Were changes in stress state responsible for the 2019 Ridgecrest, California, earthquakes?

The 2019 Ridgecrest earthquake sequence, culminating in a M7.1 event on July 5th, 2019 (UTC), and a preceding M6.4 event, occurred within the seismically active Eastern California Shear Zone (ECSZ). Understanding the mechanisms behind this sequence, particularly the role of stress perturbation, is crucial for assessing future seismic hazards. Previous studies have utilized Coulomb stress models to explain the triggering of the M6.4 quake and its influence on the subsequent M7.1 event. However, these physics-based approaches have limitations due to uncertainties in fault locations. This study employs an alternative statistics-based approach, focusing on the spatial and temporal variations of the b-value from the Gutenberg-Richter law (log₁₀N = a - bM), a measure sensitive to differential stress. Analyzing the b-value before, during, and after the Ridgecrest earthquakes, along with Coulomb stress calculations, aims to provide insights into the stress state evolution and its connection to earthquake nucleation and propagation.

Coulomb stress models have been used to explain the Ridgecrest earthquake sequence, suggesting that the M6.4 event was influenced by previous large earthquakes (1872 Owens Valley, 1992 Landers, 1999 Hector Mine) and that it subsequently loaded the area where the M7.1 quake nucleated. However, physics-based approaches using Coulomb stress transfer have not consistently outperformed statistical models in forecasting large earthquakes, partly due to uncertainties in fault locations. This study leverages the advantages of a statistics-based approach focusing on the b-value, a parameter reflecting stress changes confirmed in numerous laboratory and field studies.

The study utilizes the earthquake catalog from the Southern California Seismic Network (SCSN), processing over 10⁵ earthquakes (M ≥ 1) since 1980 within the study region. The b-value is estimated using the Entire-Magnitude-Range (EMR) technique, which simultaneously calculates the a-value and the completeness magnitude (Mc). Spatial and temporal variations in b-value are mapped using a grid-based approach and circular sampling volumes. The optimal sampling radius is determined through a sensitivity analysis. Coulomb stress changes are calculated using the Coulomb software, employing finite-fault models for the M6.4 and M7.1 quakes from Xu et al. (2020) and moment tensor solutions for larger events between the two mainshocks. The Omori-Utsu (OU) law is used to model aftershock decay, and a modified approach based on Lippiello et al. (2012) is employed to analyze spatial seismicity concentration before the M7.1 quake. The EMR method is compared against other b-value estimation techniques (MAXC and GOF) to assess the robustness of the results. The ETAS model is also used to compare the results obtained by the OU model for the temporal analysis of seismicity.

The study reveals that both the M6.4 and M7.1 earthquakes nucleated in pre-existing zones of low b-values (~0.6–0.7), indicating high stress. After the M6.4 quake, a stress transfer occurred, increasing the b-value at the M6.4 hypocenter and decreasing it at the future M7.1 hypocenter. Coulomb stress calculations confirm that the M6.4 quake, along with subsequent seismicity, increased the Coulomb stress around the M7.1 hypocenter by approximately 2 bars, bringing the region closer to failure. Analysis of the post-M6.4 seismicity using the Omori-Utsu (OU) law shows slower decay in the northern area (including the M7.1 hypocenter), suggesting slower stress decrease. The spatial analysis based on the parameter φ confirms a seismic concentration near the M7.1 hypocenter before the event. The post-M7.1 analysis reveals a zone of low b-values near the Garlock fault, where a significant amount of stress remains unreleased. This correlates with the distribution of the M7.1 slip. The spatial distribution of b-values shows a good correlation with the slip distribution of the M7.1 quake. Areas of high slip did not overlap with areas of high b-values. The b-value analysis further shows a decrease in b-values over time in an area near the Garlock fault after the M7.1 event.

The findings strongly suggest that changes in the stress state played a critical role in triggering both the M6.4 and M7.1 earthquakes. The stress transfer from the M6.4 event, confirmed by Coulomb stress calculations and the temporal variations in b-value, likely facilitated the nucleation and rupture of the M7.1 event. The identified low-b-value zone near the Garlock fault after the M7.1 event highlights a potential future seismic hazard. Although the Garlock fault has been historically quiescent, it has a history of large earthquakes. Future ruptures could propagate along either left-lateral (M6.4 type) or right-lateral (M7.1 type) faults, possibly influencing the Garlock fault. The observed decrease in b-values over time near the Garlock fault after the M7.1 event is indicative of stress accumulation.

The combined analysis of b-values and Coulomb stress changes provides compelling evidence linking stress state changes to the 2019 Ridgecrest earthquake sequence. The identification of a highly stressed region near the Garlock fault underscores the need for continued monitoring of the ECSZ to assess future seismic hazards. Future research should focus on quantitative risk assessment methods like nowcasting to better predict the probability of future large earthquakes in the region. The methodology developed in this study can be valuable for future investigations of earthquake triggering mechanisms and seismic hazard assessment.

The study's conclusions are based on statistical analysis and modeling, which inherently involve uncertainties. The b-value is an indirect measure of stress, and other factors may influence its spatial and temporal variations. While the Coulomb stress calculations provide insights, the precise locations and properties of all potential faults are unknown, introducing uncertainties in the stress estimations. The quantitative predictive power of b-value mapping for forecasting future earthquakes still needs to be further evaluated, which will require more advanced nowcasting techniques. The study does not provide a quantitative estimate of the probability of future rupture on the Garlock fault.