The challenges of deciphering the hydrological pathways in catchment systems

Hydrological pathways are flow paths through which water may move in a catchment. Investigations into hydrological pathways were probably first made in the 1960s, when hypothesized by Hewlett and Hibbert (1967) within the context of Variable source area (VSA). Since then, resolving hydrological pathways has remained a challenge despite varying approaches that have been adopted. VSA, like Horton's overland flow concept has assumed a steady-state, time invariant of catchment rainfall-runoff response and a homogenous environment. These assumptions have, however, been criticized in subsequent studies. For example, the assumption of a homogenous environment is rarely justified on physical grounds, while steady-state models are only relevant to periods when stream discharges are constant. The End Member Mixing Analysis (EMMA) model, which has also dominated hydrochemical investigations since 1990s, equally assumes time invariant in the end-member components (soil, rainfall and groundwater-in some environments) before they mix at the stream (or near stream). Subsequent studies and discussions showed that the VSA and EMMA and other previous approaches are inadequate for deciphering hydrological pathways. While EMMA was sufficient to reveal the contributions of the end-members to stream chemistry it did not help to trace the hydrological pathways. It was thus concluded that hydrological pathways are difficult to unravel because the flow paths are complex. It was also argued that though end members ought to be linearly independent, they are often not, and uncertainties often occur when tracers become strongly dampened and unresponsive. Solutes that are often thought to be conservative are not, and isotopes assumed to be conservative could be fractionated. A more recent alternative to the existing models is the use of patterns of the hysteretic relationship between stream discharge and concentration of dissolved loads in streams. Hysteretic patterns are usually nonlinear, time-variant, and often vary with locations or between storm events. Evans and Davies (1998) and subsequent studies (e.g. Rice, et al., 2004) have interpreted the hysteretic patterns in their studies using the mixing model. This approach however requires further examination, especially why mixing models, which assume linear relationship and time-invariance were used to interpret nonlinear and time variant patterns. There may be a need to compare the results of mixing models and c-Q interpretations based on nonlinear assumptions. This presentation therefore reviews previous knowledge of the hydrological pathways, and examines the importance of hysteresis in the context of pollutant transport. Its objective is to open discussion on an ongoing research whose ambition is to reveal some yet untapped information about hydrologic and biogeochemical responses to natural disturbances and human activities over a range of geophysical and vegetation settings.