Are there more than two end-members contributing to storm-events in small head-water catchments?

<p>Stream water chemistry at catchment outlets is commonly used to infer the flowpaths of water through the catchment and to quantify the relative contributions of various flowpaths and/or end-members during, e.g., storm events. For this purpose, the number and nature of these flowpaths or end-members are commonly defined <em>a priori</em> as part of the experimental design and previous knowledge, and their contributions are calculated based on the dynamics of the stream chemistry, with the inherent assumptions and uncertainties of this approach. Here, we present a methodology, which inverts this classical approach. We use the variability of the stream chemistry data to determine the minimum number of end-members needed and, more specifically, whether two end-members would be sufficient. In this methodology, we analysed the concentration-concentration relationships of several major ion combinations on the storm-event scale for multiple events, using a multi-year, high-frequency (< 60 minutes) timeseries of the major cations and anions from the outlet of two small (0.8 &#8211; 5 km&#178;) french catchments with contrasting land-use (forest and mixed farming-cropping productions). The results indicate that a large number of storm-events (up to 92%) could be interpreted as the result of only two end-members, depending on the catchment and the ion combination used. These findings might help to revise some of the perceptual understandings of flowpath or end-member contributions in catchments during storm-events. In addition, they might stimulate the discussion about the definition of end-members or flowpaths in catchments, especially with regard to variable hydrological contributions.</p>