The Discussion of the Uncertainty in the Traveltime Seismic Tomography

Summary First arrival traveltime tomography is a widely used technique dedicated for velocity model-building. One of the important aspects of tomographic methods is the validation of the results by the uncertainty analysis. Among different factors controlling seismic traveltime inversion, the starting model is one of the most critical. It is worth to mention that the effect of the starting model for the final inversion result can be quantitatively estimated for example using Monte Carlo approach. Important factor that controls traveltime inversion is the traveltime picking precision. In this work we focus on the importance of the traveltime picking precision and its quantitative effect on the final result of the inversion. We perform our uncertainty analysis using real 2D land seismic data. We analyse the influence of both: starting model and the picking uncertainty. Performed tests show that tomographic uncertainty analysis should as well account for the rays coverage to provide meaningful results. Moreover, according to the other studies errors of manually picked traveltimes prevail an asymmetric distribution. From a number of test observations we conclude that it is more likely to pick earlier rather than later arrival times. Therefore we also test how these characteristics affects the final inversion result.