How big and how often? Reconstructing the megathrust earthquake history of the Semidi section, Alaska-Aleutian Subduction Zone, USA


Instrumental records of seismicity are rarely long enough to fully capture the recurrence intervals and magnitudes of earthquakes within subduction zones (e.g. Philibosian and Meltzner, 2020). The application of geologic techniques to develop palaeoseismic records provides an alternative means to understand how often (recurrence) and how large (Mmax) earthquakes might be in a region with the ability to develop records that cover millennial timescales (e.g., Shennan et al., 2014). One region where this information is sparse is the Semidi section of the Alaska-Aleutian megathrust, the main plate-boundary fault along the subduction zone between Kodiak Island in the east and the Shumagin Islands in the west (Figure 1). This area is of great interest as it is an area that can generate large tsunami directed at the west coast of the USA (Ross et al., 2013) and hosted multiple M8 earthquakes during the 20th-21st centuries.

While there are limited palaeoseismic records in the eastern part of the Semidi section at Chirikof Island (Nelson et al., 2015) and Siktinak Island (Briggs et al., 2014) and to the west in the Shumagin Islands (Witter et al., 2014), there is an absence of records through a 230 km section which ruptured in 1938 (M8.3; e.g., Freymueller et al., 2021), and most recently as an M8.2 in 2021 (Figure 1). These earthquakes raise questions over whether centennial recurrence of M8s is the characteristic behaviour of this section of the subduction zone or if larger earthquakes have occurred in the past. Russian records of a large earthquake and tsunami in 1788 suggest a potential M8.5+ scenario but there has been no evidence of this earthquake found for the proposed western rupture and it is unclear how far the eastern rupture propagated beyond Sitkinak Island. Satellite observations of the recent M8.2 earthquake suggest significant land-level changes were observed on the Alaskan Peninsula in proximity to Chignik and Perryville; within the detection limit of geological techniques (e.g., Shennan et al., 2016) but also suggests limited to no land-level changes at sites with existing palaeoseismic records in the region. Therefore, sites around the Chignik and Perryville areas have potential to contain millennial-scale records of earthquake-related land-level changes and/or tsunami deposits that can be used to answer outstanding questions over megathrust behaviour that cannot be addressed by existing studies.


The project has two complementary components that focus on 1) the understanding of the modern environment (sediments, microfossils, geochemistry) and 2) applying that understanding to fossil cores/outcrops to develop records of land-level changes and/or tsunami inundation coupled to a high-resolution chronology. There will be a detailed investigation of the fresh- and salt-marshes, and tidal-flat environments of the coastline to document the modern distribution of microfossils (e.g., diatoms, foraminifera) as well as sedimentary (e.g., grain size) and geochemical (e.g., carbon isotopes) indicators. These analyses will be used as modern analogues for the second component of the study. The student will develop a Bayesian Transfer Function (BTF). The BTF will establish the functional relationship between the information in the microfossil assemblages and/or geochemistry and tidal elevation.

Secondly, the student will undertake detailed stratigraphic investigations at coastal environments within the Semidi section through sediment cores and outcrops to identify coastal sediments that can be used to reconstruct land-level changes and or tsunami inundation over the past few thousand years. The chosen stratigraphic sections will undergo the same analyses as described above for the contemporary samples. The developed BTF will be applied to the core to reconstruct the relationship between each sample and former tidal levels. The microfossil assemblages will supply the primary information for the reconstruction. However, the model will also incorporate secondary information in the form of the litho- and chemo-stratigraphic data to provide the BTF with an informed “prior” distribution, based on information collected as part of the first component of the project.

The final stage will involve the development of high-resolution (decadal- to centennial-resolution) age models for the chosen cores. The primary method will be radiocarbon, but this will be supplemented using radiometric (e.g., 137Cs and 210Pb), tephra and other chronologic methods. The two fossil components (BTF and chronologic model) will be combined to produce a palaeoseismic dataset for the Semidi section. This will be compared with existing records to the west (Shumagin Islands) and to the east (Chirikof Island, Sitkinak Island, and Kodiak Island).

Project Timeline

Year 1

Assessment of aerial imagery to determine the spatial distribution of suitable coastal environments within the Semidi section. Selection of initial study sites for collection of modern samples and initial sediment cores during the first fieldwork season. Characterization of the sedimentological, microfossil, and geochemical characteristics of the sampled modern sediments. This will partly be conducted during a secondment to Virginia Tech.

Year 2

Fieldwork to collect core samples from a second site and to fill identified gaps within the modern record from prior field season. Characterization of the fossil core sediments including sedimentological, microfossil, and geochemical characteristics, as well as chronologic markers. This will partly be conducted during a secondment to Glasgow.

Year 3

Finalize laboratory work and develop BTF with appropriate priors. Develop final chronology to constrain timing of land-level changes and/or tsunami inundation. Finalise a regional palaeoseismic database for the Semidi section including the newly obtained data.

Year 3.5

Completion of thesis write-up and preparation of manuscripts.

& Skills

This project is a Collaborative Studentship between Durham University and the University of Glasgow. The project will also benefit from external supervisory support from Dr. Robert Witter (United States Geological Survey, and Dr. Tina Dura (Virginia Tech,–tina.html).

During the project, the successful candidate will obtain training to develop necessary key skills, including those required for field investigations of coastal stratigraphy (Durham and Virginia Tech); organic geochemistry and analysis (Durham); microscopy and interpretation of microfossil assemblages (Durham and Virginia Tech); and radiometric dating methods, tephra analysis, and chronology development (University of Glasgow and Durham). As part of the training support, we have budgeted for the successful candidate to attend the University of Oxford “Radiocarbon Dating and Bayesian Chronological Analysis” workshop.

The project will involve two field seasons along the Alaska Peninsula in the USA to obtain data and samples for analyses. The student will also develop skills relating to lake-based palaeoseismology as part of companion projects during fieldwork. The student will also have the opportunity to present their results at national and international conferences to develop presentation and communication skills and to disseminate the results.

References & further reading

Briggs, R.W. et al., 2014. Uplift and subsidence reveal a nonpersistent megathrust rupture boundary (Sitkinak Island, Alaska). Geophysical Research Letters 41, 2289-2296.

Nelson, A.R. et al., 2015. Tsunami recurrence in the eastern Alaska-Aleutian arc: A Holocene stratigraphic record from Chirikof Island, Alaska. Geosphere 11(4), 1172-1203.

Philibosian, B., and Meltzner, A.J., 2020. Segmentation and supercycles: A catolog of earthquake rupture patterns from the Sumatra Sunda Megathrust and other well-studied faults worldwide. Quaternary Science Reviews 241, 106390.

Ross, S.L., et al., 2013. The SAFRR tsunami scenario – Improving resilience for California. U.S. Geological Survey Fact Sheet 2013-3081, 4p.

Shennan, I. et al., 2014. Late Holocene paleoseismology of a site in the region of maximum subsidence during the 1964 Mw 9.2 Alaska earthquake. Journal of Quaternary Science 29, 343-350.

Shennan, I. et al., 2016. Detection limits of tidal-wetland sequences to identify variable rupture modes of megathrust earthquakes. Quaternary Science Reviews 150, 1-30.

Witter, R.C. et al., 2014. Little late Holocene strain accumulation and release on the Aleutian megathrust below the Shumagin Islands, Alaska. Geophysical Research Letters 41, 2359-2367.

Further Information

For further information, please contact:
Dr. Simon E. Engelhart
+44 (0) 191 33 43509

Apply Now