Numerical computations of earthquake generated tsunamis have been carried by multiple
researchers in order to recreate and hopefully predict tsunamis. The Okada model is typically
used for the tsunami wave generation and the Shallow waters equation model is solved for
the wave propagation. The Okada model however requires a lot of uncertain parameters as
input. The markov chain Monte-Carlo (MCMC) methods are sample based bayesian methods
that can be used to efficiently estimate the probability distributions of unknown parameters
in high-dimensionnal problems. In this work we will explore the theory behind tsunami
generation and the Metropolis hastings MCMC algorithm and see the effectiveness of this
algorithm in inferring initial parameters for Shallow water type problems.