Our eyes can recognize the arrival of the signal by comparing the recorded amplitude in various seismic traces. In this article we tried to imitate this natural ability to find a mathematical tool instead of the eyes. Fractals are mathematical models, which are able to feel these changes.
An algorithm is introduced for the automatic picking of seismic first arrivals. It detects the presence of a signal by analyzing the variation in fractal dimension along the trace. Among different methods, the “divider method” was found to be the most suitable method for calculating the fractal dimension of a seismic trace. A change in dimension was found to occur close to the transition from the noise part of a trace to the signal plus noise part of it, that is the first arrival. The nature of this change varies from trace to trace, but a detectable change is always found to occur. The algorithm has been written in MATLAB environment and is slower than the other non-fractal methods.