MANGALA VALLES, MARS: ASSESSMENT OF EARLY STAGES OF FLOODING AND DOWNSTREAM FLOOD EVOLUTION GIL J. GHATAN and JAMES W. HEAD III

The Mangala Valles system is an ~900 km fluvially carved channel system located southwest of the Tharsis rise and is unique among the martian outflow channels in that it heads at a linear fracture within the crust as opposed to a collapsed region of chaos as is the case with the circum-Chryse channels. Mangala Valles is confinedwithin abroad, north–south trendingdepression, andbegins as a single valleymeasuring up to 350 km wide that extends northward from a Memnonia Fossae graben, across the southern highlands toward the northern lowlands. Approximately 600 km downstream, this single valley branches intomultiple channels, which ultimately lose their expression at the dichotomy boundary. Previous investigations of Mangala Vallis suggested that many of the units mapped interior to the valley were depositional, related to flooding, and that aminimumof two distinct periods of flooding separated by tens to hundreds ofmillions of years were required to explain the observed geology. We use infrared and visible images from the THermal EMission Imaging System (THEMIS), and topographic data from the Mars Orbiting Laser Altimeter (MOLA), to investigate the nature of the units mapped within Mangala Vallis. We find that the geomorphology of the units, as well as their topographic and geographic distribution, are consistent with most of them originating from a single assemblage of volcanic flow deposits, once continuous with volcanic flows to the south of the Memnonia Fossae source graben. These flows resurfaced the broad, north–south trending depression intowhichMangalaVallis formed prior to any fluvial activity. Later flooding scoured and eroded this volcanic assemblage north of theMangala source graben, resulting in the present distribution of the units within Mangala Vallis. Additionally, our observations suggest that a single period of catastrophic flooding, rather than multiple periods separated by tens to hundreds of millions of years, is consistent with and can plausibly explain the interior geology of Mangala Vallis. Further, we present a new scenario for the source and delivery of water to the Mangala source graben that models flow of groundwater through a subcryosphere aquifer andupa fracture that cracks the cryosphere and taps this aquifer. The results of ourmodel indicate that the source graben, locally enlarged to a trough near the head region of Mangala, would have required less than several days to fill up prior to any spill-over of water to the north. Through estimates of the volume of material missing from Mangala (13,000–20,000 km), and calculation of mean discharge rates through the channel system (~5 · 10 m s), we estimate that the total duration of fluvial activity through the Mangala Valles was 1–3 months.