Moment Tensors of Induced Microearthquakes in The Geysers Geothermal Reservoir From Broadband Seismic Recordings: Implications for Faulting Regime, Stress Tensor, and Fluid Pressure

Abstract

Full moment tensors of 1,421 microearthquakes in The Geysers geothermal field were calculated using waveform data from a field‐wide broadband network and the approach based on the principal component analysis. Spatial characteristics of faulting regime, stress tensor and the isotropic component (ISO) of moment tensors were investigated. The studied events form different clusters dominated by normal faults (NF) and strike‐slip (SS) faults, respectively. The SS‐dominated clusters are related to the SS stress state observed in the southwestern side of the field where two NW‐SE trending fault zones exist. Increasing proportions of SS faults were observed near the shallow and deep parts of the NF‐dominated clusters. Temperature differences between the upper and lower parts of the reservoir do not change the overall stress states of clusters. The stress ratios in the NW part are much smaller than in the SE part of the field. The retrieved ISOs range between −5% and 25% for 96% of events. The average percentages of positive ISOs are correlated with the average injection rates in different clusters, and keep an increasing trend with depth below the main injection interval. Only 10% of events show negative ISO and are mainly constrained within the depth range of steam extraction. The proportion of events with negative ISO suggests much weaker seismic responses of the steam extraction than the water injection. The spatial variations of the ISO percentage do not follow the seismicity variations as the ISO is more sensitive to the pressure changes related to the fluid injection and migration.

Analysis of Microseismicity Framing ML > 2.5 Earthquakes at The Geysers Geothermal Field, California

“…The recorded broadband, high‐density, and good azimuthal coverage continuous recordings resulted in one of the highest‐quality data sets for geothermally induced seismicity. Main noise components were found to be microseism (0.16 Hz) and anthropogenic noise (10–20 Hz; Yu et al, ).…”

Frequency‐Dependent Moment Tensors of Induced Microearthquakes

“…Consequently, the local seismicity rate was significantly enhanced Dellinger et al, 2017;Majer & Peterson, 2007). In Yu et al (2018), the field-wide variations of faulting regime, stress tensor, and ISO component were investigated using the full MTs of 1,421 induced microearthquakes recorded by a broadband seismic network. R. Johnson, 2014b;Kwiatek et al, 2015;Martínez-Garzón et al, 2014.…”

“…In order to get robust results, ranges of five parameters; the DC, CLVD, and ISO percentages; and the orientations of P/T axes are calculated for a set of eight optimal MTs of each event (see Figure S2 in the supporting information). According to the statistical distributions (see Figures 5 and 10 in Yu et al, 2018), the total variations of the DC, CLVD, and ISO percentages and of the orientations of the P/T axes are determined at the 95% confidence levels. According to the statistical distributions (see Figures 5 and 10 in Yu et al, 2018), the total variations of the DC, CLVD, and ISO percentages and of the orientations of the P/T axes are determined at the 95% confidence levels.…”

“…The spatiotemporal variation of the ISO component can directly be related to the injection and migration of fluids (Martínez-Garzón et al, 2017;Yu et al, 2018). The spatiotemporal variation of the ISO component can directly be related to the injection and migration of fluids (Martínez-Garzón et al, 2017;Yu et al, 2018).…”

“…However, the variations of the DC, CLVD, and ISO do not follow the tendencies of the RMS, noise, and the S/N ratio. The increasing uncertainty of MTs with the noise level is only reflected by more scattered distributions of the DC, CLVD, and ISO, resulting in widening of their probability density (Yu et al, 2018). c The ratio of the average range value and the width of confidence interval.…”