Developing a foundational hydrological model for the Limpopo River Basin using the Soil and Water Assessment Tool Plus (SWAT+)

This study aimed to create a foundational hydrological model for the transboundary Limpopo River Basin (LRB) in Southern Africa using the Soil and Water Assessment Tool Plus (SWAT+) model. The model is a crucial part of a larger project to develop a digital twin of the river basin. The SWAT+ model simulated streamflow and other important hydrological processes in the LRB, using various data sources such as global gridded rainfall and other weather parameters, soils, landcover datasets, and in-situ discharge measurements from seven locations in South Africa for calibration and validation. The automatic calibration routine IPEAT+, freely available as part of the SWAT+ framework, was used for model calibration. The model was run for 23 years, from 2001 to 2023, with the calibration and validation periods varying for each gauge location. The calibration of the LRB SWAT+ model primarily consists of the discharge data from South Africa with Botswana, Zimbabwe, and Mozambique still awaited from the stakeholders. The results showed that the foundational LRB SWAT+ model achieved a good performance in simulating streamflow in the South Africa part of the LRB, with Nash Sutcliffe efficiency values ranging from 0.42 to 0.69. Further calibration in other sub-basins and the incorporation of additional reservoir data are expected to improve the overall performance of the LRB SWAT+ model.

The partially calibrated and validated foundational SWAT+ model for the LRB (mainly in the South Africa part of LRB) will be the initial core of a river basin digital twin that is under development. It will integrate seasonal rainfall forecasts to simulate seasonal water availability. The fully calibrated foundational SWAT+ LRB model is expected to be available by the end of this year. It will serve as the basis for several planned applications as part of the LRB DT. The first application will focus on deriving flow patterns in the LRB catchment under unaltered natural conditions. The next application will concentrate on integrating ECMWF seasonal rainfall forecasts to predict water availability at 310 sub-basins and key channel reaches critical for environmental flows in the LRB. An updated version of the SWAT+ LRB foundational model is currently in progress. It aims to enhance the representation of reservoirs and their operations within the SWAT+ model and provide additional calibration sites within different sub-basins. This model will act as a baseline to create multiple planned applications co-developed with stakeholders.

Citation

cg.contributor.initiative: Digital Innovation