s e and the ese outflow d Noachian t sculpted the s decreased the Mars proposed ern plains istributed ated to minor www.elsevier.com/locate/icaru Episodic flood inundations of the northern plains of Mars Alberto G. Fairén,a,b,∗ James M. Dohm,c Victor R. Baker,c,d Miguel A. de Pablo,b,e Javier Ruiz,f Justin C. Ferris,g and Robert C. Andersonh a CBM, CSIC-Universidad Autónoma de Madrid, 28049 Cantoblanco, Madrid, Spain b Seminar on Planetary Sciences, Universidad Complutense de Madrid, 28040 Madrid, Spain c Department of Hydrology and Water Resources, University of Arizona, Tucson, AZ 85721, USA d Lunar and Planetary Laboratory, University of Arizona, Tucson, AZ 85721, USA e ESCET, Universidad Rey Juan Carlos, 28933 Móstoles, Madrid, Spain f Departamento de Geodinámica, Universidad Complutense de Madrid, 28040 Madrid, Spain g US Geological Survey, Denver, CO 80225, USA h Jet Propulsion Laboratory, Pasadena, CA 91109, USA Received 19 December 2002; revised 20 March 2003 Abstract Throughout the recorded history of Mars, liquid water has distinctly shaped its landscape, including the prominent circum-Chrys northwestern slope valleys outflow channel systems, and the extremely flat northern plains topography at the distal reaches of th channel systems. Paleotopographic reconstructions of the Tharsis magmatic complex reveal the existence of an Europe-size drainage basin and subsequent aquifer system in eastern Tharsis. This basin is proposed to have sourced outburst floodwaters tha outflow channels, and ponded to form various hypothesized oceans, seas, and lakes episodically through time. These floodwater in volume with time due to inadequate groundwater recharge of the Tharsis aquifer system. Martian topography, as observed from Orbiter Laser Altimeter, corresponds well to these ancient flood inundations, including the approximated shorelines that have been for the northern plains. Stratigraphy, geomorphology, and topography record at least one great Noachian-Early Hesperian north ocean, a Late Hesperian sea inset within the margin of the high water marks of the previous ocean, and a number of widely d minor lakes that may represent a reduced Late Hesperian sea, or ponded waters in the deepest reaches of the northern plains rel Tharsis- and Elysium-induced Amazonian flooding.  2003 Elsevier Inc. All rights reserved. Keywords:Mars; Tharsis floods; Oceans; Lakes of em. t is p, fea- and ters well, .g., 82; and est, the a- and an cur- lles aka, ely r the an igh- 92; ally the 1. Introduction Elucidating the hydrogeological cycle of Mars is one the main challenges in the exploration of the Solar Syst Long-term aqueous activity on the surface of the plane indicated by fluvial (e.g., Mars Channel Working Grou 1983; Malin and Edgett, 2000a, 2000b) and lacustrine tures (e.g., Squyres, 1989; Scott et al., 1995; Cabrol Grin, 1999, 2001); phased degradation of impact cra (e.g., Chapman and Jones, 1977; Craddock and Max 1993); permafrost (e.g., Lucchitta, 1981); periglacial (e Squyres, 1979) and glacial landforms (e.g., Lucchitta, 19 Kargel et al., 1995); and outflow channels (e.g., Baker * Corresponding author. E-mail address:agfairen@cbm.uam.es (A.G. Fairén). 0019-1035/$ – see front matter 2003 Elsevier Inc. All rights reserved. doi:10.1016/S0019-1035(03)00144-1 Milton, 1974; Scott and Tanaka, 1986; Greeley and Gu 1987). The outflow channels began to form as early as Noachian, and their activity extended to the Early Am zonian (Rotto and Tanaka, 1991a; Dohm et al., 2001a), even to the latest Amazonian in Elysium Planitia (Berm and Hartmann, 2002; Burr et al., 2002a, 2002b), with re rent flooding for some of the systems, including Kasei Va (Scott, 1993) and Mangala Valles (Chapman and Tan 1993). Diverse geologic and geochemical evidence collectiv point towards the previous existence of wide oceans ove Martian northern lowlands, which include: (1) the Marti dichotomy boundary, which separates the southern h lands from northern lowlands (e.g., Tanaka et al., 19 Smith and Zuber, 1996), and provides a topographic lower area in the plains for water accumulation; (2) http://www.elsevier.com/locate/icarus ecob Lápiz ecob Lápiz ecob Lápiz ecob Lápiz 54 A.G. Fairén et al. / Icarus 165 (2003) 53–67 ary nov yse rth- 1a) ity om- rlow ely an- e o ked ived ges (5) ins, ex- pu- om 99; dies vol- lian nd ry sse the er- ars 86; ker 994; l et t al. only clo- ds. on- igh clo- un- ive, (see out 999 ible tac tt et r the che ugh sted tion hy- ion a) dic 91, that un- An har- tions here ered SVs ned ed to ans, rker 95; d- f the ring avy s re- and 001; ysio- ust. ered red, rials erial anic, cott 987; 02). that ec- cott 987; ave rgel outflow channels, which either terminate at the bound (Parker et al., 1993) or fade into the northern plains (Iva and Head, 2001), including the prominent circum-Chr (e.g., Rotto and Tanaka, 1995) and recently identified no western slope valleys (NSVs, Dohm et al., 2000, 200 outflow channel systems; (3) the relatively low dens of superposed impact craters in the northern plains c pared to the southern densely cratered highlands (Ba and Bradley, 1990; Parker et al., 1993), and its extrem flat topography at the distal reaches of the outflow ch nel systems (Head et al., 1999); (4) the broad occurrenc wide age-ranging glaciers that are interpreted to be lin to magmatic-triggered flooding and associated short-l (tens of thousands of years) environmental/climatic chan (Baker, 2001; Cabrol et al., 2001a, 2001b, 2001c); and the chemical signatures reported for the northern pla including high abundances of S and Cl or the possible istence of sulphate minerals and chloride salts, making a tative andesite-rich component or weathered basalt the d inant material type in the lowlands (McSween et al., 19 Zuber, 2001; Wyatt and McSween, 2002). Standing bo of water, therefore, best explain such evidence, though canism (Head et al., 2002), tectonism (Sleep, 1994), eo modification (Malin and Edgett, 2000a), ground volatile a debris flow activity along the highland-lowland bounda (Tanaka, 1997; Tanaka et al., 2002), and/or glacial proce (Baker et al., 1991) also probably contributed to shaping Martian lowlands. The presence of large bodies of water during diff ent periods of time on the northern lowlands of M have been previously proposed (McGill, 1985; Jöns, 19 Lucchitta et al., 1986; Parker et al., 1987, 1989, 1993; Ba et al., 1991; Scott et al., 1991a, 1991b; Chapman, 1 Scott et al., 1995; Scott and Chapman, 1995; Karge al., 1995; Head et al., 1999). For example, Parker e (1987, 1989) described seven different boundaries, but two of them could be approximately traced as complete sures within and on the margins of the northern lowlan Parker et al. (1993) proposed an outer boundary, or C tact 1, located along the Martian dichotomy, as the h water mark of a primitive ocean. The later second en sured boundary, or Contact 2, which occurs several h dreds of kilometers north of the dichotomy, is less extens and would indicate a younger high-stand water mark Clifford and Parker, 2001, for a complete discussion ab the contacts). A subsequent analysis by Head et al. (1 pointed out that Contact 2 could represent the only poss shoreline, because the variations of the elevation in Con 1 do not reflect an equipotential surface. In addition, Sco al. (1995) present geologic and geomorphic evidence fo possible existence of mid-sized lakes in the deeper rea of the lowlands as recent as the Amazonian Period, tho they also indicate that a larger water body may have exi if the lakes had been earlier connected at a similar eleva Supported by published geologic, geomorphic, paleo drologic, and topographic information, and in conjunct f - s ) t s . Table 1 Absolute age estimates for the surface of Mars Epoch Absolute age range (G Late Amazonian 0.6–0.3 to present Middle Amazonian 2.1–1.4 to 0.6–0.3 Early Amazonian 3.1–2.9 to 2.1–1.4 Late Hesperian 3.6 to 3.1–2.9 Early Hesperian 3.7 to 3.6 Late Noachian 3.82 to 3.7 Middle Noachian 3.95 to 3.82 Early Noachian > 3.95 Condensed from Fig. 14 of Hartmann and Neukum, 2001. with the MEGAOUTFLO hypothesis that explains episo water stability in the lowlands of Mars (Baker et al., 19 2000; Baker, 1999), we propose a conceptual model here points out the direct impact of Tharsis-triggered flood in dations in the shaping of the northern lowlands (Fig. 1). enormous Noachian basin/aquifer system in eastern T sis has been revealed by paleotopographic reconstruc of the Tharsis magmatic complex in the western hemisp of Mars, proposed to have sourced the Tharsis-trigg outburst floods that sculpted the circum-Chryse and N outflow channel systems (Dohm et al., 2001b), entrai boulders, rock, and sediment during passage, and pond form sequentially through time various hypothesized oce seas, and lakes (e.g., McGill, 1985; Jöns, 1986; Pa et al., 1987, 1993; Baker et al., 1991; Scott et al., 19 Head et al., 1999) (Fig. 2). The dwindling of these floo waters through time, due to an inadequate recharge o Noachian drainage basin/aquifer system, is expected du the time of noted low erosion rates following the late he bombardment. Note that we use inferred absolute age cently condensed from models of impactor populations studies of martian meteorites (Hartmann and Neukum, 2 see Table 1). 2. Physiographic setting The northern and southern provinces represent a ph graphic and geomorphologic dichotomy of the martian cr The boundary clearly separates relatively young, uncrat surface materials of the lowlands from the highly crate ancient highland rock assemblages. The lowland mate are interpreted to have been emplaced by both suba and subaqueous processes, which include eolian, volc fluvial, mass-wasting, lacustrine, and marine (e.g., S and Tanaka, 1986; Tanaka, 1986; Greeley and Guest, 1 Parker et al., 1993; Scott et al., 1995; Tanaka et al., 20 The ancient highlands consist of a mélange of rocks are interpreted to mainly include lava flows, impact br cias, and eolian, fluvial, and colluvial deposits (e.g., S and Tanaka, 1986; Tanaka, 1986; Greeley and Guest, 1 Dohm and Tanaka, 1999), although glacial deposits h also been proposed (e.g., Kargel and Strom, 1992; Ka et al., 1995). E p iso d ic flo o d in u n d a tio n s o fth e n o rth e rn p la in s o fM a rs 55 Fig. 1. Chart comparing the evolutional stages of geologic activity in the Tharsis Mag- matic Complex/superplume region with major geologic features and proposed inundations (Stage information corresponds to Stage information of Fig. 2); size of solid areas roughly proportional to degree of exposed deformation and feature extent. Violet—centers of tec- tonic activity interpreted to be the result of magmatic-driven uplift and local volcanism, dike emplacement, and hydrothermal activity; orange—mountain building; blue—water; red—primarily emplacement of shield-forming and lava-field-forming lavas. Note that the commencement and/or end of activity of the components of the complex and inundations are not absolutely constrained and that features such as the shield volcanoes of Tharsis Montes and Olympus Mons could be presently active. The schematically portrayed events are those that have survived thousands and perhaps millions of years of wind and water erosion to stand out as significant windows of the Martian past, while smaller windows have been erased or highly subdued from the Martian record; the inundation representations, for exam- ple, may include several isolated events. Fig. 2. Topographic shaded relief map of the northern hemisphere of Mars constructed from Mars Orbiter Laser Altimeter (MOLA) data showing major geographic features of the northern hemisphere, including three major basins (Borealis basin= Vastitas Borealis, Utopia basin= Utopia Planitia, Isidis basin= Isidis Planitia). Also shown are Shoreline 1 (black line), Contact 1 in Arabia Terra (dashed-black line) and Shoreline 2 (dark blue line), which are based on Edgett and Parker (1997), Carr (2002), Parker et al. (1987, 1989, 1993), Clifford and Parker (2001), and Head et al. (1999); paleolakes (light blue lines), based on Scott et al. (1995); and Stage information (numbers) that reflects the geologic mapping of Tanaka et al. (2002) and correlative with Stage information of Dohm et al. (2001b, 2001c) and Anderson et al. (2001). Polar Stereographic projection; scale varies with latitude; modified from Tanaka et al. (2002). 56 A.G. Fairén et al. / Icarus 165 (2003) 53–67 r- red ans has ary al., ori- om e or t al., ma and- lt o 94), 2; w- arly cep by rel- me ges ses ins e to ore evi- ug- hore al., rim- rra, up h as gett ity) n- , tern ena due have the e 1; and arr, s to gra- ated ore- ove Con elow en- to ac- lue els a- t Vs axi- rger ther og- out d. be in, he stal dif- 990; nal pro- sed et’s tion tric uiz 2 is older to be aller ces a cep- ant, ate 986; ruc- est ents rsis ocal is- ting te cre- erra The origin of the generally smooth plains-forming Ma tian northern lowlands, the locality where Tharsis-trigge floodwaters may have ponded to form the putative oce and small water bodies, remains uncertain. However, it been proposed that the Martian highland/lowland bound began to form at least since the Noachian (Dohm et 2001b). Several hypotheses have been invoked for its gin, which include either subcrustal erosion resulting fr mantle convection (Wise et al., 1979), excavation by on more large impacts (Wilhelms and Squyres, 1984; Frey e 2002), or overturn of unstable cumulates in an initial mag ocean (Hess and Parmentier, 2001). Alternately, the l scape of the northern lowlands may have been the resu sea-floor spreading, linked to plate tectonism (Sleep, 19 during the Early into Middle Noachian (Baker et al., 200 Dohm et al., 2002; Fairén et al., 2002). In fact, although lo lands’ basement is old, perhaps dating from the final E Noachian (Frey et al., 2002), the cratered highlands (ex for the northwestern part of Arabia Terra; an area noted Zuber et al. (2000) to be related to the lowlands due to its atively thin crust) are older indeed (Frey, 2003). Thus, so hundreds of millions of years during the embryonic sta of Mars’ development may have allowed for inner proces to shape the lowlands. Of the two shorelines proposed for the northern pla (Parker et al., 1987, 1993), the inner younger one is clos an equipotential line (Contact 2, hereafter referred as Sh line 2), whereas the outer older shoreline (Contact 1) d ates from equipotentiallity (Head et al., 1999). Here, we s gest the existence of a greater shoreline (referred to as S line1 from the previously reported position by Parker et 1987, 1993), which may represent the high stand of the p itive Martian ocean (Figs. 2 and 3). In western Arabia Te the coastline may have been some thousands of km wards in the cratered terrain, which includes regions suc northern Sinus Meridiani and western Arabia Terra (Ed and Parker, 1997). Elevation variations (equipotentiall decrease from∼ 11 km using the previously mapped Co tact 1 (Head et al., 1999) to only∼ 2.5–3 km in Shoreline 1 which transects across northern Sinus Meridiani, wes Arabia Terra, northeast Arabia Terra, and north Thyrr (Tharsis and Elysium rises are not taken into account to the abundant post-Noachian activity those structures experienced). In addition, valley networks formed during Noachian in Arabia Terra surround and empty at Shorelin and there are only isolated valleys between Shoreline 1 the previous position of Contact 1 in northeast Arabia (C 2002; see Fig. 3B). Moreover, Shoreline 1 correspond the crustal thickness dichotomy, as deduced from topo phy and gravity data (Zuber et al., 2000; see Fig. 3C). The volumetric extent of the earliest and largest estim water body, in our opinion, can be best portrayed by Sh line 1 through several lines of evidence, including: (1) MOLA data, which shows rougher surfaces ab Contact 1, less rougher surfaces at all scales between tact 1 and Shoreline 2, and generally smooth surfaces b f t - - - - Shoreline 2 (Head et al., 1999). (Note that Shoreline 1 closes the previously mapped Contact 1; see Fig. 2). (2) Total water capacity below Shoreline 1 (estimated be more than 108 km3 compared to the approximated cap ity below the previously mapped Contact 1 of 9.6×107 km3 by Head et al., 1999), which lies between the minimum va for water that flowed through the Chryse outflow chann (0.6 × 107 km3; Carr, 1996a) and the maximum water c pacity estimated for regolith (5 to 20× 107 km3; Squyres e al., 1992). (3) The newly identified Noachian-Early Hesperian NS outflow channel system, with an enormous estimated m mum peak discharge rate of about 2× 1010 m3/s, which is proposed to be a significant source of water for the la (Shoreline 1) ocean (Dohm et al., 2001a) (Fig. 2). Nevertheless, magmatism and tectonism among o contributors likely modified ancient (e.g., Noachian) top raphy, including Shoreline 1, where a deviation of ab 2–3 km from an equipotential line most likely occurre Specifically, these deviations from equipotentiallity may explained by: (1) The influx of water throughout time (Cabrol and Gr 2001). (2) Local variations of the thermal structure of t lithosphere (due to variations in mantle heat flow or cru heat generation, for example) that generate elevation ferences between regions (Lachenbruch and Morgan, 1 Turcotte et al., 2002; Ruiz et al., 2003). If there were regio differences in heat flow in early Mars, the subsequent gressive decline of heat flow through time might have cau differential thermal subsidence (relaxation) of the plan lithosphere. This could have contributed to the distor of the original Noachian topography in an even kilome scale, including the equipotentiallity of Shoreline 1 (R et al., 2003). The original topography along Shoreline expected to be less deformed when compared to the Shoreline 1 because mantle temperature is expected less through time (e.g., Schubert et al., 1992). This sm heat flow following the Late Hesperian (smaller differen in heat flow or crustal heat production) would include decrease in magmatic-driven uplifts and subsidence. Ex tions include Elysium rise and Alba Patera where signific concentrated magmatic-driven activity is noted for the L Hesperian and Amazonian (e.g., Scott and Tanaka, 1 Greeley and Guest, 1987; Anderson et al., 2001). (3) Pre-oceanic tectonic structural fabrics, such as st turally controlled cliffs in the hypothesized oldest and larg ocean and post tectonically-controlled elevation adjustm to putative Shoreline 1, related to the evolution of Tha and/or other elevational adjustments associated with l volcanism, wind and/or glacial activity, groundwater d charge and surface runoff, and tectonism (including faul and impact cratering). (4) A possible Early into Middle Noachian martian pla tectonism episode, with convergent plate margins and ac tion of terranes (e.g., northwards Terra Cimmeria and T Episodic flood inundations of the northern plains of Mars 57 (A) (B) (C) Fig. 3. A: Topographic map of Mars. Note the equipotentiallity of Shoreline 1 in northern Sinus Meridiani and western Arabia Terra. Arrows indicate the possible Noachian coastline, both in (A) and (B). Marked is the area showed in detail in (C) (IEG0062T.IMG MOLA image, both shaded and colored according to topography. Produced by Tayfun Oner). B: Valley channels appear to debouch along Shoreline 1, even in northeast and southwest Arabia Terra. (In Carr, 2002.) (C) Inferred crustal thickness from MOLA data, running from the north pole (left) to the south pole (right), along the 0◦ longitude region, including Arabia Terra. Crustal thickness does not correlate with the dichotomy boundary or Contact 1, but appears to be related with Shoreline 1. This could represent the first great ocean that occupied the northern plains of Mars. (Image from MOLA Science Team). 58 A.G. Fairén et al. / Icarus 165 (2003) 53–67 ism et re- sec- sed d th u- atic- on and aleo fter 1c). ns (see ac- est, lex n the . 1) ave s. 2 ely arly ic- r- ich ism, ity, nne em- d to the in- , an de- lso ted 1a, nly can- ian ve re, the ly mi- son om- the cord , it ise be d- ring mo 2). pa- ape rom 02; uld O the wa- seas ay dur- and an, than long- ean to ould amic een 1), ived and ate to a and and ave arker ated s cy- 3), ria asia rth- ern od- heat Sirenum; Fairén et al., 2002). The proposed plate tecton for early Mars (first hundreds of millions of years, Baker al., 2002) would greatly modify the topography along Sho line 1. (5) The approximated values of the hypothesized ondary and smaller water-masses, whose associated mentary deposits and water resources could have altere primitive geomorphological features of Shoreline 1. 3. Inundation model In our model, the Tharsis Magmatic Complex (TMC)/S perplume development, which includes pulses of magm driven hydrological activity, directly influences the evoluti of the atmosphere and climate, as well as subsurface surface water processes as observed in geologic and p hydrologic records. Figure 1 reflects Stage information a Anderson et al. (2001) and Dohm et al. (2001b, 200 This model highlights Tharsis-triggered flood inundatio and their direct impact on shaping the northern plains Fig. 2), and other contributors such as magmatic-driven tivity at Elysium rise (Tanaka, 1986; Greeley and Gu 1987; Skinner and Tanaka, 2001). 3.1. Noachian to Early Hesperian (Stages 1–3): the first great ocean Incipient development of the Tharsis magmatic comp (TMC) during the Early into Middle Noachian (Anderso et al., 2001; Dohm et al., 2001b) to as late as possibly Early Hesperian (Stages 1–3 of Dohm et al., 2001b) (Fig resulted in the first inundation; an ocean that would h covered the northern plains (approximately 1/3 of the total surface area of Mars, see location of Shoreline 1 on Fig and 3). Dohm et al. (2001b) show evidence that collectiv indicate that the TMC began to form sometime around E into Middle Noachian with significant pulses of magmat driven activity during Stages 1–3, including the growth A sia, Syria Planum, and central Valles Marineris rises, wh are interpreted to be centers of magmatic-driven tecton volcanism, dike emplacement, local hydrothermal activ and associated NSVs and circum-Chryse outflow cha activity. The development of Thaumasia plateau and the placement of intercrater materials, which are interprete be the result of phreatomagmatic explosions such as in Valles Marineris region where magma/water/water-ice teractions have been proposed (Chapman et al., 1991) older wrinkle ridged materials interpreted to represent formed lava plains (Dohm et al., 2001c), may have a been influential on outflow channel activity, and rela flood inundations during Stages 1–3 (Dohm et al., 200 2001b, 2001c). In addition, valley networks are mai Noachian in age (e.g., Carr, 2002), and so excellent didates to contribute to the filling of the first great mart ocean. i- e - l d The incipient development of TMC’s Stage 1 may ha occurred from Early into Middle Noachian. Furthermo a magnetic field may have been operative on Mars in initial Early into Middle Noachian (from approximate 4.4 to 4.0 Gyr), since one of the oldest and most do nant Noachian centres of activity, Claritas rise (Ander et al., 2001), spatially corresponds with a magnetic an ally (Acuña et al., 1999, 2001), but not subsequent to Hellas impact event, since the Hellas basin does not re magnetic anomalies (Acuña et al., 1999). Importantly is difficult to determine for sure whether the Claritas r records incipient TMC development or whether it may related to magmatic activity independent from TMC. In a dition, plate tectonism may have been in operation du the Early into Middle Noachian, surviving the inner dyna for some tens of millions of years (Fairén et al., 200 If at least the first pulse of TMC magmatism accom nied plate tectonism, then the northern lowlands’ landsc could have resulted from seafloor spreading (lasting f about 4.4–4.3 to approximately 3.9 Gyr; Baker et al., 20 Dohm et al., 2002); and possible carbonate recycling wo have accounted for the reinjection of large amounts of C2 into the atmosphere. Subducted hydrous material of lithosphere would have represented an ample supply of ter for later Tharsis volcanism, and its water release to and/or lakes (Baker et al., 2002). In addition, Mars m have been more volatile-enriched than Earth, especially ing its embryonic stages of development (Craddock Howard, 2002). Therefore, until at least Middle Noachi the Martian atmosphere could have been much thicker at present, capable of sustaining a considerable and standing liquid ocean. As stated before, this primitive oc is best portrayed by Shoreline 1. Without the magnetic field protection since the Middle Late Noachian (Stage 2), the atmospheric erosion rate w have increased, as well as the loss of water to hydrodyn escape. Surface water stability, however, could have b possible during hundreds of million of years (Lundin, 200 especially if local remanent magnetic anomalies surv in the crust once the inner dynamo shut off (Jakosky Phillips, 2001). Additionally, if Mars passed through a pl tectonic phase, its termination would have contributed break-up of the atmospheric equilibrium between adding retiring carbonates; the result would be thinner, dryer colder atmospheric conditions. The ocean would thus h been reduced and ice-covered during Late Noachian (P et al., 1993), and its possible geomorphologic-associ features would be erased by subsequent major hydrou cles. An Early Hesperian pulse of Tharsis activity (Stage which includes further development of the Arsia, Sy Planum, central Valles Marineris rises, and Thaum plateau, may have triggered more flood waters to the no ern plains adding to the potentially already existing north plains ocean and/or ice body/ground ice. As the new flo waters washed over the northern plains, the additional Episodic flood inundations of the northern plains of Mars 59 uen ro- lse (ap- 91, tor es- ka, 001 ere ig- ing 91; gni- 999 an .g., ses ave eric ly hen anic rly lume is ris ave f uld mil- t es- f the rian sien The ed defin der eez- an ati- nd in nd a- c- um mor- els. n- and nd wa- ow ott 992; 95; lson aka, 994; ded es in . 2 mi- erdt, ed ese sis- an- till se- es ust, anic ase may have melted the upper layers of ice and the subseq gradients would allow the melt water to cycle into the hyd logic system (Dohm et al., 2001a). This third major pu of TMC activity may have contributed enough CO2 and other volatiles to the atmosphere to induce a short-lived proximately tens of thousands of years; Baker et al., 19 2002) climatic perturbation. Another potential contribu to the flood inundation hypothesis during the Early H perian is incipient development of the Elysium rise (Tana 1986; Greeley and Guest, 1987; Skinner and Tanaka, 2 Tanaka et al., 2002). If both Tharsis and Elysium w concurrently active during the Early Hesperian, then s nificant climatic responses would be expected, includ growth of ice sheets at the poles (Baker et al., 19 Head and Pratt, 2001), outflow events of enormous ma tudes to the northern plains (e.g., Nelson and Greeley, 1 Dohm et al., 2001a), and spring-fed discharge along unvegetated (instable) highland-lowland boundary (e Tanaka et al., 2002). When compared to today’s conditions, water and ga released during such Tharsis Superplume activity could h resulted in thicker post-heavy-bombardment atmosph conditions. The first inundation originated from a high productive aquifer system, and may have occurred w environmental conditions were more clement; the oce environment of the initialOceanus Borealismay have per- sisted from the Early to Middle Noachian into the Ea Hesperian, related to episodic, pulsating Tharsis superp activity. For example, an amount of∼ 3× 108 km3 of mag- mas has been proposed as the total release of the Thars during the Noachian, and their volatile content would h produced the equivalent of 1.5-bar CO2 and a global layer o water of 120 m thickness (Phillips et al., 2001). Water co so have extended over the lowlands during hundreds of lions of years (e.g., Clifford and Parker, 2001). 3.2. Late Hesperian/Early Amazonian (Stage 4): the las Martian ocean Between the disappearance of the Noachian-Early H perian great ocean and the temporary establishment o Late Hesperian minor sea, a cold and dry Late Hespe intermediate period occurred, perhaps related to a tran pause of the Tharsis development (Dohm et al., 2001b). first clear discontinuity in TMC development is observ between stages 3 and 4, when Thaumasia Plateau was itively established (see Fig. 1). The transition to this col climate, and the decline in crustal heat flow, caused the fr ing of the primordial ocean, which presumably developed ice cover that thickened with both time and increasing l tude (Clifford and Parker, 2001). In addition, the colder a dryer conditions might account for the significant decline the rate of valley network formation in the Hesperian a Amazonian (Fig. 4). Following the period of magmatic quiescence, a m jor pulse of Late Hesperian/Early Amazonian TMC’s a t ; ; e t - Fig. 4. Histogram showing relative ages of channels on Mars. A: Maxim stratigraphic ages of highland channels. B: Revised ages of A based on phology. C: Maximum stratigraphic ages of highland and lowland chann D: revised ages of C based on morphology. N= Noachian; H= Hesperian; A = Amazonian. (From Scott et al., 1995). tivity (Stage 4) coincides with significant outflow cha nel development (e.g., Scott et al., 1995; Head, 2002) the time proposed for Shoreline 2 formation (Clifford a Parker, 2001). This new pulse triggered yet more flood ters that incised Chryse and other circum-Chryse outfl channels (e.g., Milton, 1974; Baker and Milton, 1974; Sc and Tanaka, 1986; DeHon, 1992; DeHon and Pani, 1 Scott, 1993; Rotto and Tanaka, 1995; Scott et al., 19 Rice and DeHon, 1996; Chapman and Tanaka, 1996; Ne and Greeley, 1999) and Mangala Valles (Scott and Tan 1986; Chapman and Tanaka, 1993; Zimbelman et al., 1 Craddock and Greeley, 1994; Scott et al., 1995) and pon in the northern plains to form a sea in the deeper recess the lowlands (Shoreline 2), inset within Shoreline 1 (Figs and 3). In its later stages, it is also possible that this he spheric ocean turned into a mud ocean (Tanaka and Ban 2000) or froze (Carr, 1996b), or progressively diminish into small seas or lakes (Scott et al., 1995), none of th scenarios mutually exclusive. The water, CO2, and other volatiles released during this stage of significant Thar related effusive volcanic activity may have produced a tr sient greenhouse atmosphere. In addition, if plate tectonism occurred, its stands sometime in the Middle Noachian would lead to a sub quent stagnant-lid convection with mantle upwelling plum affecting to more concrete locations under the Martian cr because plate movement ceased. Thus, significant volc activity could be anticipated in these locations, as is the c 60 A.G. Fairén et al. / Icarus 165 (2003) 53–67 arly t of and ent the ur- re- an from in- kin- an- har- riod 987 wth tes ac- 01; ease warm end he O ma- ed ith ting ar- 91), mal an- ern be- the the 02). eat s- rr, of line ian ex- in- rsis arge et ely e of f the ach ory. ble of a hy- ules, the din, oxi- the be- mal osky nce ern suf- up- ment lete ting s of kes, tion aps ted n- ater like unts , or ctic ty. s or ans, ies e of ld and lting sis ar- for ation and en- 93) ver for both Tharsis and Elysium; such Late Hesperian and E Amazonian activity is recorded in the early developmen the giant shield volcanoes, Olympus Mons, Alba Patera the Tharsis Montes during Stage 4 of TMC developm (Dohm et al., 2001b), as well as the shield volcanoes of Elysium rise (Greeley and Guest, 1987). 3.3. Amazonian (Stage 5): temporary lacustrine environments Finally, generally cold and dry desert conditions d ing the Amazonian were punctuated by small outflow leases such as at and near Kasei and Mangala Valles the channels systems that debouch into Utopia basin the northwestern flank of Elysium rise, Hrad, Apsus, T jar, and Granicus Valles (Greeley and Guest, 1987; S ner and Tanaka, 2001) to form a number of minor tr sient seas or lakes (Fig. 2). Various isolated pulses of T sis magmatism are documented for the Amazonian Pe (e.g., Scott and Tanaka, 1986; Greeley and Guest, 1 Anderson et al., 2001), including the continuous gro of Olympus Mons, Alba Patera and the Tharsis Mon shield volcanoes (see Fig. 1). Late-stage Elysium Mons tivity during the Amazonian (Skinner and Tanaka, 20 Pounders et al., 2002) also may have contributed to rel carbonates and water into atmosphere. In any case, the periods were infrequent and short (103 to 105 years) as recorded in the low cumulative rates of erosion since the of the Noachian (Baker, 1999, 2001; Carr, 2002). High heat flux originating in magmatic activities near t surface is proposed to be responsible for the melting of C2 clathrate, causing rapid outflows of water and so the for tion of chaotic terrains (Komatsu et al., 2000); the mobiliz water-sediment mixture would create outflow channels, w ages ranging from Noachian to Amazonian. Equally, mel of ground and surface ice by volcanic activity has been gued for the origin of the channels (e.g., Baker et al., 19 the melting possibly being accompanied by hydrother circulation (Gulick, 1998). In fact, it is possible that subst tial remnants of the floodwaters that inundated the north lowlands during Noachian and Hesperian could remain neath thin mantles of dust and volcanics emplaced in Hesperian and Amazonian, as potentially indicated from latest results of the Odyssey mission (Boynton et al., 20 Nevertheless, the long-term decline in planetary h flow and sequestering CO2 in carbonates and the progre sive trapping of H2O into clays in the cryosphere (Ca 2002) would have greatly depleted the original inventory groundwater. This could well explain the apparent dec in outflow channel activity observed during the Amazon (Scott et al., 1995). Floods and water bodies are also pected to be progressively smaller with time due to a dw dling Noachian water supply (e.g., Noachian eastern Tha basin/aquifer system) and the related insufficient rech after each endogenically-driven hydrologic event (Dohm al., 2001b). So, the Late Hesperian–Amazonian extrem d ; cold and dry conditions could even include the presenc a mantled, thin frozen ocean in the deeper reaches o northern plains (Janhunen, 2002). In addition, the environmental conditions between e period ensures the loss of water from the global invent The comparatively lower martian gravity prevents a sta thick atmosphere (Baker et al., 1991) and the absence magnetic field allows energetic solar particles to force drodynamic dissociation upon near-surface water molec releasing the heavier oxygen into the soil and allowing lighter hydrogen to escape into the exosphere (e.g., Lun 2001). The released oxygen may have contributed to the dation of the regolith, hydrolysis of silicate minerals, and formation of carbonate minerals. Also, impact events are lieved to have contributed to atmospheric erosion or ther escape during the period of heavy bombardment (Jak and Phillips, 2001). 4. Discussion Contrary to the numerous works that present evide for the possible occurrence of water bodies in the north plains, Malin and Edgett (1999) suggest that there is in ficient Mars Orbital Camera (MOC)-based evidence to s port the existence of coastlines. This apparent disagree may be attributed to several factors, including incomp MOC coverage of the Martian surface, inadequate tes of the features described (actually, common expression the shorelines associated with large terrestrial paleola see Clifford and Parker, 2001), wind and water modifica over a prolonged geologic time period (millions to perh a billion years of resurfacing), ill-defined shorelines rela to rapidly diminishing water bodies shortly following inu dation through the sublimation and the transferral of w into the ground and/or to freezing, the lack of an Earth- moon to cause tidal forces responsible for the large amo of energy available for erosion at terrestrial shorelines persistant environmental conditions similar to the Antar (Mahaney et al., 2001), which would inhibit wave activi In this sense, searching for lines of boulders, scour mark gravel bars, associated to the action of ice-covered oce might evidence Artic and Antartic-like shore morpholog (Fig. 5). Since at least the Late Noachian, no evidenc warm, but episodic milder conditions in a long-term co and dry climate, are reported for Mars; so, morphologies landforms related to temperate-climate shorelines, resu from the action of waves, may be hard to found. The inundation hypothesis is not the only hypothe which offers explanations for some of the significant ch acteristics described above. An alternative mechanism producing a dense atmosphere, for example, is the vari of Martian orbital parameters (see Carr, 1990; Touma Wisdom, 1993; Laskar and Robutel, 1993) and related vironmental/climatic changes. Touma and Wisdom (19 proposed that an obliquity of the Martian rotation axis o Episodic flood inundations of the northern plains of Mars 61 OC by rthern bar ere this ved and ar- on- uld ge. ave ing lo- und ding Fer- x- ed a ms e uid mil- nd do ys- gala suf- gths e.g., the y- in- d to r et ns rted ner- ter ver- lly- igh- th con- rig- ess wet sug- tary ard- act ere. at- eads en a- ub- on- her- el- other er aka The esis e oc- lley ian 991; ott an n or Pa- ribu- lles the sses s not the Fig. 5. 026A72 Viking image (left) and subframe of SPO2-515/06 M image, evidencing Artic and Antartic-like lines of boulders (indicated arrows in the MOC image) on the Acidalia Plains (45◦ N, 7◦ W): the flat-topped mesa can so be interpreted as a possible island in a no plains frozen ocean. (Images courtesy of MSSS.) 30◦ would generate an atmospheric pressure above 25 m which would allow a total water release to the atmosph equivalent to 500 pr µm (Jakosky and Carr, 1985). But process is not necessarily different than the TMC-deri one, since it is possible that the effect of the Tharsis mass elevation on its own could have been capable of altering m tian orbital parameters (Melosh, 1980). In addition, a c currence of Tharsis activity and a change of obliquity co have resulted in a magnified environmental/climatic chan However, it is important to note that obliquity changes h there least effect in equatorial latitudes. As such, invok obliquity change to explain the formation of channels cated near the equator and formed from catastrophic gro water release (e.g., Mangala Valles and the surroun groundwater sapping channels) is highly questionable ( ris and Dohm, 2002). In addition, liquid CO2 has been proposed to help e plain the various erosional landscapes that are express the martian surface, including the outflow channel syste (Hoffman, 2000). Liquid CO2, in our opinion, raises mor queries than affords answers, as follows: how can a liq CO2 reservoir persist for tens of thousands and perhaps lions of years, especially during periods of magmatic a tectonic activity, as well as impact cratering events?; how multiple flood events occur at any one outflow channel s , - t tem such as in the case of Kasei (Scott, 1993) and Man (Chapman and Tanaka, 1993) Valles such that there is ficient recharge of the associated CO2 reservoirs?; how do such flows sustain their erosive capabilities for great len at Martian conditions? These and many other queries ( Stewart and Nimmo, 2002) have yet to be addressed by White Mars hypothesis of Hoffman (2000), though the h pothesis has provoked productive scientific debate and quiry. CO2 as a catalyst, however, has been suspecte help trigger the large catastrophic outflows (e.g., Bake al., 1991). Another point of view to the concept of episodic ocea and lakes put forth in this work has recently been purpo by Segura et al. (2002), who estimate that the heat ge ated by impacts could provide a sufficient amount of wa vapor to incorporate into the atmosphere, from the con sion of shallow subsurface or polar ice and water loca derived from the impactors themselves. As the peak in h land valley network formation is roughly coincident wi the period of heavy bombardment, Segura et al. (2002) clude that rains excavating the valley networks were t gered almost exclusively by large body impacts, a proc which frequency made the persistence of temperate and conditions impossible. On the other hand, it has been gested that impacts have contributed to a loss of plane atmosphere especially during the period of heavy bomb ment (Jakosky and Phillips, 2001). In addition, large imp events transfer a large amount of dust into the atmosph The higher the impact frequency, the more dust in the mosphere triggering a possible chained reaction that l to significant cooling of Mars during a time period wh snowfall becomes more likely than rainfall and the form tion of rivers does not occur until melt occurs and/or s glacial flows develop. Therefore, it is hard for us to rec cile the model proposed by Segura et al. (2002). Furt more, although valley network formation may have dev oped as a result of large impact events, there are many viable contributors to valley network formation and oth Noachian aqueous activity (e.g., Scott et al., 1995; Tan et al., 1998; Craddock and Howard, 2002; Carr, 2002). impact idea, therefore, does not invalidate the hypoth of a great Noachian ocean: similar processes may hav curred in the terrestrial Archean oceans. Moreover, va network formation is observed also during the Amazon Period, although in a reduced degree (Baker et al., 1 Scott et al., 1995); in fact, 10% (Carr, 1995) to 40% (Sc and Dohm, 1992) of valley networks may be younger th Noachian. Some of them are unambiguously Hesperia Amazonian, including the valleys on the volcanoes Alba tera, Ceraunius Tholus, and Hecates Tholus, and the t taries incised into the walls of the central and western Va Marineris (Carr, 2002). Thus, impact cratering during Noachian, in our opinion, represents one of many proce that reshaped the martian landscape, and certainly doe rule out the possibility of a northern plains ocean during 62 A.G. Fairén et al. / Icarus 165 (2003) 53–67 r et of ere ater ter- ra- ion ally ater /or x- ntly ping phic d in esis Late ion of For ater haps wa- us, sible tua ons his- su and ar- ghly on ord enc hy m- ards ers one by ion ean r the t its the are n nd ol- es- an e 1 lso ms the reas 02). difi- 01; nd e. port he ing: rad- ds, .g., ally ation ero- high sec- ble ity ice the 00), the 0– the ies ory nd ore- ntial re- of d in ity oded l- ern pable an out- d et ore- d et ins in Noachian by MEGAOUTFLO-related phenomena (Bake al., 1991). Clifford (1993) proposed that if the planetary inventory outgassed H2O exceeded the pore volume of the cryosph by more than a few percent, a subpermafrost groundw system of global extent would necessarily result. This in connected global aquifer would allow the downward mig tion of polar basal melt to result in the upward migrat of water at temperate and equatorial latitude, theoretic sufficient enough to replenish regions that have lost w by magmatic and tectonic activity (including impact) and sublimation. Although Clifford’s model can adequately e plain recharge of water to Tharsis, it does not sufficie address the surface morphology ranging from small sap channels, mid-size valley network systems, to catastro outflow channel systems, which are largely concentrate the Tharsis region. In our opinion, the inundation hypoth best explains the diverse evidence, especially from the Noachian until present. Now we will discuss the inundat hypothesis. We cannot accurately define the exact number MEGAOUTFLO events the planet has passed through. example, it is possible that two large-scale standing w bodies may have existed in very close periods, or per even overlapping periods, allowing almost the same ter amount to stand at similar topographic positions. Th topographic data may be inadequate to detect all pos shorelines. In consequence, this work offers a concep model that highlights the Tharsis-induced flood inundati in the northern plains that are recorded in the geologic tory. These are the events that have appeared to have vived thousands and perhaps millions of years of wind water erosion to stand out as significant windows of the M tian past, while smaller windows have been erased or hi subdued from the Martian record. 4.1. The early ocean Long-term stability of a primordial hemispheric ocean the lowlands of Mars has been recently argued by Cliff and Parker (2001), who established that the former pres of such an ocean can be deduced by considering the draulic conditions needed to explain the origin of the circu Chryse outflow channels, and extrapolating them backw in time. Furthermore, if the outflow channel flood wat were derived from a subpermafrost aquifer, as much as third of the planet’s surface would have been covered standing bodies of water and ice throughout its first bill years of evolution. Their conclusions indicate that an oc on Mars (as on Earth) may have condensed shortly afte planet was formed. The total water amount on Mars’ surface throughou history remains uncertain. But, if the average depth of Tharsis basin is between 2 and 7 km, and its measured is approximately 9× 106 km2 (Dohm et al., 2001b), the the fill volume for an average depth of 5 km is arou l r- e - a 4.5 × 107 km3, much more than the equivalent to the v ume of water required to create the secondary Late H perian ocean (1.4×107 km3, Head et al., 1999), but less th the volume required to filling the mean level of Shorelin (> 108 km3). Thus, other hydrologic activities may have a contributed water to the primitive ocean, by mechanis such as mobilization of large quantities of ground ice in southern highlands and spring-fed activity along many a of the highland/lowland boundary (e.g., Tanaka et al., 20 In this sense, there is evidence of great landscape mo cation rates on Mars during Noachian (Phillips et al., 20 Hynek and Phillips, 2001) and Early Hesperian (Irwin a Howard, 2002) due to the erosion by water at the surfac There are a large number of observations that sup the potential stability of this large body of water during t Noachian and possibly into the Early Hesperian, includ the higher erosion rates estimated for the lowlands (C dock et al., 1997); the valley networks in the highlan which major activity is documented through Noachian (e Squyres and Kasting, 1994; Carr, 2002), and which typic V or U-shaped cross-section suggests a gradual form process, transitioning from water-related to ice-related sional mechanisms (Baker and Partridge, 1986); the erosion rates during the Noachian and the extensive dis tion of Noachian terrains by valley networks, compara to terrestrial erosion rates, in a long-term fluvial activ recently suggested by Mars Odyssey-THEMIS data (R et al., 2003), dropping 4–6 orders of magnitude into Hesperian and Amazonian (Golombek and Bridges, 20 and in particular the timing of extensive denudation of highlands, approximated to be limited to an interval of 35 500 myr in the Noachian (Hynek and Phillips, 2001); or higher early heat flow (Zuber et al., 2000), which impl that a much larger fraction of the planet’s total invent of H2O would have been present as a liquid (Clifford a Parker, 2001). 4.2. The later ocean For the case of the secondary and inner ocean, Sh line 2 represents a surface of nearly constant equipote (Head et al., 1999), which is consistent with its interp tation as a paleoshoreline. In fact, the equipotentiallity this Late Hesperian ocean’s shoreline is only disrupte Tharsis and Elysium, where Amazonian volcanic activ have taken place. So, Shoreline 2 should have been er by a fluid in hydrostatic equilibrium, and it remains a most completely equipotential, due to its relatively mod age and the presence of less energetic processes ca of deformation. Additional supporting to Shoreline 2 as ancient shoreline are: debouching altitude of Chryse flow channels is close to contact 2 mean level (Hea al., 1999; Ivanov and Head, 2001); surface under Sh line 2 is smoother at all scales than that above (Hea al., 1998); the northern lowlands are covered with pla units that are mostly Late Hesperian–Early Amazonian Episodic flood inundations of the northern plains of Mars 63 987 im- ose arge 02) he sid- 2) as ater vol- ean a- ds rian u- 002; yse ave 01). ted ow ent 77; ned tion of -age ulse 1b) ice he 991; len- . act in ges bro rel- s a mer ge. la- and 2.5 le ap- enc Ely- a- rian- asin ted 8). a- 995; out- he 5; dur- noff rly ow- the s of tied nis rged ght dies em also the ar- deas f the nels clud- nic ac- 89; 1a, rian sec- ma- ater hin ent d of ons and nd lec- a- dry p- cott ent low- ilar and and age (e.g., Scott and Tanaka, 1986; Greeley and Guest, 1 Tanaka et al., 2002); the contact itself is close to the l its of the late Hesperian Vastitas Borealis Formation, wh sedimentary deposits strongly suggest the stability of a l body of water at the level of Shoreline 2 (Head et al., 20 at the time of peak of outflow channel activity during t Late Hesperian (Tanaka, 1986), and which are also con ered by some authors (e.g., Kreslavsky and Head, 200 possibly originated as a sublimation residue of a large w body; and topography of features interpreted as sub-ice canoes suggests an ice top in a level very close to the m elevation of the Shoreline 2, which is consistent with a w ter body totally or partially frozen in the northern lowlan (Chapman, 2003). Its estimated age is∼ 2–3 Gyr. (Clifford and Parker, 2001), also consistent with the Late Hespe age postulated here. A distinct episode of outflow channel activity is doc mented for the Late Hesperian (Tanaka, 1986; Carr, 2 Fig. 4), the majority concentrated on the margin of Chr Planitia. In fact, outflow channel formation seems to h peaked at this time (Head et al., 2001; Masson et al., 20 This and earlier outflow channel activity, especially rela to the incipient development of the cirum-Chryse outfl channels, is probably driven by the continued developm of Tharsis and Valles Marineris (e.g., Masursky et al., 19 Dohm et al., 2001a), and the release of water from confi aquifers due to cracking of the cryosphere by dike injec (Wilson and Head, 2003). In addition, complex networks drainage channels have been described in Hesperian terrains (Scott et al., 1995). Therefore, the fourth great p (Stage 4) of Tharsis magmatic activity (Dohm et al., 200 probably triggered a massive regional melting of ground and the heating of CO2-charge groundwater that started t discharge of groundwater to the surface (Baker et al., 1 McKenzei and Nimmo, 1999; Head and Pratt, 2001), rep ishing the lowlands with a secondary and smaller ocean 4.3. Late Hesperian to Amazonian paleolakes Cabrol and Grin (1999) have catalogued 179 imp crater lakes at Viking resolution for the whole planet; this study, a relative age was suggested for the 30 lar and these were mostly Late Hesperian to Amazonian. Ca and Grin (2001) proposed a method to narrow down the ative age for all of the 179 crater lakes, which include summary of their findings that suggests periods of war and wetter climatic conditions and environmental chan Also, Cabrol and Grin (2001) estimate that the main custrine activity is placed between 3.1 and 1.8 Gyr, rises to a maximum in the Early Amazonian (about to 2.1 Gyr). Aqueous activity continues into the Midd and Late Amazonian, though during a time when lakes pear to be less numerous. Rice (1997) presented evid for wide age-ranging aqueous sedimentary deposits in sium Basin, pointing to the stability of Noachian to Am zonian lacustrine environments. Such Noachian/Hespe ; d t l e aqueous activity is also noted for Gusev impact crater b where recurrent Ma’adim Vallis flooding may have resul in episodic lakes (Scott et al., 1993; Cabrol et al., 199 In addition, the presence of a significant volume of w ter in Utopia Basin has been proposed (Scott et al., 1 Thomson and Head, 2001), as well as the subsurface flow and surface runoff from the northwest flanks of t Elysium rise into the Utopia basin (Mouginis-Mark, 198 Greeley and Guest, 1987; Skinner and Tanaka, 2001) ing the Late Hesperian and Amazonian and surface ru in north Terra Cimmeria from the Late Noachian to Ea Amazonian (Nelson et al., 2001). Scott et al. (1995) relative age-dated highland and l land channels to determine their possible contribution to size and stability of the paleolakes. The main supplier water to the lowlands were the outflow channels that emp into Chryse Basin, and the ones that flowed into Amazo and Elysium basins. Water and sediment volumes discha by the Chryse outflow channels, several millions of km3 in a short period of time (Rotto and Tanaka, 1991b), mi have been enough to have filled large-standing water-bo within the basin. Floods from the Elysium channel syst and the valley networks around the lowlands may have contributed to a depositional sedimentary environment in northern plains. In any case, Scott et al. (1995) found that 56% of the m tian channels are Noachian in age, consistent with the i stated here about the first great ocean. Therefore, 44% o channels are younger than Noachian (Fig. 4); these chan are interpreted to have formed by several processes, in ing local hydrothermal activity due to impacts or tecto deformation (Brakenridge et al., 1985) and magmatic tivity (Baker and Partridge, 1986; Gulick and Baker, 19 Gulick, 1998; Tanaka et al., 1998; Dohm et al., 200 2001b, 2001c). Most of these channels are Late Hespe (31%), consistent with the emplacement epoch for the ondary ocean; but the wide range in age (Hesperian to A zonian) of the channels that supplied the basins with w mark long periods of intermittent presence of water wit the northern lowlands of Mars. In fact, valley developm seems to have substantially decreased when the perio late heavy bombardment ceased, including a∼ 2-km decline in the mean elevation of outflow channel source regi between the Late Hesperian and Amazonian (Clifford Parker, 2001); but it did not cease completely (Gulick a Baker, 1990), as the post-Noachian valley formation col tively point towards brief warm, wet Hesperian and Am zonian excursions from prevailing, present-day cold and desert conditions (Baker, 2001). In particular, extensive flooding in the lowlands, su ported by channel-meander features in many areas (S et al., 1995), which would have flowed over the anci sea-floor plains are described. The channels on the lands were formed during the Amazonian Period, sim to those on the flanks of volcanoes in Tharsis (Gulick Baker, 1989) or Elysium (Skinner and Tanaka, 2001), 64 A.G. Fairén et al. / Icarus 165 (2003) 53–67 man ave rob- rred ow- and a- por- in th accu ple - al., hick 97; (lat ly ore- the d to f dif- land So, ex- d/or ions nsis ob- ime r ere ing hout ach- ars; in th an; ay ers ino me Ely- ma- the om- ffer- n to and ling per tic- in by lpful son, es, tal ex- D., , C., field bits. .C., ters est- on- –236. and de- ale, le on 2. der, on, o- cient in the Medusae Fossae Formation (Scott and Chap 1991). By this time, outflow channel activity appears to h declined and become more localized around regions of p able geothermal activity (e.g., Baker et al., 1992). 5. Highland features Water-associated processes could have similarly occu in the highlands at the same time as those in the l lands (e.g., Baker et al., 1991; Baker, 2001; Head Pratt, 2001). Impact cratering that resulted in the form tion of large basins appears to be one of the most im tant processes on Mars to produce closed depressions cratered highlands. In some of these basins, massive mulation of sediments have been described. For exam Gusev impact crater (15◦, 184◦) is rounded by the rem nants of deltaic deposits from Ma’adim Vallis (Cabrol et 1996), and sediments in Argyre are hundreds of meters t and as much as kilometres thick in Hellas (Parker, 19 Moore and Wilhelms, 2001). In addition, Cassini Crater 24◦ N., long 328◦) shows well-defined terraces in its high conserved rim morphology, similar to terrestrial paleosh lines. Basin broadening by lakeshore degradation during late Noachian and early Hesperian has been propose explain these features (Parker et al., 1989). Channels o ferent ages are equally bordering some of these high paleolakes in impact craters (e.g., Scott et al., 1995). also in the highlands, when suitable conditions may have isted on Mars perhaps related to Tharsis activity, lakes an ice bodies may have occupied local topographic depress such as in impact crater/tectonic basins. 6. Conclusions The hydrogeological hypothesis presented here, co tent with available data, provides an explanation for served geological patterns of water evolution through t on the surface of Mars, and should help to motivate furthein situ investigations. The hydrological model articulated h for the water evolution on the surface of Mars isconsistent with: (1) The compelling evidence for a potentially long-last and great ocean covering the martian lowlands throug the Noachian and possibly into the Early Hesperian, re ing a duration close to several hundreds of millions of ye a Late Hesperian sea extending over the deeper areas lowlands inset within the boundary of the first great oce and a number of widely distributed minor lakes, which m represent a reduced Late Hesperian sea, or ponded wat the deeper reaches of the northern plains, related to m Amazonian Tharsis and Elysium reduced flooding. (2) The five specific stages of the Tharsis Superplu development, as well as possible contributions from the , e - , , - e in r sium rise development during the Hesperian into the A zonian, and their influence on the water evolution on surface of Mars. The modelexplains: (1) Why only two coastlines can be roughly traced as c plete closures of ancient oceanic basins; and the di ences in equipotentiallity between them. (2) Paleolake recurrent formation during Late Hesperia Amazonian, even in almost contemporary times, their wide range in level and in age. And the modelpredicts: (1) If the two Superplumes have not reached their coo threshold, future water refilling may occur in the dee lowlands, when they begin another stage of magma driven activity. Acknowledgments We thank Kenneth Tanaka for the topographic map Fig. 2, and Tayfun Oner for Fig. 3A production. Reviews Nathalie Cabrol and James Skinner were extremely he in the work of this paper. References Acuña, M.H., Connerney, J.E.P., Ness, N.F., Lin, R.P., Mitchell, D., Carl C.W., McFadden, J., Anderson, K.A., Rème, H., Mazelle, C., Vign D., Wasilewski, P., Cloutier, P., 1999. Global distribution of crus magnetization discovered by the Mars Global Surveyor MAG/GR periment. Science 284, 790–793. Acuña, M.H., Connerney, J.E.P., Wasilewski, P., Lin, R.P., Mitchell, Anderson, K.A., Carlson, C.W., McFadden, J., Réme, H., Mazelle Vignes, D., Bauer, S.J., Cloutier, P., Ness, N.F., 2001. Magnetic of Mars: summary of results from the aerobraking and mapping or J. Geophys. Res. 106, 23403–23417. Anderson, R.C., Dohm, J.M., Golombek, M.P., Haldemann, A.F Franklin, B.J., Tanaka, K.L., Lias, J., Peer, B., 2001. Primary cen and secondary concentrations of tectonic activity through time in w ern hemisphere of Mars. J. Geophys. Res. 106, 20563–20585. Baker, V.R., 1999. The MEGAOUTFLO hypothesis for long-term envir mental change on Mars. Bull. Am. Astron. Soc. 31, 1133. Baker, V.R., 2001. Water and the martian landscape. Nature 412, 228 Baker, V.R., Milton, D.J., 1974. Erosion by catastrophic floods on Mars Earth. Icarus 23, 27–41. Baker, V.R., Partridge, J.B., 1986. Small martian valleys: pristine and graded morphology. J. Geophys. Res. 91, 3561–3572. Baker, V.R., Strom, R.G., Gulick, V.C., Kargel, J.S., Komatsu, G., K V.S., 1991. Ancient oceans, ice sheets and the hydrological cyc Mars. Nature 352, 589–594. Baker, V.R., Carr, M.H., Gulick, V.C., Williams, C.R., Marley, M.S., 199 Channels and valley networks. In: Kieffer, H.H., Jakosky, B.M., Sny C.W., Matthews, M.S. (Eds.), Mars. Univ. of Arizona Press, Tucs pp. 493–522. Baker, V.R., Strom, R.G., Dohm, J.M., Gulick, V.C., Kargel, J.S., K matsu, G., Ori, G.G., Rice, J.W., 2000. Mars’ Oceanus Borealis, an Episodic flood inundations of the northern plains of Mars 65 net. ical nar of ain. dc- rface and 2. ings tian ted y ex- ous . 29. ar- nts . d lake. n: a. iolog- rs: ars. c. gical 55. cts lley ures i. 44, pli- 131. olake a ma y of and Inv. 062 an- er. nce GS of ere: e of ian 453– ad- ap wet try ns of TM 03 gh a onf. ars: net. om, Sys- ent rth- are, .R., rs: 314. om, .H., ntial akes. and . Inv. are, geo- 39. eous rabia etic dy- Wet ars: rn- th, to, 29. lica- site. egion ial glaciers, and the MEGAOUTFLO hypothesis. In: Proc. Lunar Pla Sci. Conf. 31st, p. 1863. Abstract [CD-ROM]. Baker, V.R., Maruyama, S., Dohm, J.M., 2002. A theory for the geolog evolution of Mars and related synthesis (GEOMARS). In: Proc. Lu Planet. Sci. Conf. 33rd, p. 1586. Abstract [CD-ROM]. Barlow, N.G., Bradley, T.L., 1990. Martian impact craters: correlations ejecta and interior morphologies with diameter, latitude, and terr Icarus 87, 156–179. Berman, D.C., Hartmann, W.K., 2002. Recent fluvial, volcanic, and te tonic activity on the Cerberus plains of Mars. Icarus 159, 1–17. Boynton, W., Feldman, W., Prettyman, T., Hamara, D., 2002. Subsu ice content in the polar region of Mars: comparison between north south. Eos Trans. AGU 83 (47). Fall Meet. Suppl., Abstract P11B-0 Brakenridge, G.R., Newsom, H.E., Baker, V.R., 1985. Ancient hot spr on Mars: origin and paleoenvironmental significance of small mar valleys. Geology 13, 859–862. Burr, D.M., Grier, J.A., McEwen, A.S., Keszthely, L.P., 2002a. Repea aqueous flooding from Cerberus Fossae: evidence for very recentl tant, deep groundwater in Mars. Icarus 159, 53–73. Burr, D.M., McEwen, A.S., Sakimoto, S.E., 2002b. Recent aque floods from the Cerberus Fossae, Mars. Geophys. Res. Lett 10.1029/2001GL013345. Cabrol, N.A., Grin, E.A., 1999. Distribution, classification and ages of m tian impact crater lakes. Icarus 142, 160–172. Cabrol, N.A., Grin, E.A., 2001. The evolution of lacustrine environme on Mars: is Mars only hydrologically dormant? Icarus 149, 291–328 Cabrol, N.A., Grin, E.A., Dawidowicz, G., 1996. Ma’adim Vallis revisite through new topographic data: evidence for an ancient intravalley Icarus 123, 269–284. Cabrol, N.A., Grin, E.A., Landheim, R., 1998. Ma’adim Vallis evolutio geometry and models of discharge rate. Icarus 132, 362–377. Cabrol, N.A., Wynn-Williams, D.D., Crawford, D.A., Grin, E.A., 2001 Recent aqueous environments in martian impact craters: an astrob ical perspective. Icarus 154, 98–113. Cabrol, N.A., Grin, E.A., Dohm, J.M., 2001b. From gullies to glacie a continuum of evidence supporting a recent climate change on M EOS Trans. AGU 82 (47). Abstract F694. Cabrol, N.A., Wynn-Williams, D.D., Crawford, D.A., Grin, E.A., 2001 Recent aqueous environments in impact craters and the astrobiolo exploration of Mars. In: Proc. Lunar Planet. Sci. Conf. 32nd, p. 12 Abstract [CD-ROM]. Carr, M.H., 1990. D/H on Mars: The effect of floods, volcanism, impa and polar processes. Icarus 87, 210–227. Carr, M.H., 1995. The martian drainage system and the origin of va networks and fretted channels. J. Geophys. Res. 100, 7478–7507. Carr, M.H., 1996a. Water on Mars. Oxford Univ. Press, New York. Carr, M.H., 1996b. Channels and valleys on Mars: cold climate feat formed as a result of a thickening cryosphere. Planet. Space Sc 1411–1423. Carr, M.H., 2002. Elevations of water-worn features on Mars: im cations for circulation of groundwater. J. Geophys. Res. 107, 5 10.1029/2002JE001845. Chapman, M.G., 1994. Evidence, age, and thickness of a frozen pale in Utopia Planitia, Mars. Icarus 109, 393–406. Chapman, M.G., 2003. Sub-ice volcanoes and ancient oceans/lakes: tian challenge. Global Planet. Change 35, 185–198. Chapman, C.R., Jones, K.L., 1977. Cratering and obliteration histor Mars. Ann. Rev. Earth Planet. Sci. 5, 515–540. Chapman, M.G., Tanaka, K.L., 1993. Geologic map of the MTM-05152 -10152 quadrangles, Mangala Valles region of Mars. USGS Misc. Ser. Map I-2294 (1:500,000). Chapman, M.G., Tanaka, K.L., 1996. Geologic maps of the MTM 25 quadrangle (digital compilation) and the MTM 25067 quadrangle (m ual compilation), Kasei Valles region of Mars. USGS Misc. Inv. S Map I-2398 (1:500,000). Chapman, M.G., Masursky, H., Scott, D.H., 1991. Geologic map of scie study area 2, north Kasei Valles, Mars (MTM 25072 quadrangle). US Misc. Inv. Ser. Map I-2107 (1:500,000). r- Clifford, S.M., 1993. A model for the hydrologic and climate behaviour water on Mars. J. Geophys. Res. 98, 10973–11016. Clifford, S.M., Parker, T.J., 2001. The evolution of the martian hydrosph implications for the fate of a primordial ocean and the current stat the northern plains. Icarus 154, 40–79. Craddock, R.A., Maxwell, T.A., 1993. Geomorphic evolution of the mart highlands through ancient fluvial processes. J. Geophys. Res. 98, 3 3468. Craddock, R.A., Greeley, R., 1994. Geologic map of the MTM-20147 qu rangle, Mangala Valles region of Mars. USGS Misc. Inv. Ser. M I-2310 (1:500,000). Craddock, R.A., Howard, A.D., 2002. The case for rainfall on a warm, early Mars. J. Geophys. Res. 107. 10.1029/2001JE001505. Craddock, R.A., Maxwell, T.A., Howard, A.D., 1997. Crater morphome and modification in the Sinus Sabaeus and Margaritifer Sinus regio Mars. J. Geophys. Res. 102, 13321–13340. DeHon, R.A., 1992. Geologic map of the Pompeii quadrangle (M 20057), Maja Valles region of Mars. USGS Misc. Inv. Ser. Map I-22 (1:500,000). DeHon, R.A., Pani, E.A., 1992. Duration and rates of discharge throu martian outflow system: Maja Valles. In: Proc. Lunar Planet. Sci. C 23rd, pp. 297–298. Abstract [CD-ROM]. Dohm, J.M., Tanaka, K.L., 1999. Geology of the Thaumasia region, M plateau development, valley origins, and magmatic evolution. Pla Space Sci. 47, 411–431. Dohm, J.M., Anderson, R.C., Baker, V.R., Ferris, J.C., Hare, T.M., Str R.G., Rudd, L.P., Rice, J.W., Casavant, R.R., Scott, D.H., 2000. tem of gigantic valleys northwest of Tharsis Montes, Mars: lat catastrophic flooding, northwest watershed, and implications for no ern plains ocean. Geophys. Res. Lett. 27 (21), 3559–3562. Dohm, J.M., Anderson, R.C., Baker, V.R., Ferris, J.C., Rudd, L.P., H T.M., Rice Jr., J.W., Casavant, R.R., Strom, R.G., Zimbelman, J Scott, D.H., 2001a. Latent outflow activity for western Tharsis, Ma significant flood record exposed. J. Geophys. Res. 106, 12301–12 Dohm, J.M., Ferris, J.C., Baker, V.R., Anderson, R.C., Hare, T.M., Str R.G., Barlow, N.G., Tanaka, K.L., Klemaszewski, J.E., Scott, D 2001b. Ancient drainage basin of the Tharsis region, Mars: Pote source for outflow channel systems and putative oceans or paleol J. Geophys. Res. 106, 32943–32958. Dohm, J.M., Tanaka, K.L., Hare, T.M., 2001c. Geologic, paleotectonic, paleoerosional maps of the Thaumasia region of Mars. USGS Misc Ser. Map I-2650, scale 1:5,000,000. Dohm, J.M., Maruyama, S., Baker, V.R., Anderson, R.C., Ferris, J.C., H T.M., 2002. Plate tectonism on early Mars: diverse geological and physical evidence. In: Proc. Lunar Planet. Sci. Conf. 33rd, p. 16 Abstract [CD-ROM]. Edgett, K.S., Parker, T.J., 1997. Water on early Mars: possible subaqu sedimentary deposits covering ancient cratered terrain in western A and Sinus Meridiani. Geophys. Res. Lett. 24, 2897–2900. Fairén, A.G., Ruiz, J., Anguita, F., 2002. An origin for the linear magn anomalies on Mars through accretion of terranes: implications for namo timing. Icarus 160, 220–223. Ferris, J.C., Dohm, J.M., 2002. Comparative Hydrogeomorphology of Beaver Creek, Arizona, and Abus Vallis, Mars. Icarus. Submitted. Frey, H.V., 2003. Large-diameter visible and buried impact basins on M implications for age of the highlands and (buried) lowlands and tu off of the global magnetic field. In: Proc. Lunar Planet. Sci. Conf. 34 p. 1838. Abstract [CD-ROM]. Frey, H.V., Roark, J.H., Shockey, K.M., Frey, E.L., Sakimo S.H., 2002. Ancient lowlands on Mars. Geophys. Res. Lett. 10.1029/2001GLO13832. Golombek, M.P., Bridges, N.T., 2000. Erosion rates on Mars and imp tions for climate change: constraints from the Pathfinder landing J. Geophys. Res. 105, 1841–1853. Greeley, R., Guest, J.E., 1987. Geologic map of the eastern equatorial r of Mars. USGS Misc. Inv. Ser. Map I-1802B (1:15,000,000). Gulick, V.C., 1998. Magmatic intrusions and hydrothermal origin for fluv valleys on Mars. J. Geophys. Res. 103, 19365–19387. 66 A.G. Fairén et al. / Icarus 165 (2003) 53–67 tes. ian olu- ory: n of D- shee and ert, ars: ata. m- Mars Jau- 01. . ars: n in 5. umu- 319. at- f the ian 818. con- r hy- eter ati- Na- . Geo and In: M]. 20, haw, n in gini, plo- roc. s in as a hys. tism 174, Na- res. eo- osits 174. . D., or- –130. hern eo- Sci- page eol. n, R., . 96, 77. on oc. the on, T., oore, 999. ition hys. ars. ars. ice- ium nels 8669. uth nar asins ro- ica- Geo- on -01, l mor- ons 5. erger, s. J. .M., n- hy- bek, . In: Gulick, V.C., Baker, V.R., 1989. Fluvial valleys and martian palaeoclima Nature 341, 514–516. Gulick, V.C., Baker, V.R., 1990. Origin and evolution of valleys on mart volcanoes. J. Geophys. Res. 95, 14325–14334. Hartmann, W.K., Neukum, G., 2001. Cratering chronology and the ev tion of Mars. Space Sci. Rev. 96, 165–194. Head, J.W., 2002. The hydrological cycle in the last 80% of Mars hist sources, locations, residence times, stability, duration and evolutio H2O. In: Proc. Lunar Planet. Sci. Conf. 33rd, p. 1724. Abstract [C ROM]. Head, J.W., Pratt, S., 2001. Extensive Hesperian-aged south polar ice on Mars: evidence for massive melting and retreat, and lateral flow ponding of meltwater. J. Geophys. Res. 106, 12275–12299. Head, J.W., Kreslavsky, M., Hiesinger, H., Ivanov, M.A., Pratt, S., Seib N., Smith, D.E., Zuber, M.T., 1998. Oceans in the past history of M test for their presence using Mars Orbiter Laser Altimeter (MOLA) d Geophys. Res. Lett. 25, 4401–4404. Head, J.W., Hiesinger, H., Ivanov, M.A., Kreslavsky, M.A., Pratt, S., Tho son, B.J., 1999. Possible ancient oceans on Mars: evidence from Orbiter Laser Altimeter data. Science 286, 2134–2137. Head, J.W., Greeley, R., Golombek, M.P., Hartmann, W.K., Hauer, E., mann, R., Masson, P., Neukum, G., Nyquist, L.E., Carr, M.H., 20 Geological processes and evolution. Space Sci. Rev. 96, 263–292 Head, J.W., Kreslavsky, M.A., Pratt, S., 2002. Northern lowlands on M evidence for widespread volcanic flooding and tectonic deformatio the Hesperian period. J. Geophys. Res. 107. 10.1029/2000JE0144 Hess, P.C., Parmentier, E.M., 2001. Implications of magma ocean c late overturn on Mars. In: Proc. Lunar Planet. Sci. Conf. 32nd, p. 1 Abstract [CD-ROM]. Hoffman, N., 2000. White Mars: a new model for Mars’ surface and mosphere based on CO2. Icarus 146 (2), 326–342. Hynek, B.M., Phillips, R.J., 2001. Evidence for extensive denudation o martian highlands. Geology 29, 407–410. Irwin III, R.P., Howard, A.D., 2002. Drainage basin evolution in Noach Terra Cimmeria, Mars. J. Geophys. Res. 107. 10.1029/2001JE001 Ivanov, M.A., Head, J.W., 2001. Chryse Planitia, Mars: topographic figuration, outflow channel continuity and sequence, and tests fo pothesized ancient bodies of water using Mars Orbiter Laser Altim (MOLA) data. J. Geophys. Res. 106, 3275–3295. Jakosky, B.M., Carr, M.H., 1985. Possible precipitation of ice at low l tudes of Mars. Nature 315, 559–561. Jakosky, B.M., Phillips, R.J., 2001. Mars’ volatile and climate history. ture 412, 237–244. Janhunen, P., 2002. Are the northern plains of Mars a frozen ocean? J phys. Res. 107. 10.1029/2000JE001478. Jöns, H.P., 1986. Arcuate ground undulations, gelifluxion-like features “front tori” in the northern lowlands on Mars. What do they indicate? Proc. Lunar Planet. Sci. Conf. 17th, pp. 404–405. Abstract [CD-RO Kargel, J.S., Strom, R.G., 1992. Ancient glaciation on Mars. Geology 3–7. Kargel, J.S., Baker, V.R., Beget, J.E., Lockwood, J.F., Pewe, T.L., S J.S., Strom, R.G., 1995. Evidence of ancient continental glaciatio the martian northern plains. J. Geophys. Res. 100, 5351–5368. Komatsu, G., Kargel, J.S., Baker, V.R., Strom, R.G., Ori, G.G., Mosan C., Tanaka, K.L., 2000. A chaotic terrain formation hypothesis: ex sive outgas and outflow by dissociation of clathrate on Mars. In: P Lunar Planet. Sci. Conf. 31st, p. 1434. Abstract [CD-ROM]. Kreslavsky, M.A., Head, J.W., 2002. Fate of outflow channel effluent the northern lowlands of Mars: the Vastitas Borealis Formation sublimation residue from frozen ponded bodies of water. J. Geop Res. 107 (E12), 5121. 10.1029/2001JE001831. Lachenbruch, A.H., Morgan, P., 1990. Continental extension, magma and elevation; formal relations and rules of thumb. Tectonophysics 39–62. Laskar, J., Robutel, P., 1993. The chaotic obliquity of the planets. ture 361, 608–612. Lucchitta, B.K., 1981. Mars and Earth: comparison of cold-climate featu Icarus 45, 264–303. t - Lucchitta, B.K., 1982. Ice sculpture in the martian outflow channels. J. G phys. Res. 87, 9951–9973. Lucchitta, B.K., Ferguson, H.M., Summers, C., 1986. Sedimentary dep in the northern lowland plains, Mars. J. Geophys. Res. 91, E166–E Lundin, R., 2001. Erosion by the solar wind. Science 291, 1909–1910 Mahaney, W.C., Dohm, J.M., Baker, V.R., Newsom, H.E., Malloch, Hancock, R.G.V., Campbell, I., Sheppard, D., Milner, M.W., 2001. M phogenesis of Antarctic paleosols: martian analog. Icarus 154, 113 Malin, M.C., Edgett, K.S., 1999. Oceans or seas in the martian nort lowlands: high resolution imaging tests of proposed coastlines. G phys. Res. Lett. 26, 3049–3052. Malin, M.C., Edgett, K.S., 2000a. Sedimentary rocks of early Mars. ence 290, 1927–1937. Malin, M.C., Edgett, K.S., 2000b. Evidence for recent groundwater see and surface runoff on Mars. Science 288, 2330–2335. Mars Channel Working Group, 1983. Channels and valleys on Mars. G Soc. Am. Bull. 94, 1035–1054. Masson, P., Carr, M.H., Costard, F., Greeley, R., Hauber, E., Jauman 2001. Geomorphologic evidence for liquid water. Space Sci. Rev 333–364. Masursky, H., Boyce, J.M., Dial, A.L., Schaber, G.G., Strobell, M.E., 19 Classification and time of formation of martian channels based Viking data. J. Geophys. Res. 82, 4016–4038. McGill, G.E., 1985. Age and origin of large martian polygons. In: Pr Lunar Planet. Sci. Conf. 16th, pp. 534–535. Abstract [CD-ROM]. McKenzei, D., Nimmo, F., 1999. The generation of martian floods by melting of ground ice above dykes. Nature 397, 231–233. McSween Jr., H.Y., Murchie, S.L., Crisp, J.A., Bridges, N.T., Anders R.C., Bell III, J.F., Britt, D.T., Brückner, J., Dreibus, G., Economou, Ghosh, A., Golombeck, M.P., Greenwood, J.P., Johnson, J.R., M H.J., Morris, R.V., Parker, T.J., Rieder, R., Singer, R., Wänke, H., 1 Chemical, multispectral, and textural constrainst on the compos and origin of rocks at the Mars Pathfinder landing site. J. Geop Res. 104, 8679–8715. Melosh, H.J., 1980. Tectonic patterns on a reoriented planet: M Icarus 44, 745–751. Milton, D.J., 1974. Geologic map of the Lunae Palus quadrangle of M USGS Misc. Inv. Ser. Map I-894 (1:5,000,000). Moore, J.M., Wilhelms, D.E., 2001. Hellas as a possible site of ancient covered lakes on Mars. Icarus 154, 258–276. Mouginis-Mark, P.J., 1985. Volcano/ground ice interactions in Elys Planitia, Mars. Icarus 64, 265–284. Nelson, D.M., Greeley, R., 1999. Geology of Xanthe Terra outflow chan and the Mars Pathfinder landing site. J. Geophys. Res. 104, 8653– Nelson, D.M., Farmer, J.D., Greeley, R., 2001. Hydrologic history of so Elysium basin—north Terra Cimmeria area, Mars. In: Proc. Lu Planet. Sci. Conf. 32nd, p. 2069. Abstract [CD-ROM]. Parker, T.J., 1997. Fluvial and lacustrine degradation of large martian b during the Noachian. Workshop on Early Mars: Geologic and Hyd logic Evolution, Physical and Chemical Environments, and the Impl tions for Life, Houston, TX. Abstract 3073. Parker, T.J., Schneeberger, D.M., Pieri, D.C., Saunders, R.S., 1987. morphic evidence for ancient seas on Mars. In: MECA Symposium Mars: Evolution of its Climate and Atmosphere, LPI Tech. Rept. 87 Lunar and Planet. Inst., Houston, TX, pp. 96–98. Parker, T.J., Saunders, R.S., Schneeberger, D.M., 1989. Transitiona phology in the west Deuteronilus Mensae region of Mars: implicati for modification of the lowland/upland boundary. Icarus 82, 111–14 Parker, T.J., Gorsline, D.S., Saunders, R.S., Pieri, D.C., Schneeb D.M., 1993. Coastal geomorphology of the martian northern plain Geophys. Res. 98, 11061–11078. Phillips, R.J., Zuber, M.T., Solomon, S.C., Golombek, M.P., Jakosky, B Banerdt, W.B., Smith, D.E., Williams, R.M.E., Hynek, B.M., Aharo son, O., Hauck, S.A., 2001. Ancient geodynamics and global-scale drology on Mars. Science 291, 2587–2591. Pounders, E., Anderson, R.C., Dohm, J.M., Haldemann, A.F.C., Golom M.P., 2002. Tectonic evolution of the eastern hemisphere of Mars Proc. Lunar Planet. Sci. Conf. 33rd, p. 1906. Abstract [CD-ROM]. Episodic flood inundations of the northern plains of Mars 67 ake gic and ngle ap flu- roc. odes onf. ling r in ork- the . In: .H., ky, na ran- 08 eo- ns. D- s of 397 ment CD- orial . nce roc. ter- . nce Ser. nels ntal m. 5655 sig- ilar wet? rd, ., ss, test- 107. 139– ryse rs: n. In: ief- rs. eys ins. sur- ping l Edi- n of ution 259, 02. phys. omy ars? e of a 4th, of asalt Na- f the isc. 7. L., son, .D., ars 88– Rice, J.W., 1997. Elysium basin, Mars: An intermittent or perennial l from Noachian to Amazonian time. Workshop on Early Mars: Geolo and Hydrologic Evolution, Physical and Chemical Environments, the Implications for Life, Houston, TX. Abstract 3052. Rice Jr., J.R., DeHon, R.A., 1996. Geologic map of the Darvel Quadra (MTM 20052), Maja Valles region of Mars. USGS Misc. Inv. Ser. M I-2432 (1:500,000). Rice, J.W., Christensen, P.R., Ruff, S.W., Harris, J.C., 2003. Martian vial landforms: a THEMIS perspective after one year at Mars. In: P Lunar Planet. Sci. Conf. 34th, p. 2091. Abstract [CD-ROM]. Rotto, S.L., Tanaka, K.L., 1991a. Geologic history and channeling epis of the Chryse Planitia region of Mars. In: Proc. Lunar Planet. Sci. C 22nd, pp. 1135–1136. Abstract [CD-ROM]. Rotto, S.L., Tanaka, K.L., 1991b. Chryse Planitia region, Mars: channe history, flood-volume estimates, and scenarios for bodies of wate the northern plains. Mars Surface and Atmosphere through Time W shop, Boulder, CO, LPI Technical Memo 92-01, 111-112. Rotto, S.L., Tanaka, K.L., 1995. Geologic/geomorphologic map of Chryse Planitia region of Mars. USGS Misc. Inv. Ser. Map I-2441. Ruiz, J., Fairén, A.G., de Pablo, M.A., 2003. Thermal isostasy on Mars Proc. Lunar Planet. Sci. Conf. 34th, p. 1090. Abstract [CD-ROM]. Schubert, G., Solomon, S.C., Turcotte, D.L., Drake, M.J., Sleep, N 1992. Origin and thermal evolution of Mars. In: Kieffer, H.H., Jakos B.M., Snyder, C.W., Matthews, M.S. (Eds.), Mars. Univ. of Arizo Press, Tucson, pp. 147–183. Scott, D.H., 1993. Geologic map of the MTM 25057 and 25052 quad gles, Kasei Valles region of Mars. USGS Misc. Inv. Ser. I-Map 22 (1:500,000). Scott, D.H., Chapman, M.G., 1991. Mars Elysium basin: g logic/volumetric analyses of a young lake and exobiologic implicatio In: Proc. Lunar Planet. Sci. Conf. 22nd, pp. 669–677. Abstract [C ROM]. Scott, D.H., Chapman, M.G., 1995. Geologic and topographic map the Elysium Paleolake Basin, Mars. USGS Misc. Inv. Ser. Map I-2 (1:5,000,000). Scott, D.H., Dohm, J.M., 1992. Mars highland channels: an age assess In: Proc. Lunar Planet. Sci. Conf. 23rd, pp. 1251–1252. Abstract [ ROM]. Scott, D.H., Tanaka, K.L., 1986. Geologic map of the western equat region of Mars. USGS Misc. Inv. Ser. Map I-1802-A (1:15,000,000) Scott, D.H., Chapman, M.G., Rice, J.W., Dohm, J.M., 1991a. New evide of lacustrine basins on Mars: Amazonis and Utopia Planitiae. In: P Lunar Planet. Sci. Conf. 22nd, pp. 53–62. Abstract [CD-ROM]. Scott, D.H., Rice, M.G., Dohm, J.M., 1991b. Martian paleolakes and wa ways: exobiological implications. Orig. Life Evol. Bios. 21, 189–198 Scott, D.H., Dohm, J.M., Applebee, D.J., 1993. Geologic map of scie study area 8, Apollinaris Patera region of Mars. USGS Misc. Inv. MAP I-2351. Scott, D.H., Dohm, J.M., Rice, J.W., 1995. Map of Mars showing chan and possible paleolake basins. USGS Misc. Inv. Ser. MAP I-2461. Segura, T.L., Toon, O.B., Colaprete, A., Zahnle, K., 2002. Environme effects of large impacts on Mars. Science 298, 1977–1980. Skinner, J.A., Tanaka, K.L., 2001. Long-lived hydrovolcanism of Elysiu Eos. Trans. AGU 82 (47), Fall Meet. Suppl. Abstract P31B-07. Sleep, N.H., 1994. Martian plate tectonics. J. Geophys. Res. 99, 5639– Smith, D., Zuber, M.T., 1996. The shape of Mars and the topographic nature of the hemispheric dichotomy. Science 271, 184–188. Squyres, S.W., 1979. The distribution of lobate debris aprons and sim flows on Mars. J. Geophys. Res. 84, 8087–8096. . . Squyres, S.W., 1989. Water on Mars. Icarus 79, 229–288. Squyres, S.W., Kasting, J.F., 1994. Early Mars: how warm and how Science 265, 744–749. Squyres, S.W., Clifford, S.M., Kuzmin, R.O., Zimbelman, J.R., Costa F.M., 1992. Ice in the martian regolith. In: Kieffer, H.H., Jakosky, B.M Snyder, C.W., Matthews, M.S. (Eds.), Mars. Univ. of Arizona Pre Tucson, pp. 523–554. Stewart, S.T., Nimmo, F., 2002. Surface runoff features on Mars: ing the carbon dioxide formation hypothesis. J. Geophys. Res. 10.1029/2000JE001465. Tanaka, K.L., 1986. The stratigraphy of Mars. J. Geophys. Res. 91, 158. Tanaka, K.L., 1997. Sedimentary history and mass flow structures of Ch and Acidalia Planitiae, Mars. J. Geophys. Res. 102, 4131–4150. Tanaka, K.L., Banerdt, W.B., 2000. The interior lowland plans unit of Ma evidence for a possible mud ocean and induced tectonic deformatio Proc. Lunar Planet. Sci. Conf. 31st, p. 2041. Abstract [CD-ROM]. Tanaka, K.L., Scott, D.H., Greeley, R., 1992. Global stratigraphy. In: K fer, H.H., Jakosky, B.M., Snyder, C.W., Matthews, M.S. (Eds.), Ma Univ. of Arizona Press, Tucson, pp. 345–382. Tanaka, K.L., Dohm, J.M., Lias, J.H., Hare, T.M., 1998. Erosional vall in the Thaumasia region of Mars: hydrothermal and seismic orig J. Geophys. Res. 103, 31407–31419. Tanaka, K.L., Skinner, J.A., Hare, T.M., Joyal, T., Wenker, A., 2002. Re facing history of the northern plains of Mars based on geologic map of Mars Global Surveyor data. J. Geophys. Res., Planets, Specia tion. In press. Thomson, B.J., Head, J.W., 2001. Utopia Basin, Mars: characterizatio topography and morphology and assessment of the origin and evol of basin internal structure. J. Geophys. Res. 106, 23209–23230. Touma, J., Wisdom, J., 1993. The chaotic obliquity of Mars. Science 1294–1296. Turcotte, D.L., Scherbakov, R., Malamud, B.D., Kucinskas, A.B., 20 Is the martian crust also the martian elastic lithosphere? J. Geo Res. 107. 10.1029/2001JE001594. Wilhelms, D.E., Squyres, S.W., 1984. The Martian hemispheric dichot may be due to a giant impact. Nature 309, 138–140. Wilson, L., Head, J.W., 2003. Oceans in the northern lowlands of M Assessment of dike emplacement as a mechanism for rapid releas confined subcryosphere aquifer. In: Proc. Lunar Planet. Sci. Conf. 3 p. 1186. Abstract [CD-ROM]. Wise, D.U., Golombek, M.P., McGill, G.E., 1979. Tectonic evolution Mars. J. Geophys. Res. 84, 7934–7939. Wyatt, M.B., McSween, H.Y., 2002. Spectral evidence for weathered b as an alternative to andesite in the northern lowlands of Mars. ture 417, 263–266. Zimbelman, J.R., Craddock, R.A., Greeley, R., 1994. Geologic map o MTM -15147 quadrangle, Mangala Valles region of Mars. USGS M Inv. Ser. Map I-2402 (1:500,000). Zuber, M.T., 2001. The crust and mantle of Mars. Nature 412, 220–22 Zuber, M.T., Solomon, S.C., Phillips, R.J., Smith, D.E., Tyler, G. Aharonson, O., Balmino, G., Banerdt, W.B., Head, J.W., John C.L., Lemoine, F.G., McGovern, P.J., Neumann, G.A., Rowlands, D Zhong, S., 2000. Internal structure and early thermal evolution of M from Mars Global Surveyor topography and gravity. Science 287, 17 1793. Episodic flood inundations of the northern plains of Mars Introduction Physiographic setting Inundation model Noachian to Early Hesperian (Stages 1-3): the first great ocean Late Hesperian/Early Amazonian (Stage 4): the last Martian ocean Amazonian (Stage 5): temporary lacustrine environments Discussion The early ocean The later ocean Late Hesperian to Amazonian paleolakes Highland features Conclusions Acknowledgments References