Lunar and Planetary Science XXIX 1933.pdf

THE ORIGIN OF WARREGO VALLES: A CASE STUDY FOR FLUVIAL VALLEY FORMATION ON EARLY MARS. Virginia C. Gulick 1"2,James Dohm 3, Ken Tanaka 3, Trent Hare 3, 1MS 245-3, Space Sciences Divi- sion, NASA-Ames Research Center, Moffett Field, CA 94035, email: [email protected], 2also at Dept. of Astronomy, NMSU, Las Cruces, NM, 3U.S.Geological Survey, Flagstaff, AZ g6001.

Warrego Valles (Fig. I) is perhaps the most for hundreds to thousands of kilometers is difficult to "famous" of all Martian valleys as it appears as fre- understand if the source of the water was rainfall. quently in undergraduate astronomy textbooks as it Gulick I2,3,61 and Gulick et al. 171 argue that does in scientific talks. It is almost universally cited as the localized erosion pattern at Warrego is more con- evidence for rainfall during a warm, wet early Mars. sistent with a localized water source. Two possibilities Thus understanding the origin of Warrego is central to are hydrothermally driven ground-water outflow [2,6] answering the question, "How did the ancient valley and melting of a snowpack by locally high geothermal networks form?" Did they require a substantially dif- heat flow [2,7]. The radial drainage pattern of War- ferent climate with global temperatures above freezing rego Vailes may provide insight as to its formation as or could they have formed under less clement climatic it appears centered approximately on a 35 km diameter conditions involving geothermal/hydrothermal proc- impact crater 12,3,61. However, a closer study reveals esses as both sources of heat energy and drivers of that the valleys have formed on and have delineated a ground-water outflow. much larger region of uplift (Fig 2). Warrego Valles liesalong the southern edge A possible cause of the uplift may have been of the Thaumasia plateau. It formed in the southern from a magma intrusion in the subsurface or incipient highlands but tectonic fractures dated by Tanaka and volcanic activity I2,6,71. Both the subsurface mag- Dohm [1] to be early Hesperian cross cut and have in matic event and the formation of an impact crater turn been cross cut by the valley system I2,31. This would have resulted in the formation of vigorous hy- implies that Warrego Valles was still actively forming drothermal systems. We have mapped in detail the well into the Hesperian. Warrego region with the goal of understanding how Pieri [4 I, who completed an early systematic the possible interaction of tectonic and volcanic proc- study of the valley networks on Mars, concluded that esses may have led to fluvial valley formation. In ad- the digitate pattern formed by this valley system is dition, Dohm et al. [81 and Tanaka et al. 191 have con- most consistent with formation by ground-water sap- eluded based on their 1:15,000,00 geologic map of the ping. This finding is in contrast to the often cited evi- Thaumasia Quadrangle, that valley formation within dence for a rainfall origin I51. Gulick 12,31 also ques- the entire Thaumasia region is consistent with areas of tioned the rainfall origin. She concluded that regard- inferred past hydrothermal activity. less of whether the valley system formed by surface While both the impact crater and the hypothe- runoff or ground-water sapping, the lack of similar sized magmatic intrusion may have triggered hydro- erosion along adjacent edges of the Thaumasia plateau

© ©

• m *m T im

Figure 1. WarregoValles (left) and sketch map (fight). The boundary between the southern edge of the Thaumasia Plateau and the adjacent lowlands is shown as a heavy line, dashed where approximate. Note how the valleys clus- ter over an area about 200 km across. Lunar and Planetary Science XXIX 1933.pdf

WARREGO VALLES, MARS: Gulick et al.

thermal circulation and produced fluvial valleys, the Gulick 12,6l finds that hydrothermal systems associ- magmatic intrusion is more likely responsible. ated with magmatic intrusions ffeater than 500 km _ The large impact crater from which the val- can easily drive sufficient ground water outflow to leys seem to radiate is approximately 35 km in di- form fluvial valleys. She estimates that a 500 km J ameter. From scaling relations in Melosh I10], we intrusion will produce approximately 40,000 km _ of estimate that depending on the impact velocity, the outflow over 106 years. The volume of the Warrego formation of this impact crater would have produced valleys is estimated to lie in the range of approxi- approximately 12 km 3 of impact melt, which is un- mately 400 to 1500 km 3. Assuming a conservative certain to about a factor of 2. Distributed over the ratio of water to removed material of 1000:1, a rea- final crater area, this would produce a melt pool ap- sonably-sized magmatic intrusion beneath Warrego proximately 10 m deep. Such a thin melt lens would Valles could provide sufficient ground water outflow cool quickly by conduction. Terrestrial analog stud- to drive fluvial valley formation. ies I2,31 suggest that highly integrated tributary val- In summary, although Warrego Vailes is one ley systems such as Warrego require liquid water to of the best examples of a well integrated fluvial val- flow on or near the surface for periods of several 10_ ley system that formed early in the geological history -106 years or more. Indeed Gullck et al. [lll esti- of Mars, the lack of similar erosion elsewhere along mated that only impact cratet:s greater than 100kin in the edge of Thaumasia plateau is not consistent with diameter would produce sufficient impact melt to a formation by rainfall. Instead the radial pattern of drive a hydrothermal system capable of forming flu- this valley system centered on a region of localized vial valleys [6]. There is no such impact crater in the uplift argues for a more localized water source. We immediate vicinity of Warrego Valles. conclude that this uplift was most likely the result of The region of uplift upon which Warrego a subsurface magmatic intrusion and that the esti- sits is approximately 100 km in radius. An underly- mated volume of this intrusion is sufficient to cause ing magmatic intrusion with only 10% of the radius enough hydrothermal ffound-water outflow to form would still have a volume in excess of 300 km 3,even the valley system. A possible alternative to this sce- if it were only 1 km thick. A thicker intrusion, as nario is hydrothermai ground-water outflow com- seems likely given the 5 km rise across the boundary, bined with a melting snow pack. could easily have a volume in excess of 1000 km.

Figure 2. Digital elevation model of region surrounding WarregoValles. Note that the rise from the mouth of the valley network to its source regions is approximately 5 kin. (1977) JGR 82,4016-4038. 161Gulick, V.C. (in press) REFERENCES JGR. [7] Gulick etal. (1997) Icarus 130, 68-86. [8] Dohm J. etal. LPSC (this volume). [9] Tanaka et al. I11 Dohm and Tanaka (1995) LPSC XXVI, 337-338. LPSC (this volume). [10] Melosh, H.J. (1989), Impact I21 Gulick, V.C. (1993). Ph.D. thesis, U. of AZ. I31 Cratering, Oxford U. Press. [1 ll Gulick V.C. et al. Gulick, V.C. ( in press) Geomorphology. I4] Pied, D. (1988) LPSC XIX, 441-442. (1979) Ph.D thesis, Comell. I51Masursky, H. et al.