Last updated: August 31, 2024
Article
Caught By Surprise, One Park Learned Its True Landslide Risk
Unexpectedly heavy rainfall caused damaging landslides in a national recreation area. A staff scientist showed such events were more frequent than previously thought—and more predictable.
Image credit: NPS / Jonathan Malzone
Slope failures—rock falls, slumping stream banks, and landslides to name a few—have been rare in institutional memory at Delaware Water Gap National Recreation Area. So when I began my job as the park’s physical scientist, little did I know I’d be focusing on catastrophic movements of earth and water. But in spring 2022, several intense rainstorms caused multiple slope failures and other flood-related damage throughout the nearly 70,000-acre park. By summer, there were road closures on almost every major park roadway. Suddenly, my responsibilities were much more urgent as I scrambled to help the park understand what had happened and why. Through my historical research and analyses of rainfall patterns—coupled with some key partnerships and tools—the park is now better equipped to handle the consequences of extreme weather.
An Increasing Sense of Urgency
Conservative estimates are that the Delaware Water Gap National Recreation Area will experience a 3.2 to 9.4 °F increase in annual average temperature and an 8 to 15 percent increase in annual precipitation by the year 2100. Regional climate trends predict that individual storms will become more intense. Scientists expect the kinds of storms that used to come every 50 years to now come every 30 years.
The loss of evergreen trees increases the amount and intensity of rainfall reaching the ground in winter months.
There’s more: Eastern evergreen forests are becoming more susceptible to rainfall due to invasive pests like the hemlock woolly adelgid, which has decimated stands of native hemlock forests over the last 30 years. The tree species that replace these evergreen forests tend to be deciduous, losing their leaves in the winter. Because leaf cover functions as an umbrella, catching as much as 10 to 50 percent of annual rainfall, the loss of evergreen trees increases the amount and intensity of rainfall reaching the ground in winter months. I knew it was crucial that the park learned how to adapt and respond to these climactic and ecological changes.
Aerial images courtesy of Pennsylvania Spatial Data Access
A Pattern Emerges
Geologists have reported slope failures in the Delaware Water Gap in the past, but when I started this work, the park had few written records about them. I began by collecting what information I could about how often slope failures had occurred. In the absence of official park records, I gathered data from news articles, government reports, geological publications, conversations with staff, and through direct observation. In this way, I identified 25 separate failure events along roadsides in the park since 1881.
The failures occurred many miles and years apart. This made it difficult to find anyone who knew much about them.
With that number of roadside slope failures, I was surprised the park wasn’t more familiar with them. But then, a few patterns began to emerge that told me why. First, the failures occurred many miles and years apart. This made it difficult to find anyone who knew much about them, because employees had changed their work locations or left the park through the years. Second, almost all reports of slope failures near roads coincided with extreme rainfall or other precipitation. In a river park, these types of events produce floods, downed trees, damaged infrastructure, and other hazardous conditions. In the aftermath, parks likely focus on making critical repairs, not on writing geological reports. Yet my observations told me that—infrequent or not and recorded or not—slope failures clearly posed a significant risk to park resources.
Left image
March 2022
Credit: NPS / Jonathan Malzone
Right image
December 2023
Credit: NPS / Jonathan Malzone
Broadening the Search for Data
Next, I used high-resolution elevation data to make a land surface (digital elevation) model of the park. Then I examined the model for evidence of slope failure scars and matched those areas to geologic information and historical slope failure reports. I mapped and classified the failures, finding over 500 failure scars throughout the park. These were the remains of slope failures that had occurred throughout geologic time, giving the full scope of slope stability across the park.
These failure-prone conditions are natural. They result from how the Appalachian Mountains formed and from ice ages.
I used the geologic information, map information, my fieldwork, and previous writings to create an updated list of six major geologic failure-prone conditions in the park:
- Loose talus slopes: steep piles of broken rock, easily shifted by water.
- Rocks partially composed of limestone; rainwater can dissolve the limestone cement that binds these rocks together.
- Rocks with cracks that form an unstable pattern; they could fall.
- Glacially polished rock surfaces: smooth surfaces created by ice scraping that reduce the rock’s potential to hold overlying soil.
- Low-density or soft glacial sediments; sediments become weak when wet.
- River undercutting: river erosion that steepens banks over time.
Despite being potentially hazardous, these failure-prone conditions are natural. They result from how the Appalachian Mountains formed and from ice ages that buried the area under glaciers.
The next step was to go outside and collect data for ground truthing my office research—verifying its validity (or not) based on real-world conditions. With the help of colleagues, I set up monitoring stations at areas with active slope failures. Where possible, we placed sensors to measure surface movement. I also got rainfall records from a station inside the park operated by the Western Regional Climate Center.
Image credits: NPS / Jonathan Malzone
Finally, to analyze the risk of future events in failure-prone areas, the National Park Service’s Geological Resources Division helped the park start an unstable slope management program. This is a rating system that enables quick field assessment of slope failures. Using this tool, we rated the risk of future failures in developed areas of the park. We found 37 slopes along roadways and near infrastructure that showed some risk of failure, with areas that require regular monitoring. Roads and trails critical for visitor access to the park were the most commonly threatened type of infrastructure.
Identifying the Trigger
To better understand the relationship between precipitation and slope failures, I compared the results of my office research to the rainfall records from the climate center’s station. I looked at slope failures and station records over the 19-year period from 2004 to 2023. All known slope failures during that time occurred during or shortly after storms that ranked in the top five percent for intensity, duration, and total rainfall.
Image credit: NPS / Jonathan Malzone. Data source: Western Regional Climate Center
Graph with three axes showing individual storms by their total rainfall in milimeters on the vertical y-axis, their duration in hours on the horizontal x-axis, and rainfall intensity in milimeters per hour on the third dimension z-axis.
The points representing storms are either black or gray, with thin lines descending from each to give a better sense of where it is in the three dimensional space. The black points represent common storms below the 60th percentile (n=1,360). They are tightly clustered in the front, lower left of the graph, indicating they are mostly short storms with lower ranfall rates and totals.
The gray points are more losely clustered further back, further right, and higher up. They are above the 95th percentile (n=118) of all the storms. They have totals sometimes approaching 200mm or rain, durations mostly exceeding 10 hours, and intensities of 5-25mm/hr. Four of the gray points with particularly high total rainfalls and rainfall intensities are circled in red. These are storms that are associated with known instances of slope failures in Delaware Water Gap.
Now that we know what to look for, a simple weather report can give park staff an idea of the risk for catastrophic slope failures.
A precipitation gauge at a single site is not perfectly predictive. But it gives us a reasonable metric for determining which storms require a weather alert and when to anticipate slope failures. Now that we know what to look for, a simple weather report can give park staff an idea of the risk for catastrophic slope failures due to predicted rainfall.
Finding Balance
The Middle Delaware River valley is the center of the Lenape ancestral homelands for four federally recognized Tribes: Delaware Nation, Delaware Tribe, Stockbridge-Munsee Community Band of Mohican Indians, and Shawnee Tribe. The landscapes of the valley and the river itself hold great cultural significance to the Tribes. The park collaborates with the Tribes through consultation and co-stewardship to ensure continued protection and management of traditional cultural resources.
There’s a balance between safety, tribal values, visitor needs, resource protection, and cost that parks strive to achieve. It isn’t always easy. U.S. Route 209 is a major north-south roadway that runs through the Pennsylvania side of the park, roughly parallel to the Middle Delaware National Scenic and Recreational River. It’s a heavily traveled road, connecting communities and serving as a main commuter thoroughfare, emergency access route, and gateway to some of the park's most popular recreational facilities. At one point, the roadway intersects a steep slope of loose shale 100 to 150 feet from the Delaware River. The popular McDade Trail threads through this small space. There’s an eddy in the river at this spot that causes a deep pool to form. This pool is a major historical landmark and a beloved fishing spot.
The road closure caused public frustration and reduced emergency and boater access to the river.
When an intense rainstorm in 2022 removed all but eight feet of stable material between the road and river, the park closed U.S. Route 209 for safety considerations. The road closure caused public frustration and reduced emergency and boater access to the river. Searching for solutions, the park consulted scientists and engineers from the Geological Resources Division, the U.S. Geological Survey, and the Federal Highway Administration. We eventually opened one lane of the road to traffic. A temporary traffic light managed oncoming cars, restoring access to the traveling public and maintaining safe access to the river.
Image credit: NPS / Kris Salapek
My research revealed a history of frequent slope failures in this area in the last century, which means there’s no quick fix for Route 209. “We need to [invest] in sustainable infrastructure improvements that account for projections of hydrologic and ecological changes [to reduce] costs and impacts in the long term,” said Kara Deutsch, the park’s Resource Management and Science Division supervisor. But repairing the slope and reducing the risk of recurring failures will change the area, affecting trail and fishing access, cultural sites, and scenic and natural resources.
To help address and reconcile those potentially conflicting needs, the park partnered with the National Park Service’s Denver Service Center. Together, they developed climate-resilient and cost-effective solutions, such as designing context-sensitive rock slopes, planting deep-rooted native plants, and maintaining winter leaf cover. These could help conserve the McDade Trail, recreational access, tribal interests, and wild and scenic river values. Working with the center, the park is considering a range of options to stabilize the slope and manage water to maintain resources in this narrow corridor.
Recognizing the Risk
Relying only on institutional memory can be costly in the age of climate change. Through our work connecting historical records of slope failures to concurrent precipitation events, the park came to recognize the true risk of slope failures, especially considering rapid climate and ecological change. Now we’re making new partnerships and drawing on the expertise of western parks—long acquainted with the consequences of extreme weather—to find solutions.
Relying only on institutional memory can be costly in the age of climate change.
I consult the hourly rain gauge after every storm and evaluate road conditions more than I ever thought possible as I work to identify signs of imminent slope failure across the park. But our most important tool is the unstable slope management program. It’s been the one that ties all our data together and creates a permanent record for the park—one that’s independent of whether employees remember the last record-breaking downpour.
