Kunz, Richard Peters.Xolo, Thobeka.2025-11-212025-11-2120252025https://hdl.handle.net/10413/24140Masters Degree. University of KwaZulu-Natal, Pietermaritzburg.The accurate assessment and modelling of hydrological processes relies heavily on comprehensive and reliable climate data. In South Africa, the alarming decline in the number of climate monitoring stations and the poor quality of observed data (i.e. missing records) present a significant challenge to reliable hydrological modelling. In addition, it is crucial to select climate driver stations as representative as possible of the catchment being studied. Climate driver stations are essential for capturing representative climate conditions necessary for water resources management and planning. This study assessed existing techniques used for selecting climate driver stations for a catchment. Available daily rainfall and temperature datasets were infilled and extended to create a 70-year record for quaternary catchments C41A to C41E, which are situated in the Lejweleputswa District Municipality (Free State Province, South Africa). The Inverse Distance Weighting method was used to infill rainfall data, whilst the Mean Temperature Difference method, Difference in Standard Deviation method and a ranking algorithm method were used to infill missing temperature data. Rainfall driver stations were selected using the common Driver Station (DS) method as well as the Adjustment Factor (AF) method which is closely related to the CalcPPTcor approach. Pseudo temperature driver stations were selected for each selected rainfall driver station using a revised ranking algorithm. The Dent et al. (1989) median, Lynch (2004) median, Lynch (2004) mean and Pegram et al. (2016) mean gridded rainfall datasets were compared for their performance in estimating rainfall adjustment factors using the R2 , Nash-Sutcliffe Efficiency and Root Mean Square Error statistics. Each gridded dataset was then used to verify the methods for selecting a climate driver station DS and AF methods. The ACRU model was used to simulate inflow to the Erfenis Dam, which was then compared to a dam water balance as a means of verifying which method performed better. Key findings showed that the Pegram mean gridded datasets (monthly and annual) perform better in enhancing the representativeness of station rainfall for the study catchment. The results for the AF and DS methods were inconclusive due to various challenges, i.e. having no observed streamflow for the study catchment. It is recommended that the Pegram mean grids be considered when deriving rainfall adjustment factors, which are applied to the rainfall driver station to improve the representativity of catchment rainfall. It is recommended that the DS and the AF methods be re-evaluated in another study catchment with more climate stations and a reliable streamflow monitoring network.enCC0 1.0 Universalhttp://creativecommons.org/publicdomain/zero/1.0/Climate monitoring stations.Rainfall.Temperature.Thesis