Seismic history from in situ 36Cl cosmogenic nuclide data on limestone fault scarps using Bayesian reversible jump Markov chain Monte Carlo

Abstract Constraining the past seismic activity and the slip-rates of faults over several millennials is crucial for seismic hazard assessment. Chlorine 36 ( 36 Cl) in situ produced cosmogenic nuclide is increasingly used to retrieve past earthquakes histories on seismically exhumed limestone normal fault-scarps. Here we present a new methodology to retrieve the exhumation history based on a Bayesian transdimensional inversion of the 36 Cl data and using the latest muon production calculation method. This procedure uses the reversible jump Markov chains Monte-Carlo algorithm (RJ-MCMC, Green 1995) which enables 1-exploring the parameter space (number of events, age and slip of the events), 2-finding the most probable scenarios, and 3- quantifying the associated uncertainties. Through a series of synthetic tests, the algorithm revealed a great capacity to constrain event slips and ages in a short computational time (several days) with a precision that can reach 0.1 ky and 0.5 m for the age and slip of exhumation event, respectively. In addition, our study show that the amount of 36 Cl accumulated when the sampled fault-plane was still buried under the colluvial wedge, prior its exhumation, might represents up to 35% of the total 36 Cl. Additional sampling under the colluvial is necessary to constrain this contribution.