Impact of Isotopic Heterogeneity in Shallow Systems on Modeling of Stormflow Generation

Abstract

A major uncertainty in hydrologic and geochemical modeling of stormflow generation in watersheds has been quantification of the contributions of water and solutes from different sources and hydrologic pathways to streamwater. Isotopic techniques have recently gained widespread acceptance as useful tools in the investigation of sources of stream flow, but considerable debate still surrounds the question of whether the spatial and temporal variations in the isotopic and chemical compositions of water components are negligible. At Panola Mountain, Georgia, a 2-year study of temporal and spatial variability in rain and throughfall has determined that average throughfall is enriched by 0.5% in 0^18O and 3.0% in 0D relative to rain; site-specific differences in canopy cause up to 1.2 % variation in 0^18O among collectors for the same storm; and throughfall ^18O enrichment takes place throughout the storm, not just at the beginning. Evaporative losses are greater and throughfall is generally slightly enriched in ^18O in conifer forests relative to deciduous forests. However, throughfall shows little evidence of evaporative fractionation; instead, the high deuterium-excess values suggest considerable exchange with re-evaporated waters. A 490-m^2 artificial catchment in China was used to investigate the effects of temporal variations in rain composition, and temporal and spatial variations in dominant water flowpath, on the development of isotopic and chemical heterogeneity in soil waters and groundwater. In response to changes in storm intensity, variability in the amounts of water transported via piston versus macropore flow caused a 4% range in 0^18O of groundwaters. Selective storage of early rain in shallow soils makes characterization of the isotopic composition of infiltrating rain water problematic. Seasonal and hydrologic differences in the sources of alkalinity were investigated at four watersheds at Catoctin Mountain, Maryland, by analyzing the dissolved inorganic carbon in streamwater for 0^13C. Because of short residence times, the isotopic signatures of the two primary carbon sources, calcite and soil-derived carbonic acid, do not appear to be appreciably overprinted by exchange reactions, biological recycling, or degassing; hence, 0^13C seems to be a useful semi-conservative tracer of water flowpaths and carbon sources.