A Unified Model of Crustal Stress Heterogeneity From Borehole Breakouts and Earthquake Focal Mechanisms

Observations of crustal stress orientation from the regional inversion of earthquake focal mechanisms often conflict with those from borehole breakouts, possibly indicating local stress heterogeneity, either laterally or with depth. To investigate this heterogeneity, we compiled SHmax estimates from previous studies for 57 near-vertical boreholes with measured breakout azimuths across the Los Angeles region. We identified subsets of earthquake focal mechanisms from established earthquake catalogs centered around each borehole with various criteria for maximum depth and maximum lateral distance from the borehole. Each subset was independently inverted for 3-D stress orientation and corresponding SHmax probability distributions, then compared with the corresponding borehole breakout-derived estimate. We find good agreement when both methods sample the basement stress (breakouts are close to the sediment-basement interface), or when both methods sample the mid-basin stress (sufficient earthquakes are present within a sedimentary basin). Along sedimentary basin margins, in contrast, we find acceptable agreement only when focal mechanisms are limited to shallow and close earthquakes, implying short-length-scale heterogeneity of <20 km. While the region as a whole shows evidence of both lateral and vertical stress orientation heterogeneity, we find a more homogeneous stress state within basement rock, over length scales of 1-35 km. These results reconcile the apparently conflicting observations of short-length-scale heterogeneity observed in boreholes, which sample primarily the basins, with the relative homogeneity of stress inferred from focal mechanisms, which sample primarily the basement, and imply distinct regimes for the appropriate use of each type of stress indicator. Plain Language Summary Earthquakes and other tectonic phenomena respond largely to the state of stress within the Earth's crust, but this stress state is extraordinarily difficult to measure. Two methods are commonly used to estimate at least the orientation of the stress field within the crust: the azimuth of breakout fractures observed within drilled boreholes, which tend to be <3 km deep within sedimentary basins, and the inversion of sets of earthquake focal mechanisms observed seismically, which tend to be >5 km deep within bedrock. Studies using these two techniques have repeatedly arrived at different conclusions about how variable the crustal stress field is, and particularly have disagreed about the distances over which those variations occur. The discrepancy may be due to differences in the volume of crust sampled by each estimate, and we test this by systematically comparing observations from both types of estimates using a range of different crustal volumes. We find that if the two estimates are made near enough to one another, especially in depth, they generally agree, but "how near is near enough" changes depending on the local geology. These investigations help bridge the gap between our understanding of how stress changes over different length scales.