A section of the Sunda megathrust underneath Simeulue is known to persistently halt rupture propagation of great earthquakes, including those in 2004 (Mw 9.2) and 2005 (Mw 8.6). Yet the same section generated large earthquakes in 2002 (Mw 7.3) and 2008 (Mw 7.4). To date, few studies have investigated the 2002 and 2008 events, and none have satisfactorily located or explained them. Using near‐field InSAR, GPS, and coral geodetic data, we find that the slip distributions of the two events are not identical but do show a close resemblance and largely overlap. We thus consider these earthquakes “siblings” that were generated by an anomalous “parent” feature of the megathrust. We suggest that this parent feature is a locked asperity surrounded by the otherwise partially creeping Simeulue section, perhaps structurally controlled by a broad morphological high on the megathrust.
23 November 2020 in Journal of Geophysical Research: Solid Earth
Nearly every type of slip behavior of which megathrusts are known to be capable of has been observed within a narrow ~75‐km‐wide strip of the Sunda megathrust—the Banyak Islands subsegment. The diverse list of recorded slip events starts with the great 1861 MW ~ 8.5 earthquake and includes the 1907 MW ~ 8.2 shallow tsunami earthquake, the 1966–1981 slow slip event, the great 2005 MW 8.6 Nias‐Simeulue earthquake and associated afterslip, many small (M < 6) to moderate (6 ≤ M < 7) earthquakes, and the large 6 April 2010 MW 7.8 Banyak Islands earthquake. In this paper we map out the spatial and temporal relationships between these slip events to search for a pattern and possible controls on slip behavior. We use GPS and coral geodetic data to derive a coseismic slip distribution for the relatively understudied 2010 event and find that this event fits like a puzzle piece into a low slip patch left by the 2005 event. We compare the recent slip sequence with the historical sequence associated with the 1861 event and find that the modern slip pattern could be a repeat of the historical slip pattern, possibly suggesting the importance of stationary fault properties in controlling slip. The diverse set of slip events mostly fit our expectations within depth‐dependent rupture zones; however, the overlaps between events of different slip types may complicate this conceptual model. Along strike, we suggest that a subducting fracture zone could act to diversify the slip behavior, and we explore possible mechanisms for this slip control.
16 April 2026 in Natural Hazards and Earth System Sciences
Landslides can dam rivers and require rapid response to mitigate catastrophic outburst floods. Here, we present a workflow to map landslide dam formation susceptibility at a regional scale. We define a probabilistic function that combines river valley width and landslide volume to efficiently determine the likelihood of a landslide dam or “damability”. We combine damability values with landslide susceptibility to estimate landslide dam susceptibility. The valley width measurements are automated using a new elevation threshold-based algorithm. Landslide volume is represented as a statistical distribution from mapped landslides. We validate and apply our approach to the Oregon Coast Range, USA and find that 36 % of river stretches exceed a dam potential threshold; these are in river headwaters and steeper terrain, which in this case correlate with more resistant lithologies. We also estimate volumes of the potential dammed lakes and find that most rivers with high dam susceptibility are less likely to impound large lakes because they have low drainage areas. However, widespread susceptibility, and the potential impacts from exceptionally large landslides, suggest that this hazard should be considered in the Pacific Northwest. The damability function workflow can ingest new data and be applied more broadly to assess future landslide dam hazards.
Landslides can dam rivers and require rapid response to mitigate catastrophic outburst floods. Here we present a workflow to map landslide dam formation susceptibility at a regional scale. We define a probabilistic function that combines river valley width and landslide volume to efficiently determine the likelihood of a landslide dam or ‘damability’. We combine damability values with landslide susceptibility to find landslide dam susceptibility. The valley width measurements are automated using a new elevation threshold-based algorithm. Landslide volume is represented as a statistical distribution from mapped landslides. We verify and apply our approach to the Oregon Coast Range, USA and find high susceptibility in river headwaters and generally steeper terrain, which in this case correlates with more resistant lithologies. We also estimate volumes of the potential dammed lakes and find that most rivers with high dam susceptibility are less likely to impound large lakes, because they have low drainage areas. However, widespread susceptibility, and the critical potential impacts from exceptionally large landslides, suggest this hazard should be considered in the Pacific Northwest. The damability function workflow can readily ingest new data and can be applied more broadly to assess future landslide dam hazards.
Landslides can dam rivers and require rapid response to mitigate catastrophic outburst floods. Here we present a workflow to map landslide dam formation susceptibility at a regional scale. We define a probabilistic function that combines river valley width and landslide volume to efficiently determine the likelihood of a landslide dam or ‘damability’. We combine damability values with landslide susceptibility to find landslide dam susceptibility. The valley width measurements are automated using a new elevation threshold-based algorithm. Landslide volume is represented as a statistical distribution from mapped landslides. We verify and apply our approach to the Oregon Coast Range, USA and find high susceptibility in river headwaters and generally steeper terrain, which in this case correlates with more resistant lithologies. We also estimate volumes of the potential dammed lakes and find that most rivers with high dam susceptibility are less likely to impound large lakes, because they have low drainage areas. However, widespread susceptibility, and the critical potential impacts from exceptionally large landslides, suggest this hazard should be considered in the Pacific Northwest. The damability function workflow can readily ingest new data and can be applied more broadly to assess future landslide dam hazards.