41st Lunar and Planetary Science Conference (2010) 1834.pdf

A LAKE IN UZBOI VALLIS AND IMPLICATIONS FOR LATE CLIMATE ON . J. A. Grant1, D. Buczkowski2, R. P. Irwin, III1, and K. Siebach3, 1Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, 6th St. at Independence Ave. SW, Washington, DC 20560, [email protected], 2Applied Physics Laboratory, Johns Hopkins University, 11100 John Hopkins Road, Laurel, MD 20723, 3Earth and Planetary Sciences, Washington University in Saint Louis, Saint Louis, MO 63130.

Introduction: Uzboi Vallis (centered at ~28°S, The Uzboi Lake Basin: The Uzboi basin is deep- 323°E) is the southernmost segment of the northward- est and bounding relief is lowest where the draining Uzboi-Ladon-Morava (ULM) meso-scale crater rim rises ~900 m above the floor to an elevation outflow system [1-3]. The ULM emerges from Argyre of –350 m (Fig. 1). Distributaries crossing the - basin, and the associated catchment covers ~9% of west rim of Holden crater from Uzboi [9] indicate that Mars [4, 5]. Originally, the ULM consisted of incised flooding to this level created a lake of >4000 km3. segments draining radially across the structural rings of the Holden and Ladon impact basins [6] that were separated by depositional systems within the ancient basin centers [1]. Uzboi’s connection with Argyre was interrupted by the craters and then [7, 8], and drainage into Holden basin was blocked by Hol- den crater to the north, creating a ~400-km-long en- closed basin that was flooded during the late Noachian [9].

Figure 2. HiRISE image PSP_006189_1510_RED (25 cm pixel-scale) in Uzboi basin (28.8°S, 323.9°E). Here and elsewhere, smooth dark-toned deposits (upper left) rise to the –350 m contour and are interpreted as lacu- strine deposits. A transition to rougher surfaces occurs above –350 m. North is up. Additional evidence of impoundment in Uzboi is provided by thin, dark-toned deposits below –350 m (Fig. 2), layered deposits mostly exposed in alcoves, tributaries that terminate near –350 m, possible shore- lines, layered fill within an unnamed crater (Fig. 1) that is fractured into ~5–15 m polygons (Fig. 3), and local ponding in topographic lows of alluvium that is sometimes characterized by diverse, 50–100 m poly- gons. Once filled, Uzboi lake was geologically short- lived and drained rapidly into Holden crater. The common termination of most tributaries (except Nir- gal) at about –350 and absence of recognizable deposi- tional features where tributaries debouched into the Uzboi lake or at the limit of backflooding in (Fig. 1) indicates lake level was not stable at lower elevations for a long period. Water overtopping Figure 1. MOLA topography over THEMIS daytime Holden’s rim coalesced into a dominant breach [9] that IR mosaic 2.0 of Uzboi Vallis/basin. Elevations range incised hundreds of meters as the lake drained into the from ~1300 m (red) to ~–2300 m (purple). The blue crater. Resultant fan deltas on the floor of Holden ex- within Uzboi Vallis is the area below –350 m, flooded tend more than 25 km from the breach and incorporate before the rim of Holden crater was overtopped. The meter-sized blocks whose transport requires high dis- rim of Bond crater was not overtopped, but lower Nir- charge. In addition, longitudinal grooves along the gal Vallis was backflooded. North is up. 41st Lunar and Planetary Science Conference (2010) 1834.pdf

basin floor are hundreds of meters long and average 60 m, respectively, indicating that the regional water table m wide. These grooves, potholes, and other erosional was below the –1300 m floor of Uzboi. Nirgal Vallis is bedforms common to large floods on Earth suggest the major tributary to Uzboi, but it grades to the floor that they were formed by cavitation/macroturbulence of the basin, lacks depositional features associated [10–15] during rapid draining of at least the final stag- with backflooding, and shows little evidence for post- es of the Uzboi lake. On the Earth, similar longitudinal lake discharge [9]. Hence, while Nirgal Vallis may forms are associated with discharge rates of 105 to 106 have contributed discharge to the Uzboi lake, it likely m3/sec and flow velocities exceeding 10 m/s [10–15]. existed in something close to its present form prior to filling and draining of basin. Most Uzboi basin tributaries terminate at or near the –350-m contour and imply a base level near that elevation during much of their incision. These tributa- ries show poor connections with segments low- er in the basin that were likely reactivated or wholly formed during lake draining and subsequent draw- down of the local water table. Because tributaries above –350 m did not respond to lake draining and a dramatic lowering of base level (hundreds of meters), their incision must have ended when the lake drained. Implications: The Uzboi lake may have coincided with or just followed late Noachian alluvial and lacu- strine activity in Holden crater [9]. Little groundwater Figure 3. HiRISE image ESP_013586_1490_RGB (50 contribution to the Uzboi lake points to runoff from cm pixel-scale) of polygons disrupting layered fill in regional surfaces as the primary source of water and an unnamed crater inundated during flooding (Fig. 1). may require widespread precipitation. The absence of Comparison between the volume of water in the abrupt relief (outside of craters) or regional slopes that Uzboi lake and the lesser volume represented by the might induce orographic precipitation in Uzboi’s direct upper limit of benches, fan deltas, and lacustrine depo- watershed implies that precipitation was associated sits in Holden [9, 16] indicates that lake draining was with a broader, perhaps global, but short-lived and not a single event. Wall failures may have partly abruptly ending hydrologic system in the late Noa- dammed the breach during incision and led to pulsed chian [e.g., 17, 18]. discharge. Observed post-drainage slides into the References: [1] Grant J. A., and Parker, T. J. breach support this interpretation. (2002) JGR, 107, 10.1029/2001JE001678. [2] T. J. Limited CRISM data of the lake deposits and un- Parker (1985), thesis, California State University Pon- derlying basin floor (FRT0000ABB5) reveal Mg- drelli M. et al. (2005) JGR, 110, 2004JE002335. [3] smectites (possibly saponite or hectorite) that appear Saunders, S. R., (1979) USGS Map I-1144. [4] Ba- enriched in Fe in the basin floor (possibly a nontronite- nerdt, W.B. (2000), Eos Trans. AGU, 81, Abstract saponite mixture). CRISM data of the breach into Hol- #P52C-04 [5] Phillips, R.J., et al. (2001) Science, 291, den crater (HRS0001266E) confirm more outcrops of 2587–2591. [6] Schultz P. H. et al. (1982) JGR, 87, Mg-smectite-bearing rocks but also reveal a thin (tens 9803. [7] Philippoff, A.J., et al. (2009) LPSC XL, of meters thick) near-surface layer of possible Mg- 1737. [8] Reiss, D., et al. (2009) PSS, 57, 958-974, chlorite (clinochlore). The basin-floor smectites may doi:10.1016/j.pss.2008.09.008. [9] Grant, J. A., et al., reflect long-term weathering before the lake formed in (2008) Geology, 36, 195-198, doi: 10.1130/G24340A. Uzboi Vallis. The smectites in the lake deposits likely [10] Alho, P., et al. (2005) Quat. Sci. Rev., 24, 2319- reflect erosion and redeposition of these older smec- 2334. [11] Baker, V.R. (2001) Nature, 412, 228-236. tites, given that the lake was geologically short-lived. [12] Baker, V.R. (2002) Science, 295, 2379-2380. [13] The origin of Mg-chlorite near Holden’s rim is uncer- Baker, V.R. (2009) Ann. Rev. Earth and Planet. Sci., tain, but occurrence as a near-surface layer implies that 37, 393-411. [14] Baker, V., et al. (1988) Flood Geo- it post-dates crater formation and may be related to morphology: NY, Wiley Interscience, 503 p. [15] Gup- impact-induced or other post-crater alteration. ta, S., et al. (2007) Nature, 448, 342-346. [16] Pon- Source of Water: Regional groundwater contribu- drelli M. et al. (2005) JGR, 110, 2004JE002335. [17] tions to the Uzboi lake were likely minor, because al- Wray, J. J. et al. (2009) Geology, 37, 1043-1046, doi: luvial fans in adjacent Holden crater and Ostrov crater 10.1130/G30331A.1. [18] Irwin R. P. III et al. (2005), ~200 km east of Holden grade to –1900 m and –1500 JGR, 110, doi:10.1029/2005JE002460.