An adaptive reversible jump MCMC inversion algorithm for airborne time-domain electromagnetic data

Abstract

The reversible jump Markov Chain Monte Carlo algorithm is a trans-dimensional Bayesian inversion method, which can not only obtain the best inversion solution, but also provide the uncertainty information of inversion parameters, so as to effectively evaluate the reliability of inversion results. However, in trans-dimensional Bayesian inversion, the step size of model sampling has a great influence on the efficiency and precision of inversion. To solve the selection problem of sampling step size, this paper presents an adaptive reversible jump MCMC inversion algorithm for airborne time-domain electromagnetic data. In this method, the sampling step length is automatically adjusted in the sampling process through the data fitting error of each sampling model, so as to avoid the influence of unreasonable sampling step size and improve the acceptance rate of model sampling. The validity of the proposed method is verified by inversion test on the synthetic data and compared with the regularized inversion results. The inversion results demonstrate that the adaptive reversible jump MCMC inversion algorithm can improve the acceptance rate of model sampling and the accuracy of inversion results.