Accurate water level prediction for both lake and river is essential for flood warning and freshwater resource management. In this study, three machine learning algorithms: multi-layer perceptron neural network (MLP-NN), long short-term memory neural network (LSTM) and extreme gradient boosting XGBoost were applied to develop water level forecasting models in Muda River, Malaysia. The models were developed using limited amount of daily water level and meteorological data from 2016 to 2018. Different input scenarios were tested to investigate the performance of the models. The results of the evaluation showed that the MLP model outperformed both the LSTM and the XGBoost models in predicting water levels, with an overall accuracy score of 0.871 compared to 0.865 for LSTM and 0.831 for XGBoost. No noticeable improvement has been achieved after incorporating meteorological data into the models. Even though the lowest reported performance was reported by the XGBoost, it is the faster of the three algorithms due to its advanced parallel processing capabilities and distributed computing architecture. In terms of different time horizons, the LSTM model was found to be more accurate than the MLP and XGBoost model when predicting 7 days ahead, demonstrating its superiority in capturing long-term dependencies. Therefore, it can be concluded that each ML model has its own merits and weaknesses, and the performance of different ML models differs on each case because these models depends largely on the quantity and quality of data available for the model training.
Global warming induces spatially heterogeneous changes in precipitation patterns, highlighting the need to assess these changes at regional scales. This assessment is particularly critical for Afghanistan, where agriculture serves as the primary livelihood for the population. New global climate model (GCM) simulations have recently been released for the recently established shared socioeconomic pathways (SSPs). This requires evaluating projected precipitation changes under these new scenarios and subsequent policy updates. This research employed six GCMs from the CMIP6 to project spatial and temporal precipitation changes across Afghanistan under all SSPs, including SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5. The employed GCMs were bias-corrected using the Global Precipitation Climatological Center's (GPCC) monthly gridded precipitation data with a 1.0° spatial resolution. Subsequently, the climate change factor was calculated to assess precipitation changes for both the near future (2020-2059) and the distant future (2060-2099). The bias-corrected projections' multi-model ensemble (MME) revealed increased precipitation across most of Afghanistan for SSPs with higher emissions scenarios. The bias-corrected simulations showed a substantial increase in summer precipitation of around 50%, projected under SSP1-1.9 in the southwestern region, while a decline of over 50% is projected in the northwestern region until 2100. The annual precipitation in the northwest region was projected to increase up to 15% for SSP1-2.6. SSP2-4.5 showed a projected annual precipitation increase of around 20% in the southwestern and certain eastern regions in the far future. Furthermore, a substantial rise of approximately 50% in summer precipitation under SSP3-7.0 is expected in the central and western regions in the far future. However, it is crucial to note that the projected changes exhibit considerable uncertainty among different GCMs.