A quiet Alaska fault is missing the fluids scientists expected – and it’s changing what we know about earthquake zones
- Written by Yinchu Li, Ph.D. Candidate in Marine Geology, Georgia Institute of Technology
Not all earthquake faults behave the same. Some stick and snap[1], causing earthquakes. Others move slowly over time.
For years, the leading explanation for slow-moving faults has been that high-pressure fluids along the fault lubricate it, allowing the slabs to slide steadily rather than building up stress until that stress is eventually released in a large, destructive earthquake.
But in a new study of the Shumagin Gap[2], a quiet section of the Alaska-Aleutian subduction zone – the area where one tectonic plate dives below another – my colleagues and I found that the fault does not contain enough fluid to explain why it slides slowly. Scientists may need to rethink this assumption about subduction zones around the world.
Pinning down why faults creep matters for how scientists build models of the world’s most powerful earthquake zones to assess long-term earthquake and tsunami hazards[3], from Alaska to Japan to the Pacific Northwest. Knowing how earthquakes are likely to behave is essential for helping communities decide where and how to build homes and other infrastructure so they can withstand an earthquake and tsunami.
How earthquakes happen along faults
An earthquake fault[5] is a break in Earth’s outer rock layer where two blocks of rock slide past each other. The way they slide determines what kind of shaking, if any, reaches the surface.
Some faults are “locked[6].” They do not budge until stress builds to a breaking point, then they release it all at once in a sudden rupture. This is what happens during most damaging earthquakes. Other faults “creep[7].” They glide past each other steadily, releasing stress gradually.
The biggest and most destructive earthquakes on Earth happen along subduction zones[8], where one tectonic plate dives beneath another. The Alaska-Aleutian margin[9], the Japan Trench[10], the subduction zone off Chile[11] and the Pacific Northwest’s Cascadia zone[12] are all examples. When a locked patch of a subduction fault suddenly slips, the seafloor can jolt upward and a tsunami can follow.
A quiet fault challenges a common assumption
Deep underground, fault behavior is hard to see directly, especially offshore where faults often sit beneath kilometers of seawater and sediment.
Scientists rely on measurements from GPS stations, seismometers[13] and seafloor sensors, and then build computer models of what must be happening below. For decades, the leading explanation for creeping faults has been that high-pressure fluids along the fault reduce friction, the way a film of water causes tires to hydroplane.
Testing that idea requires seeing the fluids, and that’s where our team came in.
We use marine electromagnetic imaging[15], a method that maps how easily underground materials conduct electricity. A ship tows an instrument close to the seafloor, sending electromagnetic signals into the rocks below, while other instruments on the seabed record the response. Different materials beneath the seafloor conduct electricity differently, and that shows up in the measurements. Because salty water conducts electricity very well, the method is especially good at mapping where fluids are and where they aren’t.
Researchers deploy marine electromagnetic instruments off Alaska to image fluids and rock structure beneath the seafloor.
EMAGE team/Kerry Key[16]
We surveyed a 75-mile (120-kilometer) stretch of seafloor[17] across the Shumagin Gap, a section of the Alaska-Aleutian subduction zone that has been creeping for more than a century. The Shumagin Gap had long been considered a quiet part of the margin[18], even though neighboring segments have produced magnitude 8 and larger earthquakes.
To our surprise, the fault at the Shumagin Gap was not as fluid-rich as the leading explanation would predict.
Our images show that the shallow part of the fault, closest to the ocean, has little open space in the rock for fluid[19] to occupy. And the fluid that is there is under roughly normal pressure, not the high pressure that the “slippery fluid” model predicts.
The fault surface is bumpy and rugged. The upper plate appears to be a patchwork of stronger and weaker material, and we found possible pathways where fluids may drain into the rock above the fault.
In other words, this quiet fault isn’t quiet because it’s well lubricated. Something else is keeping it stable, most likely a combination of rough fault surface, varying rock strength and, in some places, fluid.
A cross-sectional illustration of the Shumagin Gap shows a rough plate interface and limited fluid.
Yinchu Li, et al., 2026[20]
What this means for assessing earthquake risks
Our findings about this fault have consequences for assessing earthquake and tsunami hazards more broadly.
Many models lean on the idea that fluid pressure helps determine whether a subduction fault slips suddenly or creeps. If fluid isn’t the main control keeping the Shumagin Gap quiet, other quiet faults might similarly lack fluid, raising questions about how stable those faults really are.
Understanding these mechanisms matters for assessing coastal communities’ earthquake and tsunami risks. A shallow slip near a trench is what drives the most destructive tsunamis. Tsunamis from Alaska–Aleutian earthquakes have reached distant coasts before. Large earthquakes in 1946[21], 1957[22] and 1964[23] generated tsunamis that damaged the coasts of Hawaii and California.
As our results show, there isn’t a single, simple story explaining slow-sliding faults. More and better offshore data will help scientists more accurately assess earthquake and tsunami hazards around the world and help communities well beyond Alaska prepare.
References
- ^ stick and snap (www.usgs.gov)
- ^ study of the Shumagin Gap (doi.org)
- ^ tsunami hazards (www.tsunami.gov)
- ^ Yinchu Li, et al., 2026 (doi.org)
- ^ earthquake fault (www.usgs.gov)
- ^ Some faults are “locked (www.usgs.gov)
- ^ Other faults “creep (www.usgs.gov)
- ^ subduction zones (www.usgs.gov)
- ^ Alaska-Aleutian margin (oceanexplorer.noaa.gov)
- ^ Japan Trench (doi.org)
- ^ subduction zone off Chile (doi.org)
- ^ Pacific Northwest’s Cascadia zone (pnsn.org)
- ^ seismometers (www.bgs.ac.uk)
- ^ Yinchu Li, et al., 2026 (doi.org)
- ^ marine electromagnetic imaging (marineemlab.ucsd.edu)
- ^ EMAGE team/Kerry Key (egl.eas.gatech.edu)
- ^ surveyed a 75-mile (120-kilometer) stretch of seafloor (egl.eas.gatech.edu)
- ^ quiet part of the margin (earthquake.alaska.edu)
- ^ little open space in the rock for fluid (doi.org)
- ^ Yinchu Li, et al., 2026 (doi.org)
- ^ 1946 (sos.noaa.gov)
- ^ 1957 (www.usgs.gov)
- ^ 1964 (www.usgs.gov)
Authors: Yinchu Li, Ph.D. Candidate in Marine Geology, Georgia Institute of Technology
