Search

We analyze envelopes of 233 and 22 ML0.0 to ML1.8 earthquakes induced by two geothermal stimulations in the Helsinki, Finland, metropolitan area. We separate source spectra and site terms and determine intrinsic attenuation and the scattering strength of shear waves in the 3–200 Hz frequency range using radiative transfer based synthetic envelopes. Displacement spectra yield scaling relations with a general deviation from self‐similarity, with a stronger albeit more controversial signal from the weaker 2020 stimulation. The 2020 earthquakes also tend to have a smaller local magnitude compared to 2018 earthquakes with the same moment magnitude. We discuss these connections in the context of fluid effects on rupture speed or medium properties. Site terms demonstrate that the spectral amplification relative to two reference borehole sites is not neutral at the other sensors; largest variations are observed at surface stations at frequencies larger than 30 Hz. Intrinsic attenuation is exceptionally low with Qi−1 ${Q}_{\mathrm{i}}^{-1}$ values down to 2.4 × 10−5 at 20 Hz, which allows the observation of a diffuse reflection at the ∼50 km deep Moho. Scattering strength is in the range of globally observed data with Qsc−1 ${Q}_{\mathrm{s}\mathrm{c}}^{-1}$ between 10−3 and 10−4. The application of the employed Qopen analysis program to the 2020 data in a retrospective monitoring mode demonstrates its versatility as a seismicity processing tool. The diverse results have implications for scaling relations, hazard assessment and ground motion modeling, and imaging and monitoring using ballistic and scattered wavefields in the crystalline Fennoscandian Shield environment. Plain Language Summary: We analyze seismograms from earthquakes that were induced during two geothermal stimulation experiments in the Helsinki, Finland, metropolitan area, in 2018 and 2020. We process long signals including later parts of the seismograms to solve the persistent problem of separating the effects of the earthquake source process, of the bedrock, and of the ground immediately below a seismic sensor on the observed data. The high data quality allows us to measure systematic differences in some fundamental earthquake source parameters between events induced during the two stimulations. We attribute this to the effect of the fluids that were pumped into the 6 km deep rock formations. These observations are important since natural earthquakes and earthquakes induced by such underground engineering activities are governed by the same physical mechanisms. We also find that the bedrock in southern Finland is characterized by some of the lowest seismic attenuation values that have so far been measured in different tectonic environments. Last, the so‐called site effects at the instrument locations show a diverse amplification pattern in a wide frequency range, which is important for the assessment of shaking scenarios in the area. Key Points: We find lower stress drop values for events induced by the 2020 compared to the 2018 stimulation and a deviation from self‐similar scalingThe observation of a diffuse reflection at the 50 km deep Moho highlights the low intrinsic attenuation in the Fennoscandian ShieldSite effect terms between 3 and 200 Hz show diverse frequency and site dependent patterns with high‐frequency amplification [ABSTRACT FROM AUTHOR]

We estimate vertical rotation rates for 204 earthquakes that were induced by the 2018 stimulation of the Espoo/Helsinki geothermal reservoir from wavefield gradients across geophone arrays. The array‐derived rotation rates from seismograms recorded at 6–9 km hypocentral distances vary between 10−9 and 10−7 rad s−1, indicating a comparable sensitivity to portable rotational instruments. Using co‐located observations of translational and rotational motion, we estimate the local propagation direction and the apparent phase speed of SH waves, and compare these estimates with those obtained by S wave beamforming. Propagation directions generally align with the earthquake back azimuths, but both techniques show deviations indicative of heterogeneous seismic structure. The rotational method facilitates a station‐by‐station approach that resolves site specific variations that are controlled by the local geology. We measure apparent S wave speeds larger than 5 km s−1, consistent with steep incidence angles and high propagation velocities in the Fennoscandian Shield. Plain Language Summary: Earthquakes generate seismic waves consisting of both translational (back‐and‐forth) and rotational ground motion. Translational motion is routinely measured by standard seismometers, but the observation of the rotational motion requires relatively expensive and rare instruments. In this study we estimate rotational ground motion caused by earthquakes using groups of translational seismometers. The computation of rotational motion from translational seismometers has been demonstrated before, but the novelty of our study is to use high‐quality recordings of earthquakes that were induced by the creation of a geothermal reservoir at 6 km depth in bedrock. We use our measurements of ground rotation to estimate the speed and direction in which the seismic waves are travelling when they reach the seismometers. We find that the direction in which the seismic waves travel usually points back to the earthquake location, but at some seismometers the waves arrive from a different direction. At these locations, it is likely that local geological features are altering the direction of the waves. We expect that our findings will provide access to approaches for determining earthquake characteristics and Earth structure that currently require highly specialized instruments. Key Points: Wavefield gradiometry is applied to arrays of translational seismometers to estimate rotational motions from induced earthquakesHigh‐quality records of M0.0 to M1.8 events obtained within 10 km hypocentral distance facilitate array‐based resolution of rotation ratesThe obtained apparent phase speed and wave propagation directions are linked to subsurface structure [ABSTRACT FROM AUTHOR]