EGU23-11353, updated on 26 Feb 2023
https://doi.org/10.5194/egusphere-egu23-11353
EGU General Assembly 2023
© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

Stable isotopes as a tool for improving rainfall-runoff modelling in South Africa

Jared van Rooyen1, Andrew Watson1, Yuliya Vystavna2, and Jodie Miller2
Jared van Rooyen et al.
  • 1Stellenbosch, School for Climate Studies, Matieland, South Africa (jvan@sun.ac.za)
  • 2International Atomic Energy Agency, Isotope Hydrology Section, Vienna, 1400, Austria

Understanding the way in which a water budget is distributed within a hydrological system is imperative in the prediction of the systems behaviour when this water budget has changed. A complex interaction of variable flow rates, residence times and reactive transport, controls the available streamflow of a river system not only over seasonal changes, but under longer term climate fluctuations as well. Hydrological modelling techniques have been instrumental in predicting these changes by monitoring/simulating rainfall, river and groundwater contributions but are dependent on robust data collection through the maintenance of old infrastructure and the creation of new infrastructure. South Africa is a pertinent example of the decline of gauging infrastructure and a prime use case for novel stable isotope techniques as an accessory to traditional hydrological modelling in semi-gauged watersheds. Furthermore, to constrain contributions in modified systems, that include reservoirs and land use changes, isotopes present an opportunity to assess unpredictable water mobilisation in the streamflow system. In this study, stable isotope measurements of rainwater, groundwater and stream water (δ2H and δ18O), together with a tertiary mixing model were used to develop an isotope-enabled version of the JAMS/J2000 rainfall-runoff model, named J2000iso. The application was applied to the upper Berg River catchment, a catchment impacted by recent drought, but important for regional water supply. Compared to the base version, the J2000iso had 13% more simulated interflow, with 56% less variance in the ensemble results and less overall process uncertainty. The J2000iso was also more robust than the base version during a subsequent validation. The isotope-enabled models provided a means to constrain the proportion of surface runoff, interflow and baseflow considering the streamflow signal changes due to upstream reservoir operations. As many catchments in South Africa are still ungauged or impacted by upstream reservoirs, the J2000iso model provides a means to simulate hydrological processes, given the appropriate collection of isotope and auxiliary data.

How to cite: van Rooyen, J., Watson, A., Vystavna, Y., and Miller, J.: Stable isotopes as a tool for improving rainfall-runoff modelling in South Africa, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-11353, https://doi.org/10.5194/egusphere-egu23-11353, 2023.