First Break

Key Takeaways

• Refraction seismology using first-break analysis is the classical method for determining the depth and velocity of the near-surface low-velocity layer (the weathering layer) that sits above the more competent rock of the consolidated formation: in a simple two-layer model (low-velocity surface layer over a high-velocity refractor), the first-break travel time increases linearly with offset at short offsets (where the direct wave is the first arrival, traveling through the surface layer at velocity V1), then transitions to a different linear slope at the crossover distance (where the refracted wave that travels along the top of the high-velocity layer at velocity V2 and is bent upward to the surface by Snell's law becomes faster than the direct wave), and continues at the slope corresponding to the refractor velocity beyond the crossover distance; the intercept time of the refracted arrival branch (the time at which the refracted branch, extrapolated back to zero offset, intersects the time axis) and the slope of the refracted arrival branch (inverse of V2) provide the two parameters needed to solve for the depth to the refractor using the time-intercept method; in land seismic acquisition, first-break analysis is used routinely to map the base of the weathering layer (the boundary between the low-velocity near-surface material and the consolidated rock below it), providing the velocity model needed for applying refraction statics corrections that remove the time delays caused by lateral variations in the weathering layer thickness from the reflection seismic record.
• Automated first-break picking in large 3D seismic datasets has become essential because the volume of data (millions of seismic traces per survey) makes manual picking by human interpreters impractical, and multiple automated picking algorithms have been developed that use the characteristics of the seismic trace in the first-break window to identify the arrival time automatically: the simplest automated pickers use amplitude threshold criteria (the first time the trace amplitude exceeds a fixed fraction of the peak amplitude in the first-break window is taken as the first-break time), which work well for clean, high-amplitude first breaks but fail on noisy or weak arrivals where the threshold is exceeded prematurely by noise; more sophisticated algorithms use cross-correlation between adjacent traces (assuming the first break should have a consistent waveform between neighboring shots and receivers), the envelope of the trace (the instantaneous amplitude, computed as the magnitude of the analytic signal), the fractal dimension of the trace, or machine learning classifiers trained on manually picked examples from the dataset; the accuracy of automated first-break picks is typically assessed by comparing a subset of automated picks against manual picks by an experienced processor, with acceptable picking accuracy defined as less than 5-10 milliseconds RMS error for standard land seismic applications; picking errors at first breaks propagate into errors in the derived refraction statics corrections that show up as residual time shifts in the stacked seismic section that can be misinterpreted as structural features if the statics corrections are inadequate.
• First-break amplitude and waveform analysis provides information about the near-surface attenuation and the coupling of the seismic source and receivers to the earth: a strong, sharp first break (large amplitude, short rise time) indicates good coupling between the seismic source and the ground (the source impulse has been efficiently transmitted to the ground without energy loss at the source-ground interface), good coupling between the geophone and the ground (the geophone is well planted in firm soil, not in loose sand or grass that would attenuate and scatter the seismic energy), and a low-attenuation near-surface path (the first break has not been severely attenuated by highly absorptive near-surface material such as dry sand, peat, or permafrost); a weak, emergent first break (small amplitude, long rise time) indicates one or more of these problems, and the source and receiver coupling problems can sometimes be distinguished by comparing first breaks from multiple shots at the same receiver (coupling problems at the receiver appear in all shots at that receiver) or from multiple receivers at the same shot (coupling problems at the source appear in all receivers from that shot); the systematic analysis of first-break amplitude across the survey area provides a quality map that identifies zones of poor coupling that may require special source or receiver treatment (larger arrays, different coupling methods, or exclusion of noisy data) in the acquisition QC process.
• Velocity inversion using first-break travel times in the tomographic approach provides a more sophisticated near-surface model than the simple layer-cake refraction approach, by treating the near-surface velocity distribution as a continuous 3D field that is constrained by the travel times of first arrivals from many source-receiver pairs: refraction tomography inverts the first-break travel times from a 3D seismic survey (typically 10-100 million first breaks from a large land survey) to compute a continuous 3D velocity model in the near-surface that explains all the observed first-break times simultaneously, using an iterative ray-tracing and travel-time inversion process that updates the velocity model until the predicted travel times (computed by ray-tracing through the model) match the observed first-break picks within the picking accuracy; the resulting tomographic velocity model is significantly more detailed and laterally continuous than the layer models from conventional refraction analysis, capturing the lateral velocity variations caused by lithological changes, water table variations, erosional topography, and karst dissolution that produce the statics anomalies seen in reflection seismic data; the tomographic model is then used to compute static corrections (time shifts applied to each trace based on the travel time through the near-surface velocity model) that more accurately correct for near-surface effects than the simple refraction statics, improving the quality of the reflection stack in areas with complex near-surface geology.
• Marine first breaks from ocean bottom nodes (OBN) and ocean bottom cable (OBC) seismic systems provide information about the water column velocity and the shallow seafloor sediment properties that differ from the land first-break applications: marine first breaks from OBN or OBC systems show the direct wave traveling through the water column from the surface source to the seafloor receiver (at the water wave velocity of approximately 1,500 m/s), followed by the refracted wave if the seafloor sediment velocity exceeds the water velocity; the travel time of the direct water wave at each receiver provides the precise water depth at the receiver location (acoustic depth sounding), which is more accurate than bathymetric charts for detailed seafloor mapping; the first arrival in the seafloor sediments provides velocity information about the shallow seafloor sediments, which may be gas-charged (extremely low velocity, 200-800 m/s) or normally compacted clays (velocity 1,500-2,000 m/s), information that is important for drilling hazard assessment (gas in shallow sediments is a blowout risk in deepwater drilling) and for seismic processing (correcting for the anomalously low velocity in gas-charged shallow sediments that causes time pull-up and pull-down in the underlying reflection seismic data).