高密度即時強震觀測網於地震減災之運用

Abstract

地震預測目前還未被完全實現，相對突顯地震預警、速報等防、減災系統的重要性。即便近十年相關即時強震觀測系統蓬勃發展，其中仍有不少挑戰。欲突破之，有一個共通前提，那便是更密集且更廣的即時觀測。在臺灣，採用低價位加速儀建立的P-Alert觀測網，從2010年開始架設至今，已在多次地震事件中驗證其可靠性及穩定度；利用P-Alert觀測網高密度且即時的優勢，能產生較為精細的即時強地動峰值圖，更能近即時判別斷層破裂的方向。因此，本文將透過P-Alert觀測網即時資料，探討高密度即時強震網在地震防、減災上的運用。首先，本文透過快速解算各方向上即時最大地表加速度值 (Peak Ground Acceleration, PGA) 衰減關係，推演出地震事件最終PGA值可能的分布情形，此地震預警演算法稱之為「ShakingAlarm」。相較於傳統先求取地震位置的預警方法，此演算法不求取地震位置，計算流程簡單、處理時間較短，能提供更充足的時間來進行災害評估及緊急應變。當災害性地震發生後，精細的強地動峰值圖是災損評估的重要依據，其中包含PGA、地表最大速度 (Peak Ground Velocity, PGV) 及不同週期之譜加速度 (Spectral Acceleration, Sa) 三種常見的強地動峰值圖。為此，本研究建立了PGA、PGV及0.3秒、1秒週期Sa強地動峰值圖的即時繪製系統，期望為未來震後災損評估提供更佳的參考依據。此外，2022年9月17日和18日一系列襲擊花東縱谷南部的地震，在臺東和花蓮造成嚴重災損。P-Alert觀測網在此地震事件中，成功繪製了多種數值的強地動峰值圖。強地動峰值圖所指示的高震度區域與此次地震期間觀察到的損害情況相符。而在中央氣象署區域預警盲區內，P-Alert測站透過現地型預警功能提供了3到10秒的預警提前時間。且P-Alert區域預警系統分別在主震和最大前震發生後約9秒和7秒提供了第一份預警報告，估計規模分別為5.74和5.67。綜合前述，本文立基於高密度且即時的P-Alert觀測網，研發出兩個新系統，分別應用於地震預警及速報，同時驗證了P-Alert觀測網既有預警功能；揭示了高密度即時觀測網在地震防、減災領域中所能提供的寶貴貢獻和發展潛力。

Earthquake prediction has not yet been fully realized, underscoring the critical importance of earthquake early warning (EEW), rapid reporting, and other disaster prevention and mitigation systems. Despite the booming development of real-time strong motion observation systems over the past decade, numerous challenges persist. To overcome these challenges, a common prerequisite emerges: the necessity for denser and wider real-time observations. In Taiwan, the P-Alert network, composed of low-cost accelerometers, was established. Since its establishment in 2010, its reliability and stability have been verified in numerous earthquakes. By leveraging the high-density and real-time capabilities of the P-Alert network, it can generate more precise real-time shake maps and even identify the direction of fault rupture in near real-time. Therefore, this study utilizes the real-time data from the P-Alert network to explore potential applications of high-density, real-time networks in earthquake disaster prevention and mitigation. First, this study rapidly derives the time-dependent attenuation relationship of real-time peak ground acceleration (PGA) in all directions, deduces the potential distribution of the final PGA value for an earthquake, and names this earthquake early warning algorithm as ''ShakingAlarm''. Compared with traditional early warning methods that require knowledge of the location of hypocenter, this algorithm eliminates the need for such information. The calculation process is simple, and the processing time is short, providing more time for disaster assessment and emergency response. The precise shake map following a disastrous earthquake provides crucial information for assessing disaster damage, including three common values: PGA, peak ground velocity (PGV), and spectral acceleration (Sa) at different periods. Therefore, this study establishes a real-time system for generating shake maps of PGA, PGV, and Sa at 0.3-second and 1-second periods, aiming to provide a valuable reference for future disaster damage assessments. Moreover, a series of earthquakes struck Taiwan''s southern Longitudinal Valley on September 17 and 18, 2022, causing severe damage to several buildings in Taitung and Hualien. During this earthquake, the P-Alert network successfully generated shake maps with various values. The high-shaking areas on these maps align with the observed damages during the earthquake. In the regional early warning blind zone of the Central Weather Administration, the P-Alert stations can provide an early warning leading time of 3 to 10 seconds by on-site EEW method. The P-Alert regional EEW system provided the first report about 9 s & 7 s after the mainshock and the largest foreshock occurrence, respectively, with estimated magnitudes of 5.74 & 5.67. Based on the above, this study develops two new systems for early warning and rapid reporting based on the high-density and real-time P-Alert network. It also verifies the existing early warning function of the P-Alert network. These findings reveal the valuable contribution and development potential that high-density real-time networks can provide for earthquake prevention and disaster mitigation.