Tag: ArcticDEM

Heavy rain can more than double the speed of the Sít’ Kusá (Turner) Glacier in Alaska, especially during its slower “quiescent” phase, according to GPS measurements linked to rainfall.

This comes from a 2026 EGUsphere preprint (under review for The Cryosphere) by Parker Wilkerson, Ellyn Enderlin, and colleagues at Boise State University and the University of Idaho. The popular- science framing matches reports like those on Phys.org.

The study examined a temperate surge- type glacier in southeast Alaska’s St. Elias Mountains using ten on- ice GPS stations (plus off-ice references) recording surface speeds roughly every 15 seconds, combined with modeled rainfall (e.g., from ERA5).

Key findings on rain-induced speed- ups:

• Researchers identified large rain events and checked for following speed anomalies (via short- term vs. long- term speed ratios), then measured each anomaly’s timing, peak magnitude, duration, and percent change.
• During the end of a surge (2020- 2021): 9 of 16 large rain events (56%) triggered speed-ups. These averaged a peak of about 7.5 m/day (116% increase over pre- event speeds) and lasted ~66.5 hours. Accelerations often began ~12 hours after rain onset.
• During quiescence (2022- 2024): Speed-ups were somewhat less frequent (17 of 37 events, or 46%), smaller in absolute terms (~1.5 m/day), and shorter (~33 hours), but they showed a much larger relative increase, averaging 251% (more than double, and sometimes higher). Response lagged longer (~32 hours on average).

The differences point to a more efficient subglacial drainage system (better-connected channels that can handle extra water without as much pressure buildup) during quiescence. Even so, rain still caused clear accelerations in both phases, indicating a persistently inefficient component of the drainage system (isolated cavities or similar features) that lets water pressure rise and reduce basal friction, boosting sliding.

Particularly intense or prolonged rain could temporarily slow the glacier during the surge phase (possibly by forcing temporary drainage adjustments), but speeds typically recovered to pre- event levels within about two weeks.

Overall, the work supports the idea that surge- related drainage changes are not a simple on and off switch and shows that rainfall can noticeably modulate speeds on both fast- and slow- flowing temperate glaciers.

This builds on prior observations that sudden water inputs (heavy rain, extreme melt, or lake drainage) can overwhelm subglacial systems and cause short-lived speed- ups lasting hours to days. The high- resolution GPS records here provide detailed quantification across a full surge cycle at one well- studied Alaskan glacier.

The methods section (Section 2 of the preprint) is detailed and carefully designed to isolate short- term rain- driven speed changes from longer- term surge- related variability.

Here’s a structured breakdown:

2.1 Study Site

Sít’ Kusá (Turner Glacier) is a temperate surge- type glacier in the St. Elias Mountains of southeast Alaska.

The authors divide it into four main zones for context:

• Upper- northern tributary
• Icefall (surface elevation drops ~500 m between 20.5- 22.5 km from the terminus)
• Reservoir zone (from the base of the icefall to the dynamic balance line at ~7- 8 km from the terminus)
• Receiving zone (from the dynamic balance line to the terminus)

All GPS stations used for the main analysis were located in the reservoir zone.

2.2 GPS Processing

They used 12 GPS records:

• 10 on- ice stations in the reservoir zone
• 2 off- ice (“Rock”) stations for reference and noise characterization

Instrument details:

• NetR9 GPS receivers from the EarthScope Consortium
• Recorded positions every 15 seconds
• Deployed 2020- 2024, but usable records are mostly limited to late summer through early fall (July- mid- November) due to power issues and deep snow burial in winter

Processing pipeline (applied uniformly):

1. Post-processing: Canadian Precise Point Positioning (CPPP) referenced to the Yukon GPS Epoch (2002.0). Positions converted to UTM and time shifted to Alaska Standard Time.
2. Smoothing and Velocity calculation: Rolling centered moving- window linear (polynomial) fit on Northing and Easting positions.
   • Window size = 1 day
   • Required ≥90% of expected samples (5760 positions) in each window to accept the velocity estimate.
3. Time- domain filtering: Low-pass 4th- order Butterworth filter applied separately to Northing and Easting components.
   • Corner frequency selected as 0.9 cycles/day after testing (0.1- 4 cycles/day).
   • This frequency best removed atmospheric, ionospheric noise (peaks at 1, 2, 3, 4 cycles, day observed on the stationary Rock10 station) while preserving sub-daily to multi- day signals of interest.
   • Note: This also attenuated natural diurnal melt- driven oscillations, which the authors accepted because their focus was rain events.
4. Along-flow speed calculation: Magnitude of the filtered velocity vector.
5. Detrending: Critical step to isolate short- term anomalies from longer- term surge- related trends.
   • During the surge: Smoothing spline fit to the 6- day minima of the speed record (background “floor”).
   • During quiescence: Simple average of the entire record.
   • The background trend was subtracted from the speed time series.
   • Window size of 6 days was chosen after testing shorter and longer windows to avoid over- or under- fitting individual events.

2.3 Modeled Rain Data

• Primary source: ERA5 reanalysis (0.25° × 0.25° grid cell covering the terminus).
• Downscaled to the glacier surface using ArcticDEM elevations and the precipitation component of the Energy Balance Firn Model (EBFM; van Pelt et al., 2012).
• Elevation correction formula: P(z)=PERA​(1+γ100z​)
  where $\gamma =0.1$γ=0.1 (10% increase in precipitation per 100 m elevation gain) and $z$z is the elevation difference.
• Validation: Visual comparison with daily precipitation at Yakutat Airport (~55 km south). Magnitudes were broadly similar; Yakutat records typically led the modeled data (as expected for coastal- to- inland weather progression). No quantitative statistical validation was possible due to the remote location and lack of on-glacier rain gauges distinguishing rain vs. snow.

Analysis of rain impacts was restricted to 1 July- 15 November each year (GPS data availability and cessation of modeled meltwater equivalent output, after which precipitation is assumed to fall as snow or refreeze).

2.4 Analysis (Speed Anomaly & Rain Event Identification)

Speed anomaly detection:

• Applied a short- term and long- term average (STA and LTA) ratio to the detrended speeds (a method commonly used in seismology and GPS studies).
  • STA window = 0.75 days
  • LTA window = 90 days
• Peaks in the STA and LTA ratio were identified with MATLAB’s findpeaks (threshold ratio > 1.9; minimum 2- day separation).
• Additional filters and manual quality control:
  • Minimum absolute speed increase of 0.65 m/day (to avoid noise during quiescence).
  • Manual inspection removed 4 false positives and added 1 missed event → final total of 37 speed anomalies.

Large rain event identification:

• MATLAB findpeaks on the rain rate time series.
• Threshold: peak rate > 2 mm/h (average peak of selected events was 3.5 mm/h, ~95th percentile).
• Minimum separation of 1 day between events.
• Compound events (two closely spaced rains producing one continuous speed anomaly) were merged.

Linking rain and speed:

• A speed anomaly was classified as rain- induced (high sensitivity) if it occurred within a 6- day window after the start of a large rain event.
• Periods without a corresponding speed anomaly were labeled low sensitivity.
• For quantitative characterization, speed and rain series were averaged into 4-hour blocks.
• Characteristic speed profiles were created for high- vs. low- sensitivity periods in both surge and quiescence phases. Additional ice displacement attributable to speed- ups was calculated by subtracting the low- sensitivity displacement from the high- sensitivity displacement.

This careful combination of high-frequency GPS processing, noise removal, detrending, and dual thresholding (speed and rain) allowed the authors to robustly quantify differences in the glacier’s response to rainfall under contrasting drainage- system conditions.

Rain- induced speed- up events of a temperate surging glacier, AK, inferred from GPS measurements

This study examines how heavy rainfall affects the surface speed of Sít’ Kusá (also known as Turner Glacier), a temperate surge- type glacier in the St. Elias Mountains of southeast Alaska.

The researchers used high- resolution GPS records from the glacier surface, paired with modeled rainfall, to compare the glacier’s kinematic response during the end of a surge (2020– 2021) versus the subsequent quiescent (slow-flow) phase (2022– 2024).

Surge- type glaciers alternate between short periods of rapid flow (surges) and longer periods of slow flow (quiescence).

These changes are strongly linked to the efficiency of the subglacial drainage system (isolated cavities = inefficient; connected channels = efficient). The study tests how short- term pulses of rainwater interact with these systems.

Methods (summary)

• 10 on- ice GPS stations in the reservoir zone and 2 off -ice reference stations.
• GPS units recorded positions every 15 seconds (NetR9 systems).
• Data processing: precise point positioning, smoothing (1- day moving window), low- pass Butterworth filtering to remove atmospheric noise, and detrending.
• Records mainly cover late summer to early fall due to power and snowpack limitations.
• Rainfall was modeled (primarily ERA5 data).
• Speed anomalies were identified by comparing short- term vs. long- term average speeds. Only anomalies following large rain events were analyzed for duration, magnitude, percent increase, and lag time.

Key Results

During the surge (2020– 2021):

• 9 of 16 large rain events (56%) triggered speed- ups.
• Average peak speed increase: 7.5 m/day (116% above pre- event speeds).
• Average duration: ~66.5 hours.
• Acceleration typically began ~12 hours after rain onset.

During quiescence (2022– 2024):

• 17 of 37 large rain events (46%) triggered speed-ups.
• Absolute increases were smaller (~1.5 m/day).
• Average duration: ~33 hours.
• Relative increases were much larger: average 251% (more than double the pre-event speed).
• Response lag was longer (~32 hours on average).

Particularly large, prolonged rain events sometimes caused temporary slow downs during the surge phase, but speeds generally returned to pre- event levels within less than two weeks.

Interpretation

• The larger absolute, longer- lasting, and more frequent speed- ups during the surge are consistent with a less efficient subglacial drainage system (higher basal water pressures and greater sliding sensitivity).
• The higher percentage increases during quiescence occur because the background speeds are much lower.
• Speed- ups in both phases indicate that some inefficient drainage components persist throughout the surge cycle (i.e., there is not a complete binary switch between inefficient and efficient systems).
• Rain can therefore modulate the speeds of both fast- flowing (surging) and slow- flowing (quiescent) temperate glaciers.

Broader Significance

The work supports recent views that glacier surge dynamics involve more gradual or incomplete changes in drainage system configuration rather than abrupt on and off switches.

It also improves understanding of how extreme rainfall events, expected to become more common in a warming climate, can rapidly influence glacier motion and potentially contribute to mass loss or instability.

Status: Open for discussion until 5 December 2026.
Full PDF and supplement are available on the EGUsphere page.

Journal: The Cryosphere

DOI: 10.5194/egusphere-2026-6232

Preprint: EGUsphere, (discussion started 22 October 2026; under review for The Cryosphere)

Authors: Parker Wilkerson,
Ellyn Enderlin,
Yoram Terleth,
Jukes Liu,
and Jake Anderson

Abstract.

Temperate surge-type glaciers undergo periodic order-of-magnitude increases in flow speeds generally attributed to drainage system evolution, making them ideal sites to explore how the kinematic response of glaciers to rain events varies with subglacial drainage system efficiency. Here, we compare ten GPS-derived glacier surface speed records with modeled rainfall to assess short-term evolutions of the drainage system at the end of a surge (2020–2021) and during quiescent flow (2022–2024) at Sít’ Kusá, Alaska. Speed anomalies were identified using a short-term to long-term average ratio approach and all speed anomalies that followed large rain events were characterized by their duration, magnitude, and immediacy. During the surge, 9 of 16 large rain events (56%) caused speed anomalies, peaking at 7.5 m/d (116% increase) and lasting 66.5 hours on average. During quiescence, rain-induced speed-ups were less frequent (17 of 37 rain events (46%)), smaller (~1.5 m/d), shorter (~33 hours), but greater in percent increase (251%). Although the differences in speed characteristics generally indicate that the drainage system is more efficient during quiescence, speed-ups throughout the surge cycle suggest there is a persistently inefficient component of the drainage system. We find that particularly large and prolonged rain events can cause temporary slow-downs during the surge, but that speeds return to pre-event levels in less than two weeks as the drainage system reverts to its inefficient state. Our work supports recent analyses of glacier instabilities suggesting there isn’t a binary switch in drainage system configuration during surges. More broadly, our findings expand our understanding of glacier kinematic responses to rainfall, demonstrating that rain events can modulate speeds of both fast- and slow-flowing glaciers.