Extreme Precipitation Intensity-Duration-Frequency and Design-Discharge Estimates for Forest Road-Stream Crossings Design and Ecological Applications
On this page:
Developing Precipitation-Intensity-Duration-Frequency (Onsite-PIDF) and design discharge estimates using onsite long-term datasets to accommodate ongoing effects of climate change, is crucial in the design and planning of climate-resilient forest road-stream crossing structures (RSCS) and preserving resilient forest ecosystems. Despite these resources, the RSCS on US National Forest lands and other landscapes of a similar nature are frequently designed using design discharge rates (Qp) estimated using USGS Regional Regression Equations and sub-hourly and sub-daily precipitation intensities (PIs) provided by the National Oceanic and Atmospheric Administration Atlas 14 (NOAA-Atlas14). Both the NOAA-Atlas14 PI and USGS peak flow estimates have some limitations in establishing valid Qp in small, wooded watersheds due to non-linearity in natural systems. We hypothesized that recently available continuous up-to-date, long-term, and fine resolution precipitation and streamflow dataset from onsite gauging stations at the United States Department of Agriculture Forest Service's (USDAFS) Experimental Forests (EF) can be better alternatives to reliably estimate the peak intensities for the design of the RSCS. In this study we estimated onsite-PIDFs and Qp by employing Generalized Extreme Value (GEV) distribution fit to Annual Maximum Series (AMS) with high temporal resolution data (5-min to 1- hour freaquency) at six USDAFS EFs locations: Horace Justin Andrews, Oregon; Santee, South Carolina; Coweeta Hydrologic Laboratory, North Carolina; Alum Creek, Arkansas; Fraser, Colorado; and Hubbard Brook, New Hampshire. Regional frequency analysis was performed where data from multiple gauges was available. Onsite-PIDFs and Qp for various frequencies were then evaluated against NOAA-Atlas14, and USGS Regional Regression Equations (RRE) based design discharge estimates, respectively. Additionally, the model was evaluated using observed flood frequencies, and inter-model uncertainties. The Qp were also assessed using the onsite precipitation intensities for a durations equivalent to the watershed time of concentration derived in the Rational Method (RM), and the USGS RRE, using the Log-Pearson Type III and GEV distributions. Finally, the estimated Qp at various frequencies, and the USDAFS's RSCS database, were used to identify culverts and their carrying capacity, maps of the current road culverts identified as vulnerable to historical climate change based on the design discharge rates associated with 25-, 50-, 100-, and 200-yr flood occurrences. These findings including results from alternative modeling approaches are helpful for evaluating hydrologic risks in the RSCS design and restoration while maintaining their resiliency in small, wooded watersheds facing increasing extreme precipitation-based storm events.