The influence of deep groundwater discharge on spatial variability of stream discharge and chemistry

In forest catchments, large amounts of groundwater infiltrate and pass through bedrock. This deep groundwater has a long residence time, as well as high concentrations of minerals that originate from dissolved matter. Thus, deep groundwater has a great influence on stream discharge and chemistry. However, in small catchments, stream discharge and chemistry vary greatly, even in geologically homogeneous catchments. The representative elementary area (REA) concept proposed by Wood defines the REA as a threshold area, beyond which an average hydrologic response occurs. Extensive research has shown that different catchments have different REA values. However, the factors that determine the REA value and the effect of deep groundwater on the REA are not yet clear. Hence, our objective was to elucidate the influence of deep groundwater discharge on the spatial variability of stream discharge, chemistry, and REA value. Our study site was the Inokawa catchment in Japan. The steep, forested catchment area was 5 km2. Inokawa is underlain by a Tertiary formation of sedimentary rock. In one zero-order hollow in Inokawa, large amounts of water (annual average 520 mm) infiltrate bedrock. We measured stream discharge and sampled stream water, which was then analyzed using ion chromatography. In small catchments, all solute concentrations and specific discharges have large spatial variability. As the catchment area increases, variability decreases. We found a convergence of NO3-,Mg2+,Ca2+,SO42- above 1-2 km2 and confirmed the existence of an REA. However, other dissolved matter and specific discharge did not converge to certain values and increased at 3-5 km2. Thus, we could not confirm the existence of an REA at these distances. We used the End Members Mixing Analysis(EMMA) method, separating stream water into three components: soil water, groundwater, and deep groundwater. The soil water component did not show a quantitative change above 1 km2. In contrast, groundwater and deep groundwater components increased at 3-5 km2. The increase in deep groundwater was especially notable, accounting for 38% of the stream water at 5 km2. Deep groundwater discharge drastically increased below 160 m altitude. We speculate that a discontinuity of geological structure determines the flow pathway and discharge of deep groundwater. In addition, if deep groundwater does not discharge (i.e., if stream water is formed by soil water and groundwater), the predicted specific discharge and groundwater contribution to stream water converges above 2 km2. Thus, deep groundwater discharge increases the heterogeneity of spatial variability of stream discharge and chemistry and increases REA.