Using Particle Tracking to Understand Flow Paths, Age Distributions, and the Paradoxical Origins of the Inverse Storage Effect in an Experimental Catchment

Abstract Many catchment studies use estimates of stream age distributions, including transit time distributions and StorAge Selection (SAS) functions, to make inferences about hydrologic processes and dominant flow paths. To improve the basis for such inferences, this study examined how different hydrologic flow paths contribute to the age distribution of discharge and evapotranspiration (ET) at the U.S. Department of Agriculture's Mahantango Creek experimental catchment in PA, USA, using the integrated surface‐subsurface hydrology model Parallel Flow‐Common Land Model with SLIM‐FAST particle tracking. The model tracked ET and five flow paths: direct runoff, overland flow, interflow, and shallow and deep groundwater. Over a 1‐year simulation, we found that individual flow paths release generally older age‐rank storage under wetting conditions (a direct storage effect). Paradoxically, total discharge originated from younger age‐rank storage under wetting conditions (an inverse storage effect). This inverse storage effect was caused by the disproportionate increase in contributions from younger flow paths, particularly overland flow. This result suggests that where direct storage effects have been observed for individual sources of discharge (such as shallow groundwater), the same may not necessarily hold for total discharge and could in fact be reversed. Other findings include (1) a highly nonunique mapping between water age and flow path, (2) a “middle‐aged depression” in stream ages that cannot be fit by unimodal distributions (e.g., gamma and beta), (3) a taxonomy of transit time distribution shapes that occur during a typical storm, and (4) relatively young ET age distributions with a time‐invariant SAS function shape that reflects the parameterized depth distribution of root‐water uptake.