Abstract / Description of output
It has been a challenge to image velocity changes in real time by seismic travel time tomography. If more seismic events are included in the tomographic system, the inverted velocity models do not have necessary time resolution to resolve velocity changes. But if fewer events are used for real-time tomography, the system is less stable and the inverted model may contain some artifacts, and thus, resolved velocity changes may not be real. To mitigate these issues, we propose a wavelet-based time-dependent double-difference (DD) tomography method. The new method combines the multiscale property of wavelet representation and the fast converging property of the simultaneous algebraic reconstruction technique to solve the velocity models at multiple scales for sequential time segments. We first test the new method using synthetic data constructed using real event and station distribution for Mount Etna volcano in Italy. Then we show its effectiveness to determine velocity changes for the 2001 and 2002 eruptions of Mount Etna volcano. Compared to standard DD tomography that uses seismic events from a longer time period, wavelet-based time-dependent tomography better resolves velocity changes that may be caused by fracture closure and opening as well as fluid migration before and after volcano eruptions.