ABSTRACT
Streamflow generation processes in a large (1670 km²), alpine watershed in the southern Rocky Mountains of southwestern Colorado were investigated in an attempt to explain structured trends in stream chemistry. Lots of scatter was observed in the streamflow chemistry at subwatershed scales less than 300 km². However, above this critical scale, concentrations of chemical constituents in streamflow increase linearly with increasing scale. Our initial hypothesis was that these structured trends were a result of increasing contributions from deep-basin-scale groundwater flow in the watershed (termed the “increasing deep groundwater” hypothesis in previous studies). We used endmember mixing analysis (EMMA) methodologies to quantify endmember contributions to streamflow across multiple scales and at intervals of increasing watershed scale down Saguache Creek. Results indicated that groundwater contributions were important at all scales accounting for as much as 36 percent of streamflow generation during the peak of the snowmelt freshet and as much as 78 percent of baseflow. Groundwater contributions were also very large in small headwater and tributary subwatersheds implying that residence times in these small subwatersheds may be underestimated in current residence time studies. Our results also indicated that groundwater contributions do in fact increase with increasing scale; therefore, providing support for the “increasing deep groundwater” hypothesis. Our results further indicate that scaling small-scale runoff processes to larger watershed scales may be inappropriate due to topographic and geologic constraints within the watershed and due to the increasing importance of deep, basin-scale groundwater as scale increases. The groundwater and vadose zone components of streamflow generation remain poorly constrained in most runoff studies and this paper illustrates the importance of these contributions to streamflow generation in a large watershed.