| 摘要: | 位於東蘇答班達島弧(印度尼西亞)的弗洛雷斯逆衝斷層是世界上罕見的弧後逆衝斷層之一。此向南傾斜斷層的弧後區域的位置自蘇達/爪哇海溝俯衝帶延伸到帝汶海槽陸弧碰撞過渡帶。關於該斷裂帶在板塊構造發育及現今運動學上，仍有一些值得爭論的觀點。儘管弗洛雷斯衝斷層經常發生破壞性淺層地震和一些引發型海嘯，至今仍舊缺少有關弗洛雷斯逆衝斷層帶地震震源調查的研究。弗洛雷斯逆衝斷層的其中三個部分一直處於地震活躍的階段，目前的地震目錄中記錄了一些明顯達到M5.7+的地震，從西向東，分別為龍目島段，松巴哇段和西弗洛雷斯段。在斷裂帶西部的這些區段中，地震活動往往伴隨著發生中等規模的地震。
本論文使用來自全球和區域地震觀測網的地震資料，主要針對 1999 年至 2022 年發生的地震，調查了弗洛雷斯逆衝斷層沿線地震的震源特徵。我們採用MW 6.2+ 的地震作為控制，利用遠震體波和表面波進行了有限斷層破裂反演，並透過刀切抽樣法去計算模型的不確定性。此外，我們使用雙差震源重定位和統計地震學研究了龍目島和西弗洛雷斯的兩個地震序列。斷層中有三個段落是我們特別關註的：在 2018 年 7 月至 8 月龍目島地震序列中破裂的龍目島部分（1 MW 6.5 和 2 MW 6.9 事件）;在 2002 年至 2009 年松巴哇地震中破裂的松巴哇部分（5 MW 6.2 -6.6 事件）; 以及 2003 年發生 MW 6.4 正常斷層事件和 2022 年地震序列的西弗洛雷斯段。
我們針對以描述弗洛雷斯逆衝斷層沿線的特徵地震源和地震活動進行了完整的地震學分析。我們認為地震破裂位在該區域的主要分支逆衝斷層，而不是在俯衝弗洛雷斯海洋地殼的基底滑脫斷層中。我們強調了弗洛雷斯逆沖地震的逆衝破裂過程的相似性，即它們的起始破裂、破裂速度、破裂大小和靜應力下降，它們是發生在板塊邊界的逆衝型地震。
在龍目島部分，我們討論了 2018 年異常的高能地震活動度與其他重大事件的破裂行為的關係。複雜斷層破裂與地震活動的時空演變之間的機制差異表示地震-火山相互作用，並提供了對斷層破裂在這種斷層-火山環境中的行為方式的見解。此外，地震活動範圍的最西端和最東端也顯示出強大的分段屏障。在松巴哇段，我們所構建的五個中等地震的有限斷層模型表明一個非重疊的地栓區域。破裂通常沿走向或下傾方向傳播，靜應力降低。我們認為在西弗洛雷斯逆衝斷層上具有低靜態應力降的中等尺度地栓在特徵上相較於已成熟發展的東部相比還不太成熟。或者，這些位在松把哇段下傾的地震可能如地栓一般活動，其較淺的部分仍有可能發生 MW> 7.0 的破裂，類似於 1836 年的歷史地震。在西弗洛雷斯段， 2022 年地震序列也破裂了主要的張開逆衝斷層並突出了複雜的變形，但我們可以確認 2003 年的正斷層事件不座落在弗洛雷斯逆衝斷層上，相反的，它是在弗洛雷斯海早期俯衝板塊（板塊內地震事件）中，位在弗洛雷斯逆衝斷層帶下方深度更深的板塊破裂。
基於震源模型和地震活動分析結果，我們提出了一種新的地震構造模型和弗洛雷斯逆衝斷層帶的分段。研究結果顯示出弗洛雷斯衝斷層是一個複雜的不連續帶，並且本研究中所分析的三個斷層分段分別表現了不同的結果。我們為沿弗洛雷斯逆衝斷層的重大地震編制了地震尺度，這是建立震源標度所必需的。本研究提供了對弗洛雷斯逆衝斷層沿線斷層力學和地震活動度的統一處理，並闡明了龍目島、松巴哇和西弗洛雷斯段的地震板塊構造、地震成核、地震活動度，這對於進一步的地震危害分析是必要的。
;The Flores Thrust fault zone in the eastern Sunda–Banda arc (Indonesia) is one of the rare back-arc thrusting faults in the world. This southward-dipping fault is located in the back-arc region in the transition zone from the subduction zone at the Sunda/Java trench to the continent-to-arc collision at the Timor trough. There are some arguable views on the tectonic development and present kinematics of this fault zone. Despite the Flores Thrust often hosting destructive shallow earthquakes and some triggered tsunami, there is a lack of study that integrates seismic source investigations for earthquakes along the Flores Thrust fault zone. Three parts of the Flores Thrust fault have been seismically active and produced some significant M5.7+ earthquakes in the present earthquake catalog, namely, from west to east, the Lombok segment, the Sumbawa segment, and the West Flores segment. The seismicity tends to be clustered with a cascade of moderate magnitude earthquakes in these segments in the western part of the fault zone.
Using seismic data came from the global and regional seismic network, in this thesis, we investigate the source characteristics of earthquakes along the Flores Thrust fault zone, mainly for earthquakes that occurred from 1999 to 2022. We performed finite-fault rupture inversions for MW 6.2+ earthquakes constrained by the teleseismic body and surface waves and introduced jackknife resampling tests to measure the model uncertainties. In addition, we investigated two seismic sequences in Lombok and West Flores using double-difference hypocenter relocation and statistical seismology. We mainly focus on three segments, the Lombok segment that ruptured in July–August 2018 Lombok seismic sequence (with one MW 6.5 and two MW 6.9 events), the Sumbawa segment that ruptured in the 2002–2009 Sumbawa earthquakes (with five MW 6.2–6.6 events), and the West Flores segment that hosted an MW 6.4 normal-faulting event in 2003 and a seismic sequence in 2022.
We performed a comprehensive seismological analysis to characterize earthquake sources and seismicity along the Flores Thrust fault. As a result, we suggested that the earthquakes ruptured the major splay thrust fault in the zone and not in the basal detachment fault of the incipient subducting Flores sea oceanic crust. Furthermore, we highlighted the similarities of the thrusting rupture processes of Flores Thrust earthquakes, i.e., in their rupture initiation, speed, size, and static stress drops, and they are the type of thrust earthquakes that occur in a plate boundary.
In the Lombok segment, we discussed the relations of unusual, energetic seismicity in 2018 and the rupture behaviors of the significant cascading events. The mechanical discrepancy between complex fault ruptures and the spatiotemporal evolution of seismicity suggested an earthquake-volcano interaction and provides insights into how fault rupture behaves in such a fault-volcano environment. In addition, the westernmost and easternmost termination of seismicity extent indicated a strong segment barrier. In the Sumbawa segment, constructed finite-fault models of five moderate earthquakes suggest a non-overlap asperities area. The ruptures often propagated along-strike or down-dip directions with low static stress drops. We suggested that the cascade moderate-size asperities with low static stress drop may be the indication of the less mature feature of western Flores Thrust compared to its more developed eastern part. Alternatively, these earthquakes may act as asperities located at the down-dip patches of the Sumbawa segment, and its shallower section still has a potential of ruptures with MW > 7.0, similar to a historical earthquake in 1836. In the West Flores segment, the 2022 seismic sequence also ruptured the major splay thrust fault and also highlighted a complex deformation, but we confirmed that the 2003 normal faulting event was not located on the Flores Thrust; instead, it ruptured at the deeper depth, in the Flores Sea incipient subducted slab (intraslab event), below the Flores Thrust fault zone.
We proposed a new seismotectonic model and segmentation of the Flores Thrust fault zone based on our source models and seismicity analysis results. Our result indicated that the Flores Thrust is a complex discontinuous zone and that the three segments analyzed in this study behaved differently. In addition, we compiled earthquake dimensions for significant earthquakes along the Flores Thrust fault necessary to build an earthquake source scaling. This study provides a unified treatment of faulting mechanics and seismicity along the Flores Thrust fault and sheds light on the seismotectonic, earthquake nucleation, and seismicity activation in the Lombok, Sumbawa, and West Flores segments. It would be necessary for further seismic hazard analysis. |
