Hamiltonian Monte Carlo (HMC) is known to be highly efficient when sampling high-dimensional parameter spaces. This high efficiency can be attributed to Hamilton’s equations, which guide the sampling of the model space. In the case of weakly non-linear problems, this efficiency can be increased even further by linearizing the forward problem. In this study, we exploit this for the purpose of estimating source parameters of a 3.4 magnitude induced event that originated in the Groningen gas field in 2019. In total, we estimate ten earthquake parameters: centroid (three coordinate components), moment tensor (six elements), and origin time. We demonstrate that, in the absence of a sufficiently accurate centroid prior, the linearization of the forward model necessitates multiple initial centroid priors. Here, we consider two sets of initial centroid priors. The first set is based on the known fault geometry in the Groningen reservoir, whereas the second set is obtained by placing initial centroid priors on a uniform horizontal grid at a depth of 3 km (the approximate depth of the gas reservoir). In general, the results from both sets are in good agreement with each other. Most important, however, is their agreement with the geological knowledge of the area: the posterior peaks for model vectors containing a centroid near a major fault and a moment tensor that corresponds to normal faulting along a plane which has a strike almost coinciding with the strike of that major fault.