H13F-0471 1340h
Appalachian Flood Hydrology in Headwater Watersheds: Hillslope Scale Studies at the Fernow Experimental Forest
Streamflow and rainfall are combined with observations from a network of 240 crest piezometers on two headwater watersheds (0.30 and 0.14 km$^{2}$) at the Fernow Experimental Forest near Parsons, WV to examine the storm event response of forested, Appalachian watersheds. Weekly data from each piezometer, 5-minute data from over 20 piezometers, and continuous rainfall and streamflow measurements from 2003 and 2004 are used in these analyses. Watershed and piezometer response is examined in the context of antecedent conditions, rainfall accumulations, and rain rates. Piezometer nests (piezometers depths of 25, 50, and 100 cm) allow observations of the relative timing of saturation and piezometric head at different depths in the soil profile, allowing us to look for perched water tables and to distinguish saturation excess and infiltration excess runoff. Observations show pronounced heterogeneity even within the unchannelized swales of headwater watersheds. Piezometer observations yield insight into the dominate runoff production mechanisms for flood response and help to identify signatures in the watershed hydrographs due to different runoff production mechanisms. These hydrograph signatures will be used to help examine differences in watershed response for times when piezometers observations are not available, such as after historical logging of watersheds in the Fernow Experimental Forest.
http://www.princeton.edu/~nsbates/work/fernow.html
H13F-0472 1340h
Hydraulics of a Catastrophic Flood in a Small Central Appalachian Watershed
An orographic thunderstorm on 9 August 2003 produced catastrophic flooding in the 2.1 square kilometer Saul's Run watershed in the Valley and Ridge physiographic province of West Virginia. Hydraulic studies of the Saul's Run flood are based on a detailed survey of high-water marks (HWM) and analyses using a 2-D, depth-averaged unsteady flow model (Telemac 2-D). Estimated peak discharge of the 9 August 2003 Saul's Run flood at 1 sq. km. spatial scale was approximately 20 cms, resulting in a unit discharge peak of 20 cms per sq. km. Peak discharge analyses in small, high-gradient streams, like Saul's Run, are complicated by variability in the free water surface at peak discharge in both longitudinal and cross-sectional directions. Model analyses and analyses of HWM are used to characterize the longitudinal variation of 2-D flow features of the Saul's Run flood and develop procedures for rapid response monitoring of extreme floods. Model analyses are also used to examine the longitudinal variation in shear stress and unit stream power and their relation to fluvial impacts of the flood.
H13F-0473 1340h
The Role of Debris Flows in Long-term Denudation and Landscape Evolution in the central Appalachians
Four major storms spanning a 46 year period from 1949 to 1995 that triggered debris flows in the Virginia-West Virginia Appalachians provided new insights into the role of high-magnitude, low-frequency storm events in long-term denudation and landscape evolution in mountainous terrain. Storm denudation measured in five Blue Ridge Mountain drainage basins (mean=3.7 cm) was approximately an order of magnitude greater compared to four basins located in the mountains of the Valley and Ridge province (mean=0.2cm). This difference is probably the result of higher storm rainfall from the Blue Ridge storms. Long-term (10$^{3}$ yrs) denudation rates were estimated using several lines of evidence, including 1) studies of the volume of sediment deposited in, or offshore of, the Atlantic Coastal Plain; 2) findings of parallel rates of continental uplift and denudation; and 3) historic sediment-load data. Using these estimates and subtracting the denudation attributed to chemical load, the mechanical denudation rate of the central Blue Ridge is approximated as 2.4 cm/k.y. Whereas debris flows recur at a frequency of approximately one event each three years somewhere in the unglaciated terrain of the Appalachians, the return interval is much greater when only individual mountainous basins are considered. Radiocarbon dating of debris-flow deposits in mountainous first- and second-order river basins of the Blue Ridge indicates a debris-flow return interval of not more than 2000 to 4000 yr. These data on debris flow frequency, combined with measurements of storm-induced upland basin denudation, suggests that approximately half of the long-term denudation from mechanical load occurs episodically by debris-flows. Although floods of moderate magnitude are largely responsible for mobilizing sediment in low-gradient streams, the data suggest that high-magnitude, low-frequency events are the most significant component in delivering coarse-grained colluvium from mountainous hollows and channels to the lowland floodplains. In the Appalachians, and probably other mountainous terrains located in humid-temperate climates, the role of high-magnitude events on geomorphic effectiveness and landscape evolution arguably has been underestimated. The presence of coarse bedload stored in upland channels, porous regolith that mantles the slopes, and densely-vegetated terrain marginalizes the effectiveness of frequent, low magnitude storms in mobilizing sediment. In contrast, high magnitude events trigger debris flows, which incise streams, export sediment from the uplands, and deposit colluvium onto debris fans or into lowland stream channels and floodplains. In the Blue Ridge, numerous upland channels impacted by debris flows in the study areas have been slow to recover; and they continue to maintain a greater hydraulic geometry than required for frequent, low magnitude storms. Throughout much of the Appalachians, the ubiquity of specific landforms and deposits; including debris fans and levees, boulder bars and terraces, remarkably wide alluvial valleys that originate at the terminus of debris fans, and single-channel floodplains that become braided during catastrophic flooding all suggest that geomorphic work and effectiveness in mountainous terrain is achieved largely by high-magnitude, infrequent events.
H13F-0474 1340h
Vulnerability to Episodic Acidification of Streams in Shenandoah National Park
Characterizing magnitude, frequency, and duration of acid-neutralization capacity (ANC) depression in streams requires a time series of ANC values at a time scale appropriate to the resources at risk from ANC depression. For small catchments, ANC excursions on the order of hours may be critical. However, hourly observations of ANC are prohibitively expensive. On the other hand, hourly observations of discharge are relatively inexpensive. We develop transfer function time series models with outlier correction to predict hourly ANC from discharge for five catchments in Shenandoah National Park, Virginia. These models outperform simple regression. The catchments range in size, are situated on different geologies and exhibit other physiographic differences. We use the predicted hourly ANC time series to calculate annual low-ANC values for various time spans: 2, 6, 12, 24, 72, 96, and 168 hours. We then fit Pearson type III distributions to the annual low-ANC values and extrapolate those distributions to other catchments in Shenandoah National Park where Park staff monitor fish populations. This allows us to generate maps of, for example, the 72-hour, 4-year low-ANC frequency for the catchments considered.
H13F-0475 1340h
Are Brook Trout Streams in Western Virginia and Shenandoah National Park Recovering from Acidification?
Stream water composition data obtained through periodic sampling of small mountain headwater streams in Shenandoah National Park (SNP) and western Virginia were examined for evidence of change in acid-base status. Decreasing concentrations of sulfate and increasing concentrations of acid-neutralizing capacity (ANC) in stream waters of SNP during the 1988-2001 period are consistent with recovery from acidification, whereas changes in the concentration of these constituents in stream waters of the larger western Virginia region during the same period are consistent with continued acidification. Although the observed changes in the composition of streams in SNP represent an improvement over acidification trends observed in the 1980s, the degree of apparent recovery is relatively minor in relation to recovery observed in other regions, in relation to estimated historic stream acidification in SNP, and in relation to potential biological significance. In addition, a transient increase in concentrations and export of nitrate in SNP stream waters following forest defoliation by the gypsy moth complicates interpretation of trends in acid-base status and may account for much of the apparent recovery from acidification. An overview of regional trends in ANC and sulfate concentrations suggests the presence of a north-south gradient in surface water recovery from acidification, consistent with expectations based on regional differences in watershed retention of sulfur.