Collaborative Research: Improving models of interseismic locking and slow slip events in Cascadia and New Zealand

该项目将对现有的地球物理数据进行新的分析，其补充目标是促进科学认识和改进未来的地震灾害预测。俯冲带是两个构造板块碰撞的地方，一个板块下沉到另一个板块下面。 俯冲带产生了世界上最大的地震，例如2011年在日本北方造成广泛破坏的9.0级东北地震。 该项目将重点关注两个俯冲带地区：美国的卡斯卡迪亚和新西兰的希库朗吉。将对陆上和海上GPS数据进行建模，以确定作为缓慢滑动事件释放的滑动量，作为稳定蠕变，或锁定区域和不滑动区域。慢滑动事件是指类似于地震期间发生的断层上的滑动，但发生得慢得多，并且不会发出破坏性的地震波。锁定区域指的是断层中摩擦力阻止滑动的区域，允许能量积累以备将来地震时释放。与以前的建模工作相比，该项目将使用受以前地震研究约束的地球的真实三维弹性特性。该项目的产品将直接为卡斯卡迪亚和希库兰吉俯冲带的地震灾害预报提供信息，这两个俯冲带附近有主要城市（卡斯卡迪亚的波特兰和西雅图;希库兰吉的惠灵顿）。计划于2018年在希库朗吉北方边缘进行的国际海洋发现计划项目将获得重要的额外信息，该项目的目标是缓慢滑动区域。该项目将支持和培训一名早期职业女科学家，由一名更有地位的女科学家指导。 该项目还通过在新西兰和美国的公开讲座进行公众宣传，并将结果纳入美国两所大学的讲座材料中。该项目将为卡斯卡迪亚和希库朗吉（新西兰）俯冲带生成新的和改进的慢滑事件和板块锁定模型目录。将使用网络反演滤波器技术处理所有现有的全球定位系统数据，以确定缓慢滑动事件，并制作滑动和滑动率随时间演变的模型。 NIF技术基于卡尔曼滤波方法，该方法平衡了物理上合理但不准确的稳态滑移模型与噪声数据。 此外，这些NIF反演将纳入2014年中期至2015年中期Hikurangi俯冲带的Hikurangi海底震颤和慢滑调查部署的海上海底压力（OBP）数据。 该等海上BP仪器于二零一四年年底记录到一次大型海上缓慢滑动事件。这些模型包括，第一次，现实的，三维地球结构和地形和水深的影响。 将对GPS站的长期速度进行建模，以改善对每个俯冲带的板块界面锁定区域的约束。板块锁定表明，有多少预期的板块之间的运动没有在边界连续发生，因此可能在地震或缓慢滑动事件中释放。将以两种方式对全球定位系统站速度进行平均：第一种方式仅在慢滑事件之间的时间内进行平均，第二种方式通过慢滑事件进行平均，这将为每个俯冲带提供两个锁定模型。 产品将包括两个地区的新的现实三维锁定模型，两个俯冲带的慢滑事件目录和一个完整的滑动预算，每个俯冲带的不确定性作为研究时间段内位置的函数。

This project will carry out new analyses of existing geophysical data with the complementary goals of advancing scientific understanding and improving upon future earthquake hazard forecasts. Subduction zones are where two tectonic plates collide and one sinks below the other. Subduction zones produce the largest earthquakes in the world, such as the magnitude 9.0 Tohoku earthquake of 2011 that caused widespread destruction in northern Japan. This project will focus on two subduction zone regions: Cascadia in the U.S. and Hikurangi in New Zealand. Onshore and offshore GPS data will be modeled to determine the amount of slip released as slow slip events, as steady creep, or where regions are locked and not slipping. Slow slip events refer to slip on a fault that is similar to what occurs during an earthquake, but occurs much more slowly and does not emit damaging seismic waves. Locked regions refer to areas of a fault in which friction prevents slip, allowing energy to build up for future release in an earthquake. In contrast to previous modeling work, this project will use realistic, three-dimensional elastic properties of the earth constrained by prior seismic studies. The products of this project will directly inform earthquake hazard forecasts for both the Cascadia and Hikurangi subduction zones, which have major cities nearby (Portland and Seattle in Cascadia; Wellington in Hikurangi). Important additional information will be obtained for an International Ocean Discovery Program project scheduled for the northern Hikurangi margin in 2018 that is targeting the slow slip region. The project will support and train an early career female scientist who will be mentored by a more established female scientist. The project also features public outreach through public talks in New Zealand and the U.S. and the incorporation of the results into lecture materials at two universities in the U.S.This project will generate new and improved catalogs of slow slip events and plate locking models for the Cascadia and Hikurangi (New Zealand) subduction zones. A Network Inversion Filter (NIF) technique will be used to process all available GPS data to identify slow slip events, and produce models of slip and slip-rate evolution over time. The NIF technique is based on a Kalman filter approach, which balances a physically reasonable but inaccurate model of steady slip with noisy data. Additionally, these NIF inversions will incorporate offshore ocean bottom pressure (OBP) data from the Hikurangi Ocean Bottom Investigation of Tremor and Slow Slip deployment during the mid-2014 to mid-2015 time period for the Hikurangi subduction zone. These OBP instruments recorded a large offshore slow slip event in late 2014. These models include, for the first time, realistic, three-dimensional earth structure and the effects topography and bathymetry. Long-term velocities of GPS stations will be modeled to obtain improved constraints on the plate interface locking region for each subduction zone. Plate locking indicates how much of the expected motion between tectonic plates is not occurring continuously at the boundary, and could therefore be released in earthquakes or slow slip events. GPS station velocities will be averaged in two ways: first only over times between slow slip events and second by averaging through slow slip events which will provide two locking models for each subduction zone. Products will include new realistic 3D locking models for the two regions, slow-slip event catalogs for both subduction zones and a full-slip budget with uncertainties for each subduction zone as a function of location over the time period of study.