The timing of martian valley network activity: Constraints from buffered crater counting

Valley networks, concentrations of dendritic channels that often suggest widespread pluvial and fluvial activity, have been cited as indicators that the climate of Mars differed significantly in the past from the present hyperarid cold desert conditions. Some researchers suggest that the change in climate was abrupt, while others favor a much more gradual transition. Thus, the precise timing of valley network formation is critical to understanding the climate history on Mars. We examine thirty valley network-incised regions on Mars, including both cratered upland valley networks and those outside the uplands, and apply a buffered crater counting technique to directly constrain when valley network formation occurred. The crater populations that we derive using this approach allow assessment of the timing of the last activity in a valley network independent of the mapping of specific geological units. From these measurements we find that valley networks cluster into two subdivisions in terms characteristics and age: (1) valley network activity in the cratered highlands has an average cessation age at the Noachian-Hesperian boundary and all valleys that we crater counted are Early Hesperian or older. No evidence is found for valley networks in the cratered uplands of Late Hesperian or Amazonian age. The timing of the cessation of cratered upland valley network activity at the Noachian-Hesperian boundary also corresponds to a decline in the intensity of large crater formation and degradation and to the apparent end of phyllosilicate-type weathering. (2) A few valley network-incised regions formed outside of the cratered uplands on volcanic edifices, in association with younger impact craters, and on the rim of Valles Marineris. We applied our buffered crater counting technique to four such valleys, on the volcanoes Ceraunius Tholus, Hecates Tholus, and Alba Patera and on the rim of Echus Chasma, and find that each has distinctive and different Late Hesperian or Early Amazonian ages, indicating that valley networks formed from time to time in the post-Noachian period. Unlike the cratered upland valley networks, these isolated occurrences are very local and have been interpreted to represent local conditions (e.g., snowpack melted during periods of intrusive volcanic activity). In contrast to a gradual cessation in the formation of valley networks proposed by some workers, our new buffered crater counting results indicate a relatively abrupt cessation in the formation of the widespread cratered upland valley networks at approximately the end of the Noachian, followed only by episodic and very localized valley network formation in later Mars history, very likely due to specific conditions (e.g., local magmatic heating). These valley network ages and correlations are thus consistent with a major change in the near-surface aqueous environment on Mars at approximately the Noachian-Hesperian boundary. The Noachian environment supported surface running water and fluvial erosion across Mars in the cratered uplands, enhanced crater degradation, and a weathering environment favoring the formation of phyllosilicates. The Hesperian-Amazonian environment was more similar to the hyperarid cold desert of today, with valley networks forming only extremely rarely and confined to localized special conditions. Sources of water for these latter occurrences are likely to be related to periodic mobilization and equatorward migration of polar volatiles due to variations in spin-axis orbital parameters, and to periodic catastrophic emergence of groundwater.