Gregor Hillers

The green energy transition requires the development of new seismic imaging approaches to locate and characterize deposits of critical raw materials efficiently and cost-effectively. Here, we apply the Rayleigh-wave focal spot imaging approach to dense array data collected in the Kylylahti polymetallic mine area in the Outokumpu mineral district, Finland, and demonstrate the efficiency of the method in a hardrock environment. We cross-correlate one month of ambient-noise records from the 994 vertical-component stations deployed over an area of ∼10.5 km² following standard signal processing operations. The anatomy of the correlation wavefield includes spurious arrivals associated with the local mining operations and with a mine located to the SW of the array. From the narrow-band zero lag amplitude distribution of the cross-correlations we get the Rayleigh-wave focal spots which we parameterize with a Bessel-function model to estimate the local phase velocity at each station. The employed model accounts for the anisotropic background illumination and is constrained using focal spot data in the short 1λ distance range. In the 3 − 6 Hz frequency range we obtain phase velocity maps that correlate well with the Kylylahti formation, which is resolved with lower phase velocity values compared to the host rock. The lower phase velocities are associated with abundant soapstone, while higher phase velocities are indicative of the surrounding mica schist. The similarity of the phase-velocity features to the structure resolved by an independently obtained P-wave velocity model, together with the agreement between the inferred Kylylahti formation boundaries and the contours of the geological model, suggest that the focal spot method is a valuable, low-cost passive imaging tool for modern mineral exploration that can complement established approaches.

The green energy transition requires the development of new seismic imaging approaches to locate and characterize deposits of critical raw materials efficiently and cost-effectively. Here, we apply the Rayleigh-wave focal spot imaging approach to dense array data collected in the Kylylahti polymetallic mine area in the Outokumpu mineral district, Finland, and demonstrate the efficiency of the method in a hardrock environment. We cross-correlate one month of ambient-noise records from the 994 vertical-component stations deployed over an area of ∼10.5 km² following standard signal processing operations. The anatomy of the correlation wavefield includes spurious arrivals associated with the local mining operations and with a mine located to the SW of the array. From the narrow-band zero lag amplitude distribution of the cross-correlations we get the Rayleigh-wave focal spots which we parameterize with a Bessel-function model to estimate the local phase velocity at each station. The employed model accounts for the anisotropic background illumination and is constrained using focal spot data in the short 1λ distance range. In the 3 − 6 Hz frequency range we obtain phase velocity maps that correlate well with the Kylylahti formation, which is resolved with lower phase velocity values compared to the host rock. The lower phase velocities are associated with abundant soapstone, while higher phase velocities are indicative of the surrounding mica schist. The similarity of the phase-velocity features to the structure resolved by an independently obtained P-wave velocity model, together with the agreement between the inferred Kylylahti formation boundaries and the contours of the geological model, suggest that the focal spot method is a valuable, low-cost passive imaging tool for modern mineral exploration that can complement established approaches.

Characterizing the physical properties of frozen lakes, such as ice thickness and water depth, is essential for diverse applications, including ice safety assessment and climate change monitoring. Typically, these properties are measured through labor-intensive field surveys, limiting their spatial and temporal coverage. Alternatively, frozen lake environments can be investigated using seismology. Seismic waves propagating on floating ice travel at velocities that depend on physical properties of the ice–water system, including the elastic properties and thickness of the ice and the water depth. Because seismic waves are frequently generated and can propagate through entire lake systems, they have greater spatiotemporal coverage than field measurements. In this study, we investigate the use of wind-generated flexural waves recorded by a distributed acoustic sensing (DAS) system to infer ice thickness and water depth along a 1000 m fiber-optic cable installed on Lake Pääjärvi, southern Finland. We detect flexural waves in the 0.01–0.5 Hz frequency band, measure their phase-velocity dispersion curves, estimate ice thickness and water depth along four cable intervals, and compare the estimates with field measurements and the lake bathymetry. The ice-thickness estimates are comparable to field measurements, with differences ranging from 2% to 34%, whereas the water-depth estimates underestimate bathymetry values by 22%–60% while resolving the increasing water-depth trend along the cable. We find that water-depth estimates tend to become less robust as the water depth increases because the measured dispersion curves lack sufficiently low wavenumbers to resolve the water column. We illustrate that flexural waves can refract toward the shore as they approach shallow water. This effect must be considered when estimating true phase velocities from apparent velocities measured along the cable to avoid bias in the estimated ice thickness and water depth. Our results demonstrate the potential of DAS to leverage ambient signals for monitoring frozen lakes.

The green energy transition requires the development of new seismic imaging approaches to locate and characterize deposits of critical raw materials efficiently and cost-effectively. Here, we apply the Rayleigh-wave focal spot imaging approach to dense array data collected in the Kylylahti polymetallic mine area in the Outokumpu mineral district, Finland, and demonstrate the efficiency of the method in a hardrock environment. We cross-correlate one month of ambient-noise records from the 994 vertical-component stations deployed over an area of ∼10.5 km² following standard signal processing operations. The anatomy of the correlation wavefield includes spurious arrivals associated with the local mining operations and with a mine located to the SW of the array. From the narrow-band zero lag amplitude distribution of the cross-correlations we get the Rayleigh-wave focal spots which we parameterize with a Bessel-function model to estimate the local phase velocity at each station. The employed model accounts for the anisotropic background illumination and is constrained using focal spot data in the short 1λ distance range. In the 3 − 6 Hz frequency range we obtain phase velocity maps that correlate well with the Kylylahti formation, which is resolved with lower phase velocity values compared to the host rock. The lower phase velocities are associated with abundant soapstone, while higher phase velocities are indicative of the surrounding mica schist. The similarity of the phase-velocity features to the structure resolved by an independently obtained P-wave velocity model, together with the agreement between the inferred Kylylahti formation boundaries and the contours of the geological model, suggest that the focal spot method is a valuable, low-cost passive imaging tool for modern mineral exploration that can complement established approaches.

Inversions of dynamic multi‐scale multi‐physics subsurface system properties rely on the effective sampling of high‐dimensional parameter spaces and are thus notoriously difficult to perform. The associated models are defined by complex nonlinear relationships with potentially stochastic components, which lead to analytically intractable likelihoods that require challenging forward simulations of the system response. This paper presents the application of the Bayesian optimization for likelihood‐free inference (BOLFI) algorithm to invert the properties of a geothermal reservoir. We use the viscoelastic damage rheology simulator HydroPED to model an enhanced geothermal system (EGS) with controlled reservoir properties. Our results demonstrate the effectiveness of combining BOLFI and the computationally expensive simulator HydroPED to estimate an informative posterior distribution from a comparatively small number of forward evaluations. In this pilot, we use BOLFI to invert for ambient principal stress values using synthetic target data. We parameterize the 2‐km scale reservoir with ∼16′000 elements and perform tests with 200 and 500 HydroPED forward simulations that each take a wall‐clock time of 1.5 hr using an 8‐core CPU node to simulate one day of a hydraulic stimulation. Using potency, which encodes the spatial distribution of energy release within the simulator reservoir as a target function for the inversion, yields better results compared to a simpler event count target function. The approach we use in this study can be extended to a wide range of seismic and hydraulic target observations to constrain ambient stress and local EGS model parameters.

The green energy transition requires the development of new seismic imaging approaches to locate and characterize deposits of critical raw materials efficiently and cost-effectively. Here, we apply the Rayleigh-wave focal spot imaging approach to dense array data collected in the Kylylahti polymetallic mine area in the Outokumpu mineral district, Finland, and demonstrate the efficiency of the method in a hardrock environment. We cross-correlate one month of ambient-noise records from the 994 vertical-component stations deployed over an area of ∼10.5 km² following standard signal processing operations. The anatomy of the correlation wavefield includes spurious arrivals associated with the local mining operations and with a mine located to the SW of the array. From the narrow-band zero lag amplitude distribution of the cross-correlations we get the Rayleigh-wave focal spots which we parameterize with a Bessel-function model to estimate the local phase velocity at each station. The employed model accounts for the anisotropic background illumination and is constrained using focal spot data in the short 1λ distance range. In the 3 − 6 Hz frequency range we obtain phase velocity maps that correlate well with the Kylylahti formation, which is resolved with lower phase velocity values compared to the host rock. The lower phase velocities are associated with abundant soapstone, while higher phase velocities are indicative of the surrounding mica schist. The similarity of the phase-velocity features to the structure resolved by an independently obtained P-wave velocity model, together with the agreement between the inferred Kylylahti formation boundaries and the contours of the geological model, suggest that the focal spot method is a valuable, low-cost passive imaging tool for modern mineral exploration that can complement established approaches.

Ambient seismic noise interferometry is widely used for surface wave imaging. It typically requires synchronous station recordings and assumes uniform noise source distributions. Higher‐order correlations, such as the re‐correlation of the reconstructed direct waves , enable imaging with asynchronous data and can improve an incomplete source distribution. We show with field data and simulations that surface wavefields are, however, highly sensitive to the original source distribution and even amplify the effects associated with directional incidence. This can lead to systematic errors in the obtained velocity estimates and the downstream subsurface images. A range of auxiliary station distributions tested for the re‐correlation process can not mitigate this bias but introduce additional wavefield distortions. We find that both local and far‐field imaging approaches using higher‐order correlation wavefields are affected by significant and systematic velocity estimation errors. Our results show that the re‐correlation of direct waves is not an all‐purpose correlation wavefield enhancement technique, which highlights the need for a careful consideration of source effects for improved imaging.

SUMMARY: Tomographic reconstruction involves fundamental trade-offs between resolution, data coverage and parametrization. Conventional approaches commonly rely on regularized pixel-wise or cell-based parametrizations. These regularizations are valuable and can themselves be geology-informed when their assumptions match the target setting. However, they often encode structural information only indirectly through generic penalty terms and regularization weights. As a result, they can be less suited to questions in which the objective is to test a specific structural hypothesis. We develop a Bayesian framework for tomographic reconstruction and uncertainty quantification that formulates such hypotheses explicitly and assesses them through posterior uncertainty. As an illustrative application of the framework, we consider ambient noise tomography. The phase velocity reconstruction is reformulated to address a targeted geological hypothesis by introducing a parametrized structural prior. The prior is constructed from a Whittle–Matérn latent Gaussian field, combined with spectral dimensional reduction via a truncated Karhunen–Loève expansion and a differentiable nonlinear pushforward mapping that promotes approximately piecewise-constant velocity structures. The prior parameters, including the correlation length, smoothness, pushforward slope and velocity bounds, provide interpretable hyperparameters related to expected lateral scale, structural regularity, transition sharpness and plausible phase-velocity contrasts. In this formulation, the prior is used as a parametrized structural assumption informed by geological context, rather than as hard geological knowledge. The resulting parametrization substantially reduces dimensionality, making both maximum a posteriori estimation and Hamiltonian Monte Carlo sampling computationally feasible for the synthetic and observed data sets considered here. Synthetic experiments demonstrate accurate recovery of prescribed velocity contrasts and show that posterior uncertainty patterns reflect acquisition geometry and data coverage. Application to an $\sim$1000-sensor dense array data set produces structures consistent with the imposed geology-informed prior assumptions, while providing spatially resolved uncertainty estimates that identify regions of robust inference and remaining ambiguity. The framework offers a proof-of-concept formulation for goal-oriented, uncertainty-aware seismic tomography when prior information supports an explicit structural hypothesis that can be encoded through the proposed Whittle–Matérn latent-field parametrization.

The green energy transition requires the development of new seismic imaging approaches to locate and characterize deposits of critical raw materials efficiently and cost-effectively. Here, we apply the Rayleigh-wave focal spot imaging approach to dense array data collected in the Kylylahti polymetallic mine area in the Outokumpu mineral district, Finland, and demonstrate the efficiency of the method in a hardrock environment. We cross-correlate one month of ambient-noise records from the 994 vertical-component stations deployed over an area of ∼10.5 km2 following standard signal processing operations. The anatomy of the correlation wavefield includes spurious arrivals associated with the local mining operations and with a mine located to the SW of the array. From the narrow-band zero lag amplitude distribution of the cross-correlations we get the Rayleigh-wave focal spots which we parameterize with a Bessel-function model to estimate the local phase velocity at each station. The employed model accounts for the anisotropic background illumination and is constrained using focal spot data in the short 1λ distance range. In the 3 − 6 Hz frequency range we obtain phase velocity maps that correlate well with the Kylylahti formation, which is resolved with lower phase velocity values compared to the host rock. The lower phase velocities are associated with abundant soapstone, while higher phase velocities are indicative of the surrounding mica schist. The similarity of the phase-velocity features to the structure resolved by an independently obtained P-wave velocity model, together with the agreement between the inferred Kylylahti formation boundaries and the contours of the geological model, suggest that the focal spot method is a valuable, low-cost passive imaging tool for modern mineral exploration that can complement established approaches.

The multi-scale composition, structure, and dynamics of seasonal ice floating on freshwater lakes are influenced by ambient conditions. Here we describe a comprehensive geoscience experiment for lake system imaging and monitoring of spatiotemporal ice property variations. We explore the resolution of meteorological and environmental driving mechanisms that can include the quantification of methane degassing from boreal lakes. The project centerpiece is a seismic array of 210 geophones arranged in an aperiodic tiling configuration that was deployed in February 2025 on the ~25 cm thick ice of Lake Pääjärvi in southern Finland. The 10-km scale lake array is complemented by three dense circular arrays, 31 land-based sensors, eight broadband seismometers, three accelerometers, a rotational seismometer, a Distributed Acoustic Sensing system with a 1 km-long fibre optic cable, an underwater echosounder, a microphone, a Ground Penetrating Radar (GPR) survey, water chemistry measurements, manual ice thickness sampling and ice coring, and meteorological observations. We observe the strongly dispersive QS flexural mode and the weakly or non-dispersive QS₀ and HS₀ modes excited by hammer shots, icequakes, and environmental sources and reconstruct the average propagation using beamforming and noise correlations. Propagation speed estimates for the three modes range approximately between 20–100 m s⁻¹, 3000–3400 m s⁻¹, and 1650–1800 m s⁻¹, respectively. High values for the Poisson's ratio ν = 0.42 and Young's modulus E = 8.59 GPa reflect the overall competent characteristics of the ice referred to as teräsjää (steel ice). Seismic activity in the 0.03–0.2 Hz band increases during high wind speed episodes, and signals above 0.1 Hz correlate with rapid air-temperature cooling events. The GPR profile images the spatial ice variability across the lake that is compatible with the in-situ measurements, and we show that seismo-acoustic observations can be inverted for similarly compatible thickness estimates. The geochemical water and ice sample analysis suggests Lake Pääjärvi is a source of methane, and localized ebullition can potentially be resolved from echosounder data. This synthesis demonstrates that the application of environmental seismology concepts can form a bridge between bottom-up ebullition monitoring and remote-sensing approaches.