Ground sensors, satellite radar, and repeat photographs are recording measurable shifts at six national parks faster than the infrastructure built on them was designed to handle. At Yellowstone, the caldera subsided at roughly 2 to 3 centimeters per year through 2024 while new uplift appeared north of the Norris Geyser Basin starting in July 2025. In Alaska, the Pretty Rocks landslide has been accelerating along the only road into Denali National Park, driven by permafrost thaw that is undermining the slope beneath it. Across the system, geologists are watching arches crack, glaciers shrink, desert basins flood, and volcanic summits collapse on timescales that challenge the assumption parks are static scenery.
Permafrost and hydrothermal forces are reshaping park roads first
The parks changing fastest share a pattern: the forces doing the work are underground and invisible until a road buckles or a trail drops away. At Denali, the Pretty Rocks landslide has been moving downhill along a stretch of the park road at mile 45.4, fed by thawing permafrost that turns formerly frozen soil into a sliding mass. The National Park Service has documented accelerating movement and seasonal closures, and a separate NPS climate-adaptation case study outlines engineering options including a potential reroute or bridge to bypass the unstable zone entirely. Because the park road is the sole vehicle corridor into the interior, any permanent loss of alignment would cut visitor access to the most popular wildlife-viewing areas.
Yellowstone presents a different underground driver with a similar surface result. The annual observatory report documented caldera-wide subsidence of roughly 2 to 3 centimeters per year, a rate tracked by continuous GPS stations and synthetic aperture radar. At the same time, the area north of the caldera rim has been experiencing what the USGS calls the Norris uplift anomaly, a localized zone where the ground is rising rather than sinking. An InSAR-derived map covering October 7, 2024 through October 2, 2025 shows that uplift in this zone began around July 2025. Roads crossing the Norris area sit directly above these competing deformation signals, and maintenance crews already contend with thermal damage to pavement from hydrothermal features beneath the surface.
A testable question follows from both parks: if the primary change driver is permafrost thaw or hydrothermal deformation, those parks should show the earliest permanent loss of current road or trail alignments. Comparing 2024 through 2027 maintenance logs against baseline deformation rates at Denali and Yellowstone would reveal whether infrastructure failures are accelerating in step with the geologic data.
Glaciers, arches, and flood basins registering visible loss
The other four parks on geologists’ watch lists are changing through processes that leave more obvious surface evidence, though the pace still outstrips what most visitors expect. At Glacier National Park, the USGS repeat photography project has been documenting named glaciers for decades, and updated photo pairs now include 2025 images for Grinnell and Swiftcurrent. Side-by-side comparisons show ice retreating visibly between annual frames. In central Alaska, NPS Inventory and Monitoring staff conducted mass-balance fieldwork in 2024 on Kahiltna Glacier in Denali and Kennicott Glacier in Wrangell-St. Elias, using standard glaciological benchmarks developed in cooperation with the USGS Alaska Science Center. Updated results from those surveys have not yet been published beyond the 2024 field season.
At Arches National Park in Utah, the forces are subtler but no less consequential. The NPS initiated monitoring of select arches in 2013 and installed crackmeters on specific formations in 2015 to detect incremental widening that precedes rockfalls. The instruments record changes too small for a visitor to notice but large enough to signal structural fatigue over years. No refreshed crack-width data or rockfall probability figures have been released publicly since the instruments were installed, leaving a gap in the public record.
Death Valley offers the most dramatic recent example of rapid surface change. Hurricane Hilary struck the park from August 19 through 21, 2023, scouring channels and reworking alluvial fans across the basin. Floodwaters filled Badwater Basin, the lowest point in North America, creating a temporary lake visible in NASA Earth Observatory satellite imagery. The NPS documented biological and physical effects including impacts on Salt Creek. That single storm reshaped terrain that had been stable for years, illustrating how a desert park can change overnight when extreme precipitation arrives.
At Hawaiʻi Volcanoes National Park, the 2018 summit caldera collapse at Kīlauea lowered the crater floor by hundreds of meters in a matter of months. The sequence followed withdrawal of magma from the summit reservoir and a series of major deflation events, culminating in repeated collapse episodes that enlarged Halemaʻumaʻu crater and damaged roads, trails, and viewing areas along the rim. Park managers closed large sections of the summit district for safety while USGS scientists installed new instruments and re-leveled existing benchmarks to track ongoing deformation. Although surface activity at Kīlauea has fluctuated since, the 2018 collapse remains a benchmark example of how volcanic systems can rapidly reconfigure park landscapes and infrastructure.
Tracking motion from space and on the ground
What links these six parks is not just the pace of change but the way it is being measured. Satellite-based interferometric synthetic aperture radar (InSAR) can detect ground motion of millimeters over wide areas, allowing scientists to map uplift and subsidence patterns that would otherwise be invisible. At Yellowstone, InSAR time series underpin both the caldera subsidence estimates and the recognition of the Norris uplift zone. Continuous GPS stations anchored in bedrock add a second, independent line of evidence, recording vertical and horizontal motions in near real time. Those data streams are distilled into maps, plots, and explanations that the public can access through USGS Yellowstone volcano updates.
On the ground, repeat photography provides a simple but powerful tool for communicating change. At Glacier, scientists and photographers return to fixed camera points to capture modern images that can be precisely aligned with historical photos taken decades earlier. The resulting pairs and time-lapse sequences make ice loss tangible for non-specialists, turning abstract mass-balance numbers into visible retreat lines on valley walls. Similar approaches are being used informally at Death Valley and Hawaiʻi Volcanoes, where before-and-after images of flood-scoured roads or collapsed crater rims help convey the scale of single events.
Instrumentation at Arches and Denali is more specialized but serves the same purpose. Crackmeters on sandstone spans translate minute expansions into digital records that can be analyzed for seasonal cycles, long-term trends, or sudden jumps that might precede failure. In permafrost terrain, borehole thermistors and inclinometers track both ground temperature and slope movement, tying surface slumps at Pretty Rocks back to thawing layers at depth. Together, these tools give managers an early warning system for hazards that could threaten visitors or cut off key corridors.
Implications for park management and visitors
The accumulating evidence from these six parks points toward a future in which maps, guidebooks, and even long-standing assumptions about access routes will need regular revision. If Denali ultimately reroutes or bridges around Pretty Rocks, the iconic bus journey into the park interior will change, along with how wildlife viewing is managed. At Yellowstone, continued deformation near Norris could force more frequent road repairs or realignments, especially where hydrothermal features intersect pavement. In glacier parks, shrinking ice will alter streamflows, habitat, and the timing of peak visitation as classic viewsheds evolve.
Managers are already experimenting with adaptation strategies. These range from relocating trail segments away from unstable slopes and closing high-risk overlooks to redesigning culverts and bridges to handle more intense floods like those seen in Death Valley. Communication is another critical piece: interpretive programs increasingly emphasize that parks are dynamic, not frozen in time, and that some beloved features may change or disappear within a single generation.
For visitors, the message is twofold. First, safety guidance is likely to become more variable, with short-notice closures in response to landslides, rockfalls, or volcanic unrest. Checking current conditions before a trip will matter more than relying on past experience. Second, the same forces that complicate access also create rare opportunities to witness geologic change in action – from new lakes in desert basins to reshaped craters at active volcanoes. The challenge for the National Park Service and its scientific partners will be to balance that sense of awe with a clear-eyed assessment of risk as the ground beneath these iconic landscapes continues to move.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
