Research

Who am I, and what do I do? -

I am a tectonicist - I study the messy but beautiful spaces that develop between Earth's rigid tectonic plates. These areas can be as large and old as collisional mountain belts, or as small and young as a single fault plane. I am drawn to this area of geology because I am keenly interested in understanding how Earth's great dynamic systems interact.

Like many geologists, I am a storyteller first and foremost. My research projects usually come into being when I encounter a new and unexplained dataset that has the potential to tell a great story. As a field geologist by training, I understand that the Earth has been kind enough to leave us many clues, but we have to be willing to follow them even if the terrain is rough. In the modern information age, most of the best clues that can really change our perspective on tectonics come in the form of geophysical observations. Thus, I tend to work with a lot of different types of observational data. I find that each time I learn a new way to look at the Earth, I discover new stories to tell. In this sense I am very similar to my students - I am on a journey of discovery where I don't often have mastery over the subject I am studying!

On this page you will find plain language summaries - geological stories - of my work.

Current Funded Projects

As Principal Investigator: Remote sensing mapping of the Palu Valley landslides (Funded by MOE/AcRF Tier 1, 2021-2022).

As Principal Investigator: Living Together: How do fault-volcano interactions control hazards in the Lesser Sunda Islands? (Funded by EOS, 2021-2023)

Research Fellow: Dr. Harisma Andikagumi

Research Associate: Syed Idros Bin Abdul Rahman

Published Research Projects

As first author, corresponding author, or as a significant contributing coauthor.

Below are plain language summaries of studies that I either led or helped to shape in a significant way.

Stress shadows on megathrust faults

Geologists use highly accurate GPS measurements made over time to see how fast the Earth's surface is moving. This is particularly useful near an active fault, where the GPS velocities can reveal whether the fault is frictionally 'locked and loaded' and can produce an earthquake, or is slipping quietly but steadily. Our team, led by Dr. Eric Lindsey (@planet_mech), developed a new approach to modeling GPS data which takes into account a process called 'stress shadowing'. This method changes the way we interpret GPS velocities in the shallowest areas of the most dangerous class of faults - subduction zones.

How volcanic heat affects earthquake ruptures

Every earthquake starts as a microscopic break that grows over time, increasing in rupture area and magnitude as long as the host fault is capable of breaking. The size of an earthquake is determined by factors that stop the rupture from propagating. A team led by Dr. Karen Lythgoe examined an earthquake sequence that occurred in 2018 in Lombok, Indonesia, including two magnitude 6.9 earthquakes. We observed the aftershocks in 3D using a dense seismic array. We showed how the host fault has been heated by the nearby Gunung Rinjani volcano, causing a dramatic change in the part of the fault that can host earthquake ruptures.

Rice irrigation and earthquake-triggered landsliding

Some of the worst natural disasters occur when an initial hazard event triggers multiple disasters of different types. This is called a cascade. A major cascade occurred during the 2018 magnitude 7.5 Palu earthquake, which triggered abrupt tsunamis, liquefaction, and unexpectedly widespread landsliding of the gently sloping floor of the Palu Valley. We used satellites to map the horizontal motion of the Palu Valley landslides, and compared our map with the agricultural land use in the valley. The landslides occurred where there was intensive rice irrigation. This created a shallow water table that allowed liquefaction to occur during the earthquake. Our analysis suggests that mixed agriculture areas (irrigated fields mixed with tree crops) are more resistant to landsliding than densely irrigated areas.

Himalayan sediments and the 2010 Mentawai Islands tsunami earthquake

Every decade or so, a large tsunami will strike Sumatra or Java, Indonesia, without significant ground shaking being felt beforehand. Submarine earthquakes that produce unexpectedly large tsunamis are particularly dangerous because the tsunami typically arrives without warning. These types of tsunamis are not well understood, but seem to require unusually high amounts of fault slip near the seabed. We used seismic reflection data to map the 3D geology of the area that produced the 2010 Mentawai Islands tsunami. The region that seems to have produced the big waves is where the subducting geology is special: a thick layer of mud has been thrust beneath the active fault plane. This mud was eroded from the Himalayas, more than 3000 km away!

Stress shadows

(Part 1)

The predecessor paper to Lindsey et al., 2021. In this study, we explored whether GPS data can tell whether shallowest part of subduction faults are frictionally locked or not. This is important because this area of a fault is often closest to large populations (for example, India) or can produce great tsunamis (for example, Japan). We showed that this approach, while common, is severely limited by stress shadowing.

Backstop faulting in the Sumatran forearc

Every earthquake tells a story. Newly developed methods in seismology allow us to go back and re-examine old earthquake data, sometimes allowing us to write a new story. Our team, led by Dr. Xin Wang, looked at two earthquakes that happened near the Mentawai islands of Sumatra. By precisely relocating the earthquakes and aftershocks, we showed that the fault that slipped in both events was different than previous interpretations. We proposed two different explanations; perhaps future earthquakes will tell us which is more correct!

Re-imagining the Great Sumatran Fault

There are basically two types of faults on Earth: plate boundary faults, which lie along the edges of the rigid tectonic plates, and everything else. Plate boundary faults are special because their slip rate over time must match the rate determined by plate motions. The Sumatran Fault was once thought to be a non-plate-boundary fault because estimates of its slip rate didn't match plate rates. In this study, we compared geological estimates of slip rates (from volcanic tuffs that were cut by the fault) with GPS measurements of plate motion. We found that the Sumatran fault is in fact a plate boundary fault - BUT this requires the very strong oceanic crust near Sumatra to be breaking up. This contradicts the rule of thumb that continental crust is weak and oceanic crust is strong!

Estimating volcanic hazard in SE Asia

Southeast Asia is surrounded by a great ring of active volcanoes, which can produce dangrous eruptions and disrupt air travel. While some have been studied fairly well, we have little to no information about the vast majority. Dr. Patrick Whelley led this study, which uses the concept of a proxy volcano to improve our understanding of the probabilities of different types of eruptions in this region. I re-mapped the volcanoes of Indonesia using modern remote sensing datasets and topography; during this process, we discovered dozens of likely volcanoes that are not in any database. We classified the volcanoes based on their shape and size, and used proxy records to estimate eruption likelihoods.


Whelley et al., 2015: The frequency of explosive volcanic eruptions in Southeast Asia, Bulletin of Volcanology

Measuring Earth's ancient magnetic field

Earth's magnetic field shields the surface from damaging solar radiation. Early in Earth's history, before the freezing of the inner core, a sustained magnetic field was not possible. However, by the time early life had evolved, the inner core had formed and a stable magnetic field could exist. In this study, we measured the magnetic remanence of some of Earth's oldest rocks that have never been significantly metamorphosed - heated up - since their formation. Our study area allowed us to sample both above and below a very special surface - one of the oldest erosion surfaces known! This surface records the first time (that we know of) when rocks were exposed to air over a wide region, and is associated with early microbial mats called stromatolites. Our study showed that the magnetization of the rocks on different sides of the erosion surface is different - suggesting that the magnetic remanence from 3.5 billion years ago is still preserved.

How magnetized rocks record the recent breakup of Greece

When rocks form, they sometimes trap Earth's magnetic field direction. If they are then rotated tectonically, the trapped field direction is also rotated. By comparing the present day magnetic field with the trapped field, we can estimate how and when these tectonic motions occurred. In this study, I worked with Benjamin Weiss from MIT and Manolis Vassilakis from the University of Athens. We collected oriented rock samples from sedimentary beds of of different ages across central Greece. We showed that even relatively young beds (a few million years) are all rotated by almost the same amount. This happened during the latest tectonic breakup of Greece, when the many beautiful interior seaways of this region formed.

Stretching a rock magnetization

We all know that compasses line up their needles to Earth's magnetic field, always pointing toward the modern magnetic pole. Similarly, ancient rock magnetizations can record the direction to the magnetic pole when the rock formed. What many people don't realize is that the magnetic field also has a downward or upward tilt, which depends on latitude. By measuring this tilt in rocks, we can estimate the latitude the rock formed at, which is useful for reconstructing past tectonic movements. In this study, I measured hundreds of samples of sedimentary rock whose layering formed perfectly horizontally. I compared the magnetic tilt to previous studies from nearby igneous and metamorphic rocks, which had much lower tilts. I showed how tectonic stretching of these hot rocks (like pulling hot taffy) could result in abnormally low tilt, by 'stretching' the magnetization. This solved a long debate over whether these stretched rocks had moved great distances northward - the have not.

Extensional tectonics in Greece

The Aegean region has experienced many different tectonic phases over the last several hundred million years. This has created a great mosaic of complex geology, much of which still needs to be untangled despite many decades of dedicated field geology by innumerable geologists. For my Ph.D. thesis work, done under the advisement of Leigh Royden, Benjamin Weiss, and Clark Burchfiel at MIT, I focused on understanding the geology of a small part of the island of Evia, where unusual folded sedimentary basins developed while areas nearby were experiencing rapid tectonic extension. I used different methods to date the sedimentary deposits (lake and river sediments), and mapped the fault systems that likely were active when the basin formed. I showed that the main structures in the region are broad folds oriented parallel to the crustal stretching direction - similar to features seen in western North America.

For the following section, I have contributed to each study in some way, often by bringing a regional geological or tectonic perspective to a geophysical study. The primary research and writing for these studies was done by my collaborators.

Examining Sumatran Fault earthquakes from space

Special radar satellites can produce maps of ground motion in response to geological events, such as earthquakes or volcanic activity. In this study, led by Dr. Rino Salman at ASE/NTU, we used this technique to study a number of earthquakes that have occurred on the Sumatran mainland over the last decade or so. We determined that regional geology plays a major role in where these earthquakes start and stop.

The 2015 Sabah earthquake

Unexpected earthquakes can have intense impacts on local communities and the people who are visiting them. In 2015, an unusual earthquake struck the area of Gunung Kinabalu, Sabah, East Malaysia. Gunung Kinabalu is one of the tallest mountains in the entire region. In addition to the casualties in the local population, this earthquake had devastating consequences for tourists including schoolchildren visiting from Singapore. This study, led by Dr. Wang Yu, brought together many different observations about this earthquake. We showed that the fault that ruptured cannot be clearly mapped at the surface, although similar extensional faults do exist nearby.

Slip rate of the southern Sumatran Fault

My 2017 paper on the Sumatran Fault predicted higher slip rates along the southernmost fault segments than had previously been estimated. In this study, led by Danny Hilman Natawidjaja (Indonesian Institute of Sciences, LIPI), we measured lateral offsets of rivers that are cut into huge volcanic ejecta deposits from the eruption of Ranau Caldera (now Ranau Lake). Because the offsets all occurred after the eruption, we can use the eruption age to calculate long-term fault slip rates. We dated the tephras using the 14C radiocarbon method on trees buried by the ash (a bad day to be in South Sumatra), and mapped offsets with remote sensing and field study. We showed that the slip rate is in fact consistent with the estimates of Bradley et al. (2017).