Rainfall-Runoff simulation is the backbone of all hydrological and climate change studies. This study proposes a novel stochastic model for daily rainfall-runoff simulation called Stacked Long Short-Term Memory (SLSTM) relying on machine learning technology. The SLSTM model utilizes only the rainfall-runoff data in its modelling approach and the hydrology system is deemed a blackbox. Conversely, the distributed and physically-based hydrological models, e.g., SWAT (Soil and Water Assessment Tool) preserve the physical aspect of hydrological variables and their inter-relations while taking a wide range of data. The two model types provide specific applications that interest modelers, who can apply them according to their project specification and objectives. However, sparse distribution of point-data may hinder physical models’ performance, which may not be the case in data-driven models. This study proposes a specific SLSTM model and investigates the SLSTM and SWAT models’ data dependency in terms of their spatial distribution. The study was conducted in the two distinct river basins of Samarahan and Trusan, Malaysia, with over 20 years of hydro-climate data. The Trusan basin’s rain gauges are scattered downstream of the basin outlet and Samarahan’s are located around the basin, with one station within each basin’s limits. The SWAT was developed and calibrated following its general modelling approach, however, the SLSTM performance was also tested using data preprocessing with principal component analysis (PCA). Results showed that the SWAT performance for daily streamflow simulation at Samarahan has been superior to that of Trusan. Both the SLSTM and PCA-SLSTM models, however, showed better performance at Trusan with PCA-SLSTM outperforming the SLSTM. This demonstrates that the SWAT model is greatly affected by the spatial distribution of its input data, while data-driven models, irrespective of the spatial distribution of their entry data, can perform well if the data adequacy condition is met. However, considering the structural difference between the two models, each has its specific application in a water resources context. The study of catchments’ response to changes in the hydrology cycle requires a physically-based model like SWAT with proper spatial and temporal distribution of its entry data. However, the study of a specific phenomenon without considering the underlying processes can be done using data-driven models like SLSTM, where improper spatial distribution of data cannot be a restricting factor.
The escalating challenge of water scarcity demands advanced methodologies for sustainable water management, particularly in agriculture. Machine learning (ML) has become a crucial tool in optimizing the hydrological cycle within both natural and engineered environments. This review rigorously assesses various ML algorithms, including neural networks, decision trees, support vector machines, and ensemble methods, for their effectiveness in agricultural water management. By leveraging diverse data sources such as satellite imagery, climatic variables, soil properties, and crop yield data, the study highlights the frequent use and superior predictive accuracy of the Random forest (RF) model. Additionally, artificial neural networks (ANNs) and support vector machines (SVM) show significant efficacy in specialized applications like evapotranspiration estimation and water stress prediction. The integration of ML techniques with real-time data streams enhances the precision of water management strategies. This review underscores the critical role of ML in advancing decision-making through the development of explainable artificial intelligence, which improves model interpretability and fosters trust in automated systems. The findings position ML models as indispensable for real-time, data-driven management of agricultural water resources, contributing to greater resilience and sustainability under the dynamic pressures of global environmental change.
This study introduces a systematic methodology whereby different technologies were utilized to download, pre-process, and interactively compare the rainfall datasets from the Integrated Multi-Satellite Retrievals for Global Precipitation Mission (IMERG) satellite and rain gauges. To efficiently handle the large volume of data, we developed automated shell scripts for downloading IMERG data and storing it, along with rain gauge data, in a relational database system. Hypertext pre-processor (pHp) programs were built to visualize the result for better analysis. In this study, the performance of IMERG estimations over the east coast of Peninsular Malaysia for the duration of 10 years (2011–2020) against rain gauge observation data is evaluated. Moreover, this study aimed to improve the daily IMERG estimations with long short-term memory (LSTM) developed with Python. Findings show that the LSTM with Adaptive Moment Estimation (ADAM) optimizer trained against the mean square error () loss enhances the accuracy of satellite estimations. At the point-to-pixel scale, the correlation between satellite estimations and ground observations was increased by about 15%. The bias was reduced by 81–118%, was reduced by 18–59%, the root-mean-square error () was reduced by 1–66%, and the Kling–Gupta efficiency () was increased by approximately 200%. The approach developed in this study establishes a comprehensive and scalable data processing and analysis pipeline that can be applied to diverse datasets and regions encountering similar domain-specific challenges.