Sotar Synm Rntaidt. ^iL 33. Ne. 6,199?, pp. 41S-4S2. Tmalaudinm AitnmaidditMi Kimli^ Hit 31. No. 6. ?99S, pp. 4SJ~SI}. Origbul Kuiiiai Ttn Cepyr?fht O 1993 by Marthnko. Baiiitvsky, Hoffmaa\, Hani?tii CooL Ntnhm. Geology of the Common Mouth of the Ares and Tiu Valles, Mars A. G. Marchenko*? A. T. BasUevsky*, R Hoftinann**, E. Hfluber**, A. C. Cook***, and G. Neukum** ? Vemadsky Institute of Geochemistry and Analytical Chemistry, Russian Academy of Sciences, ul. Kosygina ?9, Moscow, ?17975 Russia ** DLR Institute of Planetary Exploration, 12489 Berlin, Gerrriany *** Center cf Earth and Planetary Studies, National Air and Space Museum, ^?askinglon. DC, USA Received June 23, 1998 Abstract?Since the Mars Paif^nder landing site is located at the mouth of the Ares and Tiu valles, this region attracts keen scientific interest. In order to better undcretand experinienta] data ffom this small area of the sur- face of Mars, which has been iavesti|ated by the rover, and the properties of the malciials occurring here, it is necessary to answer how, from where, and when this material was transponed. To answer these questions, we performed photogeological mapping and counted impact craterin the region. A photogeological analysis of 320 TV images of the studied area, thermal-inertia maps, and digital elevahon models were used in mapping. Our results, and those published by other scientists, allow us to distinguish several principal stages in the geological history of ?ie mouth of the Ares and Tiu valles: the destruction stage of an ancient plateau and three stages of fluvial activity. Deposits transported by water flows to the Mars Pathfinder landing site presumably consist of fragments of the material of ancient Martian highlands (impact breccias and lavas), of younger sedimentary or volcano-sedimentary material of ridged plains, of impact-crater materials, and of eoUan products. The main constituents should be as old as approximaxely 4 Gyr (highland materials) and 3.5 Gyr (ridged-plain materials). It is very hkely that the fluvial reworking of this ancient material took place between 3.6 and 2.6 Gyr ago and, possibly, even lattr, between 2,3 and ) .4 Gyr ago. 1. INTRODUCTION 1.1. Martian Outflow Valleys In 1970, images obtained by Aionnej'9revealed val- leys on the Martian surface, which are similar in mor- phology to valleys and dry river channels on the Earth (McCauley et al., 1972). The similarity is, however, incomplete; for example, many Martian valleys, in con- trast to their terrestrial counterparts, lack drainage basins. Several morphological varieties of Martian vaJ- leys, such as valley networks, outflow valleys, runoff valleys, and ?ttted valleys, are distinguished (Sharp and Malin, 1975; Carr. 1996). The Arcs and Tiu valles studied in this work are of the outflow-vallQr class. About 15 Martian valleys of this class are known at present (Baker, 1982); they all arc large (up to several thousand kilometers long) and almost completely lack- ing in tributaries. In the recent geological epoch, water on Mars' surface could exist only in the form of ice and vapor, because the atmospheric pressure at the planet surfte is only 5-10 mb. The outflow valleys were foimed 3,6-0.5 Gyr ago after a period of heavy nveieor- itic bombardment, when atmospheric conditions were most likely similar to the cuiient state (Masursky et al., 1977; Scott et al., 1979; Wise et al., 1979; Baker and Kochel. 1979). Large river valleys on the Earth evolve over millions of years, mainly under the control of atmospheric pre- cipitation. Environments at the surface of Mars and the lack of drainage basins related to out?ow valleys exclude this mechanism for their formation. Glacier movements represent another process forming valley- shaped landfonns on the Eaith. Some researchers (Luc- chittaeio/., 1981; Lucchitta, 1982; Kai^cle/a/., 1995) believe this prtwess was similarly responsible for the origin of the Martian outflow valleys. However, gla- ciers of the Earth usually move along the fluvial (water- formed) valleys, transforming them into glacial troughs (Hamblin, 1975). In addition, the growth, movement, and thawing (tHit not evaporation) of glaciers can pro- ceed under terrestrial, but not under Martian atmo- spheric conditions. According to some scientific ideas, a possible alter- native is the catastrophic (geologically instantaneous) origin of outflow valleys, which was so quick that there was no time for water evaporation. In this case, we should assume an instantaneous release of a huge vol- ume of water. The outflow valleys could have been formed as a consequence of catastrophic outflows of groimd water to the surface (Baker and Milton, 1974; Baker and Kochel, 1978; Carr, 1979). There is evidence for the idea that the Martian lithosph?re may include horizons saturated with ice or water (see, for instance, Kuz'min, 1980). The morphology of head areas of outflow val- leys is indicative of plentiful sinkholes. Several pro- cesses could provoke water outflows on the surface. According to a hypothesis by Carr (1979), this was a 0038-0946/98/32O&0425$20.0O O 1998 MAHK Hayu/lnteiperiodica Publishing 426 MARCHENKO et al. pressure increase in water-bearing horizons due to freezing. Some scientists (McCauley et ai, 1972; Masursky et ai, 1977) assume water discharge to be a result of the thawing of pennafrost under the influence of gcothermai heating. Others believe that outflow val- leys and their deposits were formed by the movement of a substance rich in solid material, resembling mud flows or lahar? (Nunrunedal and Prior, 1981; Tanaka, 1988, 1997). Features similar to the Martian outflow valleys aie well known on the Earth (Milton, 1973; McCauley t? al., 1972). They are relatcid in origin to catastrophic water release ftom glacier-dammed lakes (Bretz et at., 1956; Bretz, 1969; Butvilovskii, 1993; Rudoi, 1995) and to ice thawing in response to volcanic eruptions under gla- ciers (e.g.. Thorarinsson. 1957). These geomorpholog- ical features are rare terrestrial counterparts of the Mar- tian outflow valleys, although the water-discharge val- ues in the last case should be 10-100 times greater (Baker and Milton. 1974; Baker, 1982). J.2. What is Stimulating Interest in the Mouth of the Ares and Tiu Valles? Whatever the formation mechanism of outflow val- leys?in particular, the Ares and Tiu?was {cata- strophic fluvial, related to mudflows, or glacial), it obviously might involve the transport of the materials derived from geologically different regions of Mars cut by these valleys to the mouth of the flows. The Ares and Tiu valles begin in regions with a chaotic morphology of the ground-collapse type, then they run across an ancient, heavily cratered highland region, and finally open to the lower-laying Chryse Planitia, where they merge (Fig. la). Each of these valleys is about 1500 km long and 25-100 km wide. Thus, they are as long as the largest rivers of the Earth, e.g., the Amazon River, but are notably wider. Moreover, processes unrelated to the valley origin proper could also operate in their common mouth, being responsible for the influx of various mate- rials, e.g., for sedimentation in temporary lakes and sea basins (Parker et al., 1989, 1993; Baker et al., 1991; Scott?/fl/., 1991;Craddockrto/.. 1997), eolian accu- mulation (Kuzmin and Greeley, 1995a, 1995b), subse- quent lava eruptions (Robinson et al.. 1996). and later mudflows (Jons, 1984), The high probability of detecting diverse geological materials in the Ares/Tiu mouth was the main reason for choosing this area for a detailed study with a niKxl- erate-sized rover brought by Mars Pathfinder (Golombek et aL, 1995, 1997b). In order to interpret the geological and geochemical characteristics of this area, we need to Imow how, from where, and when these materials were transported to the area in question. To find answers to these questions we performed a geological mapping of the Ares/Tm mouth, located from 15" to 30'' N and 26" to 36" W (Fig. la) and sug- gested a scenario of its geological evolution. 1.3. Data Used and Investigation Methods Geological mapping was performed using the pho- togeological analysis of TV images and some other data (Marchenko, 1996; Marchenko etal., 1996, 1997, 1998). We used 320 images in the visible spectrum range obtained by the Viking Orbiter, thermal-inertia and rock-abundance maps compiled on the basis of the thermal-mapping data obtained by this spacecraft (Christcnsen. 1986; Christensen and Kieffer, 1989; Edgett and Christensen, 1997), along with a color mosaic of the images shot by this spacecraft through various photo-filters. The resolution of the Wang images is 36-821 m per pixel. The resolution of the themial-mapping data used in this smdy is 30 km. In addition, a semiautomatic analysis of stereoscopic images (Thomhill et al., 1993) was used to construct topographic maps of several areas and to evaluate the altitudes of some relief features. When the distin- guished age (stratigraphie) units of the surface were marked by a statistically significant number of impact craters, their calculated density was used to check the age interpretation. The impact-crater count was per- formed on the basis of 55 images. In our studies, we refer to the global and regional stratign^hy of Mars elaborated recently by other researchers. Scon and Carr (1978) were first to recog- nize three major stratigraphic-age units of this planet: Noachian (ancient), Hesperian (middle), and Amazo- nian (young). They also evaluated the crater densities characteristic of these units. The absolute age of the units is a point of controversy. Neukum and Wise (1976) defined the ages of the Noachian-Hesperian and the Hesperian-Anuzonian boundaries as 3.8 and 3.6 Gyr, whereas Hartmann et al, (1981) suggest that they are 3,5 and 1.8 Gyr old, respectively. A map by Scott and Tanaka ( 1986), comprising the region under discussion, was used as a basis for our more detailed mapping. This map depicts more detailed stratigraphie subdivisions than that of Scott and Carr (1978), and each of the units is characterized by its impact-ciater density. In our study, we distinguished several new stratigraphie units missing in the map by Scott and Tanaka (1986) and suggested a new interpre- tation to some others recognized before. Most detailed among the geological-geomorphologic maps compris- ing the studied region is that of Rotto and Tanaka (1995), although the stratigraphie and morphologic units of valley floors and eolian deposits, especially interesting for us, are missing there. 1.4. The Layout of the Paper In section 2, we describe the distinguished geological (stratigraphie) units shown in our map. Some physical parameters of their surface are presented in section 3, and SOLAR SYSTEM RESEARCH Vol. 32 No. 6 1998 GEOLOGY OF THE COMMON MOUTH OF THE ARES AND ITU VALLES, MARS 427 3 Z < E IS as 3! ? S ? u 0D III a = ^ ?IB u e- ^ - ^ tt 1 l@ > ^ Oil it n D ?g,(S ^? ID c is ii: S M? r I e IIH ? I LU Z ?. E 3 1^ D 't irae i [3b-?I ;? [le De f I o u a. C? C fi. d C 1" ?g g. '3 I il -g g %1 g s ?SE lA IM ?S-d S? SOLAR SYSTEM RESEARCH \tol. 32 No. 6 1998 428 MARCHENKO et al. EF FP DP CFT IB LibeR?d CFAC^?^ pp SP SlP FC W?bm IC Fig. 2. Stratigraphie scheine proposed for ibe studied region. RighL subdivisions preda?ng v^ley fonnaiion (CP, RP); center units of the Ares and Tiu Valles floor (StP, SP, FC, IB, PP. CFA, CFT. DP) ind fractured plains (FP); left impact craters (IC) and eotian materials (EF); the (vertical) time scale is arhitrary. For symbol expl?aiion sec Fig. Ic and text our results of the impact-crater count are included in section 4. Correlation between these results and the data of photogeological analysis is discussed in section S. The model ages for the distinguished stratigraphie units and events of their subsequent reworking are the main topics of section 6. In section 7, on the basis of our investigation, we present a scenario of the geological histoiy of the area studied. Finally, section 8 is devoted to a general discussion of the materials from the Mars Pathfinder landing site in the context of this study. 2. DISTINGUISHED MATERIAL UNITS 2.1. Structure ofThis Section On the basis of the photogeological analysis of itn^es, crater counting, and previous publications, we distinguished 14 major geological subdivisions (strati- graphic units) in the region studied (Figs, lb, Ic, Id) and suggested their formation history (for the resulting stratigraphie scheme, see Fig. 2). The distribution of the images obtained by the Viking Orbiter (Figs. 4-15), which illustrate the moiphology of these units in the region at hand, is shown in Fig. 16. Sometimes we refer to images that are missing from the set presented here. In this case, we indicate the image number according to the system of the V?fanf Orbiter catalog, e.g., 864A07. The impact-crater densities calculated for the identified stratigraphie units are listed in the table and are shown in Fig. 3. The standard crater-density values character- istic of three major subdivisions of Martian stratigra- phy axe quoted after Scott and Tanaka (1986). The stratigraphie units are described, beginning with their oldest ones. We distinguish three major groups of units: {1 ) a group predating the formation of the Ares and Tiu valles; (2) the unit group of these valleys; and (3) the youngest group consisting of only the unit of fractured plains. The materials of impact craters and eolian deposits are considered to be formed over a time period longer than any of the mentioned age groups. The first and the second group are divided into subgroups, but within the latter, it is difficult to determine age relations between the units: they are different in morphology; however, it is not clear whether this is a result of differ- ences in particular formation processes or of subse- quent resurfacing. Within some units (valley floors, islands, and benches?see below) we also distinguish morphological types of the surface. 2.2. Ancient (Valley-Predaiing) Materials The valley-predating geological subdivisions repre- sent four groups of the highland-terrain materials and the younger material of ridged plains. Highland stratigraphie units, Cratertd plateau (CP, image 864A07) is a characteristic feature of the southeastem part of the region under study. It occupies an area with large, ancient, heavily destroyed impact craters. The plateau surface also exhibits random wrin- kle ridges. The surface between the craters is spotted. Splotches are dark and rounded, up to S-10 km in diameter. It is likely that the material in the splotches represents dark eolian deposits in the heavily destroyed craters. Several faint valleys in the plateau are 2-6 km wide and seemingly fluvial in origin. The CP material presumably includes the Middle Noachian impact brec- cias and die Hcspcrian-Noachian lavas (Rotto and Tanaka, 1995). During the Late Noachian time, before the origin of the Ares and Tiu valles, the plateau was destroyed in the north by slope-collapse events and under the influence of water-saturatod mud?ows (Squyrcs and Kasting, 1994; Tanaka, 1995). This stage of erosion appears to be related to the origin of the Great Northern Plain (McGill and Dimitriou, 1990). Later flows eroded the entered plateau (CP) in the west to form the Ares ValUs. This plateau is likely the oldest in the region, because there are no data suggesting that its material overlies any other rocks. Our crater-counting results indicate a Noachian age for the plateau (Fig. 3). Knobby plateaux (KP, Fig. 4) arc typical of the south and southeast. They are quite similar to the cratered plateau (CP), but differ from the latter by the presence of numerous small knobs. Knobs are up to hundreds of meters across; they are larger in the north than in the south, and isometric in sht^. Usually, these features show chaotic distribution patterns, but are locally arranged into chjuns and rings. It was suggested that knobs were created during the intense destmction of the northern highland margin at the end of the Noa- SOLAR SYSTEM RESEARCH Vol. 32 No. 6 1998 T?blt. GEOLOGY OF THE COMMON MOUTH OF THE ARES AND ITU VALLES, MARS Areas of impact-crater counting 429 Comtinates Total Number of craters Site Viking Or- Total area. Map symbol of the Cuiitplex number (of diameter > 1km) Model abso- fcirrr images km^ complex-aiea aiea,km^ of perlO^km^tl stan- lute age, Gyr center craters* dard deviation 1 669AS7-90 574135 ICKipni 26.4N,30.8W 279.27 13 8490 ?5500 0.7 3.97 SP 26.7N.30.5W 3802.68 151 1540 ?630 1.56 3.56 IB2 26.2N,30.1W 292.55 13 - 0.87 CFr{seeI+2) 26.2N,30.4W 758.76 36 2380?1757 1.06 2.89 SC* - 609.09 - - - 2 669A9I-94 4532.15 StP 25.9N. 28.9W 106.50 6 - - SP 26.9N, 30.8W 92.95 7 23920 ?15985 3.55 4.06 SP (see 2 + 3) ?61N.28.6W 196.46 9 - - IBi 26.8N, 29.2W 332.15 16 - 1.47 U-'l{seel+2) 26.4N. 29.6W 913.22 57 3620 ?1968 1.51 3.61 ?g 26.5N, 29.0W 1939.58 115 1031 ?729 1.36 W 26.1N,29.0fW 654.83 63 4490?2611 3.56 SC ? 296.46 - ? - 1+2 O-T 93 1.30 3.24 3 669A95-96 2123.79 StP 25.8N,28.9W 35.32 3 - - SP (see 2 + 3) 26.6N,28.0W 1914.91 92 2780 ?1205 0.94 1.52 3.95 W 26.7N.28.7W 141.72 13 13650 ?9790 - 5C ? 31.84 - ? - 2 + 3 SP 101 0.94 1.52 3.93 4 669A67-70 6802.31 IC 28.8N.34.9W 736.73 26 - UO FP 28.7N.35.3W 5557.69 246 2498 ?668 2.01 3.61 f>P 28.5N.35.1W 311.47 12 - 0.71 SC ? 196.42 - - - 5 034A89-92 6095.43 IC' 24.4N,32.1W 183.97 4 - 0.38 ICWahoo" 23.6N.33JW 361.56 18 5010 ?3695 0.96 3.10 ICWahoo* 23.7N.33.0W 1586.92 22 1060?815 0.51 StP 24.0N.33.0W 1055.61 35 3425 ? 1800 1.21 3 98 PP(see5 + 6) 24.4N.32.9W 992.45 26 _ 0.68 IB? 24.6N,3Z2W 535.92 11 3710 ?2630 0.34 CFT 24.5N.315W 1112.15 31 2961?1628 0.86 3 63 SC ^ 266.85 _ _ J'V^ 6 0MA83-84 3341.94 IC Mut and Cave ? 203.59 2 - - PP (see 5 + 6) 22.5N,35.1W 2861.16 74 1770 ?780 0.95 2.22 2.97 3.74 CFT 21.8N.35.2W 151.50 6 1.40 E-3 2.83 DP 23.0N.35JW 125.69 4 - - 5 + 6 PP 100 0.87 2,22 3.67 SOLAR SYSTEM RESEARCH ^tol. 32 No. 6 1998 430 MARCHENKO et al. Table. (ConuL) Coocdinaiesof Total Number of cniets Site Viking Or- Totaiarea, Map symbol the Complex number (of diameter >1 km) Model abeo- biter iniaga ton^ oomptex-aiea area,kni^ of per 10^ km^?l stan- lutcagcGyr center enteis* dard devittioo 7 ?33A4I-44 114S3.61 IC 24.4N.27.6W 285.18 5 13420 ?6850 4.17 KP 22.5N, 27.0W 55.82 3 - - SP 24.2N, 27.2W 1492.97 19 5580 ?1932 3.88 SP(see7+9) 23.2N.26.8W 9617.59 151 2640?523 1.76 3.56 3.75 SC - 32.05 - ? _ 8 0Q8A30-3I 4252.72 tCWahoo, ?jecta 23.8N, 34.8W 35.90 3 27855 ?27855 1.93 SlPandSP 24.0N, 34,8W 560.90 31 1783 ?1783 1.05 2.69 IBj 24.0N,35,IW 93.78 5 - 3.19 DP(see6 + 8) 23.7N.35.5W 2655.40 119 1030 ?616 1.51 FP 24.2N,35.4W 817.40 45 3030 ?1910 2.20 SC - 89.34 - ? _ 6 + S DP 123 1.58 9 864A06-12 162170.5 NPL, 18.0N.24.0W 12767.21 79 6082 ?690 3.87 4.05 4.17 NPL? 14.0N.26.0W 16633.05 66 3915 ?486 3.80 3.99 4.08 CP 16.4N, 27.0W 28273.56 142 4940?41S 3.82 3.95 KP 17.5N. 27.7W 16627.46 76 4480?5I7 3.80 3.94 ICSoochow 16.8N. 28.9W 4989.83 13 2577 ?719 3.74 3.89 ICRibe 16.6N, 29.2W 426.22 2 - 3.94 Soochowand 15 3.77 Ribe 3.87 ICZuni 19.3N. 29JW 2359.59 5 2099?944 3.48 ICBaios 21.7N, 29 JW 1413.03 5 3505 ?1574 3.82 IC Libertad 23.3N, 29.4W 2845.54 7 2393 ?914 3.43 RP 20.5N. 27.7W 44051.29 145 3230 ?271 3.69 3.79 m, 15.0N.28.0W 9907.57 47 4676 ?687 3.76 3.89 CFA4 13.0N.28.ffW 14433.95 19 1640 ?378 3.52 3.75 CFA3 15.0N.29.9W 1891.70 4 2080 ?1046 3.09 CFA4 + CFA3 23 3.48 3.71 SP(s?7 + 9) 22.9N.26.3W 7999J8 27 3270 ?636 3.70 3.88 SC ? 551.10 - - - 7 + 9 SP 178 3.57 3.79 10 004A49-54 aod 8912.14 RP 20.8N. 30.4W 523.83 25 7330 ?3740 2.69 3.79 0a3A43 IBj 20.9N. 31 JW 1373.63 65 5096 ?1926 2,97 3.75 CFA 20.5N. 31 JW 6622.52 201 2869 ?658 1.69 (see 10+11) 3.69 3,74 SC - 392.16 473 - - SOLAR SYSTEM RESEARCH Vol. 32 No. 6 1998 GEOLOGY OF THE COMMON MOUTH OF THE ARES AND TTU VALLES. MARS 431 Table. (Coned.) Coordinales Total Number of crateri Site Vikif?g Or- Total area. Map symbol of the Complex number (of diameter >1 km) Model abso- biter images kn>? complex-area aiea,kni^ of perld^km^tl stan- lute age. Gyr colter craters* dard deviation n 004A41'48 9872.70 IC ? 475.11 4 - - IB, I8.7N.33.2W 21.22 4 - 3.62 CFA+CFT I9.5N.32.8W 8695.65 296 2166 ?498 1.57 (sec 104 11) 3.46 3.74 SC _ 680.72 _ ? - 10 + 11 CFA + CFT 497 1.62 3.59 3.74 12 ?WA19-23, aWA36, 8860.92 IB, 18.2N.33.6W 625.18 33 4799 ?2770 3.23 4.00 004A40 CFT 18.6N.34.3W 7936.51 340 1323 ? 407 2.25 SC - 299.23 486 - - 'Diaraeier greaierthan six pixeU. Clusiers of secondary cixters. 'Crater rim, trater (loor, '?jecta, central rise, and rim. chian time (Tanaka, 1995). That is probably why they are larger in size northwanl. where the erosion level was deeper. Rings of knobs apparently represent the destroyed rims of impact craters (RK, see below), whereas their chains are reinnants of partially wrecked wrinkle ridges. The boundary between the units knobby (KP) and cratercd (CP) plateaux is indistinct. We inter- pret the KP material as products of the terminal Noa- chian erosion of the cratered plateau (CP). Later flows that created the Ares and Tiu valles eroded the KP material. Hills and mesas (H, Fig. 4) inside ridged and smooth plains (RP and SP, respectively?see below) represent ?equenl and specific features of the surface. Hills resemble the similar landforms of knobby pla- teaux, but are larger (1-2 km across on average). Flat- topped mesas up to 10 km in diameter have steep slopes and presumably are remnants of an ancient dismem- bered surface. The hills of unit H are more renwte nom one another than their analogues in the kn(^by pla- temix. A zone of hills and mesas, known as Oxia Colles, is about 160 km wide and extends from the southwest to the northeast in and outside the studied area. Mesas and hills of unit H seem to be renmants of an ancient cra- tered plateau (Rotto and Tanaka, 1995; Tanaka, 1995). Boundaries between them and other materials are con- cealed under slope deposits, and the age rel^onships of these materials remain obscure. It is only established in a few localities that foothill areas are overlain by flow-shaped deposits of very young fractured plains (FP, see below and in Fig. S). Rings of knobs (RK, Fig. 6) are seen in the images as rings or semicircles. T^^ese landforms are mostly concentrated inside the knobby plateaux imd ridged plains, and only one ring is located much further to the north, in smooth plains (see below). It is evident that rings of knobs are remnants of the rims of large, ancient impact craters of the Noachian plateau (Rotto and Tanaka, 1995; Tanaka, 1995), and the material is thus the same in units RK, OR and KP. The northernmost ring is located at a latitude of 30'' N. According to Grumpier (1995), this position means that the material of the ancient plateaux is also present under the younger, smooth plains; in this case, the distribution area of the fonmer extends northward up to 30" N. Ridged plains (RP, Fig. 4) are located in the east of the studied region, north of the KP arca. The boundary between units KP and RP is seen in the images as an escarpment or steep slope facing the RP area. Wrinkled ridges of submeridional orientation are characteristic features of these relatively bright plains. As already mentioned, within the ridged plains, there are relic hills (H) of a more ancient material. In addition, within these plains, several indistinct, valley-like landforms and elongated hills were detected, which are of submeridi- onal orientation and lune presumably composed of younger material of faint channels and islands (FC, see below). Because the ridged plains (RP) display wrinkle ridges at their surface, as is typical of many volcanic plains of Mars, the Moon, and Venus, some researchers believed that the ridged-plain material originated nom lava eruptions (Scott and Tanaka, 1986; Rotto and Tanaka, 1995). However, neither boundaries nor vents of lava flows have been detected in the region. Waiters (1988), as well as Rice and De Hon (1996) suggested that the wrinkle ridges could also be formed over a sed- imentary substratum. We also share the opinion of SOLAR SYSTEM RESEARCH Vtol. 32 No. 6 1998 432 10000b 1000 100 10 MARCHENKO et ai. Cralered Plateay t T -I I ? I 1 III! Anuzonian .1..L?.11?1J L ^>o/10*km* 100000t 10000 1000 SyWHi Smooth Plains \ \ NoachUn H Hespeiiao 0.1 Af>o/10*kiii2 100000 1 10 Crster diameter D, km System 10000 100 0.1 100000 _l ' ' ' 1 im 1 10 Cnter diameter D, km 10000: 1000: 1000 100 System - 1 : I^sctuiednaiiu \ 4 ? : ? NoKhian \\ Hetpehan ' Cnter diameter ?), km ai 1 10 Cnter diameter/), km Fig. 3w Some iHu]tt of the cntoT'deBiity count for four diffmtit lurfacc ?cai (only one dixribution cum it shown Ux each unit). The Noacl?iia, Hesperiao, and Amazooian density langei for cnien greaier than 2 km in diameter are fnmi Scoi^ Tmzka. ( 1997), who considers the ridged-plain material to be ancient and sedimentary in origin. This material could be transported from an ancient southern plateaux affected by intense Late Noachian erosion. In the north, it is overlain by younger deposits, and this situation was also noted by Scott and Tanaka (1986). In our stiati- gr^hic scheme, we attribute this younger material to the lithologie units of smooth (SP) and spotted (StP) fluvial plains. In images of nwdcrate r?solution (e.g.. 633A41). ?ie boundary between the older (RP) and younger (StP and SP) materials is undetectabte. We think it corresponds to the boundary between the daric and bright areas in the low-iesolutioa image S64A04 (Fig. 7). The Ares and Tvx valles are incised into the RP material, and the latter was thus eroded by the flows that have formed these valleys. Prior to the deep>tnci- sion state, the flows likely meandered over the RP sur- fece producing foint channels and islands (FC). The SOLAR SYSTEM RESEARCH Vol. 32 No. 6 1998 GEOLOGY OF THE COMMON MOUTH OF THE ARES AND ITU VALLES, MARS 433 calculated density of impact craters suggests the Late Noachian-Early Hesperian age of the RP matenal. 2.3. Materials of the Ares and Tiu Valles In this section, we describe the materials character- izing the floor, benches, islands, and deltaic deposits of the Ares and Tiu valles. According to the fonnation age, we divide the material units into ancient, middle, young, and diachronous. The latter were formed in the course of the whole fluvial history of the region. Ancient fluvial material. This group includes spot- ted (StP) and smooth (SP) plains, and also faint fluvial channels and islands (FC). Spotted plains (StP, Figs. 7 and 8) are relatively dark and display brighter splotches, whose diameters vary from a few kilometers to tens of kilometers. Within these plains, there are low ring rims, lacking ?jecta zones. They seem to be the rims of impact craters par- tially overlain by the material of the plains. Lighter splotches may represent small islands of older material, most likely, of the ridged plains (RP) bordering the spotted plains (StP) in the south, or of the cratered pla- teau (CP). If the lighter areas are actually remnants of this type, then the dark plains in between represent the StP material itself. This combination of dark and lighter splotches may be indicative of a relief with depressions representing sags formed after thawing of the frozen ground or may be eolian in origin, as it is especially evi- dent in the case of pitted plains (PP, see below). The StP material is detected precisely in the mouth of the Ares and llu valles (Fig. 1 b) and may be deltaic in origin. The spotted plains were eroded by the flows responsible for the carving of the Ares and Tiu valles (Fig. 7). Eastward of the Crater Wahoo (Fig. 8), the spotted plains (StP) are bordered by a steep slope or scarp facing the pitted planes (PP); in other words, the StP material is located higher in the relief than the sur- face of the latter. According to our photocUnometric measurements, the altitude diflference between the StP and the PP surface level is less than 80 m. The scarp morphology suggests its erosional genesis. The boundary between the spotted (StP) and the neighboring smooth plains (SP, see below) is vague. As semiburied craters are detected in both types of plains, we may assume a similar fonnation style under sub- aquatic conditions. The material of both may represent deposits of the intermediate (StP) and the distal (SP) part of the alluvial fans produced by catastrophic floods, as shown in the schematic published by Rice and Edgett (1997). The unit of smooth plains (SP, Fig. 6) characterizes relatively dark smooth areas sculptured with rare wrin- kle ridges. A relief of this type is detected in the north- east of the region. Young impact craters in the smooth plains (SP) are surrounded by a bright halo and rays, which presumably mark ?jecta of a lower and older material lighter in color and representing, most proba- Fig. 4. Cratered plateau (CP), hills (H), and ridged plains (RP) as seen in Viking Orbiter image 864A08 (north is ai the top in all images). Fig. 5. Tm Vallis floor (CTF), fractured (FP) and smooth (SP) plains as seen in Vtkins OrA??er image 669A91. Mod- flow material from the fractured plains covers the valley floor, smooth plains, and the basal parts of some hills. SOLAR SYSTEM RESEARCH Vol. 32 No. 6 1998 434 MARCHENKO et al. Fig, 6. Rin^ of knobs (RK) and smooth plains (SP) as seen in Viking Orbiterimagt 670A33. Greater sizes of the hills in the south may indicate either the northward-growing thickness of the overlying material of smooth plains or the intensification of ancient erosion in this ditection. Bright rays around an impact crater located westward are seen at the left. CFT M " ' .^. ? ' SP ? ..V ?? A '? ??? ? 0 ' ? ? Ci o ,v 9 ?? : ^OK;' "' ^''A\fc.jMii WWA '' 1 ma3xVBy.^^^m i/.m^'^: . ?? ** . fc*?^.; Hg. 7. Spotted (StP). smooth (SP), and ridged (RP) plains, as well as the Tiu Valus floor as seen in WJUng Orbiierimigp 864A04; Crater Libertad is in the left lower comer. bly, the material of ridged plains (RP) or cratered pla- teaux (CP). Ancient craters are partially buried in the smooth plains (SP) by the material of the latter. Ruins of one very ancient crater are seen at 26? N, 26? E (Fig. 6) as a semicircle of knobs (RK material) that is open toward the north and suggests either a thickening of the overlying deposits in this direction, or the previous destruction of the northern sector of the crater rim. Within the smooth plains (SP), the smallest craters that are partially buried but still visible are about 3 km in diameter. Rims of these craters are 80-100 m high (Pike and Davis, 1984), and the layer of the SP material should be less than 80-100 m thick. The smooth-plain material was previously consid- ered volcanic (Scott and Tanaka, 1986) or se?mentary in origin (Rice and Edgett. 1997). In the studied region, we did not l?nd any signs of its volcanic genesis, e.g., the outlines of lava flows or volcanic vents. The forma- tion of these plains under subaquatic conditions (Parker ?taL, 1989; Scott f/o/., 1991;Tanaka, 1995) seems to be more realistic. Relations between this material and the Tiu Valus floor (CTF, see below) suggest the ero- siona) incision of the latter into the SP material (images 670A2? and 670A27), and we may state that the Tiu Vailis is younger than the smooth plains. The boundary between fractured (FP) and smooth plains is indistinct, but the material of the former is deposited over the Tiu Vailis floor, and they should be younger than the smooth plains. The cumulative-density plot of impact craters (Fig. 3) illustrates our data for one of five SP areas for which we performed crater counting. Judging from all the data obtained, the area in question is of the Late Hesperian age, whereas four other SP areas are of the Late Noachian-Early Hesperian age. The imit of faint channels and islands (PC, Fig. 9) characterizes areas with indistinct furrows and elon- gated rises, which are about 10 km wide, and decorates the surface of the cratered plateau (CP) and of ridged (RP), smooth (SP), and spotted (StP) plains, where the material is similar in brighmess but distinct in morphol- ogy. The FC materials are eroded by valley-forming flows and buried under the FP tnatcrial (see below) and ?jecta of the Libertad impact crater The FC deposits may represent products of the initial stage of fluvial activity in the region, when water flows did not carve their certain waterways and flooded the plains widely (Nelson and Greeley, 1996, 1998). Fluvial materials of intermediate age. The strati- graphic units in question are characteristic of the Ares Vailis floor (CFA) and of pitted plains (PP). We divided the CFA surface into plains of several morphological types (CFi^). The Ares Vailis floor (CFA) is seen in the southern and central parts of the region. Some investigators, e.g., Rotto and Tanaka (1995), consider valley floors a mor- phological, rather than a geological, subdivision, ascrib- ing to them solely the erosion genesis. On the Earth, SOLAR SYSTEM RESEAKCH Vol. 32 No. 6 1998 GEOLOGY OF THE COMMON MOUTH OF THE ARES AM) ITU VALLES, MARS 435 however, valleys of catastrophic water outflows host their own deposits (Rice and Edgett, 1997; Butvilovskii, 1993). In the mouth of the Ares and Tiu Valles, where the flow energy should be considerably decreased, the deposition of sediments is inevitable. The idea that catastrophic-flood deposits occur at the floor of the Ares and Tiu Valles and in their mouth is supported in addition by the results of the analysis of Mars Pathfinder panoramas (data on the size range of coarse clastic material; sec Golombek et al., 1997a). Four morphological varieties of CFA plains are as fol- lows: most widespread smooth plains with occasional fur- rows (CFj, Fig. 10); hilly plains (CF^, Fig. 11); plains with frequent, yardang-like ridges (CFj, see image 003 A57); and fretted plains (CF4, Fig. 12). These mor- phological varieties of the Ares floor plains could be related in origin to differences in hydrodynamic condi- tions between different parts of the valley-forming flows, to variable environments of subsequent modifi- cation of the plains, to the substratum diversity, or to a combination of all these factors. In some works (e.g., Costard and Kaigel, 1995; Kuzmin and Greeley, 1995a), the eolian deflation and thermokarst are sup- posed to be the main reworking mechanisms that have affected the valley floor after exsiccation and, possibly, is in progress now. Plains CF, are smooth; they display rare longitudi- nal erosion grooves and hills reseiiib?ng those of the H-unit?the classic type of outflow valley floor. Appar- ently, the hills were formed in the course of incision into older units under the influence of catastrophic flows of water (Baker, 1982). Their surface is locally covered with dark and bright streaks and spots with vague boundaries. These are likely traces of subsequent eolian redistribution of the material (see section 2.6). Plains CF2 are spotted by frequent hills, also resem- bling the H-hills and varying in diameter from hun- dreds of meters to a few kilometers. In this case, knobs may represent either mesas or huge rock blocks that were dragged by the valley-fomiing flows. According to estimates by Komatsu and Baker (1997), the cata- strophic flows responsible for the origin of the Ares Vallis were able to drag blocks as large as 10-100 m across. As a rule, hills under consideration are larger and seem to be of erosional, rather than fluvial, genesis. According to Tanaka (1988), the hills in the mouth of the Tiu and the nearby Simud Vallis may represent, however, deposits of the debris flow. Ridges of the yardang type are characteristic of the next plain category, CF3. They are usually as wide as 100-200 m, occasionally up to 2 km, and as k>ng as 30- 40 km. Baker (1982) believes that some of the most elongated ridges in the Ares Vallis floor are true yardangs formed by eolation in a relatively son fluvial material. Actually, ridges with the CFj morphology may exemplify these landforms; however, in our case, they are elongated parallel to the valley walls, i.e., dis- cordantly relative to the bright and dark wind streaks y*' ^ ^\ : -W " '' ^ i- i'- 't T- '?'? '??'^???;pm K||->, -^i .^uS| HP^' ' Pf 7. ?'-. Fig. 8. Spotted (StP) and pined (FP) plaiiu as xtn in Vitinj Orbiier ?nage 034A89. A panially buried crater is located in the upper image area. (Figs. Id, le). Accordingly, their eolian genesis, at least under the current wind regime, seems doubtful. It is likely that these morphological areas of the Ares Vallis floor were created by fluvial incision into a material dif- ferent from the rocks characteristic of the valley floor where other morphological varieties occur. The surface of CF4 plains is fretted, displaying an alternation of ridges and depressions of irregular shape. The ridges are from hundreds of meters to a few kilo- meters wide and up to 10-15 km long; the depressions are up to 10 km across. Rotto and Tanaka (1995) assume that areas of the CF4 type are composed of a material resistant to fluvial erosion, whereas Costard and Kargel (1995) believe that the respective system of ridges and depressions is of tiiermokarst or eolian ori- gin. In addition. Costard and Baker (1995) suggested the glaciofluvial genesis of this relief It is also possible that the cauldron-type subsidence also contributed much to the formation and deepening of the Ares Vallis floor (Sharp and Malin, 1975). According to our mea- surements, the valley in the region is 6(X)-12(X) m deep. The Ares Vallis is incised into the material of the cratered (CP) and knobby plateaux (KP), or into the ridged (RP) or spotted (StP) plains. Accordingly, we may assume that its incision down to the present level took place after the formation of ancient fluvial depos- its (StP and SP). A catastrophic flood of that time could also be responsible for the Tiu Vallis formation, because both valleys have tributaries from the Hydaspis Chaos, which are approximately concurrent in age (Rotto and Tanaka, 1995). At present, however, the Tiu Vallis floor is incised into that of Aies Vallis and thus should be younger. The crater-counting results for the Ares Vallis floor suggest its Hesperian age. Pitted plains (PP, Figs. 8 and 11) characterize rela- tively bright surface areas with abundant dark, flat-bot- tomed depressions approximately 30 m deep (Costard SOLAR SYSnSM RESEARCH \bl. 32 No. 6 199S 436 MARCHENKO et al. Ftg. 9. Faint channels and islands (FC) inside smooth plains (SP) as seen in Mfcing Orbiier image 0O3A16. Fluidiicd ejecca from Crater Ubenad are scattered in ihe right uppei comer. F1|t 10> Islands and benches (IB) in the T^u Vallis as seen in Viking Orbiter image (W3A14. The valley floor is smooth (CFi). and Kargel, 199S). Joining locally, these depressions form branching, gorge-like features opening into dark plains (DP, see below). In the western part of the region, pitted plains exhibit bench-like forms and islands located in the common mouth of the Ares and Tiu Valles (Fig. lb). Treiman (1995. 1997) reported alter- nating bright and dark horizons in the scarps of bright and other plains and plateaux located in the mouth of the Ares and Tiu Valles. Some researchers (Scott and Tanaka, 1986; Rotto and Tanaka, 1995) argued for the volcanic origin of this material; however, in morphol- ogy, the pitted plains better correspond to the karst areas of the Earth. Costard and Kaigel (1995) believe that abundant thermokarst cauldrons in the pitted plains (PP) were formed over a ground saturated with ice, but they also do not exclude the possibility that these caul- drons are products of deflation. The material saturated with ice could be of fluvial, lacustrine, or glacial origin. Tanaka (1997) suggested that it was transported mainly from the Maja ValJis, located further westward. According to Kuz'min (1996), the unit under consider- ation characterizes the fluvial deltaic deposits of the Ares and Tiu Valles, and this viewpoint seems to be most reasonable. Islands of the pitted plains (PP) extend southwest- ward to 20? N, 38? E, crossing the mouth of the Ares, liu, and Simud valles. Apparently, the deltaic sedimen- tation that deposited the PP material was spread over a vast area. As mentioned above, the spotted plains (StP) are bordered by a scarp facing the pitted plains (PP). It is possible to suggest two models for scarp formation: either the mudflow forming Tiu Vallis reached the lower level of the PP material a?er the erosion of the StP material, or the PP material was deposited during the accumulation stage that followed the erosion period. In the latter case, the pitted plains (PP) are younger than ancient fluvial deposits and, being of flu- vial origin as well, they characterize the subsequent st^e of erosion and accumulation in the valley. The pit- ted plains (PP) were eroded by the Tiu Vallis flow, and, probably, the boundary between the Tiu Vallis and the Ares Vallis was simultaneously destroyed. The crater- count results suggest the Late Hesperian-Early Ama- zonian age for the pitted plains (PP). YouDg fluvial materials. This group includes the Tiu floor (CTF) and dark plains (DP). Plains of several morphological types (CF,_2 and CFs) are distinguished in the CTF area. The Tiu Vallis floor (CTF) is spread over a greater area of the studied region than that of the Ares Vallis floor. It exhibits two morphological varieties of plains characteristic of the latter smooth, more spacious plains CF, and hilly plains CF2. Plains of a new type, netted at the surface and sponed with bright depres- sions (CFj, Fig. 13), are also distinguished in the Tiu Vallis floor. These plains are spotted in images and dis- play distinct bright depressions of irregular or elon- gated shape and also dim bright spots. The diameter of the depressions varies from hundreds of meters to 20 km. They are about 20 m deep (Golombek et al., 1997b) and may represent a result of fluvial or eolian erosion that removed the dark upper horizon and exposed the underlying bright material. The latter may also correspond to younger eolian products accumu- SOLAR SYSTEM RESEARCH Vol. 32 No. 6 1998 GEOLOGY OF THE COMMON MOUTH OF THE ARES AND TTU VALLES, MARS 437 lated here (Kotmtsu et al., 199S). Eolian sand is thought to represent also the material of the dark upper layer (Edgett and Christensen, 1994; Kuzmin and Gree- ley, 1995a, 1995b). In the opinion ofTanaka(1988), the nuaterial of these plains represents debris-flow deposits. Locally, the CFj plains incorporate mesa-type rises of bright material up to S km in diameter (images 006A19 and 003A20). One of them, decorated with a caldera- like depression, is less than 100 m high, according to our measurements. Rises of this type may be mesas of deltaic pitted plains (PP). An area of chaotic valley-like depressions is clearly seen within the CFj plain in images 006A17 and 0O6A34. These iandforms presum- ably appeared as a result of the thawing of the perma- frost zones. In some places of CF, plains, bright depres- sions are located close to teardrop-shaped islands and are elongated in the same direction, thus being formed by the same flows (images 003A50 and 004A17). Other bright lineaments, which may mark scarps or fractures, are discordant relative to the flow direction (images 004A16 and 004A79). Thus, the origin of CFj plains is not yet clear. They might be the result of the particular- ities of fluvial erosion that operated in these areas, and from subsequent thermokarst and eolian reworking. The Tiu Vallis is incised into the Ares Vallis floor, as well as in several other units, such as knobby plateaus (KP) and plains of the ridged (RP), spotted (StP), smooth (SP), and bright (PP) types. The fractured-plain material (FP, see below) lies over the Tiu Vallis floor (CTF, Fig. 5). According to the crater-counting results, the Tiu Vallis floor is Late Hesperian-Early Amazonian in age (Fig. 3). Dark plains (DP, image 24A55) are located down- stream of a bench composed of pitted-plain (PP) mate- rial. It is possible that these plains (DP) represent only a part of vast fluvial plains located in the mouth of the Tm and Simud Valles. Morphological varieties CF[ and CF2 of the Ares Vallis floor (see above) are also charac- teristic of the dark plains (DP). As the surface of the lat- ter is darker than the Tiu Vallis floor at the same lati- tude, it appears to be covered with a dark eolian mate- rial (in general, the further north, the darker the surface of the region). This situation is also seen in the Wahoo Crater adjacent on the east and displaying a dark floor (in the region at hand, dark streaks are situated east of the craters; see section 2.6). The northern dark plains seem to be incised mto the material of spotted and smooth plains; they are overlain, in turn, by the material of frac- tured plains (image 008A51 ) and are older than it. Diachronous fluvial isIaDds and benches (IB, Fig. 11). Fluvial islands and benches everywhere are in contact with the valley-floor materials. South of 25? N, the slopes of islands and benches are steep or stepwtse (morphological variety IB,). Near 20? N, 30.5" W, Kuzmin and Greeley (1995a, 1995b) detected eight slope steps, and near 23" N, 30? W, we distinguish four steps. Northward of 25? N, the slopes of islands and rPP Fig. II. Erosion Ttmnants (islands) composed of the pitted- plain material (PP) as seen in Wking Orbiterimags 003A24. The valley floor is decorated with many hills (CF?). Flg. 12. Fractured monotonous valley floor (CF4) as seen in Viking Orbiter image 864A12. benches become less steep and steps are lacking (?mage 670A16, morphological variety IBj), Features of the IB, type appear to be formed by the incision of valley-forming flows (Baker, 1982). Two mechanisms for step formation were suggested. First (Kuzmin and Greeley, 1995a, 1995b), this may be a result of the erosion of layered deposits left by previous flood events and partially duhcrusted (Treiman, 1995, SOLAR SYSnSM RESEARCH Vol. 32 No. 6 1998 438 MARCHENKO et al. ,^\fi,:^M Flg, 13. Fractured valley floor with bright cauldrons (CF5) K Km in Vf/t?ij Orbittr image 004A62. Fig, 14. Fractured plains (FP) with confocmabty striking ridges and fractures as seen in VUdng Orbiter image 632A12, 1997), Second, they may reflect the pulsating incision of flows (De Hon, 1989) into a homogeneous, nonlay- eied ground. We measured the hypsometric parameters of two typical islands crowned with the Bok and Gold craters and of a small island situated in between. The slopes are stepwise at the rear side of the islands and steep at their front side, where the top surface is bordered by escarpments. The fiat top surface of the islands is ele- vued 30-60 m above the neaiby plains. Baker (19S2) believes that the escarpments are characteristic of those island tops which were partially affected by the last flood event. As for the northern islands and benches displaying smooth, gentle slopes (morphological variety IB2), they could be formed in the zone of accumulation and weak erosion in the valley mouth. According to Baker (1982), their gentle slopes mean that the isiani? were con^iletely floodied during the last stage of fluvial activity. It is also likely that islands and benches were taking shape during all episodes of catastrophic floods, when flows incised into t?ie material of plains and plateaux. In this case, their material represents a mixture of eroded rocks and sediments deposited by the flows. Benches of the IB type are inset into ridged (RP), spotted (StP), bright (PP), and, possibly, smooth (SP) plains. The material of fractured plains (FP) locally overlies the basal horizons of the northernmost islands (image 669A93), thus being younger than the IB unit. 2.4. Fractured Plains (FP) Postdating Valley Formation Fractured plains characterize areas displaying nar- row fractures, trenches, and ridges in their central parts (Figs. 5 and 14). Fractures and ridges are about a half- kilometer wide, whereas trenches arc up to 2-4 km across. The length of these features ranges from a few to 40-50 km. Inside the fractured plains (FP), hills comparable in size with those of the H-type are sur- rounded by benches. Ridges and fractures are often ori- ented conformably (Fig, 14), or the latter may cross the former. In addition, some fractures cross ?jecta of a young impact crater (Fig. 15), whose age is estimated to be 1.5 Gyr on the basis of counting results for smaller impaa craters at its surface. Benches around the hills were classed with either wave-abrasion platforms (Parker et al., 1993), or slope-subsidence forms (Luc- chitta, 1984). The observed orientation of fractures and ridges, longitudinal to transverse relative to the valleys, suggests they might be formed in the course of exsicca- tion, sagging, and freezing-thawing cycles, which occurred in combination or separately in flie water-sat- urated material (Jons, 19S6; McGill and Knobs, 1992; Parker et al., 1993). Ridges may represent a result of the water-rich material squeezing from below, when the upper layer was already frozen (Tanaka, 1997). Luc- chitta et al., (1986) argue for a certain similarity between these forms and some ridges of the Antarctic shelf and suspect they originated under subglacial envi- ronments. Other scientists aigue for coastal and fluvial origins of the ridges (Parker et ai, 1993; Scott, 1982). Between the FP and valley-floor materials, there are frequent lobate flows, whose orienta?on suggests that they penetrated into the Ares and Hu Valles from the north. Jons (1984,1985,1986) related them in origin to the material incursions from a hypothetical mud ocean that existed, according to his opinion, in the northern SOLAR SYSTEM RESEARCH Vol. 32 No. 6 1998 GEOLOGY OF THE COMMON MOUTH OF THE ARES AND ITU VALLES, MARS 439 plains. Tanaka (1995, 1997) suggested that this male- rial (tike that of the valley floors) was transported from the plains by catastrophic floods, whose frontal parts rose up the slopes by inertia and then rushed back. Parker et al., (1993) interpreted these lobate flows as analogues of teirestrial alluvial fans of subaquatic mud- flows. Under Martian conditions, these features could be produced when flows saturated with a clastic mate- rial rushed into an ancient sea or ocean. Thus, the FP material is of unclear origin. In our opinion, it is sedimentary and originally was saturated with water or ice (Lucchitia ero/., 1986; Tanaka, 1995). According to crBter-counting results, this material is of the Late Hesperian-Early Amazonian age. The FP material was formed either at the stage of fluvial activ- ity that deposited the Tiu floor sediments, or, more likely, later, when a catastrophic flood event brought the material from another valley located somewhere nearby. Only the eolian deposits and fresh impact cra- ters are younger than the FP material. 2,5, Materials of Impact Critters Impact craters (Fig. 7) formed during the whole geo- logical history of the region. In the map (Fig. lb), we depicted 76 craters with diameters greater than 4-5 km; 17 ofthem are more than I6kmacross, and4are more than 40 km across (the largest, the IClpini Crater, is 70 km in diameter). Some craters are relatively fresh; others, heavily destroyed. The former dominate among smaller craters, whereas larger ones are frequently ruined. We divided the crater materials into the ancient-crater complex (ICo), rim-and-floor complex of young craters (IC), and complexes of dry (ICd) and fluidized (ICO ?jecta. Craters with fluidized ?jecta indicating the pres- ence of solid HjO and other volatiles in the substratum (Carr et al., 1977) are detected almost everywhere. Costard (1994) counted the number of craters with flu- idized ?jecta per 5? x 5" cell for a vast area covering our region toge?ier with others. According to his evalua- tion, craters of this type represent from 13 to 53% of the crater population in a cell. Figure 2 shows the stratigraphie position of two large craters: Libertad (IC and ICf subunits) and Wahoo (ICo unit). Ejecta from the Libertad Crater overlie the FC, SP, and StP deposits, whereas the younger Tiu Val- lis is incised into these ?jecta. Consequently, the impact event was intermediate in time between the first and the last stage of ?uvial activity. The same situation is char- acteristic of a nameless adjacent crater located north- east: its ?jecta rest on the FC material, being overlain, in turn, by ?jecta from the Libertad Crater. TTic Wahoo Crater apparently escaped the Late Noachian stage of intense denudation. Its ?jecta underlie ancient fluvial deposits (SP and StP), and the crater formation thus predates the accumulation period of the latter. The Soochow and Zuni craters (IC and ICf subunits) are superposed onto the ridged plains (RP) and seem to have been formed after them, but prior to the termina- F^ 15. Fractured plains (FP) as seen in \^kmg Orbiter image 669A69. A fresh impact crater is located in the lower rigiit corner and a young fracture in the center cuts the ?jecta from this crater. tion of erosion in the Ares Vallis, because they are eroded. The Concord and Ore craters (IC and ICf sub- units) may have been formed between two stages of flu- vial activity in the Ares and Tiu Valles, since they are located on the erosion surface of benches, and in turn, were eroded by valley-forming flows. In addition to primary impact craters, we detected small secondary craters in the region, which appeared after impacts of ?jecta from the primary impact struc- tures. The secondary craters are concentrated in the meridional zone between 18*-24'' N and 31 "-33? W. Kuz'min (1996) believes that they were produced by ?jecta from the 50-km-wide crater distinct in morphol- ogy and located at 7.5? N, 32.9? W. 2.6. Eolian Materials Eolian materials (Figs. Id and le) are the constitu- ents of several features of the Martian surface, such as: bright wind streaks (Wsb), which extend southwest- ward and range from a few to 100 km wide and from a few to 50 km long, depending on the size of the screen behind which they accumulated; dark areas (Wsd), which are from a few to 2(X) km across and represent wind streaks extending firom west to east and in the west-southwestern direction, and also dark deposits on the floors of valleys and of the largest impact craters, in their southwestern sectors; and bright to dark, narrow wind streaks (Wsn, image 524A26), which are variably oriented and up to a few tens of kilometers long and wide. The last features may indicate a diversity of wind SOLAR SYSnSM RESEARCH Vol. 32 No. 6 1998 440 MARCHENKO t? al. Flg. 14?, Mosaicof?nages shown in Figs. 4-15. diiec?ons, as they were formed in different epochs. In the region considered, we also detected two areas where elongated ridges resembling yardangs are up to a few kilometers long and e;(tend either south-south- west (the Wahoo Crater floor, image 34A90) or west- southwest (?jecta of the Kipini Crater, image 669A86). In our opinion, these are purely erosional, but not litho- logie, formations. The dark eolian material most likely corresponds to sand that was derived from the Acidalia Planitia and covered the valley floors (Edgett and Christensen, 1994; Kuzmin and Greeley, 1995a). If this assumption is correct, the abundance of sand may grow northward. Actually, the color images of the region show that, in contrast to southern areas, all materials located north of 20"^ N, with the exception of a few crater rims and bright streaks, are dark and bluish at the surface, which is characteristic of eolian sand (see below). The tow resolution of images did not allow us to detect eolian dunes in the region, where they might be widespread. Small dtmes were pictured direcdy at the Mars Pathfinder landing site (Matijevic et ai, 1997). Many dark spots inside craters may also correspond to sandy dunes (Arvidson, 1974), Rice and Edgett (1997) believe that the entire northern part of the region (the distribution area of FP and SP materials) is covered with sandy deposits of catastrophic outflows, which have experienced eolian reworking and incorporate small, isolated dunes. The elongated dark splotches in the low-resolution images presumably mark dime clus- ters located nonh of 27? N, within the spotted and frac- tured plans. Wind streaks formed by barchans, which are well known on the Earth and on Mars, may also occur in the region under consideration. In the stratigraphie aspect, the eolian materials are younger than the surface they overiie. 2.7. Stratigraphie and Age Relations between the Units The revealed stratigraphie succession of material units and respective landforms can be used as an indi- cator of principal stages in the geological history of the region. The first events at the surface of the ancient highland were the destruction of the cratered plateau (CP) and the formation of the ridged plains (RP). Ejecta from the large Zuni and Soochow impact craters overlie the ridged plains but are eroded by flows that created the Ares Vallis, and this indicates that the timespan between the ridged-plain formation and the next resur- facing stage was signiflcant, because great impact events are fairly rare. Our data, and those of others (Kuzmin and Greeley, 1995a, 1995b; Nelson and Gree- ley, 1996, 1998; Rotto and Tanaka, 1995; Tanaka, 1995), suggest that catastrophic floods in the Are&Tiu mouth were recurrent. In our opinion, this is evident, first of all, from several generations of fluvial deposits and landforms, which are incised into one another, suc- cessively overlie one another, and were, therefore, formed at different stages of erosion and accumulation. Faint fluvial channels and islands (PC) along with spot- ted (StP) and smooth (SP) plains originated first, prior to the ^pearance of major valleys. The stratigraphie position of the Libertad and another nameless crater located nearby (see above) shows that both impact events took place between the first and the third outflow and. consequently, these outflows were separated by a considerable time interval. The second outflow (possi- bly a simple continuation of the first one) was respon- sible for the erosion of older deposits, created the Ares floor surface (CFA), and deposited the material of the pitted plains (PP). The escarpment between the spotted and pitted plains might be formed by the incision of the valley into its own deposits. The third outflow from the Tiu Vallis eroded these deposits and the Ares Vallis floor; however, it is difficult to determine the length of time between this and the second outflow. Anyway, the hypothesis of two catastrophic flood events separated in time seems to be quite reasonable. The fractured plain material (FP) lies over the young fluvial deposits of the Tiu Vallis floor (CIT) and dark plains (DP), being prob- ably even yotmger. In summary, we may suggest the following sequence of geological events in the region: the fonna- SOLAR SYSTEM RESEARCH Vbl. 32 No, 6 199S GEOLOGY OF THE COMMON MOUTH OF THE ARES AND TTU VALLES, MARS 441 tion of the cratered-plateau material (CP) ?*- its ero- sion and the partial accumulation of the r?dged-plain nnateria] (RP) above the plateau ?? the first cau- strophic flood, which left behind the material of small islands and valleys (CF) and the deltaic deposits of smooth (SP) and spotted (StP) plains ?- the second flood, which deposited the material of the Ares Vallis floor (CFA) and pitted plains (PP) ??- the thitd flood and (he fontution of the Tiu floor deposits (CTF) ?* the appearance of the fractured-plain material (FP). Eolian processes and impact cratering at the Martian surface have apparently lasted throughout the geologi- cal history of ?ie region. 3. SURFACE CHARACTERISTICS INFERRED FROM IMAGES OBTAINED WITH DIFFERENT FILTERS AND FROM THERMAL-INERTIA DATA 3.1. Surface Color It is well Icnown that various regions of the Martian surface are colored dark-gray or bright- to dark-red, presumably depending on the local composibon of the surface material, (Arvidson ei ai, 1989; Mustard, 1995). As assumed, the dark-gray areas mark cither rel- atively fresh outcrops of mane rocks and sand or very thin, palagonite-like dust covers overlying them. The bright-red areas are assumed to be covered with oxi- dized palagonite-like material, including eolian dust The dark-red surface is thought to characterize a mix- ture of a dark-gray and a bright-red material or to con- sist of immature weathering products formed over a basaltic basement (Mustard, 1995). In Older to reveal the characteristics of distinguished nnaterial units, we studied images obtained with differ- ent Alters, analyzing the brighmess ratio between the red and violet spectral ranges and correlating this parameter with a thermal-inertia map (Christcnsen and Kieffer, 1989). The same approach was used by Soder- blom ei al., (1978) and Arvidson et ai, (1982), who studied other areas. The original images of the same area (Sodeiblom et ai, 1978) were made by the \^king Orbiter with three filters; violet (0.45 ? 0.03 fim, image 666A52), green (0.53 ? 0.05 |jm, image 666A56), and red (0.59 {jm, image 666A58). These images allowed us to char- acterize most units distinguished by us, although their resolution is low, only 0.82 km per pixel. The bright- ness-distribution histogram for the red spectral range shows two distinct peaks, one corresponding to surface areas located north of 20? N and another, south of this latitude. This may indicate that materials in the two parts of the region are different in composition. In order to analyze the images in greater detail, we selected 48 small areas in which the surface color of one or another unit is approximately the same over a mosaic of the red, green, and violet images superimposed. Some units with varying surface color were studied in their several areas. For instance, we investigated five areas of the Tiu Vallis floor notably contrasting in color one to another. For each of the studied areas, we measured the radiance factors in all the three spectral ranges, recon- structed their spectra, and determined the radiance ratio between the red and the violet spectral range using a technique described by Arvidson et ai, ( 1982). In the plot illustrating the variation of the red-range radiance factor versus the red-to-violet brightness ratio (Fig. 17), one can distinguish several color groups of mattrials. Tlie darkest blue material is characteristic of deposits in the impact craters and northern areas of the valley floors. This material most likely corresponds to eolian deposits, and the dark color suggests that they arc nnafic in composition. The bright-red color is typi- cal of wind stre^ and plateaux located in the south and may characterize eolian dust. The dark-red areas in the south of the Ares and Tiu Valles and in the neighbor- ing plateaux may indicate a duricrusted soil or a mix- ture of dark and bright materials (Arvidson et ai, 1989; Mustard, 1995). In Fig, 17, the material that is in an intermediate position between these groups of materi- als may characterize a surface composed of two or three components. Three principal color groups are per- fectly distinct in the plot of three-component spectra (Fig, 18). In the violet spectral range, their radiance fac- tors are almost identical, whereas the dark-gray mate- rial is least bright in the green and the red range. The bright- and dark-red materials are also different in these spectral bands. Our analysts also icvealed that the rims of aiK?ent impact craters are colored brighter in red than surrounding areas, as was first noted by Soderblom et al., (1978). 3.2 Thermal-Inertia Dala The thermal-inertia parameters of the ground may be used for the evaluation of the material-grain size (Edgett and Christensen, 1997). Our analysis of these parameters was based on the thermal-inertia maps pub- lished by Christensen and Kieffer (1989) and Edgett and Christensen (1997). In the first map, the thermal inertia is represented by contours with an interval of 10"^ cal/(cm^ s*^* K), and in the second, as discrete val- ues with a resolution of 0.5? of latitude and longimde. In the region smdied, the magnitude of thermal inertia, by and large, increase from south to north. Accordingly, on the basis of correlation between these values and the grain size of the material (Edgen and Christensen, 1994), we may assume that medium-giained sand is dominant in the south, whereas coarse sand is wide- spread in the north. However, this general trend is irrel- evant in two situations. ( 1) In the south, the thermal-inertia magnitude in the Ares Vallis ground is greater than in the nearby pla- teaux (CP and KP) and plains (RP). This indicates that the valley floor is covered with coarse or very coarse sand, whereas medium- to coarse-grained sands are widespread in plains surrounding ?ie incised valley. SOLAR SYSTEM RESEARCH Vol. 32 No. 6 1998 442 MARCHENKO et al. 0.20 fe 0.18 I o I 0.16 S O? 0.14 0.12 Bii^tRed. "Bright" toot Q . South b[ ''r'?. f p : DiikR?l E)ar?cGT?y Nonb 2X1 2.4 2.8 Red/Violet ratio 3.2 ^ [>?il eoliui ?pM( ^ witti?) inqMct traten , Dark ?pou within ' dutUKl floon C Crucrnoon p 0?(etrimt tDdejtcu JP FUctuTcd PUini blaod within Fncnired PlAin* I Bright touki within FnctuRd Pliint f Clunneli'flocn b^ Bright itRttu wd spott ^ wimin chkiBK] floon P Pined PUim r Spotted Ptiins g SswotbRUtis Brigbicraierhaloet * within Smooth Plaiii* ^ Ridged Pliins . Bright Ptuewi ' M?tni?l PDailt PUietu Material ence between eoliao and surrounding deposits is insig> ni?cant. 3.3. Rock Abundance Using the thermal-mapping data by the taking Orbiter, Chnsten&en {1982, 1986) estimated the rela- tive abundance of rock fragments greater than 10 cm across occurring on the surface. His results arc repre- sented in a digitei] map with a resolution of I " in latitude and longitude, which comprises the region of our study. In the tegion, this material is distributed inegularly: rock fragments occupy 0-5 to 21 -25% of a given reso- lution element Rocks aie quite abundant in the spotted (StP). bright (PP), and daric (DP) plains. Their esti- mated abundajice is decreased inside the two largest dark areas of eoUan {WSd) material (29? N. 34? W and 28.5? N, 28.5? W; Figs. 1 d, 1 c); in a dark streak located at 26? N. 35.5? W; and in floor areas of the Kipini, Lib- ertad, and the nameless crater situated at 15? N, 27? W. Within frequent narrow, bright streaks, rock fragments are locally more abundant (1^-25%), and these areas seem to be heavily affected by deflation. Broad, bright streaks show a variable density of rock-fragment popu- lations and are interpreted as areas of a weak deflation and dust accumulation, which do not alfect this density. A moderate to low abundance of rock fragments is characteristic of cratered (CP) and knobby (KP) pla- teaux and of ridged plains (RP). F^ 17. Red-ratlinDce factor versus the red-to-violet ratio for different areas of the surface. Areas situated ttear the Mftrs Pathfinder landing site are marked by Ujtiares. This situation, noted ?rst by Betts and Murray (1993), may reflect either the origmally coarser grain size of fluvial deposits or be a consequence of deflation diat was most active in the valley and removed the fine- grained material from the fluvial sediments. (2) Spotted (StP), dark (DP), and bright (PP) plains located approximately in the center of the region dis- play abnormally high thermal-inertia magnitudes of (12-15) X 10-^ cal/(cm^ s" K), characteristic of coarse sand or gravel (Edgett and Christcnsen, 1994). It is pos- sible that fine particles were removed from the surface of these plains, or they were originally composed of materials with these granulometric parameters. On the Earth, such coarse deposits are characteristic, for exam- ple, of midfan facies of catastrophic flows (Rice and Edgett, 1997). As for eolian deposits, which seem to be present in the region, they have no indications in the thermal-iner- tia map by Edgett and Christensen (1997). Possible explanations for this fact are as follows: the thermal mapping was of insufGcient resolution; the bright dust covers are thinner than the layer, a few centimeters thick, affected by diurnal thermal fluctuations in the ground (Arvidson ?t ai, 1989); or the grain-size dtffer- 3.4. A Comparison of the Surface Color, Thermal-Inertia Daia, and Rock Abundance A comparison of these parameters allows us to dis- tinguish three types of areas in the region: gray to dark- gray areas with high thermal-inertia nutgnitudes and variable amount of rock fragments; red to dark-red areas marked by low to moderate thermal-inertia mag- nitudes and irregular distribution of rock fragments; and red areas displaying high thermal-inertia magni- tudes and abundatice in rock fragments. Gray to dark-gray material of the surface cone- sponds to the coarse sand and gravel (Fig. 19) and asso- ciated rock fragments distributed unevenly. According to their morphology, these dark streaks and splotches are attributed to sand-accumulation areas. It is also likely that dark eolian deposits in the Libertad, Wahoo, and Kipini craters predominantly consist of sand, because the estimated rock abundance here is a mini- mum. Sandy eolian deposits of mafic composition may also be typical of dark splotehes in the valley floors and in the fractured aik-grey Wahoo) violet green red Ft^ 18. TtiTM-compoiieiit spectra for surfac? mater??t most contFuting in color. Hesperian-Early Amazonian for fractured plains. Such uncertainties may be due to the effect of subsequent resurfacing in a given area, the partial preservation of older crater populations, and the mere stochastic varia- tions in the crater density. In most cases, the areas of distinguished imits are too small, and the number of craters exceeding 2 km in diameter is insufficient to reliably define the unit posi- tion in the Martian stratigraphie scale. To better under- stand age relations in this case, we compared the count- ing results for craters greater than 1 km in diameter (they are detectable in all stratigraphie units) between different units and then verified these relations, based on statistical results with age relations obtained from a stratigraphie analysis. This approach is used below to elucidate the principal resurfacing stages in the region. 5. PRINCIPAL RESURFACING STAGES IN THE REGION The age succession of stratigraphie units distin- guished and mapped in the region was used as a basis to depict the most significant resurfacing stages. This succession is substantiated by the revealed stratigraphie reliuions between surface materials and, independently, by the results of the impact-crater counting. Final results of this combined approach are shown in Fig. 21. At the left of Fig 21, we plotted the crater-counting results for all stratigraphie units except the relic hills (H), rings of knobs (RK), and faint fluvial channels and islands (FC), the areas of which are too small and dis- play an insufficient number of craters. This plot also demonstrates our counting results for two complexes of cratered (NPL,) and eroded (NPL2) plateaux distin- guished by Rotto and Tanaka (1995) outside the area of our map. The crater density for old cratered plateaux (CP) or knobby plateaux (KP) may be underesdmated SOLAR SYSnSM RESEARCH Vol. 32 No. 6 1998 444 MARCHENKO et al. Termal inertia .y 14 h 12 10 8- f ? Ubenid f r 2^ 2.4 2.8 RedA^iolet ratio 3a F^ 19. Vw?Mioiu in the ihermaJ-inertia tnagnitudes [to units 10"^ eal/(cin^ s**' K); after Chmtensen and Kief?cr (1989)] venu? red-u>-violei br?ghmets ratiot for difTereot areai of ibe surface (symbols as in Fig. 17; indtcsted rock typet from Edgeti and Chriswnwn, 1994). because of the preferential destruction of small craters within these long-existing complexes. Along the abscissa axis, units are arranged in their stratigraphie succession (Fig. 2) and divided into three groups (from left to right); (1) rims and ?jecta of large impact craters; (2) fluvial deposits (StP. SP. PP. IB. CFA, CFT, DP) and fractured plains (FP); and (3) val- ley-predating units (NPL,, NPLj, CP. KP, RP). At the right of Fig. 21, the relative age sequence of subdivisions also corresponds to that shown in Fig. 2. The revealed stratigraphie relations between the units suggest that the main stages of the Martian sur- face reworking were the erosion in cratered plateaux and three catastrophic floods in the valleys (see above). The stratigraphie relations cannot elucidate, however, how long the timespans between these stages were. At the same time, using the crater-counting data, we may evaluate the intervals separating different geologic events (the sequence of which is suggested by the suc- cession of material units; see the right side of the plot) in terms of impact-crater densities acctunulated on a given area after its last resurfacing (left side of the plot). This evaluation is possible, of course, only if the inter- vals were suf?ciently long to ensure a statistically meaningful difference between the densities of crater populations. In Fig 21, the suges of resurfacing inferred nom stratigraphie relations and expressed in units of the cumuUuive crater density are marked by gray horizontal bands. The age position of stages (their interval in the scaled left ordinate) is estimated as described below. Stage I: Erosion of the Cratered Plateau (CP) and Formation of Ridged Plains (RP) Judging ftom the stratigraphie relations between the cratered plateaux and ridged plains, the former are older than the iMtcr. Consequently, the band marking this stage should be placed above the crater-density value for the plateau (with due account for errors) but below the value characterizing the ridged plains. In view of a high error in one of two values estimated for the ridged plains, the crater density for the ?rst band is assumed to be equal to that vnthin the overlapping range of two estimates obtained for ridged plains (approximately 3300-4200 craters greater than 1 km in diameter per 10^ km^). Stage II: Fluvial Reworking Postdating the Ridged Plains but Predating the Incision of the Ares and Tiu Valles The stratigraphie relations (the right part of Fig. 21) show that the oldest fluvial deposits of the region that created the spotted (StP) and smooth (SP) plains were acctunulated a?er the ridged-ptain formation. The for- mation of islands and benches (IB) in the Ares and Tiu Valles Boors presumably corrunenced at the same time. Accordingly, the horizontal band (second nom below) marking the second stage of resurfacing and the f?rst stage of valley incision is placed at the level corre- sponding to the overlapping range of crater-density val- ues for ancient fluvial plains (StP and SP) but above the value characteristic of ridged plains (RP), i.e., within the interval of 2700-3300 craters greater than 1 km in diameter per 10^ km^. In this case, one of the SP and one of the IB area remain below this interval. Judging from the abundance of partially buried and eroded cra- ters in these areas, this departure in the plot may indi- cate that they incorporate craters originated on an older surface (CP, KP, or RP). Another SP area plots above the indicated density band. However, it is located near the vague boundary with the younger FP surface and may thus belong to it The time interval between the first and the second stage is probably short. We show it tentatively, without attributing any age signi?cance to its width. Stage III: The Incision of the Ares Valtis and Accumulation of PP material Our stratigraphie analysis indicates that the Ares Vallis is incised into ancient fluvial deposits (SP, StP, FC). The deposits of pitted plains (PP) might be formed at the same time, although the PP surface could be heavily eroded by younger flows, which ciestioyed some craters and "rejuvenated" the surface. Accord- ingly, the horizontal band marking the age position of this stage is plotted at the level of the CFA unit above the previous stage (the range of 2200-'27(X) craters grcmer than 1 km in diameter per 10' km^). Apparently, SOLAR SYSTEM RESEARCH Vol. 32 No. 6 1998 GEOLOGY OF THE COMMON MOUTH OF THE ARES AND ITU VALLES, MARS 445 the time gap between the second and the third stage was also not long. Stage IV: The Tiu Vallis Incision The crater-density band for this stage is bordered by the overlapping level of the values estimated for the CTF and DP fluvial materials deposited after the for- mation of the CFA and PP surface areas. In addition, this band should be placed below the crater-density val- ues for the fractured plains (FP), whose niaterial over- lies the CTF and DP areas, i.e., within the range of 1300-1400 craters greater than 1 km in diameter per 10* km^. We do not reject the fact that all three groups of material insignificantly differ in age. One area of fractured plains (FP) turns out to be b?low this band, because, as we think, some ancient craters remained unburied here by a thin sedimentary cover characteris- tic of this area. This situation explains why the crater- density values are greater here and the area looks older than other surface areas reworked at the last stage. We should also add that all suges of impact crauring were paralleled by eolian processes, as at the present. Thus, the combined analysis of stratigraphie rela- tions, together with the results of impact-crater count- ing (Fig. 21). reveals four principal stages in the evolu- tion of the Martian surface in the Ares-Tiu mouth. Three of them might be separated by perceptible timespans. These are suges I and III, separated by impact events that formed the Zuni and Soochow cra- tets; stages II and IV, limiting the formation interval of the Libertad and the nameless crater; and stages III and rV, because the surface of their materials is character- ized by notably different values of cumulative crater density. 6. AGE ESTIMATES FOR STRATIGRAPHIC UNITS AND RESURFACING STAGES The absolute ages of the distinguished stratigraphie units were estimated by means of calibrating the crater- counting results with the use of the reference curve of crater density versus the crater-accumulation time as obtained by Neukum and Hiller (1981). As they dem- onstrated, the characteristic inflection points at the curve mark the episodes of the most intense destruction of craters, i.e., the resurfacing stages of the Martian siu"- face. When applied to our crater-density curves charac- terizing the studied areas, this approach allows us to evaluate the absolute model ages for the formation and resurfacing stages of the distinguished stratigraphie units. Our results are presented in the table and illus- trated in Figs. 22 and 23. The estimated age values are highly dispersed for many individual areas (table); moreover, for the stud- ied region as a whole they vary even more, from 4.25 to 0.5 Gyr (Fig. 22), This dispersion apparently reflects the combined effect of two factors: the real existence of several discrete episodes in the resurfacing history Fig, 20. Areas selected for impaci-crater counting (num- beied as in the table). (inflection points at the curves) and stochastic varia- tions related to the high random errors caused by a small volume of data samples. The wide error intervals for the calculated crater- density values (Fig. 3) result in great uncertainties of age determinations for particular events. We believe, however, that these random errors should be statisti- cally compensated. If this is correct, then individual age values for a given event should form clusters surround- ing the true value on the time axis. Actually, in Fig. 22 we can distinguish a group of data points clustering around 3.7 Gyr (it is delimited by a dashed line and includes all valleys and older areas) and another group less distinct and concentrated around 1.5 Gyr. The his- togram plotted for the same data (Fig. 23) demonstrates their bimodal or even polymodal structure more clearly. In the histogram, the oltter, first group of values demon- strates a distinct peak in the range 3.75-4.0 Gyr, whereas in the younger, second group there are three minor peaks marking intervals of 2.25-2.0, 1,75-1.5, and 1,0-0.75 Gyr. In our opinion, these peaks delimit the age ranges for the resurfacing episodes in the region. SOLAR SYSTEM RESEARCH Vol. 32 No. 6 1998 446 MARCHENKO et al. N^ 10* km^; D > 1 km 1000 younger S older 10000 Wiboo i 1 J ?^, KipLnJ 25N,27W Craters Craters Ares/Tiu materials Older materials Arcs/Till materials Older materials Fig. 21. Stages of resuifacing in the region. Left: ordiaates are the cumulative densities of craters greater than 1 km in diameter (solid diamonds); vertical ban show starK?ard deviations; extrapolated values are shown by open diamonds. Right sick of the diagram (light gray rectangles) is given according to Fig. 2. Horizontal hands mark the principal stages of resurfacing. Figure 22 also demonstrates that age estimates for all areas of the oldest stratigraphie units (KP, CP, NPL,, NPLi) plot within the oldest interval of the first group (4.25-3.75 Gyr). Values for complex IB, fall within only this group. Many other data characterizing either younger areas or those affected by later resurfacing processes (StP, SP PP, CFA, CFT, FP) plot within both the older and the younger groups, whereas all estimates for dark plains (DP) belong only to the second group. In the case of ridged plains (RP), which seem to be in the intermediate stratigraphie position between ancient and younger areas, three quite similar values are in the center of the first group, and one falls between this and the second group. We interpret this diversity in age detenninations as a combination of values characteriz- ing the mapped stratigraphie unit proper, the older underlying materials, and Ute later resurfacing events? for instance, by persistently active eolian accumulation and erosion, which destroy small craters. For the oldest surface areas (NPL,, NPLj, CP, KP), the crater count turned out to be possible only in low- resolution images. Accordingly, son? craters could remain undetected, and age determinations in this case could be younger than the true age of the unit. Another disadvantage of these images is an a fortiori deficiency in the quantity of small craters; consequently, it is impossible to date the resurfacing episodes using the inflection points of the cumulative crater-density curves. In fact, in the second age group, data related to the oldest stratigraphie units are laclcing as if these units did not experience later transformations. The cratered plateau (CP) representing the oldest surface area in the region was apparently formed about 4 Gyr ago; its surface was eroded 3.8-3.7 Gyr ago and was partially overlapped by the ridged plain (RP) mate- rial about 3.7-3.6 Gyr ago. It is difficult to state forcer- tain what peaks in the right part of the histogram (Fig. 23) may mark the particular stages of fluvial activity described above. Tliese stages may be restricted to intervals of 3.6-3.5, 3.5-3.3, 3.0-2.6, 2.3- 1.9, and 1.6-1.4 Gyr. The fractured plains (FP) could have been created 1.4-0.6 Gyr ago. The youngest resur- facing stages may be related to episodes of eolian activ- ity and local thermokarst reworking. The ages of the Ares and Tiu Valles were estimated to correspond to 3.65-2.5 and 3.5-2.0 Gyr (Neukum and Hiller, 1981). Using the same approach to the interpreta- tion of the crater-density distribution, Robinson et al., (1996) concluded that some areas of the Ares Vallis were inainly formed prior to 3.5 Gyr, and their later reworking took place between 2.21 and 1.66 Gyr. Our age estimates characterizing the formation period and subsequent resurfacing stages of the Ares and Tiu Valles are thus concoidant in general with the quoted data. In summary, let us compare the dating results with the outlined stratigraphie relations. According to the stratigraphie data, ?ie studied region went through sev- eral stages of resurfacing. The estimated absolute ages of individual stratigraphie units suggest this scenario, virtually, for each of them. 7. A RECONSTRUCTION OF THE GEOLOGICAL HISTORY OF THE REGION The results of this study and the data published before allow us to atg:ue for the following scenario of geological history in the Ares and Tiu Valles region. SOLAR SYSTEM RESEARCH \tol. 32 No. 6 1998 GEOLOGY OF THE COMMON MOUTH OF THE ARES AND ITU VALLES, MARS 447 Gyr !? 2- 3- 4- yi t Craters Fluvial units Old plateaux and plains Fig. 22. EsiiiDated absoluie-age values (solid circles) for the distinguished stratigraphie units, arranged along the horizontal axis in the same succession as in Figs. 2and2LInside the vertical columns, dots corresponding to the same area are placed strongly above one another, dots aie slightly displaced in the horizontal direction with respect to one another to denote different areas of the same unit The lithologie unit of the cratercd plateau (CP) was formed here, as in many other regions of Mars, diuing the Noachian time (about 4 Gyr ago; see Fig 24a). This plateau is thought to be composed of impact breccias and ancient lavas (Rotto and Tanaka, 199S). The Ares and Tiu Valles open into the impact depression of the Chiyse Planitia impact basin that was also formed at the Noachian time (Schultz et o/., 1982). A consider- able subsidence in the northern hemisphere of Mars and the formation of northern lowlands is attributed to the end of the Noachian period (McGill and Dimic?ou, 1990). In the studied region, these events created an escarpment at the northern boundary of the cratercd plateau, and this structural feature subsequently retreated southward imder the influence of slope-col' lapse events and water-saturated mudflows. As a result, the ancient cratered plateau was under destruction in the north, leaving behind the knob-plateau unit (KP), hills (H), and remnants of ancient craters in the form of rings of knobs (RtC). This was resurfacing stage I of the region. It is likely that, during this time (3.8-3.7 Gyr ago), there were aiso fluvial valleys or their systems in the region, but their traces are destroyed. Close to this period (3.7-3.6 Gyr ago), the surface between the hills was cov- ered with the material of ridged plains (RP, Hg. 24b), i.e., with lavas and/or sedimentary materials transported from the eroded plateaux. Afterward, but still at the end of the Noachian time and during die first half of the Hesperian period, a giant catastrophic outflow arose from an underground reser- voir located far to the south. Originally, its flows mean- dered over the plains, ertxled them along a wide front, and deposited their material in faint valleys (FC); later, they were concentrated in the present-day valleys to form their upper benches. The outflow deltaic sedi- ments represent the material of spotted (StP) and smooth (SP) plains (Fig. 24c). The sedimentary cover, almost uniform in thickness (dozens of meters), over- lies the impact craters, thus in^lying that deltaic mate- rials were probably accumulated over a vast area of a temporary sea. This was resurfacing stage II of the region. After a time interval of imknown duration, there was another cycle of catastrophic floods in the Hesperian Noachian Hesperian Amazonian Fig. 23. Histogram of age estiinates (the same as in Fig. 22); main epochs of Martian stratigraphy nom Hartmann (1981). SOLAR SYSnEM RESEARCH Vol. 32 No. 6 1998 448 MARCHENKO et al. '..Al',,.. .H \fi-J . ->,- *i- Jftt,..-^, MATERIALS: ??AAAAA - - o?d Plains Channet Floois Deltas H Eoltan Flg. 24. Schemes iUiusirating the reconstruction of the geological history of the jtudied region. Present-day position of larige impact ersten and valley boundanes b shown in all figures. SOLAR SYSTEM RESEARCH \bl. 32 No. 6 1998 GEOLOGY OF THE COMMON MOUTH OF THE ARES AND ITU VALLES, MARS 449 period?resurfacing stage ID. This event deepened the valleys, and deltaic deposits of the preceding stage wem ettxled by newly bom flows. We consider this stage as the fommon period of pitted plains (PP) with frequent rounded cauldrons. Some researchers beheve that these cauldrons are similar to tcnestrial thetmokaist land- forms?alases (Costard and Kargel, 1995). This stage was terminal for the fluvial reworking of the Ares floor material (CFA, Fig. 24d) and might be initial for the Tiu Vallis formation. Moreover, the Tm Vallis floor is incised into the floor of the Ares Vallis, and the crater density in the former is less than in the latter. This sug- gests that the Tiu Vallis was affected by a later episode of fluvial activity separated from the described one by a considerable time gap. During the next Late Hesperian-Early Amazonian period (resurfacing stage IV), the incision of the Tiu Vallis was further in progress, the material of pitted plains was heavily eroded, and remnants of the latter are represented now by benches and islands. Likely, the Tiu Vallis of that time was the only route for run-off, and fluvial processes completed the formation of the valley-floor deposits. Plains with fiinpws, wrinkle ridges, and fractures (FP) may also be related in origin to the last stage of flu- vial activity, as these landforms probably appeared as a result of exsiccation, freezing, and consolidation of sediments (Jons, 1986; McGill and Hills, 1992; Parker et aL 1993; TanaJta, 1997). In the opinion of Tanaka (1995,1997), the material of these plains was disturbed by return flows and invaded the terminal areas of the valleys (Fig. 24e). It is also possible that the FP mate- rial was brought from another outflow valley, located outside the region. Eolian and slope processes along with impact cia- tering developed throughout the entire geological his- tory of the region (Fig. 24f); they are active even at present 9. CONCLUSION Our purpose in this work was to answer the ques- tions of how, from where, and when various materials appeared in the exploration region of the Mars Path- finder rover (Fig. la). This region is at the boundary between the lithologie units of the Ares (CFA) and Tiu (CFT) Valles (the latter being younger than the former) that wc have mapped. According to the morphology, this part of the Tiu Vallis floor is classed with smooth floor plains (type CF,). We believe that both valleys and their lithologie units were related in origin to the fluvial activity of sev- eral catastrophic floods. Deposits in the mouth of the valleys should consist of two groups of materials transported by the flows. First, they should incorporate a more ancient material derived from the cratered and knobby plateaux, from hills, and from o?ier units (not studied by us) located to the south of the region. This is the material of ancient highland terrains (presumably, impact breccias and lavas). Second, a younger sedimentary, volcanic, or volcano-sedimenta^ material of ridged plains should be present. The material transported by water-saturated flows must be influenced by mechanical reworking, i.e., the fragment? must be rounded, sorted, etc. Judging from the first published results obtained by the Mars Pathfinder rover, these features are actually visible (Golombek et ai, 1997a; Matijevic et al., 1997). The material of ?jecta from impact craters, nearby and dis- tant, should also be present at the surface; however, it should not bear signs of fluvial reworking. In addition, the material at the rover-operation site must include dark eolian sand and bright dust. When were the deposits studied by the rover formed? According to our appraisal, the lithologie units of plateaux were formed about 4 Oyr ago and those of ridged plains approximately 3.5 Gyr ago. Thus, they are very ancient, and the fluvial reworking of these units took place first between 3,6 and 2.6 Gyr ago and then, presumably, between 2.3 and 1.4 Gyr ago. When the fluvial activity ceased, the floor deposits experi- enced wind erosion, which was apparently not very intense (Golombek et al., 1997a; Matijevic et ai, 1997; Smith ???j/., 1997;Basilevsky ??o/., 1998), and, possi- bly, thermokarst resurfacing. Finally, all these forma- tions were locally overlain by eolian dust and sand, and also by ?jecta from impact crateis. ACKNOWLEDGMENTS We are grateful to G.A. Burba and V.P. Shishkin for their assistance in the preparation of the maps, REFERENCES Arvidson. R.. Wind-Blown Streaks, Splotches and Associ- ated Craters on Mars: Statistical Analysis of Mariner 9 Hiotogn^s, Icarus. 1974, vol. 21, pp. 12-27. Arvidson, R., Guinness, E.A., and Zent, A.R. ClassiticatJon of Surface Units in the Equatorial Region of Mars Based on Viking Orbiter Color, Albedo, and Thermal Data, J. Geophys. Res., 1982. vol. 87. pp. 10149-10157. Arvidson, R.. Guinness, E.A.. Dale-Bannister. M.A.. Adams, J.. Smith, M., Christcnsen, P.R., and Singer, R.. Nature and Distribution of SurAcial Deposits in Chryse Planitia and Vicinity, Mars, /. Geophys. Res., 1989, vol. 94, pp. 1573-1587. Baker, V.R. and Milton, D.J., Erosion by Catastrophic Floods on Mars and on Eanh, Icarus, 1974. vol. 23, pp. 27-41. Baker, V,R. and Kochel, R.C., Morphological Mapping of Martian Outflow Channels, Proc. Lunar Set. Conf. IX, 1978.pp.3181-3192. Baker, V.R. and Kocbel, R.C., Martian Channel Morphology: Maja and Kasei Valles, J. Geophys. Res.. 1979, vol. 84, pp. 7961-7983. Baker, V.R., TTie Channels of Mars. Austin: Univ. of Texas Press, 1982. SOLAR SYSTEM RESEARCH Vol. 32 No. 6 1998 450 MARCHENKO et al. Baker, V.R., Strom, R.G.. Gu?ck, V.C, Kafgcl. J.S.. Komatsu, C. and Kaie, V.S., Ancient Oceans, Ice Sheets and the Hydrological Cycle of Mars, Nature, 1991, vol. 352. pp. 589-594, Basilevsky, A.T., Markiewicz, W.J.. and Keller, H.U., Mor- phology of Rocks within and nearby Rock Garden: Mars Palhfii?Ur Landing Site, Abstr. Lunar PUmei. Sei. Conf. XXIX. 199g.no. 1378 (CD-ROM). Betts, B.H. and Murray, B.C., Thermal Studies of Martian Channels and Valleys Using Termoskan r^ta, Abitr. Lunar Planet. Sei. Conf. XXIV, 1993. pp. 103-104. Bnrtz. J.H.. Smith, H.T.U., and Neff, G.E., Channeled Sea- bland of Washington: New Dala and Inlcrpretadons, Bull Ceoi Soc. Am., 1956. vol. 67, pp. 957-1049. Breu, JH., The Lake Missoula Floods and the Channeled Scabland, / Geol, 1969, vol. 77, pp. 505-543. Butvilovskii, V.V, Paleogeografiya poslednego oledenentya iGoloisena Altaya: Sobytiino-katastroficheskayamodeV (Last Glaciation and Holocene Paleogeography of Altai: Calastrophic-Evcnt Model). Tomsk: Tomsk. Univ., 1993. Carr, M.H., Crum(rfer. L.S., Cutts, J.A., Greeley. R.. Guest, J.E., and Mastirsky, H., Martian Impact Craters and Emplacement of Ejecta by Surface Row, / Ceo- phys. Res., 1977, vol. 82, pp. 4055-4065. Carr, M.H., Formaron of Martian Flood Features by Release of Water from Confined Aquifers. J. Ceophys. Res., 1979, vol. 84. pp. 2995-3007. Can. M.H., Water on Mars, New York: Oxford Univ Press, 19%. Christensen, RR., Martian Dust Mantling and Surface Com- position: Interpretation of "niermophysical Properties, J. Geophys. Res.. 1982, vol. 87, pp. 9985-9998. Christensen, RR., The Spatial Distribution of Rocks on Mars, Icarus, 1986, vol. 68, pp. 217-238. Christensen, P.R, and Kieflfer. H.H.. Moderate Resolution Thermal Mapping of Mars: the Channel Terrain around Chryse Basin. / Geophys. Res.. 1989, vol. 94, pp. 8233- 8238. Costard, F? Unusual Concentration of Rampart Craters at the Mouths of Outflow Channels, Mars, Abstr. Lunar Planet. Set. Conf. XXV, 1994, pp. 287-288. Costard, F and Baker, V, Thermokarst Landforms and Pro- cesses in Ares Valus, Mars, Abstr Lunar Planet. Sei. Conf. XXVI, 1995, pp. 285-286. Costard, F.M. and Kargel, J.S., Outwash Piaios and Thermokarst on Mars, Icarus, 1995. vol. 114, pp. 93- 112. Craddock, R.A., Grumpier, L.S., Aubcle, J.C.. and Zimbel- man, J.R., Geology of Central Chryse Pla?ida and the Viking 1 Landing Site: Implications for the Mars Path- finder tAis^f}n,J. Geophys. Res., 1997, vol. 102, no. ?2, pp. 4161-4183. Cnimpler, L.S., Geologic Mapping Traverse of the Highland- ttvLowland Transition in an Area Adjacent to the Man Pathfinder Keg^on, Lunar Planet. last. Tech. Rep., 1995, vol. 95-1,pp. 10-11. Dc Hon. R.A? Flood Surge in Mwtian Outflow System: Epi- sodic Flow, Abstr Lutuir Planet. Sei. Conf. XX. 1989, pp. 230-231. Edgett. K.S. and Christensen, P.R., The Particle Size of Mar- tian Aeolian Dunes, / Ceophys. Res., 1991, vol. 96, pp. 22765-22776. Edgett, K.S. and Christensen, P.R., Mars Aeolian Sand: Regional Variations Among Dark-Hued Crater Floor Features, / Geophys. Res., 1994, vol. 99, pp. 1997- 2018. Edgett, K.S. and Christensen, P.R., Rocks artd Aeolian Fea- tures in the Mars Pathfinder Landing Site Region: Viking Infrared Thermal Mapper Observations. J. Geophys. Res., 1997, vol. 102, pp. 4107^116. Golombek, M.R, Parker, TJ.. and Moore, HJ., Mars Paih- ,^n(JerLanding Site Characteristics,/Ifrin: LunarPlanet Sei. Cot\f. XXVL 1995, pp. 481-482. Golombek, M.R, Cook, R.A., Economou, T, FoUcner, W.M., Haldemann, A.F.C., Kallemeyn, P.H., Knudsen, J.M.. Manning, R.M.. Moore, HJ., Parker, TJ., Rieder. R., Schofield, J.T., Smith. RH., and Vaughan, R.M., Over- view of die Mars Pathfinder Mission and Assessment of Landing Site Predictions, Science, 1997a, vol. 278, pp. 1743-1748. Golombek, M.R. Cook, R.A., Moore, HJ., and Parker, T.J., Selection of the Mars Pathfinder Landing Site, J, Geo- phys. Res., 1997b. vol. 102. no. E2. pp. 3967-3988. HambUn, W.K., The Earth's Dyrutmic Systems, Minneapolis: Burgess, 1975. Hartmann, W.K., Strom, R.G., WeidenschiUing, SJ., Bal- sius, K.R., Woronow, A., Dence, M.R.. Grieve, R.AJ'.. Diaz, J., Chapman. C.R., Shoemaker, E.M., and Jones K J-, Chronology of Planetary Volcanism by Comparative Studiesof Planetary Cratering, in &uiiJrte Volcantsmon the Terrestrial Planets, New York: Peigamon, 1981, pp. 1049-1128. Jons, H.-R, Sedimentary Basins and Mud Flows in the Northern Lowlands of Mars, Abstr Lunar Planet. Sei. Conf XV, 19M, pp. 4n-in. Jons, H.-R, Late Sedimentation and Late Sediments in the Northern Lowlands on Mars. Abstr. Lunar Planet. Sei. Conf XVI, ?9&S. pp. 414-415. Jons, H.-P., Arcuate Ground Undulations. Getifluxion-like Features and 'Tront Tori" in the Northern Lowlands on Mars?What do They Indicate? Aoi/r Lunar Planet Sei Conf XVH, 1986, pp. 404-405. Kargel, J.S., Baker, V.R., Beget, J.E., Lockwood, J.E. Pcw?, Tl., Shaw. J,. and Strom, R.G,. Evidence of Ancient Continental Glaciation in the Martian Northern Plains, / Geophys. Res., 1995, vol, 100, no. ?3. pp. 5351-5369. Komatsu, G., Baker, V.R., and Johnson, J.R., Ares Vallis; Flood Geontorphology o? Mars Pathfinder Landing Site. Abstr. Lunar Planet. Sei. Conf XXVI, 1995. pp. 779- 780. Komatsu, G. and Baker. V.R.. Paleohydrology and Flood Geomoiphotogy of Ares Vallis, J. Geophys. Res., 1997, vol. 102. no. E2. pp. 4151^160. Kuz'min. R.O., Determination of the Occurrence Depth of lce>Bearing Rocks on Mais from Morphology of Young Craters, Dokl. Akad. Nauk SSSR. 1980, vol. 252, no. 6. pp. 1445-1448. Kuzmin, R.O. and Greeley, R., Geology o? Mars-Patf^inder Landing Site Ellipse, i? Abstracts i^Papers Submitted to SOLAR SYSTEM RESEARCH Vol. 32 No. 6 199S GEOLOGY OF THE CONfMON MOUTH OF THE ARES AND ITU VALLES, MARS 451 ihe 22 Russian-American Micmsymposium on Compar- ative Planetohgy, Moscow, 1995a, i^. 51-52. Kuzmin. R.O. ?nd Greeley, R.. Geology of the Landing Site Ellipse. Lunar Planet. Inst. Tech. Rep., 1995b. vol. 95- 01, pp. 20-21. Kuz'min, R.O., private communication, 1996. Lucchitta. B.K-. Anderson. D.M., and Shoji, H., Did lee Streams Carve Martian OutRow Channels?. Nature, 1981. vol. 290. no. 5809. pp. 759-763. Lucchitta, B.K., Ice Sculpture in the Maitian Outflow Chan- nels. J. Geophys. Res., 1982, vol. 87, pp. 9951-9973. Ulcchina, B.K.. Ice and Debris in the Fretted Terrain, Mars, Proc. Lunar Planet. Sei. Conf. XIV, 1984, pp. B409- B418. Lucchitta. B.K., Ferguson, H.M., and Summers. C, Sedi- mentary Deposits in the Northern Lowland Plains. Mais. J. Geophys. Res., 1986, vol. 9IE, pp. 166-174. Marchenko. A.G.. Geologic Mapping of the Mouth of Ares Valus, Mars: Preliminary Resulu, in Abstracts of Papers Submitted to the 24th Intentaiional Microsymposium on Planetology, Moscow, 1996. pp. 62-63. Marchenko, A.G., Basilevsky. A.T, Neukum, G.. and Hoff- mann. H.. Crater Counting in Thn? Areas of Ares Vallis, Space Planet. Set. (Ann. Geophys.), 1996, vol. 14, Suppl. in. p. C792. Marchenko, A.G., Basilevsky. A.T. Neukum, G.. Hauber, E., HofTmann, H., and Cook, A.C., Mapping of the Mouth of Ares Vallis, Mars, Abstr. Lunar Planet. Sei. Conf. XXV/ll. 1997. H>. 865-866, Marchenko, A.G.. Basilevsky, A.T, Neukum, G., Hauber, R, Hoffmann. H.. and Cook, A.C. The Study of the Mouth of Arcs and Tiu Valles, Mars, Abstr. Lunar Planet. ScL Conf XXIX, 1998, no. 1174 (CD-ROM). Masunky. H., Boys, J.M., Dial, A.L.. Shaubcr, G.G.. and Strobe!, M.E., Ctassi?cation and Time of Fomiation of Man?an Channels Based on Viking Data, J. Geophys. Res., 1977, vol. 82, pp. 4016-4038. M^jcvic. J,R? Crisp, J., et ai. CharacteriiatJon of the Mar- tian Surface Deposits by the Mars Pathfinder Rover, Sojoumer, ia>ne#, 1997. vol. 278. pp. 1765-1768. McCauley. J.F.. Carr, M.H., Cutts. J.A., Hartmann, W.K. Masursky, H., Milton, DJ., Sharp. R.P., and Wil helms, D.E.. Preliminary Mariner 9 Report on the Geol- ogy of Mars./ciinu, 1972, vol. 17, pp. 289-327. McGill, G.E. and E}?[nitT?ou, A.M., Origin of the Maitian Global Dichotomy by Crustal Thinning in the Late Noa- chian or Early Hesperian. J. Geophys. Res., 1990. vol. 95, pp. 12395-12605, McGill, G.E. and Hills, L.S., Origin of Giant Martian Poly- gons. J. Geophys. Res., 1992, vol. 97, pp. 2633-2647, Milton, D.J., Water and Processes of Degradation in the Mar- tian Landscape, /. Geophys. Res., 1973, vol. 78, pp. 4037^*047. Mustard. J.F., Are Dark Red Soils and Bright Dust on Mars Related through Mineralogie Phase Changes? Abstr Lunar Planet. Sei Con/; XYW, 1995. pp. 1021-1022. Nelson, D.M. and Greeley. R.. (3eologic Sequence of Forma- tion and Evidence for Shcctwash Erosion at the Ares Val- lis Region. Mars. Abstr. Lunar Planet. Sei, 1996. pp. 945-946, Nelson, D.M. and Greeley. R., Xante Tcira Outflow Channel Geology at the Mars Pathfinder Landing Site, Abstr. Lunar Planet. Sei. Cor^. XXIX, 1998. no. 1158 (CD- ROM), Neukum. G. and VTise, D.U., Mars; a Standard Crater Curve and Possible New Tune Scale, Science, 1976, vol. 194, pp. 1381-1387. Neukum, G. and Hiller, K., Maitian Ages, J. Geophys. Res., 1981, vol. 86, pp. 3097-3121. Nummedal, D. and Prior, D.B.. Generation of Maitian Chaos and Channels by Debris Flows, Icarus, 1981, vol. 45, pp, 77-86. Paricer, TJ., Saunders. R.S., and Schneeberger, D.M., IVana- tional Morphology in the West Dcutcronilus Mensae Region of Mars: Implications for Modification of the Uplands/Lowlands Boundary, Icarus, 1989, vol. 82, pp. 111-145, Parker. TJ? Gorsline, D.S., Saunders, RS., neri. D.C, and Schneeberger, D.M.. Coastal Geomorphology of the Martian Northern Plains, J. Geophys. Res., 1993. vol. 98. pp. 11061-110878. Pikt, R.J. and Davis, P.A.. Toward a Topographic Model of Martian Craters from Photoclinom^ry, Abstr. Lunar Planet. Sei Conf. XV, 1984, pp. 645-646. Rice, J.W? Jr. and De Hon, R. A,, Geologic Map of the Darvel Quadrangle, Maja Valus Region, Mars, USGS Map 1-2432,1 : 500000,1996. Rice. J.W,, Jr, and Edgett, ICS., Catastrophic Flood Sediment in Chrysc Basin, Mars, and Qulncy Basin, Washington: Application of Sandar Facies Model, J. Geophys. Res., 1997, vol. 102. pp,4l85-?200. Robinson, CA., Neukum, G., Hoffmaim. H.. Marchenko. A.. Basilevsky, A.T.. and Or?. G.G.. Suggested Geological Development for Ares Vallis. Mar^,/4ir5ri: Lunar/'^et. Sei Conf XXV?, 1996. pp. 1083-1084. Rotto. S, and Tanaka, K.L., Geologie/Geomorphologie Map qf the Ckryse Planitia region of Mars, USGS Map 1-2441.1:5000000,1995. Rudoi. A.N., Geomorphological Effect and Hydraulics of the Late Pleistocene Jdkulhlaups of Ice-Dammed Lakes, Altai, Geomorfologiya, 1995, no, 4, pp. 61-76. Schultz, PH., Schultz, R.A., and Rogers. J? The Stniaure and Evolution of Ancient Impact Basins on Mars, J. Geophys. Res., 1982, vol. 87. pp. 9803-9820, Scott, D.H. and Carr. M.H., Geologic Map of Mars, USGS Map 1-1083. 1 : 25000000,1978. Scott. D.H., Schaber, C.G., Horstman. K.C. and Dial. A.L., Jr., Lava Flows of Tharsis Montes, NASA Teeh. Memo, {"il^, vol, 80339. pp, 237-238, Scon. D.H., Meander Relics: Evidence of Extensive Flood- ing on Mars, NASA Tech. Memo, 1982, vol. 85127, pp. 216-218. Scott. D.H. and Tanaka, K.L, Geologic Map qfthe Western Equatorial Region of Mars, USGS Map M802-A., 1 : 15000000,1986, Scott. D.H., Rice. J.W., Jr.. and Dohm. J,M? Palcolakes and Lacustrine Basins on Mars, Abstr. Lunar Planet. Sei Conf. XXII, 1991, pp, 1203-1204, Scon. D,H.. Dohm, J.M., and Rice, J,W., Jr., Map of Mars Showing Channels and Possible Pateolake Basins, USGS Map l'246i, 1:30000000, 1995. SOLAR SYSTEM RESEARCH Val, 32 No, 6 1998 452 MARCHENKO et al. Shaip, R.P. and Malin, M.C., Channels on Mus, Bull. GeoL Soc. Am., 1975, vol. 86, pp. 593-609. Smith, P.H., Belt J.F. m., Bridges, N.T.. etaL. Results from the Mars Patlffinder Camera, Scitnce, 1997. vol. 278, pp. 1758-1764. Sodeiblom, L.A., Edwaids, K.. Eliason. E.M., Sachez, EM., and Cfaaiette, M.P? Color Variations on the Martian Sur- imx, leans. 1978, vol. 34, pp. 446-464. Squyrcs, S.W. and Kasting, J J^., Early Mars: How Wann and How Wet?, Science, 1994, voL 265, pp. 744-749. T?maka, KX., Chaotic Material and Debris Flows in the Simud-Tiu Valley Outflow Systtmof Mais,Abjtf; Ltmar Planet ScL Conf. XIX, 1988, pp. 1175-1176. Tanaka. K.L., Geologic History of Chryse Basin. Mars, Lunar PUmet. Insi Tech Hep., 1995. voL 95-01, pp. 39-40. Tknaka, KX., Sedimentary History and Mass Flow Struc- tures of Chryse and Acidalia Planitiae. Mars, / Geophys. Res,, 1997. vol. 102. no. E2, pp. 4131-4149. Th?rar?nsson, S., The J?kulhlaup From the Katla Area in 1955 Compared with Other JOkulhlaups in Iceland, yiiatu//, 1957. vol.7,pp, 21-25. Thomhill, CJ}., Rotheiy, D.A., Murray. J.B., Cook, A.C., Day. T., MuUer, J.P., and Iliffe, J.C.. Topogr^ihy of Apollinaris Patera and Ma'adim Vallis; Automated Extraction of Digital ElevatioQ Models. / Ctophys. Res., 1993. vol. 98, no. E12, pp. 23581-23587. Tteiman, A.H., Hardpan and Otbo' Diqgenetic "Rock" in the Catchment of Ares Vallis and Surrourtding Areas, Lunar PUmet Inst Tech. Rep., 1995, DO. 95-01, pp. 28-29. Tteiman, A?., What Is Mars' Highlands Crust?, in Con^r- ence wi Early Mars: Geologic and Hydrologie Evolu- tion, Physical and Chemical Environments, aitd Implica- tions for Life, Houston, 1997, pp. 77-78. Waiters, T.R., Wrinkle Ridge Assemblages on the Terrestrial Planets, / Geophys. Res., 1988. vol. 93, pp. 10236- 10254. Wise, D.U., Golombdc, M. P, and McGill, G.E., Tharsis I^ovince of Mars; Geologic Sequence, Geometry, and a Deformation Mechanism, Icarus, 1979, vol. 38, pp. 456-472. SOLAR SYSTEM RESEARCH Vol. 32 No. 6 1998