Short-term hydrological responses of a forested hillslope during rainstorms, at Panola Mountain Research Watershed, Georgia, USA
Ratcliffe, Elizabeth B.
PublisherUniversity of Bristol
MetadataShow full item record
A small-scale filed experiment was conducted on a hillslope plot within the Georgia Piedmont, USA, with the aim of elucidating the hydrological processes which generate storm runoff and its chemistry. Intensive hydrometric and chemical sampling enabled the collection of detailed observations of hillslope processes during rainstorms. The passage of water was traced through a one-dimensional profile in the hillslope, where rainfall, throughfall, forest floor soil water, soil water at 15, 40, 50 and 70 cm depths, groundwaters and streamwaters were monitored, either manually or automatically. Chemical samples for each water type were also collected. From analysis of hydrometric data, several hydrological flowpaths were detected that contribute water to storm runoff. Direct channel rainfall is operative in all storms, although its detection is difficult. Overland flow is in operation at some locations on the hillslope, specially in topographic lows. Macropore and mesopore flow occurred and may lead to groundwater displacement. Groundwater ridging also occurred. Each flowpath was found to vary in its operation, according to a series of controls, namely seasonality, antecedent moisture conditions, rainfall magnitude, duration and intensity, and the timing between rainstorms. Conservative tracers (chloride and temperature) were employed to investigate the contribution of 'old' and 'new' water to storm runoff. The variation in chloride concentrations in samples collected either sequentially or manually at each flowpath was monitored throughout storms. Rainfall, comprising 'new' water, was found to exhibit a distinct chloride chemistry. Most samples contained < 20 eq/1 Cl-. A similar trend was observed for samples of throughfall and forest floor soil water. Groundwaters and matrix soil waters contained two or three times greater chloride concentrations than in the 'new' waters, due to evaporative mechanisms. Hence, 'new' water could be distinguished from 'old' water on the basis of chloride chemistry. Similarly, the temperature profiles of 'new' and 'old' waters were significantly different. During the summer, rainfall ('new' water) is warmer than groundwater ('old' water), and during the winter, the reverse is true. Hence, both chloride and temperature were instrumental in distinguishing 'old' from 'new' waters. Direct channel rainfall, overland flow and macropore flow were important flowpaths for the rapid transport of 'new' water through the system during the growing season. Overland flow contributed some 'old' water during the dormant season. Although macropore flow allowed rapid transit of 'new' water to depth, this led to a groundwater displacement mechanism, which ultimately led to the rapid contribution of 'old' water to storm runoff. The combination of hydrometric and tracer data enabled a conceptual hydrological model to be developed of the responses of the hillslope to storm events.