Shear-Wave Velocity Models of the Subsurface Critical Zone at Mpala, Kenya, from Ambient Noise

In this study, we develop 1D shallow subsurface velocity profiles from one year of ambient-noise measurements collected at the Mpala Research Centre, a nature and wildlife preservation area in Laikipia County, Kenya. The Mpala Research Centre is located in a rangeland shared by wildlife, livestock, and people, and experienced extreme drought in 2023. Water access in such times depends largely on shallow wells that are influenced by subsurface hydrogeology. In addition, there is growing interest in seismic wildlife monitoring based on interpreting ground-coupled animal vibrations whose amplitudes and dominant frequencies are shaped by local site effects. Both issues point to the need for information on the shallow subsurface structure. Mapping site specific shear-wave velocities (Vs) provides a common foundation to (i) relate stratigraphy to shallow groundwater availability and (ii) correct for spatial variability in amplification that biases wildlife signal detectability.

Motivated by these needs, we develop the first locally constrained Vs models in the Mpala area based on continuous seismic data (Jan 2023–Jan 2024) across a 15 broadband station array using ambient noise HVSR and passive seismic interferometry. The HVSR and Rayleigh wave dispersion measurements from the two methods are jointly inverted. The dispersion curves’ frequency band (≈ 2–9 Hz) provides depth sensitivity of ∼70–430 m, while HVSR constrains near-surface impedance contrasts. Across the station network, we estimate three consistent velocity contrasts in the upper 100 m, first at ∼ 1–4 m (Vs ∼ 280–500 m/s), then at ∼5–20 m (Vs ∼ 345–900 m/s), and finally at ∼ 15–50 m (Vs ∼ 1160–2600 m/s). The resulting Vs models support well siting and inform how to account for local site-amplifications effects for monitoring and modeling ground-coupled wildlife sensing at Mpala. For future work we recommend multi-scale seismic array configurations with both locally denser and targeted wider spacing to more reliably estimate both shallow depths and the overall geological structure. Improved dispersion-curve measurements help reduce uncertainty arising from the limited dispersion band, uneven station pair coverage, and capture possible lateral heterogeneity.