Constraints on Slow Slip Behavior in Cascadia Through the Integration of PBO Borehole Strainmeters, GPS Time Series, and Tremor Locations

慢滑事件和非火山震动为沿断裂深部伸展的应变释放模式提供了新的见解。这些地震的特点是持续时间长(~天)，震级大(~M6)。我们还没有完全了解板块界面上的什么性质或条件可以防止地震中通常观察到的快速滑动。慢滑和震颤还限制了孕震区的下缘，这是一个对地震危险性研究有用的变量，而事件历史是探索孕震区应力随时间变化的载荷研究的输入。现有的慢滑大地测量研究大多集中在GPS观测上。在这项工作中，我们将钻孔应变仪与GPS数据相结合，以更全面地描述卡斯卡迪亚俯冲带上的慢滑行为谱。利用扩展网络逆滤波对GPS和应变仪时间序列进行了随时间变化的反演。通过将应变计结合到我们的分析中，我们可以解析较小的慢滑事件，并提取关于传播特性的更多信息，而不是单独使用GPS。从我们扩展的慢滑事件目录中，我们提取了源参数，并测试了较小的慢滑事件(~M5+)是否符合对较大的慢滑事件(~M6+)所推导的相同的标度关系。应变仪数据还使我们能够更好地比较慢滑的时空相关性和地震的位置。我们正在测试地震是否标志着慢滑的前缘，或者地震是否嵌入了滑动区。最后，我们正在评估缓慢滑动的向上倾斜锥度，并查看地面观测是否与地震所暗示的突然上升边缘一致。我们对应变仪数据的分析使我们能够更好地评估应变仪时间序列的校准和噪声特性。我们将具体使用GPS导出的滑移模型来独立验证应变仪的校准。

Slow slip events and non-volcanic tremor provide new insight into the mode of strain release along the deep extension of faults. These events are noteworthy for their long durations (~days) and large magnitudes (~M6). We have yet to fully understand what properties or conditions on the plate interface act to prevent the rapid slip that is typically observed with earthquakes. Slow slip and tremor also constrain the lower edge of the seismogenic zone, a variable that is useful for seismic hazard studies, and event histories are an input into studies that explore the time-dependent loading of stress on the seismogenic zone. Most of the existing geodetic studies of slow slip have focused on observations made by GPS. In this work, we integrate borehole strainmeters with GPS data to more fully characterize the spectrum of slow slip behavior on the Cascadia subduction zone. Time-dependent inversions of GPS and strainmeter time series are performed using the Extended Network Inversion Filter. By incorporating the strainmeters into our analysis, we resolve smaller slow slip events and extract greater information about the propagation characteristics than by using GPS alone. From our expanded catalog of slow slip events, we are extracting the source parameters and testing whether the smaller slow slip events (~M5+) obey the same scaling relationships as has been inferred for the larger ones (~M6+). The strainmeter data also allow us to better compare the spatial and temporal correlation of slow slip and the location of tremor. We are testing the idea of whether tremor marks the leading edge of slow slip or whether tremor is embedded within the slipping zone. Lastly, we are evaluating the up-dip taper of slow slip and seeing if the surface observations are consistent with an abrupt up-dip edge, as suggested by the tremor. Our analysis of the strainmeter data allows us to better evaluate the calibration and noise characteristic of the strainmeter time series. We will specifically use the GPS-derived slip models to independently validate the strainmeter calibrations.