JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 102, NO. E2, PAGES 4185-4200, FEBRUARY 25, 1997 Catastrophic flood sedimentsin Chryse Basin, Mars, and Quincy Basin, Washington: Application of sandar facies model James W. Rice Jr. Departmentof Geography, Arizona StateUniversity, Tempe Kenneth S. Edgett Departmentof Geology, Arizona State University, Tempe Abstract. Viking visible and thermalinfraredobservations and terrestrial catastrophic flood deposits provide clues to identify the outflow channel sediments that went into Chryse Basin on Mars. On Earth, sandar (outwash plains formedby coalescence of many j6kulhlaup floods)are described in termsof three laterallyadjacent facies: proximal, midfan,and distal. The Missoula Flood sediments deposited in QuincyBasin, Washington, comprise a miniature analog of Chryse Basin. The terminology and general characteristics of the sandar facies model are appliedto QuincyBasin,although the depositional environment and clast sizes are somewhat different(higher-energy flood, larger clasts, subaqueous rather than subaerial deposition). For example, the EphrataFan (a deposit of boulders, cobbles, and pebbles) formsthe midfanfacies analog; a downfan sandy deposit (reworked into a dune field) comprises the distal facies analog. In Chryse Basin the midfanis defined by a heterogeneous rocky(0-25%), intermediate-albedo (0.21-0.26), intermediate thermal inertia (260-460 J m -2 s -ø'5 K-•) surface, while thedistal facies has a low albedo (0.14-0.16) and higher thermal inertia (340-700 J m -2 s -ø's K-X).The Chryse midfan unit hasrocks andwindblown dustexposed at the surface. The sand of the distalfacies in Chryse/Acidalia is reworked by the wind, as in QuincyBasin. The Viking 1 and Mars Pathfinder landingsites are locatedon the midfan unit. Observations that can be made at the Mars Pathfinder site might help in reevaluating whether or not Viking 1 landed on outflow channel sediments. Introduction Mars Pathfinder and Chryse Basin The selection of a site at the confluence of Ares and Tiu Valles (19.5øN, 32.8øW) for the July 4, 1997, landingof Mars Pathfinder has prompted new interest in understanding the nature and results of the massive floods which emanated from the outflow channels around the margins of the ChryseBasin [Golombek et al., 1995a,b]. The Ares Vallis landingsite was advocated by several investigators at a 1994 workshop[Rice, 1994a;Kuzmin et al., 1994]. A summaryof the landing site selection process anditsgeologic characteristics isfoundin this issue [Golombek et al., this issue]. In this paper,we synthesize a combination of observations derived from Viking infrared thermal mapper (IRTM) data, previous studies of Viking images, and terrestrial analog cata- strophicflood deposits in North America and Iceland. We conclude that the outflow sediments exposed at the surface in the Chryse and southern Acidaliaregions of Mars havesimple, distinct,interpretablesedimentary faciesunits. Chryse Basin is a topographically low region ot•ciallynamed Chryse Planitia(Figure 1). At its lowest, the basin is between 2 and 3 km below the Martian datum, with an uncertaintyof _+1 km [U.S. Geological Survey, 1991]. Many of the major Martian outflow channels (Mawrth, Ares, Tiu, Simud,Shalba- tana,Maja, and Kasei Valles) debouch into Chryse Basin[e.g., Copyright 1997by the AmericanGeophysical Union. Paper number 96JE02824. 0148-0227/97/96JE-02824509.00 Cart, 1979;Baker et al., 1992]. The outflow channels are con- sidered to be the result of catastrophic floodsthat were larger and more powerful than those whichformedsimilarterrain on Earth, suchas the ChanneledScabland in Washington[Baker and Milton, 1974' Baker, 1978]. Photogeologic mappingof the Chryseregion has been car- ried out by several investigators overthe past two decades. The most detailed Viking image-based regional mapswhich cover Chryse Planitia wereprovided by Greeley etal. [1977], Scott and Tanaka [1986], and Rotto and Tanaka [1995]. All three maps show disagreement asto the lateral extent of sediments depos- ited by the circum-Chryse channels; for example,Scottand Tanaka [1986] map outflow channel material asbeing exposed at the surface in the Viking 1 landing siteregion, while Greeley et al. [1977] considered this surface to be a mare-like basalt unitwith a possible thin cover of floodsediment, and Rotto and Tanaka [1995] considered the material to be a deposit of flood sediment. Some have suggested that the volcanic plains and flood sediments are nearlycontemporaneous and are difficult to distinguish from eachother in the vicinity of the Viking 1 lander in Chryse Basin [e.g., Binder et al., 1977]. Others have proposed that as a result of the outflowchannel floods, the entire Chryse Basin (plus Acidalia and perhaps the entire northern plains) wasonceunder water (lakes/oceans) [Parker et al., 1989, 1993; Scottet al., 1991, 1995;Baker et al., 1991], while others consider the Viking 1 site to have a volcanic surface [/trvidson et al., 1989a]. Taken together, recent work by Tanaka[thisissue], Craddock et al. [thisissue], and Crumpler [this issue] suggests to us thata consensus might be emerging that 4185