In search of the vibroseis first arrival

Abstract

The first step in correcting for time delays of land seismic data due to low-velocity weathered layers is to pick the first-arrival times of the refracting energy. But doing so for vibroseis data can be difficult, as the seismic wavelet is often ringy and uncompact, resulting in cycle-skipped picks. Even when we manage to pick a waveform feature consistently, it is not clear where the first-arrival time is in relation to it. I present a novel method that shapes the seismic wavelet to a Ricker wavelet whose peak is located at the true arrival time, so the time of the first arrival is unambiguous. Further, the arrivals are less ringy and their energy more focused, so that they are less likely to cycle skip or be overwhelmed by random noise. The result is more accurate and consistent first-arrival picks.

A minimalist approach for improved time imaging of a noisy vibroseis data: A case study from Jaisalmer Basin, Rajasthan

“…The usual methods of automation of first break picking using energy ratio often fail in the case of correlated vibroseis data. The correlated vibroseis data is zero-phase, so there is always a possibility of precursor energies (Trickett, 2022), and hence difficult for auto-pickers to identify the actual first breaks based on the energy ratio. For this reason, we first converted the zero-phase wavelet to a minimum-phase wavelet before the first break picking.…”

“…However, the method relies heavily on the accuracy of the first break picks. The picking of the first breaks in the vibroseis data is a difficult problem (Trickett, 2022). Further, the scenario becomes more complicated if we encounter velocity reversals in the near-surface creating a shingled pattern of first breaks (Sun et al, 2013(Sun et al, , 2020.…”