MRI: Offshore Earthquake Monitoring at Subduction Zones Using Autonomous Underwater Vehicles and High-Speed Optical Telemetry For Data Retrieval

地球的俯冲带，即一个构造板块覆盖另一个板块的地方，是地球上最大、最危险的地震发生地。最近的例子是2004年苏门答腊9.1级地震和2010年日本东北9级地震。除了地面震动带来的巨大破坏外，这两次地震还引发了毁灭性的海啸。这些事件提醒我们，地球上最大地震的断层运动发生在大陆架下的近海，这是一个具有挑战性的监测区域。在这些大地震的震源地区，地震仪器的稀缺限制了地震学界回答基本科学问题和实时评估灾害的能力。对俯冲带进行持续的海上监测不仅可以提高对这些断层的认识，还可以潜在地识别增加的短期风险。在2010年东北地震和2014年智利皮萨瓜9级地震发生之前，大量的前震序列都位于离岸较远的地方。这些序列表明，在几个星期的时间尺度上，俯冲带更有可能发生大的破裂。为了充分了解这种基本的断层行为，并可能利用对它的了解来降低风险，需要从大地震正上方的海上快速获取高质量的地震数据。地震学界通常在海上部署地震仪，但数据获取需要基于船上的仪器恢复，这是一项漫长而昂贵的任务，在潜在的前震序列中可能不可能或不可取。地震学是一门快速发展的科学：地震在几秒钟内就能探测到，震级估计在几分钟内就能在推特上发布，断层面在几小时内就能确定，研究论文通常在大地震发生后几周内就能提交。只有将这些数据快速输入数据中心，才能充分发挥海上地震数据的潜力。伍兹霍尔海洋研究所在光学遥测和海洋机器人技术方面的最新进展，加上商用现成的低功耗地震传感器、数据记录器和原子钟的可用性，为海底地震仪（OBS）阵列的多年部署提供了能力，这些地震仪（OBS）能够将高频、精确定时的地震数据传输到岸上，而无需OBS恢复数据的延迟时间为数小时到数天。该项目将开发这样一个系统，将whoi设计的光学调制解调器与OBS和远程（高达286海里）REMUS自主水下航行器（AUV）集成在一起，该调制解调器具有每秒20兆的遥测速率。这些经过验证的速率可以在几分钟内遥测4个通道（地面运动和压力）上一周的高速率（100 Hz）地震数据（或不到2小时的数据）。此外，准确的定时对地震定位至关重要，但地震仪缺乏GPS定时的优势。光学链路还将允许测量OBS时钟相对于AUV携带的gps同步时间信号的偏移量，精度约为1微秒。这些技术将使OBS阵列的多年部署成为可能，根据需要进行数据卸载和时钟检查，而无需每年恢复/重新部署巡航，每次实验节省数十万美元。这种能力将特别适用于监测俯冲带所需的密集近岸阵列。

The earth's subduction zones, where one tectonic plate overrides another, are the sites of the planet's largest and most dangerous earthquakes. Recent examples are the 2004 magnitude 9.1 Sumatra and the 2010 magnitude 9 Tohoku, Japan earthquakes. In addition to massive destruction brought about by ground shaking, both of these earthquakes generated devastating tsunamis. These events are a reminder that fault motion in our planet's largest earthquakes happens offshore under the continental shelf, a region that is challenging to monitor. The scarcity of seismic instrumentation in the source regions of these great earthquakes has limited the seismological community's ability to answer fundamental scientific questions and to evaluate hazards in real-time. Sustained offshore monitoring of subduction zones would not only lead to improved understanding of these faults but also potentially allow the identification of increased short-term risk. Extensive foreshock sequences located far offshore preceded both the 2010 Tohoku event and the recent 2014 magnitude 9 Pisagua Chile earthquake. These sequences demonstrated that there are detectable time periods on the scale of weeks when subduction zones are more likely to produce a large rupture. To fully understand this basic fault behavior and possibly utilize an understanding of it to reduce risk, rapid access to high quality seismic data from offshore directly above the great earthquakes is needed. The seismological community routinely deploys seismographs offshore, but data access requires ship-based instrument recovery, a lengthy and expensive task that may not be possible or desirable during a potential foreshock sequence. Seismology is a fast moving science: earthquakes are detected within seconds, magnitude estimates are tweeted within minutes, fault planes are determined within hours, and research articles are often submitted within weeks of a major earthquake. The full potential of offshore seismic data can only be realized if these data are rapidly ingested into data centers. Recent advances at Woods Hole Oceanographic Institution in optical telemetry and marine robotics, coupled with the availability of commercial off-the-shelf low-power seismic sensors, data loggers and atomic clocks offer the capability for multi-year deployments of arrays of ocean-bottom seismographs (OBS) that are capable of delivering high-frequency, accurately-timed seismic data to shore with data latencies of hours to days without OBS recovery. This project will develop such a system by integrating a WHOI-designed optical modem capable of telemetry rates of 20 Mbits/second with an OBS and with a long-range (up to 286 nautical miles) REMUS Autonomous Underwater Vehicle (AUV). These proven rates allow telemetry of a week of high-rate (100 Hz) seismic data on 4 channels (ground motion and pressure) in minutes (or a year of data in less than 2 hours). Moreover, accurate timing is critical for earthquake location but OBS lack the benefit of GPS timing. The optical link will also allow measurement of the offset of the OBS clock relative to a GPS-synchronized time signal carried by the AUV to a precision of ~1 microsecond. These technologies will make possible multi-year deployments of OBS arrays with on-demand data offload and clock-check without the need for annual recovery/re-deployment cruises, saving hundreds of thousands of dollars per experiment. This capability will be particularly suitable for the dense near-shore arrays needed to monitor subduction zones.