The Evolution of MASW: From the 1999 Kansas Geological Survey to Modern Practice

In 1999, Choon B. Park, Richard D. Miller, and Jianghai Xia of the Kansas Geological Survey (KGS) published a seminal paper in the journalGeophysicsTitled "Multichannel analysis of surface waves." This publication formalized the Multichannel Analysis of Surface Waves (MASW) method, a seismic approach designed to estimate the shear-wave velocity (Vs) of the shallow subsurface. By utilizing the dispersive nature of Rayleigh waves—where different frequencies penetrate and travel through different depths—the researchers established a strong framework for non-destructive geotechnical site characterization.

Before the introduction of MASW, the geophysical community relied heavily on the Spectral Analysis of Surface Waves (SASW), which utilized only two receivers. The shift to a multichannel approach allowed for the simultaneous recording of a wavefield at multiple points along a linear array, significantly improving the signal-to-noise ratio and the ability to differentiate between various types of seismic waves. Today, the method is a standard tool in civil engineering, environmental studies, and seismic hazard assessment, having evolved from simple one-dimensional profiling to complex three-dimensional subsurface imaging.

Timeline

• Pre-1990s:Dominance of Spectral Analysis of Surface Waves (SASW) using two-channel recording systems; frequent issues with spatial aliasing and ambient noise interference.
• 1999:Publication of the MASW method by Park, Miller, and Xia, introducing the phase-shift method for calculating dispersion curves from multichannel data.
• Early 2000s:Expansion of MASW from active source (hammer or weight drop) to passive source methods, utilizing microtremors for deeper penetration.
• 2005–2010:Development of 2D MASW profiling through the "roll-along" technique, allowing for the construction of continuous shear-wave velocity cross-sections.
• 2015–Present:Integration of high-density 3D arrays, automated inversion algorithms using global search heuristics, and full-waveform inversion (FWI) research for higher resolution.

Background

The fundamental principle of MASW is the dispersive property of surface waves in a heterogeneous medium. In a vertically layered earth, the velocity of a Rayleigh wave depends on the frequency of the wave; higher frequency components have shorter wavelengths and travel through the shallower layers, while lower frequency components have longer wavelengths and are influenced by deeper material properties. By measuring the phase velocity of these different frequencies, geophysicists can construct a dispersion curve.

Prior to 1999, SASW was the primary method for surface wave analysis. However, SASW was susceptible to "receiver-mode" errors, where the presence of body waves or higher-mode surface waves could not be easily distinguished from the fundamental mode Rayleigh wave. The introduction of multichannel recording allowed for the application of wavefield transformation techniques. By transforming data from the time-space (t-x) domain into the frequency-phase velocity (f-v) or frequency-wavenumber (f-k) domain, analysts could visually identify and isolate the fundamental mode of the Rayleigh wave from noise and reflections, a process that significantly increased the accuracy of the resulting shear-wave velocity profiles.

The 1999 Innovation: The Phase-Shift Method

The KGS researchers introduced the phase-shift method as a more efficient alternative to the frequency-wavenumber (f-k) transform previously used in deep seismic exploration. The phase-shift method involves the application of a specific phase shift to the Fourier-transformed data of each channel, followed by a summation. This process scans through a range of potential phase velocities for each frequency, and the velocity that results in the maximum constructive interference (the highest amplitude) is identified as the phase velocity for that frequency. This approach offered superior resolution and was computationally lighter than its predecessors, making it feasible for the limited processing power of late-90s field laptops.

The Transition from 1D Profiling to 2D and 3D Imaging

In its original iteration, MASW was primarily used to create 1D Vs profiles directly beneath the center of a geophone array. While useful for localized site class determination (such as calculating the Vs30 for building codes), 1D profiling offered limited information regarding lateral variations in geology. The move to 2D imaging was achieved by adopting seismic reflection techniques, specifically the "roll-along" method. By shifting the entire array of geophones and the source by a set distance, researchers could generate a series of 1D profiles that, when interpolated, provided a vertical cross-section of the subsurface.

Modern practice has moved toward 3D MASW, which utilizes a grid of geophones rather than a single line. This allows for the characterization of complex geometries, such as buried paleochannels, karst features, or localized voids. The data processing for 3D MASW is significantly more intensive, requiring the handling of azimuthal variations in wave propagation. In 3D models, researchers can visualize the volumetric distribution of elastic moduli, providing a detailed view for large-scale infrastructure projects like dam foundations or airport runways.

Manual Picking vs. Automated Inversion

For much of the history of MASW, the "picking" of the dispersion curve—identifying the trend of maximum energy on a frequency-velocity plot—was a manual task requiring significant expertise. An analyst had to distinguish the fundamental mode from higher modes and numerical artifacts. Once the curve was picked, it was subjected to an inversion routine, usually a linearized least-squares approach, to produce the final Vs profile. This manual intervention introduced a degree of subjectivity into the results.

Contemporary practice has shifted toward automated and semi-automated inversion routines. These advancements are characterized by several key features:

• Global Search Algorithms:Instead of relying on a single starting model (which can lead to local minima), modern software often employs Genetic Algorithms (GA), Simulated Annealing, or Particle Swarm Optimization to explore a wider range of possible subsurface models.
• Multimodal Inversion:Automated routines now frequently invert both the fundamental mode and higher modes simultaneously. Higher modes carry essential information about deeper layers and can help constrain the inversion process, leading to more stable results.
• Uncertainty Analysis:Modern routines provide quantitative estimates of uncertainty, showing the range of models that fit the observed data within a certain tolerance. This allows engineers to understand the reliability of the Vs estimates.

The following table summarizes the key differences between the early implementation of MASW and modern standards:

Geotechnical Applications and Infrastructure Testing

The practical application of MASW has expanded beyond simple soil classification. It is now a critical component of non-destructive testing (NDT) for aging infrastructure. By analyzing the dispersion curves of surface waves induced on engineered materials, practitioners can detect delamination in bridge decks, assess the integrity of dam cores, and locate voids beneath concrete slabs. The ability to measure shear-wave velocity is particularly valuable because Vs is directly related to the small-strain shear modulus (Gmax), a fundamental parameter for calculating soil settlement and structural response to seismic loading.

Furthermore, MASW is frequently used in urban environments where traditional invasive methods, such as drilling or cone penetration testing (CPT), are restricted. Passive MASW, which records ambient vibrations from traffic or machinery, allows for deep subsurface profiling (up to 100 meters or more) without the need for large impulsive sources, making it ideal for dense city centers. The development of wireless geophone nodes has further simplified these urban surveys, removing the logistical challenges associated with long, bulky cables in high-traffic areas.

Technical Challenges and Current Limitations

Despite its evolution, MASW faces persistent challenges. One of the primary issues is the "non-uniqueness" of the inversion problem; multiple subsurface models can theoretically produce the same dispersion curve. This necessitates the use of independent data—such as borehole logs or CPT data—to constrain the inversion and ensure geological plausibility. Additionally, the presence of stiff layers over soft layers (velocity inversions) can complicate the identification of the fundamental mode, leading to potential inaccuracies if higher modes are misidentified.

Current research at the Surface Wave Hub and similar institutions focuses on refining Full-Waveform Inversion (FWI) for shallow applications. Unlike standard MASW, which only uses the dispersion characteristics, FWI utilizes the entire seismic record—including amplitudes and phases—to create high-resolution images. While computationally demanding, FWI represents the next logical step in the evolution of surface wave analysis, promising to bridge the gap between geophysical modeling and geotechnical engineering requirements.