Groundwater Level Trend Analysis and Prediction in the Upper Crocodile Sub-Basin, South Africa

Disasters related to climate change regarding water resources are on the rise in terms of scale and severity. Therefore, predicting groundwater levels (GWLs) is a crucial means aid adaptive capacity towards disasters resources. In this study, Gradient Boosting (GB) regression modelling approach for GWL prediction as function rainfall antecedent used. A correlation analysis carried out from 2011 2020 demonstrated that monthly GWLs can be predicted by rainfall. The study also sought understand long-term effects events over area through Mann–Kendall (MK) trend analysis. total 50% stations revealed declining trends, while 25% had no trends other showed an increasing trend. Again, results were used justifying trends. GB predictive model performed satisfactorily all stations, with MSE values ranging 0.03 0.304 MAE varying 0.12 0.496 validation period. R2 ranged 0.795 0.902 overall based projected levels, future using derived study.