基于高频雷达资料同化的海啸预警：以2022年汤加火山海啸为例

Data assimilation using high-frequency radar for tsunami early warning: A case study of the 2022 Tonga volcanic tsunami

王宇晨 博士

日本海洋研究开发机构（JAMSTEC）

2022.10.07（星期五）15:00，腾讯会议号：524-7569-5009

报告摘要：

高频雷达可以观测海面洋流速度，为海啸预警提供数据。SeaSondeR是日本津轻海峡东部的一套高频雷达系统，在2022年汤加火山海啸事件中观测到了海啸引起的洋流信号。由于汤加海啸存在大气耦合，形成机理较为复杂，传统的海啸预警方法并不适用。

本研究利用资料同化手段实现海啸预警，该手段利用离岸海啸观测复原海啸波动场，无需提供火山源信息。本研究利用2022年1月15日09:00(UTC)开始的数据进行同化，在14:00和15:00分别对沿岸海啸进行预测，并与函馆和下北两个潮位站进行比对。研究发现，在14:00发出的预警，对未来两小时和六小时的预测精度分别为91%和67%；在15:00发出的预警，对未来两小时和六小时的预测精度分别为63%和70%。在此次海啸事件中，高频雷达在资料同化中的作用优于传统的海底压力计（S-net）。因此，高频雷达可以在海啸预警中发挥重要作用。

High-frequency (HF) radar monitors the sea surface current velocity and provides information for tsunami early warning. SeaSondeR, an HF ocean radar system in the eastern Tsugaru Strait, Japan, measured the tsunami-induced current velocity during the 2022 Tonga volcanic tsunami. This event was triggered by the eruption of the Hunga Tonga-Hunga Ha’apai volcano on January 15, 2022. As an air-coupled tsunami, the generating mechanism was complex, making it difficult to predict coastal tsunamis using traditional early warning methods.

We adopted the tsunami data assimilation approach, which reconstructs the tsunami wavefield using offshore data and does not require source information, to forecast the coastal tsunami waveforms. Observations from the HF radar and offshore bottom pressure gauges (OBPGs) were used as the input for tsunami data assimilation. The assimilation process started at 09:00 (UTC, hereafter) and forecasts were made at 14:00 and 15:00. The surface current velocity recorded by the HF radar reached the maximum (~0.25 m/s) at 13:00, which corresponded to a negative phase of ~2 cm sea level variation observed by OBPGs.

The forecasted waveforms were compared with the observed waveforms at Hakodate and Shimokita tide gauges. The assimilation results using OBPG accurately forecasted the tsunami waveforms at Shimokita, especially for the next 2 h after the forecast. However, the forecast underestimated the waveforms at Hakodate.    The assimilation results using HF radar matched well with the observations at both Shimokita and Hakodate. Furthermore, for quantitative analysis, we adopted an accuracy index that considers the maximum amplitude in the next 2 h and 6 h after the forecast. At 14:00, the accuracy indices were 91% and 67% for the next 2 h and 6 h, respectively. At 15:00, it was 63% and 70% for the next 2 h and 6 h, respectively. The accuracy indices of the forecast using HF radar were higher than those using OBPG. Thus, we demonstrated the applicability of HF ocean radar system in tsunami data assimilation. This study is the first to apply tsunami data assimilation to early warning using real HF radar observations. The HF ocean radar system could be a good supplement to OBPG for monitoring tsunamis and providing information for data assimilation.

报告人简介：

王宇晨博士2016年获得北京大学物理学学士学位（本科导师：胡永云教授、林金泰教授），2018年获得东京大学地球与行星科学硕士学位，2021年获得东京大学地球与行星科学博士学位。2019年4月至2021年9月，任日本学术振兴会(JSPS)特别研究员。2021年10月至今，任日本海洋研究开发机构（JAMSTEC）Young Research Fellow，主要从事海啸和地震的预警与减灾相关工作，具体研究内容包括 “基于资料同化（data assimilation）方法的海啸早期预警”“不同震源类型对港湾海啸共振 (tsunami resonance）的影响”“海底水压计的布站优化”等。曾获“2020年国家优秀自费留学生奖学金”，“2021年东京大学理学系研究科研究奖励赏”等荣誉。近五年内以第一或通讯作者身份在Geophysical Research Letters、 JGR: Solid Earth、 Seismological Research Letters等国际地球科学期刊上发表论文16篇，并多次担任Nature Communications、 JGR: Solid Earth、JGR: Oceans等期刊审稿人。