Seasonal microfractures beneath Alaska’s Barry Arm reveal hidden instability signals

The Barry Landslide rests on a steep slope above Barry Arm in Prince William Sound. The mass is large, containing between 500–700 million m³ of fractured rock, which is equivalent to 17.6–24.7 billion feet³. This scale makes the slope one of the most hazardous in coastal Alaska.

The landslide has been creeping for decades, although movement remains slow. If the mass were to fail rapidly, it would fall directly into the fjord and could generate a powerful tsunami capable of impacting communities such as Whittier. Cruise vessels and kayakers regularly visit the area, which increases exposure to a rapid event.

The supporting ice of Barry Glacier has retreated significantly during the past century. The loss of this natural buttress has reduced slope stability and increased the likelihood of sudden collapse. These changes motivated the installation of seismic, radar, infrasound, and weather instrumentation beginning in 2020.

Establishing a baseline of natural seismic behavior became essential due to the complexity of the environment. Multiple glaciers, active tectonics, and rockfalls continually generate seismic noise that must be separated from any signals related to slope deformation. Only through careful classification can unusual patterns be recognized.

What researchers discovered in the seismic record

Scientists identified a distinct category of very short seismic events at the Barry Arm Top station, which sits on the landslide. These events lasted one to four seconds and had high clarity compared to the surrounding noise. Their waveforms often contained clear P and S phase arrivals, which indicates a subsurface source.

Unlike glacier calving signals or earthquakes, these events were not detected at the opposite station across the fjord. The absence of infrasound signals further showed they were not occurring at the surface. Their characteristics pointed toward brittle fracturing within rock or basal ice close to the instrument.

The team manually labeled all events from a training day on September 19, 2023. They then optimized the STA to LTA detection algorithm using a five-hertz high-pass filter to emphasize sharp impulses. After tuning the parameters, the detector identified thirty-five of thirty-nine manually observed events while keeping false detections low.

Applying the detector across a three-year dataset produced 32 877 short impulsive events. Winter periods contained gaps due to heavy snow blocking telemetry. Even with these gaps, the seasonal pattern was unmistakable. Event rates climbed in late summer, rose steadily into midwinter, and shut down abruptly between February and March.

How the seasonal cycle rules out other explanations

The timing of the seismic bursts did not match rainfall patterns. Precipitation peaks in early fall, but the burst of seismic activity lagged by several months. SI events also continued long after rain events had ceased, which makes rainfall-driven pressure changes an unlikely cause.

Air temperature showed the opposite trend. SI activity grew as temperatures fell and weakened as temperatures rose. Summer melt periods did not generate seismic cracking, which would be expected if meltwater were directly responsible for the events themselves.

Slope motion data collected by ground-based radar showed movement in late summer and early fall. In 2022, the slope moved between 30–65 mm (1.2–2.6 inches) per day and totaled about 7 m (23 feet) of displacement. In 2023, the slope moved only about 1 m (3.3 feet). SI events increased during the onset of motion but continued rising long after deformation ceased.

Glacier dynamics also failed to explain the trend. Cascade Glacier typically moves between 400–600 m (1 300 to 1 970 feet) per year, with only small seasonal fluctuations. Those fluctuations peak in spring, which contrasts strongly with the autumn rise and midwinter shutdown of SI activity.

Locating the hidden source beneath Cascade Glacier

Polarization analysis allowed the team to calculate back azimuth and incidence angles for events with rectilinearity values above 0.5. This produced 11 685 reliable measurements. Most events originated from azimuths between 10–60 degrees clockwise from north with incidence angles between 30–70 degrees from vertical.

The angles indicate that the events arrive from the southwest, which points toward Cascade Glacier rather than the landslide mass itself. Apparent S P times and assumed shallow bedrock velocities suggest source distances between 1–2 km (0.6–1.2 miles).

Repeating waveform families provided a second line of evidence. One dominant family appeared every year, while others were restricted to only certain seasons. This recurrence shows that the fractures originate from stable locations that activate repeatedly under similar conditions.

The steep incidence angles may indicate deeper sources beneath the glacier, although near-surface velocity contrasts can modify apparent angles. The mountain is composed of fractured metamorphic rock, which adds further complexity. Still, the geometry supports a source region beneath or near the base of Cascade Glacier.

Why freeze-thaw cracking provides the strongest explanation

As summer meltwater input declines in late August, the hydraulic system behind the landslide begins to contract. Flow pathways beneath the glacier constrict as temperatures drop. Remaining water in microcracks cools until it begins to freeze.

When water freezes, it expands by roughly nine percent in volume. This expansion generates significant pressure capable of forcing open small fractures in both glacier ice and bedrock. Each opening event produces a short high-frequency seismic pulse identical to the SI events recorded at Barry Arm.

This freeze-driven cracking continues throughout autumn and peaks in early winter. When the entire subglacial system freezes solid by February, cracking ceases, matching the abrupt seasonal shutdown observed in the seismic record.

A study from Norway documented similar signals in an unstable slope where cracking was linked to temperature oscillations. The similarity strengthens the interpretation that Barry Arm SI events arise from seasonal freezing within rock or basal ice.

Although it is not possible to determine whether the fractures occur primarily in rock or in lower glacier ice, both environments experience the same freeze-controlled hydraulic shutdown. In either case, the events reflect subsurface processes relevant to slope hydrology.

Implications for monitoring a dangerous landslide

While SI events do not correspond directly with slope movement, they appear to trace the seasonal behavior of the hydraulic system behind the landslide. If the hydraulic system remains active later than usual, it could delay the shutdown of SI events. This delay may indicate elevated water pressure, which is a known factor in destabilizing slopes.

A reliable understanding of normal seasonal patterns creates a foundation for detecting anomalies. A year without early autumn activity or a year with an unusually prolonged winter cycle could help identify changes in subsurface water flow that warrant closer monitoring.

The Alaska Earthquake Center is testing a regional detection system for landslides at Barry Arm. This system will issue alerts for real-time slope failures. If future research validates the relationship between SI activity and hydraulic behavior, SI event tracking may become a component of next-generation warning systems.

Barry Arm is one of the best instrumented landslide hazards in the world. Its seismic network, weather stations, infrasound arrays, and radar measurements form a comprehensive dataset that is rare in a glaciated fjord. As similar hazards emerge across alpine regions, the methods developed here may be applied more widely.

References:

¹ Study Searches for Landslide Clues in Seismic Signals from Alaska’s Barry Arm – SSA – December 2, 2025

² Searching for Seismic Precursors—The Barry Landslide Hazard – Gabrielle K. Davy et al. – GeoScience World – November 19, 2025 – https://doi.org/10.1785/0220250205 – OPEN ACCESS

Reet Kaur

I’m a science journalist and researcher at The Watchers, contributing to the Epicenter edition, where I cover peer-reviewed scientific research and emerging discoveries across Earth and space sciences. With a background in astronomy and a passion for environmental science, I’ve worked in shark and coral conservation in Fiji, conducting reef and shark-behavior research, contributing to mangrove restoration, and earning PADI Open Water and Coral Reef Certifications. I bring a blend of scientific rigor and storytelling to illuminate the discoveries shaping our planet and beyond.

This author does not have any more posts.