Verifying Subsurface Velocity Models: Passive Source vs. Borehole PS Logging

Determining the shear-wave velocity profile of the subsurface is a fundamental requirement in geotechnical engineering and seismic hazard assessment. The parameter $V_{s30}$, representing the time-averaged shear-wave velocity in the upper 30 meters of the earth, serves as the standard metric for site classification under various building codes, including those established by the International Building Code (IBC) and the California Building Code (CBC). Engineers and geophysicists primarily use two methodologies to derive these profiles: invasive borehole geophysical logging (PS logging) and non-invasive passive source surface wave methods, most notably Refraction Microtremor (ReMi).

While borehole PS logging provides direct measurements of seismic wave travel times at specific depths, passive source methods like ReMi use ambient seismic noise—originating from traffic, industrial activity, and atmospheric conditions—to characterize the dispersion of Rayleigh waves. The choice between these methods involves a trade-off between the high vertical resolution and logistical complexity of drilling versus the cost-effectiveness and broad lateral coverage offered by surface wave arrays. The integration of these techniques, supported by lithological data from the United States Geological Survey (USGS), ensures the calibration of velocity models necessary for infrastructure resilience.

At a glance

• Methodology:Borehole PS logging uses a source at the surface or in an adjacent hole to measure arrival times; ReMi uses a linear array of standard 4.5 Hz or 10 Hz geophones to record ambient vibrations.
• Depth of Investigation:Borehole methods are limited by the depth of the drilled hole, whereas ReMi can reach depths of 100 meters or more depending on the length of the geophone array.
• Cost Structure:ReMi typically costs 50% to 80% less than borehole logging because it eliminates the need for drilling permits, heavy machinery, and environmental remediation.
• Regulatory Context:Site classification benchmarks established by the California Geological Survey (CGS) in the early 2000s popularized $V_{s30}$ as a mandatory design parameter for critical infrastructure.
• Data Analysis:Borehole data is analyzed via direct time-distance slopes; ReMi requires spectral analysis (p-f transform) and the inversion of dispersion curves to generate a velocity model.

Background

The emphasis on accurate subsurface velocity modeling intensified during the late 1990s and early 2000s as seismic design codes transitioned toward performance-based engineering. The California Geological Survey (CGS) and the Southern California Earthquake Center (SCEC) conducted extensive benchmarking studies to standardize the classification of soil and rock sites. These studies revealed that ground motion amplification is highly dependent on the velocity contrast between near-surface sediments and the underlying bedrock. Consequently, the $V_{s30}$ measurement became a required entry for the ShakeMap and HAZUS platforms used in disaster mitigation.

Historically, the "gold standard" for seismic site characterization was the downhole or crosshole seismic test. These methods involve lowering an instrument package into a cased borehole to record waves generated by a mechanical impact at the surface. While accurate, the invasive nature of these tests often made them prohibitive for smaller projects or urban environments where drilling is restricted. The development of the Refraction Microtremor (ReMi) method by researchers at the University of Nevada, Reno, in the early 2000s provided a non-invasive alternative that leveraged the increasing computational power available for processing ambient noise wavefields.

The Mechanics of Passive Source ReMi

The ReMi method operates on the principle of Rayleigh wave dispersion in heterogeneous media. In a stratified subsurface, waves of different frequencies travel at different phase velocities; high-frequency waves are confined to the shallow, lower-velocity layers, while low-frequency waves penetrate deeper into higher-velocity materials. By recording ambient microtremors, geophysicists can identify these phase velocities across a broad frequency spectrum.

Data acquisition involves laying out a linear array of geophones, typically 12 to 48 sensors, spaced at intervals of 5 to 10 meters. Unlike active-source refraction surveys, which require a sledgehammer or explosive source, ReMi records the background "hum" of the site for 15 to 30 minutes. This ambient noise is processed through a slowness-frequency (p-f) transform, which maps the wavefield into a power spectrum where the dispersion curve of the fundamental mode Rayleigh wave is identified. The analyst picks the lower envelope of the energy peaks to define the dispersion curve, which is then subjected to a mathematical inversion to produce a one-dimensional shear-wave velocity profile.

Comparison with Borehole PS Logging

Borehole PS logging, specifically suspension logging or downhole travel-time measurement, provides a direct observation of the seismic wavefield. In suspension logging, a probe contains both the source and the receivers, allowing for measurements at discrete intervals (often 0.5 to 1.0 meters). This results in high-resolution data that can identify thin layers of soft clay or dense sand that might be smoothed over by surface wave methods.

However, borehole methods have significant limitations. They provide a point measurement that may not be representative of the wider site if lateral heterogeneity is present. Furthermore, the cost of drilling a 30-meter borehole in urban settings can be exorbitant, often exceeding the cost of the entire geophysical survey. ReMi, conversely, provides a site-averaged velocity profile. Because the array spans a significant horizontal distance, the resulting $V_{s30}$ value reflects the average conditions across the building footprint, which is often more useful for structural engineering purposes.

Table 1: Technical Comparison of Subsurface Characterization Methods

Inversion Algorithms and Empirical Reliability

The reliability of any surface wave method hinges on the inversion process—the procedure of converting a frequency-velocity dispersion curve into a depth-velocity model. This is an inherently non-unique problem, meaning multiple different soil profiles could theoretically produce the same dispersion curve. To resolve this ambiguity, analysts employ inversion algorithms based on least-squares optimization or global search methods like simulated annealing.

The empirical reliability of these algorithms is frequently verified by reconciling the results with lithological logs found in USGS databases or local geotechnical reports. For instance, if a USGS log indicates a transition from Holocene alluvium to Pleistocene terrace deposits at 12 meters, the inversion algorithm can be constrained to allow a velocity increase at that specific depth. This "constrained inversion" significantly reduces the uncertainty in the final $V_{s30}$ value. Research throughout the 2000s demonstrated that when properly constrained, ReMi results typically stay within 10% to 15% of borehole measurements, which is well within the tolerance required for site classification under the IBC.

What sources disagree on

Disagreement persists regarding the efficacy of passive source methods in very quiet, rural environments or in sites with extreme velocity inversions (where a hard layer overlies a soft layer). Traditional borehole logging advocates argue that ambient noise in rural settings may lack the low-frequency energy required to reach depths of 30 to 100 meters, potentially leading to an overestimation of the shear-wave velocity. Conversely, ReMi proponents suggest that even in quiet areas, atmospheric pressure changes and distant oceanic waves provide sufficient microtremor energy for deep penetration.

Another point of contention involves the identification of higher-mode Rayleigh waves. Standard ReMi analysis assumes the fundamental mode is dominant; however, in complex geological stratigraphies, higher modes can carry significant energy. If an analyst misidentifies a higher mode as the fundamental mode, the resulting velocity model will be erroneously fast. Borehole logging avoids this specific interpretive risk by measuring direct arrivals, though it remains susceptible to "short-circuiting" through the borehole casing or grout if the installation is poor.

Applications in Infrastructure and Non-Destructive Testing

The practical application of these methods extends beyond simple site classification. In the assessment of existing infrastructure, such as bridges and tunnels, surface wave analysis is used to detect voids or zones of delamination. By analyzing the dispersion curves of induced waves (active source) alongside microtremor data, engineers can identify anomalies in the subsurface that might compromise foundation integrity.

In urban planning, microtremor wavefield data is increasingly used for microzonation—the mapping of seismic hazard across an entire city. Because ReMi can be performed quickly along street medians or in parking lots without the need for drilling, it allows for the high-density data collection required to produce detailed subsurface maps. These maps help identify "blind zones" or buried utilities and voids that might not be captured in sparse borehole records, providing a more detailed understanding of the shallow subsurface environment.

Conclusion

The verification of subsurface velocity models through the comparison of passive source and borehole methods has established a strong framework for geotechnical investigation. While borehole PS logging remains essential for projects requiring high vertical precision or direct lithological sampling, the passive source ReMi method has proven to be a reliable, cost-effective alternative for $V_{s30}$ classification. By leveraging the benchmarks set by the California Geological Survey and utilizing USGS databases for model constraint, geophysicists can provide the accurate seismic profiles necessary for the safety of modern engineered structures.