Print PageBioenergy crop production raises legitimate concerns for alterations in rainfall-runoff relationships and downstream water quality. Bioenergy crop production will occur mostly on marginal agricultural or existing forestry lands and likely require incremental increases in fertilizer and herbicide applications as well as mechanical harvest frequency to achieve desired production goals. These marginal lands often feature greater slope, shallower soils, and poor nutrient retention. Thus, increased sediment, nutrient and pesticide discharges are likely. Process-based models are needed to estimate regional effects based on the results from small watershed scale or even plot scale treatment studies. Such models need to account for known mechanisms of runoff production and allow for a dynamic matrix of relevant land management decisions. Practical models must be less complex and less computationally intensive than fully distributed pixel-based models, but they need to incorporate more sophisticated hillslope process representation than does the commonly-used SWAT model. Development, calibration, and validation of such models require data on internal hillslope state variables and fluxes in addition to streamflow data. The inadequacy of streamflow data to constrain model parameterization is well known. We hypothesize that practical process-based watershed models can be calibrated and constrained with watershed studies that include time series measurements of soil moisture, groundwater, variable source areas, and interflow. We have instrumented three such watersheds at the Savannah River Site, and we are planning to implement intensive silviculture for bioenergy crop production in the fall of 2011. These watersheds feature gentle topography, loamy sand topsoils over an argillic horizon, and broad low gradient valleys typical of timber production areas in the Upper Coastal Plain of the Southeast.
Results/Conclusions
Soil and hydrologic measurements in these small Upper Coastal Plain watersheds indicate high variability in topsoil depth, soil hydraulic properties, and the spatial locations of perching and interflow movement. Perching of shallow groundwater within and above the argillic layer is common in these watersheds. However, significant interflow events are rare, occurring only a few times per year during large events. Streamflow production is dominated by slow groundwater flow paths. Overland flow under forest canopies has not been observed. The Oregon State catchment model has been found to represent these processes well, and a modified HSPF model including a riparian aquifer representation is being developed for application to bioenergy production issues.
