44 48 229-23 87 238 240 234 253 212 280 92 18 2 210 227 17.0 72 53 224-25 86 228 230 225 240 216 254 106 14 1 314 250 15.3 47 16 35-16 108 255 258 252 268 209 280 1 314 250 16.0 57 21 56-15 108 252 254 250 258 207 271 56 22 2 248 251 11.3 10 53 186-21 106 268 273 266 287 201 290 2 248 251 12.6 9 60 223-22 110 271 278 272 291 205 295 150 32 2 252 254 10.4 28 56 69-21 107 264 266 265 277 204 280 2 252 254 11.7 4 52 77-25 87 261 264 258 281 206 291 2 264 261 11.3 8 71 122-21 107 234 239 232 192 289 1 337 264 13.7 25 24 49-15 108 260 261 257 276 206 292 1 339 265 11.6 11 60 120-15 110 257 261 254 280 205 293 2 275 267 11.4 10 53 178-21 81 255 258 253 274 198 290 2 275 267 12.7 9 67 220-24 82 251 258 254 277 203 295 91 19 2 277 268 10.0 27 60 88-27 124 240 244 260 204 275 1 345 269 1.2 NA 61 147-16 109 180 180 187 174 181 191 Resolution = (0.0052 ß range)/cos e. NA means not applicable. of the spectral differences varied for each individual cloud. The The series of four IRTM observations, spanning 5.1 hours, variations appear to be due to actual variations in the infrared of the Solis Planurn dust cloud allow study of the short-period opacity of the clouds as well as to differing observational thermal changes and apparent motion of a dust storm. The incidence and emission angles. 20- tm brightness temperature maps for each of the four obser- -10 o -10 o.15 o -15 o.20 o -20 o -25 o -25 o.30 ø.30 o -35 o 110 ø I05 ø I00 o 95 o 90 o 85 o 80 o 75 o 11.5 LH Fig. 5a.35 o I10 ø 105 ø 100 ø 95 ø 90 ø 85 ø 80 ø 75 ø Fig. 5b 12.8 LH.10 ø -10 o.15 o.15 ø.20 o.20 ø.25 ø.25 o.30 o.30 o 110 ø 105 ø 100 ø 95 ø 90 ø 85 ø 80 ø 75 ø 110 ø 105 ø 100 ø 95 o 90 o 85 o 80 o 75 ø 14.6LH 16.6LH Fig. 5c Fig. 5d Fig. 5. Diurnal variation of L,227 dust cloud as seen in the 20-/ m brightness temperature maps. Local time refers to a common point at 95øW longitude. See Table I for the geometry associated with each map.

PETERFREU 4D ^ 4D KIEFFER: M^ S VOL^TILES 2859

2860 PETERFREUND AND KIEFFER: MARS VOLATILES _ 0120. i, i i i I i LATITUDE ---.15.78,, I,,, I,! 15'! 10' 105' 2 I I I ' LATITUDE ' I I t =-21.09 0. o! 15'! "! * v -- 105' LONGITUDE Fig. 7. Elevation profiles across Claritas Fossae. The data are from 12.6-cm radar observations in 1973 [Downs et al., 1975]. vations are shown in Figure 5. Peak temperatures in the region surrounding the cloud are observed at about 13 local hours, yet brightness temperatures within the area of the cloud are highest 1 hours earlier and decrease through the four observations. The thermal depression moved considerably to the south-southeast during this period. The velocity of the cloud, calculated from the displacement of the T.0 contours fitted to a 500-km circle, was 19 m s- between the first two observations, 15 m s - between the second and third, and 14 m s- between the third and fourth observations. The outline of the cloud seen in orbiter imaging at 11.6 local hour the next day (Figure 3b) agrees quite well with the 265 K contour of the 11.5 local hour T.0 map. Thus during the intervening 19 hours between the thermal and the visual observations, either the cloud had moved back to its original location or a new cloud had been regenerated in the same position. Between Ls260 and L 265, a period of 9 days, five I RTM observations of the C!aritas Fossae region recorded dramatic changes in the local dust activity. These changes are exhibited well in maps of T:0 (Figures 6a-6e). These clouds occurred 2-3 weeks prior to the onset of the second global dust storm, when the atmospheric opacity was near its interstorm minimum, although it was still 3 times that of its primary mission level (,- L 120) [T. Z. Martin et al., 1979]. On days 2, 7, and 9, local dust clouds are evident, while days I and 5 are relatively cloud free. The orbiter cameras did not observe this region during this time period, so that the evidence of the dust cloud activity is limited to the I RTM data set. The three clouds in this series, as well as four other dust clouds shown in Figure 2 and described in Table 1, occurred directly over C!aritas Fossae, which is a topographic ridge trending roughly north-south at approximately 110øW longitude. The comparison of the thermal maps of the local clouds with elevation profiles, produced from 1973 12.6-cm radar passes [Downs et al., 1975], show that the clouds are indeed centered over the ridge, local thermal minima occurring on both sides of the ridge (Figure 7). The other major topographic feature in this region is Arsia Mons (-10 ø, 120øW), a large shield volcano at the south end of the Tharsis plateau. The slope from the base of the volcano extends for a great distance to the south and southeast with regional slopes ranging from 1.2 ø to 0.1 ø [Saunders et al., 1978]. As day I was cloud free, the presence of the dust cloud on day 2 implies that it either formed rapidly or moved in from the east, where coverage was not available the previous day. The pattern of wind streaks and model calculations [Leovy and Mintz, 1969] suggest, however, that westerly winds should be predominant at this season. The rapid occurrence and dissipation of local yellow clouds has been noted from earthbased observations [L. J. Martin and W. A. Baum, 1969], suggesting that the first explanation is reasonable. The thermal characteristics of the day 2 cloud are quite similar to those of the L 227 dust cloud discussed above. During the primary mission the Claritas Fossae region exhibited a noted albedo contrast, with a bright area centered at -20 ø, 107øW [Kieffer et al., 1977], the same location as that of the clouds on days 7 and 9. From the albedo maps alone on those 2 days it is difficult to detect any unusual brightness features. On day 2 the variation in visual brightness is evidenced by a 20% brightening over the area of the cloud. In general, the albedo signatures of the local dust clouds were difficult, if not impossible, to distinguish from the underlying surface. It should be noted that the images of dust clouds in Figure 3 are highly contrast enhanced. The cloud on day 7 appears to be optically thinner than the day 2 cloud, as the spectral variations and the thermal contrast of the cloud in T:0 is less on day 7. The absence of a cloud on day 5 strongly suggests that the cloud on day 7 is a completely different cloud that has formed in the time between days 5 and 7. The cloud on day 9, however, may be the day 7 cloud slightly shifted to the east. At L 268, over Claritas Fossae, the one example of a dust storm at night was found. The spectral signature was reversed from the daytime storms, with T7 now cooler than Toby 13 K. This is what would be expected for temperature conditions at night, when the atmosphere and clouds are warmer than the surface. DISCUSSION The limited region of occurrence of local dust clouds in the southern tropical latitudes of Mars was one of the unexpected results of the Viking observations, as earth-based observations have commonly cited numerous other locations than the Claritas Fossae-Solis P!anum region. Whether or not local dust clouds will cluster in one particular region each season is not clear. However, it is apparent that the dust clouds will occur preferentially in particular regions over many seasons. The surface characteristics of the Claritas Fossae-Solis Planum region provide some insight into why these clouds preferentially occur in such a region. Dramatic differences in albedo boundaries and wind streaks in this region between 1972 and 1976 suggesthat significant redistribution of surface material has occurred in this region [Veverka et al., 1977]. Viking observations of this region from before and after the dusty period also show that dramatic changes in surface albedo and color features have occurred. Thus in the context of the local dust clouds this region has a thin veneer of fine-grained material that is mobile owing to the common occurrence of winds that exceed threshold velocities. There is a low thermal inertia region including the Tharsis plateau and extending west to the Elysium region and a relatively high thermal inertia region centered in Solis Planum (Figure 8). The minimum thermal inertia occurs on Arsia Mons, where diurnal temperature variations of approximately 150 K occur during the season of peak solar insolation, and a local maximum in Soils Planum, where diurnal variation is 60 K. The gradient in thermal inertia appears to be greatest across the Claritas Fossae-Solis Planum region, where it fol-

PETERFREUND AND KIEFFER: MARS VOLATILES 2861. o o o o o ' -0 O. 0 0.i :'7i' 'i

2862 PETERFREUND AND KIEFFER: MARS VOLATILES lows the topographic gradient (see Figure 2), and the combination could produce strong slope winds. The Claritas Fossae ridge would act essentially as a barrier across these winds, causing strong local turbulence capable of raising the dust into the atmosphere. The motion of the Ls 227 local dust cloud and three other clouds, shown in Table 1, ranged from 14 to 32 ms-', based on the displacement of the brightness temperature isotherms. The direction of motion consistently aligned with prominent wind streaks on the surface. The orientations of the wind streaks are predominantly parallel to the regional slope, thus implying the presence of downslope winds. The variation of wind streak form and contrast across the dust storm period suggests that Claritas Fossae-Solis Planurn is indeed a region of active aeolian processes. Some general inferences may also be drawn about local dust cloud height and about the infrared opacity near the cloud top by using simultaneous observations of brightness temperature at several wavelengths. An upper limit on the height of local dust clouds is placed by the absence of any noticeable change in T, 5 over the clouds. Measurements of T 5 over the cloud shown in Figure 4 were about 30 K colder than the top of the local cloud measured in the other thermal bands. The average temperature at the 0.5- mbar level at this season, latitude, and local hour is 211 K [T. Z. Martin and H. H. Kieffer, 1979]. A contrast in T, of 4 K, 4 times the noise level, would result if the lowest 14% of the T, weighting function were to sense a source at 237 K, the approximate temperature of the top of the local dust cloud. This places an upper limit on the cloud height of approximately 17 km. The appearance of vigorous activity exhibited by the local dust clouds in Viking orbiter imaging suggests that the lapse rate within the clouds is near adiabatic. A reliable estimate of absolute height based on cloud top temperature cannot be made because of the absence of information on the temperature difference across the boundary layer, or even the temperature of the surface where cloud shadowing may be important. However, assuming an adiabatic lapse rate (4.5 K/km) and no boundary layer yields a cloud height of about 8 km for the cloud observed at Ls227. The lowestemperatures occur about 250 km behind the apparent front of this cloud, so that just the assumption of a monotonic relation between temperature and height places the highest part of the cloud well away from the forward edge. The general similarity of the 9-, 11-, and 20-#m brightness temperatures for the local dust cloud is surprising, since the opacity, of the dust constituting the global pall varies considerably in this wavelength range [see T. Z. Martin et al., 1979]. This observation implies that the atmosphere has only a small temperature change over the altitude range including the first optical depth of the dust clouds at these three wavelengths. Over the L 227 dust cloud, the spectral contrast, excluding T7, is decreased from the surrounding region. T. Z. Martin et al. [1979] estimate the relative opacities of the global dust storms as r7 " 0.2r9 and r r:0 " 0.6r9. Using these values (and ignoring any opacity above the cloud)yields dt/dr " 5 K in the cloud top. Correcting for the global opacity by the crude estimate that one half of the regional spectral contrast is due to dust above the level of the top of the local cloud yields dt/dr, " 3 K. Assuming an adiabatic lapse rate implies an absorption coefficient in the 9-#m band of about 1 km - (1.1 and 0.7 for these two above cases). The spectral contrast of this dust cloud is typical of those shown in Table 1, indicating that such an absorption coefficient is representative of dust clouds'identi- fied with the I RTM. The contribution of dust into the atmosphere by the local clouds appears to be an important mechanism for triggering the major storms [Gierasch and Goody, 1973; Leovy et al., 1973]. While it appears that the initiation of the local dust clouds is tied to surface-related thermal and topographic conditions, the dust injected into the atmosphere by the local dust clouds is responsible for an increased diurnal temperature variation in the atmosphere [T. Z. Martin and H. H. Kieffer, 1979] which is the driving mechanism for the global dust storms [Leovy and Zurek, 1979]. The Viking observations have substantially increased the data base pertaining to dust clouds and storms on Mars. It remains to be determined from future Viking observations and/or future Mars missions whether the observations made during this year are typical of a Martian season or reflective of an anomalous year. Acknowledgments. This research was supported by the National Aeronautics and Space Administration, Viking Project Office, and by NASA-JPL contract 952988 to UCLA. REFERENCES Briggs, G. A., W. A. Baum, J. Barnes, and P. J. Gierasch, Viking orbiter observations of dust in the Martian atmosphere, d. Geophys. Res., 8d, this issue, 1979. Capen, C. F., A Martian yellow cloud, Icarus, 22, 345-362, 1974. Capen, C. F., and L. J. Martin, The developing stages of the Martian yellow storm of 1971, Lowell Observ. Bull., 7, 211-216, 1971. Capen, C. F., and L. J. Martin, Survey of Martian yellow storms (abstract), Bull. Amer. Astron. Soc., 4, 374, 1972. Chase, S.C., J. L. Engel, H. W. Eyerly, H. H. Kieffer, F. D. Pa!!uconi, and D. Schofield, Viking infrared thermal mapper, Appl. Opt., 17, 1243-1251, 1978. Downs, G. S., P. E. Reichley, and R. R. Green, Radar measurements of Martian topography and surface properties: The 1971 and 1973 oppositions, Icarus, 26, 273-312, 1975. Gierasch, P., and R. M. Goody, A model of a Martian great dust storm, J. Atmos. Sci., 30, 169-179, 1973. Hanel, R., B. Conrath, W. Hovis, V. Kunde, P. Lowman, W. Maguire, J. Pearl, J. Pirraglia, C. Prabhakaba, B. Schlachman, G. Levin, P. Straat, and T. Burke, Investigation of the Martian envi- ronment by infrared spectroscopy on Mariner 9, Icarus, 17, 423-442, 1972. James, P., G. A. Briggs, A. Spruck, and J. Barnes, South polar retreat seen by Viking, J. Geophys. Res., 84, this issue, 1979. Kieffer, H. H., T. Z. Martin, A. R. Peterfreund, B. M. Jakosky, E. D. Miner, and F. D. Pa!!uconi, Thermal and albedo mapping of Mars during the Viking primary mission, J. Geophys. Res., 82, 4249-4291, 1977. 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., and R. W. Zurek, Thermal tides and Martian dust storms: Direct evidence for coupling, J. Geophys. Res., 84, this issue, 1979. 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: Mari- ner 9 television experiment progress report, Icarus, 17, 373-393, 1972. Leovy, C. B., R. W. Zurek, and J. B. Pollack, Mechanisms for Mars dust storms, J. Atmos. Sci., 30, 749-762, 1973. Martin, L. J., The major Martian dust storms of 1971 and 1973, Icarus, 23, 108-115, 1974. Martin, L. J., and W. A. Baum, A study of cloud motions on Mars, final report, part B, 8 pp., Lowell Observ., Flagstaff, Ariz., 1969. Martin, T. Z., and H. H. Kieffer, Thermal infrared properties of the Martian atmosphere, 2, The 15-ttm band measurements, J. Geophys. Res., 84, this issue, 1979. Martin, T. Z., A. R. Peterfreund, E. D. Miner, H. H. Kieffer, and G. E. Hunt, Thermal infrared properties of the Martian atmo- sphere, 1, Global behavior at 7, 9, 11, and 20 tim, J. Geophys. Res., 84, this issue, 1979.

PETERFREUND AND KIEFFER: MARS VOLATILES 2863 Michaux, C. M., and R. L. Newburn (Eds.), Mars Scientific Model, Sect. 4.2, Doc. 606-1, Jet Propuls. Lab., Calif. Inst. of Technol., Pasadena, 1971. Pollack, J. B., D. Colburn, R. Kahn, J. Hunter, W. Van Camp, C. E. Carlston, and M. R. Wolf, Properties of aerosols in the Martian atmosphere, as inferred from Viking lander imaging data, J. Geophys. Res., 82, 4479-4496, 1977. Pollack, J. B., D. S. Colburn, F. M. Flaser, C. E. Carlston, D. Pidek, and R. Kahn, Properties and effects of dust particles suspended in the Martian atmosphere, J. Geophys. Res., 84, this issue, 1979. Ryan, J. A., and R. M. Henry, Mars atmospheric phenomena during major dust storms, as measured at the surface, J. Geophys. Res., 84, this issue, 1979. Saunders, R. S., L. E. Roth, and C. Elachi, Topographic control of volcanism and surface morphology in the Arsia Mons region of Mars, in Lunar and Planetary Science, vol. IX, part 2, Lunar and Planetary Institute, Houston, Texas, 1978. Snyder, C. W., The missions of the Viking orbiters, J. Geophys. Res., 82, 3971-3983, 1977. Tillman, J. E., R. M. Henry, and S. L. Hess, Frontal systems during passage of the Martian polar hood over the Viking Lander 2 site prior to the first 1977 dust storm, J. Geophys. Res., 84, this issue, 1979. Veverka, J., P. Thomas, and R. Greeley, A study of variable features on Mars during the Viking primary mission, J. Geophys. Res., 82, 4167-4187, 1977. (Received May 4, 1978; revised July 31, 1978; accepted October 4, 1978.)