Question 8 of 10: Seismic Refraction — Survey Design, Case History and Noise

Question 8: Seismic Refraction — Survey Design, Case History and Noise (16.66% of paper)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

Geological circumstances for strong refraction arrivals. Refraction relies on total internal (critical) refraction along a boundary between a slower upper layer and a faster lower layer, so the essential requirement is a NORMAL (velocity-increasing-with-depth) sequence — e.g. dry soil over saturated soil over competent bedrock — with a sufficiently strong velocity CONTRAST at each interface that the critically refracted head wave carries enough energy to be picked well above ambient noise. A LOW-VELOCITY layer buried beneath a higher-velocity layer (a velocity inversion, such as weathered/fractured rock beneath a well-cemented near-surface crust) defeats the method outright, since a ray cannot critically refract back up from a slower layer and the hidden layer is never detected, so confirming a normal sequence (from any available borehole or outcrop control) before committing to a refraction survey is essential.

Survey specifications. Geophone spacing and total spread length must be chosen so that the crossover distance (where the refracted arrival first overtakes the direct wave) falls comfortably within the spread — as a rule of thumb, spread length should be at least 3–5 times the expected depth to the target refractor, with geophone spacing fine enough (e.g. 2–5 m for a shallow engineering target) to define each linear travel-time segment with several points. Both forward AND reverse shots (source at each end of the spread, ideally with an added mid-spread shot) are essential to detect and correct for any dip on the refracting interface, which a single forward shot alone cannot distinguish from a change in the lower layer's true velocity.

Case history — plotting and interpretation. A seismic refraction survey was run across a proposed highway cut to determine depth to rippable versus non-rippable (blast-required) rock. Forward and reverse shot travel-time data were plotted on a time-distance ($t$–$x$) graph, revealing two clearly linear segments per shot: a low-slope direct-wave segment near the source (slope $=1/V_1$) and a lower-slope refracted-wave segment beyond the crossover distance (slope $=1/V_2$), with $V_2>V_1$. Fitting straight lines to each segment gave $V_1\approx450\text{ m/s}$ (loose overburden) and $V_2\approx3200\text{ m/s}$ (competent, likely non-rippable rock) from the forward shot, cross-checked against a similar pair of slopes from the reverse shot to confirm the refractor's true velocity and detect any dip (a difference in apparent velocity between the forward and reverse shots indicates a dipping interface, whose true dip and velocity are then recovered from the pair of apparent velocities via the standard reciprocal method). The depth to the refractor at each geophone was then computed from the intercept time and the two velocities, producing a depth profile along the proposed cut that directly flagged the sections requiring blasting.

Sources of noise. Ground roll (low-velocity, high-amplitude surface Rayleigh waves) can mask the weaker refracted first arrivals, especially at short offset; traffic, construction equipment and wind-shaken vegetation near the spread add broadband ambient noise; poor geophone-ground coupling (loose or frozen soil) degrades signal amplitude and adds spurious high-frequency noise; and a nearby parallel utility trench or buried pipe can generate a strong, spatially localized diffracted arrival that is easily mis-picked as a genuine refraction first break if not recognized from its inconsistent moveout.