Glacier Bay National Park & PreservePage 60 of 97
Where Landslides Could Send Tsunamis Into Glacier Bay’s West Arm
In 1958 a rockslide in Lituya Bay, triggered by a magnitude 7.8 earthquake on the Fairweather Fault, sent a tsunami up the opposite shore to 1,719 feet (524 m) above sea level and killed two people in a small boat. Lituya Bay was then part of Glacier Bay National Monument. Glacier Bay National Park and Preserve (GBNPP) still has the ingredients for such events, and in 2017 about 540,000 people visited it on cruise ships and tour boats. This article from Alaska Park Science (2019), by USGS scientists Jeffrey A. Coe, Robert G. Schmitt and Erin K. Bessette-Kirton, makes a first map of where landslides could enter the West Arm of Glacier Bay.
Map of Glacier Bay National Park and Preserve and the study area around the West Arm. Figure 1 of the article.
Why the bay is exposed
- A plate boundary. The park lies along a transform boundary about 745 miles (1,200 km) long where the Pacific and North American plates slide past each other. The onshore Fairweather Fault takes up much of the motion, about 1.5 inches (43 mm) a year.
- Rapid ice loss. The area has deglaciated extensively in the last 250 years; in 1860 the whole West Arm was filled with ice. Rebound gives the region one of the highest uplift rates on Earth, up to about an inch (30 mm) a year.
- Steep, narrow inlets. Glacial scour followed by deglaciation and uplift left long, narrow inlets with steep rock walls, and at least nine tidewater glaciers end in the West Arm. Narrow inlets can amplify waves and cause unpredictable oscillations that last for hours after the first wave.
What has happened, and what triggers it
Rock avalanches are landslides of fragmented rock larger than 1 million cubic yards (1 million cubic meters) that can travel more than 0.6 mile (1 km) at up to 220 miles an hour (100 m/s). Known events include:
| Event | Detail |
|---|---|
| Lituya Bay, 1958 | rockslide of about 39 million cubic yards (30 million m³) |
| Tidal Inlet slide | 6.5 to 14.5 million cubic yards (5 to 11 million m³), moving 1–1.5 inches (3–4 cm) a year between 2002 and 2004, and moving since at least 1892; a failure could make waves over 33 feet (10 m) high near the inlet's mouth |
| Lamplugh rock avalanche, June 28, 2016 | about 92 million cubic yards (70 million m³), the largest historical one in the park; it landed entirely on Lamplugh Glacier, not in the water |
| Taan Fiord, October 2015 (Wrangell-St. Elias National Park) | 99 million cubic yards (76 million m³); its tsunami ran about 623 feet (190 m) up the shore |
Since 1958 there have been at least 90 earthquakes above magnitude 4 within 62 miles (100 km) of the study area, but there is no evidence that any triggered a rock-slope failure there. All recent rockslides and rock avalanches have had climatic triggers: at least 24 climatically induced rock avalanches occurred between 1984 and 2016, and they have grown in size and travel distance. The authors list four possible factors: thawing mountain permafrost, glacier thinning that removes support from rock slopes, more precipitation, and rock weakened by crustal strain. A 2018 study by Coe and colleagues hypothesized that permafrost degradation is the main driver of the increase. The peak cruise season, June to August, is also when climatic slides are most likely.
How they mapped it
They treated slopes steeper than 25° as possible source areas, since the 24 mapped rock avalanches began on slopes steeper than 26°. To estimate how far slides could travel, they used travel angles from those avalanches, which ranged from 9° to 31° with a mean of 19°. A routing model, Flow-R, ran slides across combined land and seabed elevation data at 30 m resolution, using 9°, 19° and 31° to mark high, medium and low relative probability of impact. The model does not treat water and land differently and covers only slides that start on land.
Potential landslide source areas in the West Arm study area. Figure 4 of the article.
What they found
Large areas of moderate to high impact probability lie in Johns Hopkins, Tarr, Rendu and Tidal Inlets. Others are the eastern and western shores of Queen Inlet, the eastern side of Gilbert Peninsula, the western side of Gloomy Knob, and the eastern shore of the West Arm between Tidal Inlet and Tlingit Point. The maps are meant to prioritize where more detailed study is needed.
Modeled relative probability of impact from landslides that start on land. Figure 5 of the article.
While checking seabed features the authors found a large rockslide near the junction of Johns Hopkins and Tarr Inlets, which they call the Inlet Junction rockslide. It happened after 1892, when a glacier's toe covered the site. Its minimum volume is about 105 million cubic yards (80 million m³), the largest known landslide in the park. It traveled about 7,315 feet (2,230 m) while dropping about 1,410 feet (430 m); the authors suggest that moving over water-saturated soft sediment may explain the long travel, and whether it caused a tsunami is unknown.
What comes next
The authors suggest visiting the moderate-to-high areas to map geology at a scale of 1:20,000 or larger; estimating volumes and modeling runout for the most susceptible slopes, helped by new lidar data; monitoring steep coastal slopes with radar or photogrammetry to detect slow movement before a failure; and modeling the tsunamis to learn what size of boat or ship waves could threaten.
Sources
- Jeffrey A. Coe, Robert G. Schmitt and Erin K. Bessette-Kirton (U.S. Geological Survey), “An initial assessment of areas where landslides could enter the West Arm of Glacier Bay, Alaska and implications for tsunami hazards,” Alaska Park Science 18(1): 26–37 (2019), National Park Service: https://www.nps.gov/articles/aps-18-1-4.htm; rewritten in hubnx's own words.