Dating rupture events on alluvial fault scarps using cosmogenic nuclides and scarp morphology

Abstract

Scarp morphology evolution has commonly been used to estimate the timing of fault-scarp rupture events. However, rates of scarp degradation depend strongly on the geomorphic diffusivity, a parameter that is difficult to constrain independently. This difficulty may lead to large uncertainties in the estimated ages of rupture events. In this study, we have coupled the accumulation of the cosmogenic nuclide 36Cl to a model for scarp morphology in order to constrain the value of the geomorphic diffusivity and thus determine a more accurate rupture chronology than a rupture history based solely on scarp morphology. We measured depth profiles of 36Cl accumulated in situ within alluvial sediments beneath the surface of the Socorro Canyon fault scarp in central New Mexico. The material analyzed consisted of ∼150 individual gravel clasts from each depth interval, amalgamated into a single sample. The alluvium was sampled in three vertical profiles ∼4 m deep. The first profile was ∼1.5 m downslope of the fault plane on the hanging wall, the second ∼1.5 m upslope of the fault plane on the footwall, and the third ∼27 m upslope of the fault plane on the footwall. The third profile, which served as a control, and which soil geomorphic evidence indicated was beneath a stable surface, showed a simple exponential 36Cl profile from which a depositional age of 122±18 ka was calculated. The second profile (on the footwall) showed a 36Cl deficit relative to the control profile, indicating net erosion. The first profile (on the hanging wall) showed a 36Cl excess relative to the control profile, indicating net deposition. Stratigraphic evidence in the hanging wall indicated two rupture events prior to the Late Holocene. We modeled the accumulation of 36Cl in the vicinity of the scarp, simulating erosional redistribution using a diffusion equation for scarp morphology. The model accounted for redistribution of 36Cl along with the sediment and the effects of the changing fault-scarp morphology on the 36Cl production. By matching both the observed hanging-wall and footwall 36Cl profiles to profiles calculated by the model, as well as the observed topographic profile, we obtained ages of 92+16−13 and 28+18−23 ka for the two ruptures. The results of our study indicate that cosmogenic nuclides can be useful in constraining rupture chronologies of fault scarps in alluvium.