Search for: All records

Flood peak magnitudes and frequency estimates are key components of any effective nationwide flood risk management and flood damage abatement program. In this study, we evaluated normalized peak design discharges (Qp) for 1,387 hydrologic unit code 16 to 20 (HUC16-20) watersheds in the White Mountain National Forest (WMNF), New Hampshire and in five Experimental Forest (EF) regions across the United States managed by USDA Forest Service (USDA-FS). Nonstationary regional frequency analysis (RFA) and single site frequency analysis (FA) with long-term high-resolution observed streamflow data along with the deterministic Rational Method (RM) and semi-empirical United States Geological Survey regional regression equation (USGS-RRE) were used. Additionally, a hydrologic vulnerability assessment was performed for 194 road culverts as a result of extreme precipitation-induced flooding on gauged and ungauged watersheds in the Hubbard Brook EF (HBR) within the WMNF. The RM outperformed the USGS-RRE in predicting Qp in the gauged and ungauged HUC16-20 watersheds of WMNF and in three other small, high-relief forest headwater watersheds—Coweeta Hydrologic Lab EF’s watershed-14, and watershed-27 in North Carolina and HJ Andrews EF’s watershed 8 in Oregon. However, the USGS-RRE performed better for larger watersheds, such as the Fraser EF’s St. Louis watershed in Colorado and the Santee EF’s watershed 80 in South Carolina. About 31 %, 26 %, and 56 % of the culverts at the HBR site could not accommodate the 100-yr Qp estimated by RFA, RM and USGS-RRE, respectively. Based on the chosen RIs and techniques, it is determined that except for one culvert with diameter = 0.91 m (36 in.), none of the culverts with diameter of 0.75 m (30 in.) or larger are hydrologically vulnerable. Our results suggest that the observation based RFA works best where multiple gauges are available to extrapolate information for ungauged watersheds, otherwise, RM is best-suited for smaller headwater watersheds and USGS-RRE for larger watersheds. Results from the hydrologic vulnerability analysis revealed that replacing undersized culverts with new culverts of diameter ≥ 0.75-m will improve flood resiliency, provided that the structure is geomorphologically safe (with minimal effects of debris flow, erosion, and sedimentation) and allows for both bank-full discharge and necessary fish passage within that design limit. This study has implications in managing road culverts and crossings at Forest Service and other forested lands for their resiliency to extreme precipitation and flooding hazards induced by climate change. 

Many currently forested areas in the southern Appalachians were harvested in the early 1900s and cleared for agriculture or pasture, but have since been abandoned and reverted to forest (old-field succession). Land-use and land-cover changes such as these may have altered the timing and quantity of water yield (Q). We examined 80 years of streamflow and vegetation data in an experimental watershed that underwent forest–grass–forest conversion (i.e., old-field succession treatment). We hypothesized that changes in forest species composition and water use would largely explain long-term changes in Q. Aboveground biomass was comparable among watersheds before the treatment (208.3 Mg ha⁻¹), and again after 45 years of forest regeneration (217.9 Mg ha⁻¹). However, management practices in the treatment watershed altered resulting species composition compared to the reference watershed. Evapotranspiration (ET) and Q in the treatment watershed recovered to pretreatment levels after 9 years of abandonment, then Q became less (averaging 5.4 % less) and ET more (averaging 4.5 % more) than expected after the 10th year up to the present day. We demonstrate that the decline in Q and corresponding increase in ET could be explained by the shift in major forest species from predominantly Quercus and Carya before treatment to predominantly Liriodendron and Acer through old-field succession. The annual change in Q can be attributed to changes in seasonal Q. The greatest management effect on monthly Q occurred during the wettest (i.e., above median Q) growing-season months, when Q was significantly lower than expected. In the dormant season, monthly Q was higher than expected during the wettest months.