Pretty Rocks Landslide

Overview

The Pretty Rocks landslide intersects the Denali Park Road near its midpoint at Mile 45.4 and displaces 100 yards (90 m) of the full width of the road. In recent years this landslide has evolved from a minor maintenance concern to one that has caused substantial road restrictions. To learn more about the bridge construction project that began at Pretty Rocks in 2023, read about the Polychrome Area Plan and follow the Pretty Rocks Construction blog for updates.

The Pretty Rocks landslide has been active since at least the 1960s, and probably since well before the Denali Park Road was built through this area in 1930. Before 2014, the landslide only caused small cracks in the road surface and required moderate maintenance every 2–3 years. However, in 2014, road maintenance crews noticed a substantial speed up. By 2016 the movement had increased further, a slump had developed in the road, and a monitoring program was begun. The rate of road movement within the landslide evolved from inches per year prior to 2014, to inches per month in 2017, inches per week in 2018, inches per day in 2019, and up to 0.65 inches per hour in 2021.

 

Geotechnical investigations near the collapsed portion of the road at Pretty Rocks in summer 2022.

NPS Photo

 

NPS road maintenance staff made an extraordinary effort in Spring 2021 to repair the road at Pretty Rocks. By August 2021, these road maintenance efforts were no longer sustainable in the face of the landslide’s accelerated movement.

NPS Photo

By late August, 2021, park managers recognized that adding hundreds of truckloads of gravel into the slumping road to maintain a drivable surface was no longer tenable or safe, and they enacted a road closure west of Pretty Rocks. Since then, the movement of the landslide has continued to carry that section of road farther down the hillside and the road closure has remained in effect.

Pretty Rocks is one of several known landslides in the Polychrome area and one of the more than 140 mapped unstable slopes along the entire Park Road (Capps et al., 2017). The Polychrome Area Plan was developed with this in mind and addresses solutions not just for Pretty Rocks but also for the nearby Bear Cave Landslide. Even though Pretty Rocks is currently our most problematic location, the National Park Service is working with the Federal Highway Administration to manage the park road proactively and holistically in order to restore and maintain access along the entire length of the road.

 

Polychrome Area Plan

 

Composite photo of the Pretty Rocks Landslide area from 2015. Approximate landslide outline indicated by red dashed line.

NPS Photo

 

Time-Lapse Videos

A time-lapse camera is located on the west side of the landslide. The videos below show the continuous nature of the landslide over the course of months and years.

 

Duration:

22.9 seconds

A time-lapse video of the Pretty Rocks landslide and construction site from June 17, 2024 to November 8, 2024. Note how the movement of the landslide speeds up and peaks in September due to summer temperatures warming the ice-rich permafrost underground, and then slows down as winter arrives. The busy construction season concluded in October after the first four trusses of the bridge were erected on the east side of the slide. No audio.

 

• 

  Pretty Rocks Time-lapse (July 25, 2021 to April 18, 2022)

  Time-lapse video of the Pretty Rocks landslide from July 25, 2021 to April 18, 2022, looking to the northeast. From July 25th until September 2nd, road maintenance crews were spreading up to 100 truckloads of gravel per week within the landslide to maintain the road.

  Duration:

  14.967 seconds

 

• 

  Pretty Rocks Time-lapse (September 10, 2021 to April 18, 2022)

  Time-lapse video of the Pretty Rocks landslide from September 10, 2021 to April 18, 2022, looking to the southeast.

  Duration:

  14.967 seconds

 

Climate Change as Causal Agent

Based on climate data from 1950 to 2010, Denali National Park has experienced a temperature increase of 7.7°F ± 2.0°F (4.3°C ± 1.1°C) per century, the highest of all national parks (Gonzales, et al, 2018). A recent shift in mean annual temperatures in the area to near 32°F (0°C) (Swanson et al., 2021) combined with heavy rainfall events is believed to be causing permafrost to thaw, resulting in the recent acceleration of many landslides in Denali. In the case of Pretty Rocks Landslide, climate change has likely taken a problem previously solved by minor road repairs and made it difficult to overcome with short-term solutions. Temperature and precipitation increases are likely to continue, and the rate may worsen, over the coming decades.

While these changes were predicted by modelers (Rupp and Loya, 2009), the magnitude of the change has been greater than expected. For example, the area where Eielson Visitor Center is located (Mile 66 [km 106]) reached an average mean annual temperature of 32.4°F (0.2 °C) from 2015-2019 (Sousanes, 2020); this exceeds the temperature of 29.8°F (-1.2°C) projected for 2040 by Rupp and Loya (2009).

 

Geology of Pretty Rocks Landslide

We have learned about the subsurface of Pretty Rocks through multiple borings (drilling) in and around the landslide. While conditions are variable, a representative cross section is shown in the following scientific illustration by Laurel Mundy. All borings within the landslide contain permafrost, or ground that is at or below freezing for two or more years. The shallowest material, the active layer, thaws seasonally and is therefore not permafrost. Permafrost can be essentially ice free (ex. granite high on Mount McKinley), ice poor, or ice rich.

Some of the substrate here is ice poor, meaning that ice volume is equal to or less than the pore space between debris. Some of the substrate is ice rich, meaning that ice volume exceeds the pore space between debris. Pretty Rocks contains abundant ice – we have drilled through up to 15’ (4.6 m) of 85% ice directly beneath the road. The ice-rich permafrost typically is positioned directly above the debris and bedrock interface. Slope inclinometers (instruments that measure displacement) indicate that most movement is occurring at this interface.

 

Scientific illustration of the surface and subsurface of the Pretty Rocks landslide

Illustration by Laurel Mundy

While it’s appropriate to describe the substrate of Pretty Rocks landslide as ice-rich permafrost, it is most accurately described as a rock glacier. Rock glaciers are closely related to glaciers, but they contain much more rocky debris with ice in the pore spaces. Rock glaciers typically only move a few inches (several cm) per year, as was the case with Pretty Rocks prior to 2014.

However, Pretty Rocks and some other rock glaciers around the world have begun to advance quickly, with climate change as a major cause (Müller et al., 2016). Denali has many rock glaciers, likely thousands of them, and efforts are underway to identify and classify them. Some of these rock glaciers are currently advancing beyond their historic margins.

 

Conceptual model of the evolution of a rock glacier that illustrates the Pretty Rocks rock glacier accurately. Arrows show sediment transport. t0, t1 and t2 show rock glacier surfaces at different times resulting from variations in environmental factors such as warming and a decrease of sediment-ice

Müller et al., 2016 (unmodified)

Many that learn about Pretty Rocks ask a reasonable question, “Why don’t you just scrape off or blast into the material that is coming down and rebuild the road on bedrock?” The borings help us answer this question. The substrate that is now sliding was deposited horizontally about 55-million-years ago in volcanic eruptions. Subsequent folding and faulting inclined the material to approximately 60°, as shown in the following scientific illustration by Tracey Faber. The blue highlighted layers are mostly competent rock. The orange layer is mostly volcanic ash that has altered to clay through time.

The clay is very unstable on steep slopes like those at the road. The clay is also mostly frozen, but only barely – it is within 1°F (0.6°C) of thawing. Based on borings and surface mapping, the clay is known to extend hundreds of feet into the subsurface and laterally to the north-northeast all the way to the East Fork of the Toklat River just north of East Fork bridge. Where this same layer intersects the East Fork River, landslides similar to the one at Pretty Rocks are occurring but do not intersect the road. This long, vertical layer of unstable clay is the reason that we cannot simply dig the undesirable material out and put the road on competent bedrock farther in on the cliff face.

 

Scientific illustration of the Pretty Rocks area. Blue highlighted areas are mostly competent rock. The orange layer is mostly unstable clay. Pretty Rocks landslide is located where the road (white line) intersects the orange. The main branch of the East Fork River is near top right.

Tracey Faber

Monitoring System for Pretty Rocks Landslide

The Pretty Rocks landslide monitoring system is located on a stable bedrock ridge below the road and above the western side of the landslide. The robotic total station is a surveying instrument that uses a laser to accurately measure distances and angles from about 35 small prisms (mirrored reflectors) installed within and around the landslide. After staff complete initial measurements manually in the field, the total station repeats the measurements on a defined interval (typically every hour). The weather station includes a shielded thermometer, anemometer (wind speed and direction), and tipping-bucket rain gauge. We create time-lapse videos from photos taken every few hours.

 

NPS Photo

Above: Annotated image of the Pretty Rocks near-real-time monitoring system. To improve visitor safety and understand the movement of the Pretty Rocks landslide, we built a near-real-time monitoring system. The system has five main components: 1) a robotic total station; 2) a weather station; 3) a time-lapse camera; 4) a solar-power system; and 5) a satellite internet system

Because of the large amount of displacement that was occurring in the fall of 2021 and the road closure, we installed a second time-lapse camera looking down the axis of the road to create additional time-lapse videos. The solar-power system utilizes solar panels to charge numerous batteries that power the system during low-light times. The satellite internet system relays all the data from the above system to a computer at park headquarters.

This data is then processed by software and posted to a web interface that allows staff to view the data remotely at any time. Further, weather and movement threshold alarms alert staff automatically when thresholds (ex. heavy rain, increases in movement) are exceeded.

 

References

Capps, D. M., R. Rosenberg, A. Collins, D. A. Anderson, S. Hooper, H. Rogers, B. Collins, and E. Bilderback. 2017.
Geohazards Risk Assessment of the Denali National Park Road, in Proceedings North American Symposium on Landslides, Roanoke, VA. Association of Environmental & Engineering Geologists.

Gonzalez, P., F. Wang, M. Notaro, D. J. Vimont, and J. W. Williams. 2018.
Disproportionate magnitude of climate change in United States national parks. Environmental Research Letters 13: 12.

Müller, J., Vieli, A., and Gärtner-Roer, I, 2016.
Rock glaciers on the run – understanding rock glacier landform evolution and recent changes from numerical flow modeling, The Cryosphere, 10, 2865–2886, 2016. Creative Common Attribution 4.0 License.

Rupp, S. and W. Loya. 2009.
Projected climate change scenarios for Denali National Park & Preserve: University of Alaska Fairbanks, Climate change summary reports of National Parks, Preserves and Monuments.

Sousanes, P. 2020.
Analysis of temperatures from the two stations in Denali National Park and Preserve that best characterize the Pretty Rocks location along the Denali Park Road. Fairbanks, Alaska: National Park Service.

Swanson, D. K., Sousanes, P. J. & Hill, K. 2021.
Increased mean annual temperatures in 2014-2019 indicate permafrost thaw in Alaskan national parks. Arctic, Antarctic, and Alpine Research, 53, 19.