Uncertainty Based Multi-Step Seismic Analysis for the Near Surface Imaging

Near surface seismic surveys are often designed for surface waves and seismic tomographic analysis. In recent years, seismic imaging methods are more frequently used in this scale. Recognition of the near surface structures using a single method is insufficient, because of ambiguity of inversion problem. As a solution authors propose multi-step approach, where several different seismic methods are used in particular order, to receive optimal model. Multimethod approach, allows for utilize whole spectrum of recorded data, even the parts that are treated as a noise in other techniques. Novelty presented in this work is approach to results uncertainty, calculated at each step. The uncertainty of the results allows for correct results transfer between processing steps. Assuming that every consecutive applied method is more certain, authors were able to obtain accurate velocity field for seismic imaging, as a main information received from previous steps. Basing on an information from multiple methods, seismic stack in depth domain with estimated uncertainty was created as a final result.