Lateral Preferen al Flow Paths in a Glaciated Watershed in the Northeastern USA

Despite observa onal evidence of lateral preferen al fl ow paths in northeastern U.S. watersheds, their eff ects on the sources of runoff remain unclear. An intense fi eld survey was undertaken during the 2007 growing season to determine the sources of stream runoff from a 2.51 km2 watershed in the Catskill Mountains, New York State. Lateral preferen al fl ow paths are caused by groundwater springs and soil piping in this region. A two-component hydrograph separa on using δ18O showed that event water (rain water) was a signifi cant source of runoff during nine rainfall events (from July to October). With these rainfall events, 14 to 37% of the volume and 18 to 49% of the peak streamfl ow was a ributable to event water. Further, end-member mixing analysis (EMMA), using δ18O, Si, and dissolved organic carbon (DOC), showed that saturated areas accounted for 2 to 24% of the total volume and 4 to 59% of peak streamfl ow but that groundwater was the dominant source of runoff volume during all events. Field surveys of saturated areas also suggested that near-stream areas were insuffi cient to generate the observed stream chemistry during rainfall events larger than 8 mm. A connec on with the hillside saturated areas was therefore required to explain the results of the hydrograph separa ons, which were corroborated by the ming of the transient (perched) groundwater and overland fl ow. The hydrometric measurements confi rmed that hillside lateral preferen al fl ow paths rapidly transported water to near-stream satura on areas during runoff events under rela vely dry antecedent condi ons. A qualita ve comparison with conven onal techniques for distribu ng variable satura on areas (VSA) using surface topography and soil transmissivity (i.e., topographic index and soil topographic index), which do not consider the eff ects of lateral preferen al fl ow paths, demonstrated that typical parameteriza ons (on the order of <10−1 m) would not have the spa al resolu on to represent the measured lateral preferen al fl ow paths (on the order of <10−3 m). Overall, the results suggest that the lateral redistribu on of water from hillside areas reduces the infl uence of surface topography and channel topology on the sources of stream runoff , a fi nding that is consistent with recent ones from other landscapes where glacial soils have coevolved with the terrestrial hydrology.