Flow partitioning modelling using high-resolution electrical conductivity data during variable flow conditions in a tropical montane catchment

Abstract

Tracer-aided hydrological models (TAHMs) are one of the most powerful tools to identify new (event) and old (pre-event) water fractions contributing to stormflow because they account both for streamflow and tracer mixing dynamics in model calibration. Nevertheless, their representativeness of hydrograph dynamics is often limited due to the unavailability of high-resolution conservative tracer data (e.g., water stable isotopes or chloride). Hence, there is a need to identify alternative tracers yielding similar flow partitioning results than “ideal” ones while requiring fewer financial resources for high-frequency monitoring (e.g., sub-hourly). Here, we compare flow partitioning results of a TAHM calibrated using high-frequency electrical conductivity (EC) and water stable isotope (18O) data collected during 37 rainfall-runoff events monitored during variable hydrometeorological conditions in the Zhurucay Ecohydrological Observatory, a tropical alpine catchment located in southern Ecuador. When the model was calibrated using the sampling resolution of stables isotopes (6-hours to 1-hour), no statistically significant differences of pre-event water fractions (PEWFs) using both tracers for model calibration were found. PEWF differences between both tracers for 89% of the events were < 20% regardless of the events’ antecedent moisture and rainfall conditions. Model transfer functions were also similar suggesting that catchment internal processes inferred using both tracers are comparable. Events presenting larger differences (n = 4; up to 27% PEWF difference) had no samples collected during peak flow. Calibration of the model using EC data collected at sub-hourly intervals (every 5-minutes) showed a significant increase in model performance as compared to the frequency of collection of isotopic data. Similarity in flow partitioning results can be attributed to a quasi-conservative nature of EC due to the presence of organic-rich riparian soils (peat-type) overlying compact bedrock across the catchment. Findings also highlight the importance of capturing rapidly occurring catchment mixing processes though high-temporal frequency monitoring of tracer data. Our study encourages the value of assessing the use of alternative tracers, such as EC, to identify fast occurring rainfall-runoff processes, while lowering the costs needed to implement and sustain tracer data collection for long time periods.