New lidar imagery shows the full extent of Helene’s impacts

by Philip S. Prince

Over the last year, numerous collections of aerial imagery have shown Helene’s effects on western North Carolina mountainsides. This imagery has been useful in understanding the extent of the storm’s impacts on the landscape, but remaining tree cover and the natural irregularity of the landscape make it difficult to fully appreciate changes created by landsliding and flooding. Recently, long-awaited post-Helene lidar data has become available. Processing that data into imagery illustrating land surface by color coding loss and gain fully captures the storm’s effects. ALC Principal geologist Stephen Fuemmeler has produced county-wide change maps for Buncombe and Watauga Counties, and the results speak for themselves. The upper image below from the Big Ivy area of Buncombe County, North Carolina, shows Google Earth aerial imagery of debris flow landslides and tree blowdown. The lower image shows the same view with color-coded lidar surface change, highlighting debris flow scour and deposition as well as smaller landslides and the “divots” left by fallen trees.

The lidar change imagery is a basic point-by-point elevation comparison from data before and after the storm. Areas lower after the storm are indicated by yellows to oranges (brown if really extreme); these are areas from which material was removed by landslide processes or water erosion. Blues highlight areas where the land is now higher from transported soil and rock debris piling up. The entire landscape is often faintly tinted due to miniscule mismatches in positioning pre- and post-data, so the combined loss (scour or down-dropping) and gain (piling up) pattern, the shape and context of identified changes, and the raw appearance of the new lidar imagery are all important for interpretation. The following GIFs give an idea of what the change patterns look like for a couple of landslide styles, starting with debris flow landslides that scour from an initiating slide and produce a deposit downslope.

The lightest yellows indicating minor change aren’t portrayed in the diagram to keep things simple. The basic pattern is a long, red-orange track where colluvial soil is now missing from being scoured away as it fluidized within the debris flow. The scoured soil ends up in a big, blue deposit where it piled up at the bottom of the slope. This pattern is nicely visible in the first change image in the post. A tremendous number of Helene’s slides fluidized due to soil saturation but didn’t scour the slope. These “blowout” slides were incredibly numerous, and have been of great interest to ALC over years of landslide mapping because they appear to associate with the most extreme rainfall events. A blowout change pattern is illustrated in the following GIF, which also includes a scouring debris flow (image center) for comparison.

Because blowouts don’t scour the slope, they appear as closed red or orange shapes, with the colored loss area produced by the scar on the slope. Many have a blue deposit pile somewhere downslope, but others don’t, suggesting the liquefied soil spread too thin for detection or reached a channel to be transported by floodwater. Fieldwork suggests that some blowout deposit material accumulated where transported woody debris lodged on standing tree trunks and began to trap soil.

Both of the conceptual models above are well represented by the Vanderpool debris flow and surrounding slopes in Watauga County, west of Boone.

In addition to the scoured track of the debris flow and its substantial deposit, a number of small blowouts are visible. Construction changes on the lower slopes indicate work done since 2017 and are not storm-related. The yellow arrow points out a small translational slide…more on this later. While the debris flow shown here was a monster that impacted downslope properties, blowouts were the big story in Watauga County, just as they were in 1940. They dot the image of the Sugar Grove area below. A small amount of deposit appears visible well below the red scars, reflecting the extreme mobility of the fluidized soil as captured on video in the area.

Blowout-style landslides are all over the Buncombe and Watauga landscape (the only counties I have reviewed so far), and many more have been observed in other impacted counties during fieldwork. Because blowouts don’t scour a track down the slope, they can be hard to detect from aerial imagery. The following example from the headwaters of Garren Creek in Buncombe County shows how well these slides hide in the remaining forest cover. The scar is an impressive 60 feet wide. It is faintly visible in the aerial imagery once you know it’s there. This failure would have sent an impressive wave of liquefied soil, along with the few trees sitting on the failed area, downslope at high speed. No deposit, with the exception of a couple of small lumps, is visible, indicating the fluidized soil made it to the stream below (out of the frame of the image). The “vanished” material also confirms the scar’s landslide origin. No spoil is present despite the large missing volume, and no well-traveled access path (or road!) is present to mobilize machinery or personnel onto the steep slope.

Debris flows and blowouts weren’t the only slope movements triggered by Helene (remember the yellow arrow in the Vanderpool example?). Movement of intact landslide masses occurred in many locations. Due to their limited displacement, these slide movements are difficult to detect without the aid of lidar change analysis….unless one is in your backyard. Intact slides produce their own pattern of surface elevation loss and gain, with the downslope toe of the slide uplifting to various degrees as upper portions of the slide drop down along headscarps and internal scarps.

Many of these larger, slightly moving slides were reactivations of older slide features developed in huge colluvial accumulations around the flanks of steep slopes. The following images from the south side of Watch Knob, west of Swannanoa at the mouth of Bee Tree Creek, show a couple of large, intact slides. The GIF compares 2017 (pre-storm) lidar, 2025 post-storm lidar, and lidar change detection. Headscarps locations are indicated by yellow arrows; the headscarps appear as the second panel fades in. Note construction-related change at lower left, below one of the big slides.

In addition to smaller landscape details, the physical scale of the larger debris flows triggered by Helene is captured by the lidar change products. The image below shows part of the eastern prong of the Craigtown debris flow complex. The combined total length of the orange scoured tracks approaches 1.75 miles, with consistent scour depths of up to 4 feet along that entire length. Intermittent scour (with limited local deposition) continued to homes that were struck at the base of the mountain. The sheer volume of material that moved here during the three pulses over 15 minutes is staggering…and this image only shows part of it.

Lidar change imagery is also useful for understanding how landslide material moves in a storm like Helene. Some fluidized slide debris, like that from Craigtown or many of the blowouts, is incredibly mobile on both open slopes and in headwater channels. Other debris flows show less scour and mobility, despite developing from substantial initial landslides. The slide below occurred in the Big Ivy community. The upslope scar is 92 feet wide and exceeds 8 feet in depth at its center. Very limited slope erosion occurred as the initial slide debris traveled downslope, and impressive piles of deposited material developed where the landscape flattened slightly atop older, accumulated slide deposits. Whether the landscape or soil/rock characteristics reduced runout of this debris flow is unknown, but could possibly be determined through modeling and field study.

The debris flows below, from Buncombe County just west of Shope Creek, showed similarly short runouts, possibly due to topography and geology of the local ridge-forming rock from which colluvial soil is derived. The image uses the same scale change as above.

In addition to informing future modeling and hazard assessment, lidar change imagery also highlights areas that may need monitoring in future storms. The Watauga County slides at the center of the image below deposited material in a steep topographic draw, where it may be prone to saturate and fail again or be entrained by a future debris flow. Numerous examples of slide material ended up in potentially unstable locations can be identified easily, and these areas may need remediation attention.

Several geologists could devote the rest of their careers to this new lidar data and never run out of things to learn. You’ll see change results from Buncombe, Watauga, and other counties in posts on this page, and we’ll also be using them to inform fieldwork and planning in the coming months and years. As for the currently estimated 2,000 + landslides from Helene, it’s definitely on the “+” side of that…by many, many times!

What happened to mountain slopes during Helene? Geologists are gathering information to prepare for the next storm

The Toodies Creek debris flow in Yancey County tragically claimed the lives of residents living along the small stream which the debris flow followed.

In the nine months since Helene’s arrival in western North Carolina, geologists have worked steadily to better understand how to reduce future landslide-related impacts on life and property. While landslide themselves cannot be prevented from happening under extreme precipitation conditions, decision making during, and particularly before, a storm can save lives and reduce damage to infrastructure and personal property. Understanding what made certain landslides more damaging than others requires both extensive fieldwork and study of remote sensing data, like lidar imagery. ALC Principal Geologist Jennifer Bauer and Project Geologist Philip Prince recently presented some of the findings of their post-Helene work in US Geological Survey seminars. Video recordings of the talks are linked below. If you have ever wondered what a geologist sees in one of Helene’s thousands of landslide scars, these videos will give a glance of how we do our work day-to-day.

Jennifer Bauer

Jennifer’s talk focuses on the use of landslide mapping and modeling to understand and (more importantly) communicate landslide hazard before storms hit. Understanding landslide potential in a given landscape requires that geologists understand the landslide history of a landscape. Landslide inventories involving both lidar imagery analysis and lots of boots-on-the-ground fieldwork help geologists learn what has happened in the past.

Helene certainly was not the first landslide-producing storm in the western North Carolina Blue Ridge. Geologists study lidar imagery (left) and field-verify observations to determine where past slope failures occurred. This information illustrates the type of topography that may produce future slide events.

Once the geologic details that can produce landslides are understood (slope shape, slope steepness, soil type, etc.), models of potential landslide hazard zones can be developed. As Jennifer’s talk shows, the overwhelming majority of Helene’s landslides came from mapped hazard areas, but not every hazard area produced a landslide…this time. Hazard mapping can show mountain residents areas that are potentially dangerous is storms (don’t worry; it’s not everywhere-not even close) and help with decision making regarding where to live and how to prepare for the next big event.

Improving understanding of where debris flows will go as they move downslope–and how wide they may be–will be a focal point of landslide research in coming years. The teal color shows existing modeling of a debris flow path; the yellow outline shows the actual extent of the affected area.

Philip Prince

Philip’s talk is centered around the geologic details of Helene’s debris flow landslides. Debris flows are fast-moving, fluidized landslides that can travel long distances very quickly. Often called “mudslides,” debris flows actually carry huge amounts of rock and boulder debris and tremendous numbers of trees, so a debris flow impact is much more damaging than what might result from mud alone.

A debris flow conceptual model (top) attempts to illustrate the thin soil “birthplace” of many western North Carolina debris flows. Starting as small landslides which fluidize due to soil saturation, debris flows pick up more and more material moving downslope. The lower photography shows what a debris flow starting point (initiation zone) looks like in the field. Note the smooth bedrock surface exposed by the initial slide.

Philip illustrates where debris flows start in the landscape and how they accumulate so much material on their path downslope. A large debris flow could cover a football field with a few feet of mud, rocks and trees, but even small debris flows are surprisingly destructive. By understanding what type of geologic materials and slope settings produced debris flows during Helene, we can better understand what areas may be hazardous in the next event. Planners can also what parts of the landscape may be more susceptible to landslides when disturbed for building, as well as what areas at the foot of the mountains might be reached by debris flows.

This soil consisting of stacked, shingled rock fragments doesn’t look like much, but it is actually the material which slid to initiate one of the deadly Craigtown debris flows in Fairview, North Carolina. Understanding the distribution and behavior of soils like this is critical to improving understanding of potential debris flow behavior.