Nienke Blom , Nicholas Rawlinson

We investigate the effect of errors in earthquake source parameters on the tomographic inverse problem and propose mitigation strategies for avoiding artefacts caused by such errors. Source errors can be quite significant: mislocation errors can amount to tens of kilometres both horizontally and vertically, while fault plane uncertainties can be of the order of tens of degrees. To analyse how these errors affect seismograms and hence seismic tomography, we conduct a perturbation study involving a number of synthetic inversions that feature several types of source errors. It is well known that source location errors can detrimentally influence the recovery of seismic structure, but we find that also errors in the moment tensor can give rise to significant errors: even those moment tensor components to which computed waveforms are least sensitive can result in significant wavefield perturbations and corresponding timeshifts. The effect of such errors on tomographic images is consequently pronounced. All source error types that we investigate result in considerable artefacts in recovered Earth models, which for realistic errors are similar in amplitude to what is typically imaged using observational data. In some cases these artefacts extend over thousands of kilometres.

We present a number of mitigation strategies to minimise the formation of source error artefacts in tomographic images. By removing a region around the source and discarding measurements on or near nodal planes (where amplitudes are small), the adverse effects of many types of source error can be greatly reduced. The most problematic type of error is a horizontal misocation, because its imprint remains visible even when taking such mitigation measures into account. Moreover, its effect cannot be easily distinguished from the effect of several other types of source errors or in fact from any Earth structure that has a degree-2 pattern with respect to the source. Horizontal event location uncertainty must therefore be carefully evaluated, and preferably compared to ‘ground truth’ data where this is available. Failing that, suspicious events may need to be removed altogether from the inverse problem in order to avoid their imprint contributing significantly to the imaged structure.

Although this study focuses on (adjoint) waveform tomography, a large part of the results are equally valid for any other type of imaging method that is based on time- and/or phaseshift measurements. We therefore expect that the concerns and mitigation strategies presented here have broader applicability.