Sprinkler irrigation, a widely used method, faces criticism for its inefficient resource usage, particularly due to water losses from wind drift and evaporation. Estimating these wind drift and evaporation losses (WDEL) poses challenges as it is time-consuming, expensive and often marred by a high level of error. As an alternative, the use of empirical models has gained attention. These models have been explored in various geographical contexts worldwide, yet their application in developing countries remains limited. This study aimed to evaluate the suitability of existing empirical models for application in different geographical settings with similar climatic conditions. Six WDEL models in total were chosen, and they were each assessed between July and September of 2023. WDEL models were assessed using the conventional catch-can approach. The observed WDEL during the study averaged around 6.35%, with a range of 1.32% to 19.24%. Model evaluation revealed the WDEL_U3 performed the best in the climatic conditions of Mkulazi Estate, with the lowest Mean Absolute Error (MAE) of 4.05%, Percentage Error (PE) of 1.29%, and Root Mean Square Error (RMSE) of 5.77%. On the other hand, the performance of the WDEL_WSRH was not satisfactory. The study emphasized on the importance of using existing WDEL prediction models to avoid tedious processes of WDEL evaluations, as well as the need of considering local meteorological conditions and changing irrigation schedules to reduce WDEL. Addressing WDEL is critical for optimizing water consumption efficiency in sprinkler irrigation, and the selection of predictive models should be carefully addressed based on local variables to increase estimate accuracy.

Keywords: Wind Drift, Evaporation, Sprinkler, Efficiency