Hurricane Helene Case Study: Hydrology Analyses for Virginia Creeper Trail Reconstruction
Powerful remnants of Hurricane Helene impacted the Southern Appalachians in September 2024, with up to ~24 inches of rain inducing large-scale flooding. Up to ~12 inches of rain led to the Virginia Creeper Trail, a popular rails-to-trail, being heavily impacted by flooding along Whitetop Laurel and Green Cove Creeks. Nineteen bridges were severely damaged or destroyed (Figure 1) and portions of the embankment eroded and nonfunctional (Figure 2). Flood magnitudes were extreme at some locations.
The Flood Potential Portal was utilized to understand how floods vary in space and time in this area as quantified using streamgage data; this tool is detailed in River Research and Applications. U.S. Geological Survey streamgage data indicate that this region on the northern edge of the flood event experienced large and notable peak discharges from the remnants of Hurricane Helene, but was generally not unprecedented. However, streamgage data for the S.F. Holston River, downstream of Whitetop Laurel Creek, indicates that watersheds on the leeward side of the Blue Ridge Mountains (from hurricane remnants that track along the axis of the Appalachians) experienced more exceptional flooding. It is possible that flooding characteristics of this northwest edge of the Blue Ridge Mountains are shifting. Revised flood-frequency relationships were developed using the Flood Potential Portal Streamgage Analysis module, using data through water year 2024 that included Helene peak discharges. This streamgage analysis tool provides enhanced functionality compared to other applications.
Given the absence of streamgage data for Whitetop Laurel Creek, high water marks from the U.S. Geological Survey were utilized to estimate peak discharges. These preliminary data indicate that Hurricane Helene induced flooding that was unexpectedly large, exceeding the size of floods that are expected and most infrastructure is designed to safely pass. In some places, the flooding was extreme, as systematically identified using the Flood Potential Method. Flood extreme index (Ef) values were high in the Whitetop Laurel Creek watershed; this index is defined as Ef = Q/Qefp (Q = peak discharge; Qefp = expected flood potential discharge, the central tendency of record peak discharges across zones of similar flood response). Ef values of up to 2.07 were experienced, meaning that flooding occurred that was more than twice the ~100-year scale (1% annual exceedance probability) magnitude typically used for infrastructure design and floodplain management.
Using a preliminary (unpublished) version of the six-level Flood Severity Scale, Whitetop Laurel Creek experienced flooding at the Fl-4 and Fl-5 severity levels. This scale was developed using more than 8200 streamgages across the contiguous United States, for a total of more than 367,000 data points. Descriptions of the Flood Severity Scale levels are:
- Fl-1: Small to moderate magnitude flooding, with consistent out-of-channel flow (> 2 year return interval / 50% probability of annual exceedance). 0.210 ≤ Ef < 0.538.
- Fl-2: Moderate to large magnitude flooding. 0.538 ≤ Ef < 0.883.
- Fl-3: Design flood / base flood discharge event (~100-year return interval / ~1% annual exceedance probability). 0.883 ≤ Ef < 1.27.
- Fl-4: Exceeds design and base flood discharge, with infrastructure damage, extensive floodplain inundation, and some extreme magnitudes. 1.27 ≤ Ef < 1.73.
- Fl-5: Major flood, with extreme magnitudes and catastrophic consequences. 1.73 ≤ Ef < 2.67.
- Fl-6: Extreme flooding, with extensive catastrophic consequences. 2.67 ≤ Ef.
Helene-induced flooding was more severe to the south, with the Fl-6 level experienced across a wide area in the vicinity of Asheville, North Carolina. This can be visualized through an animation within a preliminary version of the Flood Status Portal, illustrating flood severity increasing from Fl-2 to Fl-6 as the hurricane remnants moved across the Appalachians.
Reviewing this portion of the Southern Appalachians using the Flood Potential Portal Mapping module reveals an area highly susceptible to flooding from hurricanes. The most substantial recorded event was induced from a hurricane in August of 1940. Previously, a powerful flood event was induced from hurricane remnants in July of 1916. Floods have inherently large magnitudes in this area, at the 78th percentile compared to all of the United States. On average floods are twice as large in this part of the Blue Ridge than in the neighboring Valley and Ridge physiographic province. However, portions of the Blue Ridge to the South, between Ashville and Boone, North Carolina, experience floods 3 times larger, on average; this area was dramatically impacted by Helene flooding.
Design flood and base flood discharges are the size of large floods we can expect, and should plan and design for. The use of multiple methods for these predictions is a best practice, to avoid systematic bias that may be present in any single prediction method. Recommendations for design flood discharges at the 100-year scale were quantified for the Virginia Creeper Trail reconstruction based on analyses performed in the Flood Potential Portal Watershed Analysis module. This decision support system provides predictions of peak flood discharges using three approaches: (a) the Flood Potential Method, including trend reporting for flood magnitudes, flood frequency, and flashiness; (b) Index flood-frequency method; and (c) USGS regional regression flood-frequency equations, as computed in StreamStats. Potential prediction bias was assessed though comparison with flood-frequency relationships at nearby streamgages, with the median Watershed Analysis module peak discharges recommended for utilization.
The full report — Virginia Creeper Trail Flooding from Hurricane Helene: Hydrology Analyses for Reconstruction — is available here.
Steven E. Yochum, PhD, PE, PH is a hydrologist and civil engineer with the US Forest Service.
Acknowledgements: Tyler Wible and Zack Mondry are greatly appreciated for their comments and suggestions for improving this article and the original report.
