We examine earthquake hypocenters, focal mechanisms, and the state of tectonic stress in the Lower St. Lawrence Seismic Zone (LSZ), a paleorift zone in eastern Canada. The largest earthquake recorded in the region is the 1999 Côte–Nord MN 5.1, which was followed by ∼80 aftershocks of MN>1⁠. It is not known if the region is capable of producing hazardous Mw>6 earthquakes, similar to the Charlevoix Seismic Zone ∼250 km upriver. Focusing on 2015–2020, we apply a machine‐learning‐based phase picker to detect 72 earthquakes in addition to the 150 catalog earthquakes in the same region over this time span. We produce an updated 1D, gradient velocity model via a Monte Carlo search using a uniform VP/VS=1.77⁠, which we computed with the Wadati method. We refine hypocenter estimates using the triple‐difference method, with sP depth phases as additional constraints on earthquake depth. We estimate focal mechanisms for >100 earthquakes with automatically picked P‐wave first motions and absolute value P‐SV‐SH amplitude ratios, and we use the focal mechanisms to invert for the state of tectonic stress. Grid searches and Bayesian analysis allow for robust uncertainty estimates of focal mechanisms, which in turn allow for uncertainty estimates of the stress tensor. The recovered west‐northwest–east‐southeast σ1 is consistent with previous estimates and with a stress tensor controlled by glacial isostatic adjustment, although a contrast between deep and shallow focal mechanisms suggests that these stresses may be concentrated in the lower crust. Epicenter lineations up to ∼40 km long may be indicative of sizable faults in the LSZ capable of generating Mw>6 earthquakes, but hypocenter and focal mechanism uncertainties are too high to say so definitively, thus pointing to a need for denser station coverage, including ocean‐bottom seismometers.

中文翻译：

---

下圣劳伦斯地震带的地震深度、震源机制和应力

我们研究了加拿大东部古裂谷带下圣劳伦斯地震带 (LSZ) 的地震震源、震源机制和构造应力状态。该地区记录的最大地震是 1999 年的 Côte-Nord MN 5.1，随后发生了大约 80 次 MN>1 的余震。目前尚不清楚该地区是否能够产生危险的 Mw>6 地震，类似于上游约 250 公里的夏洛瓦地震带。以 2015-2020 年为重点，我们应用基于机器学习的相位选择器来检测同一时间跨度内同一地区的 150 次目录地震之外的 72 次地震。我们使用统一的 VP/VS=1.77⁠通过蒙特卡罗搜索生成更新的一维梯度速度模型，我们使用 Wadati 方法计算该模型。我们使用三重差分法改进震源估计，以 sP 深度相位作为对地震深度的附加约束。我们通过自动选择的 P 波第一次运动和绝对值 P-SV-SH 振幅比估计 >100 次地震的震源机制，并使用震源机制来反演构造应力状态。网格搜索和贝叶斯分析允许对震源机制进行稳健的不确定性估计，这反过来又允许对应力张量进行不确定性估计。恢复的西-西北-东-东南 σ1 与先前的估计一致，并与由冰川均衡调整控制的应力张量一致，尽管深部和浅部震源机制之间的对比表明这些应力可能集中在下地壳。长达约 40 公里的震中线可能表明 LSZ 中存在相当大的断层，能够产生 Mw>