It is increasingly evident that the global community is recognizing the urgent need for a comprehensive reassessment of the distribution, management, and utilization of freshwater resources. Despite significant technological advancements and the proliferation of information in the modern era, millions of individuals continue to succumb annually to preventable waterborne diseases, while hundreds of millions more endure severe health conditions. Notwithstanding substantial investments and efforts, by the close of the twentieth century, approximately 2.4 billion people—surpassing the global population of 1940—remained without access to sanitation services comparable to those available to the majority of citizens in ancient Rome. As noted by Panjabi (2014), more than one billion individuals still lack access to safe and sufficient drinking water.

Human activities have had a profound impact on the Earth as our collective habitat. In particular, land use change and management affect hydrology, which in turn determines flood risk, water resources (for human and environmental needs), and pollutant transport and dilution. It is becoming increasingly evident that land and water management are inextricably linked (Wheater & Evans, 2009).

The combined impact of rising temperatures and prolonged drought conditions over the past decade in California, along with court-imposed restrictions on water allocations in the Sacramento River Delta, has significantly reduced the availability of water resources in Southern California. In response to these challenges, water agencies have implemented a range of conservation measures and irrigation regulations, with a primary emphasis on restricting outdoor water use (Lund, 2016).

The City of Los Angeles, the most populous in the United States and the largest among the 88 municipalities in Los Angeles County, has enacted a series of initiatives aimed at reducing residential water consumption. Between 2008 and 2010, three successive phases of water restrictions were introduced to curb usage and promote conservation. The first phase, which was voluntary, commenced in FY 2008. The second phase, which was mandatory, commenced in FY 2009. The third and final phase, which was mandatory and combined with increased prices and decreased total household allotments, commenced in FY 2010 (Mini et al., 2015).

The integration of artificial intelligence (AI) has the potential to enhance the identification of complex patterns within extensive datasets related to dynamic water distribution. This is accomplished through the continuous incorporation of fundamental scientific processes at a micro level. The synergy between these advanced technologies can illustrate how public trust may be strengthened through the implementation of automated and conditional “smart contracts” for water usage, based on secure and immutable data. Furthermore, it plays a vital role in enhancing the accuracy and validation of local water usage data, a key element in global ecosystem changes studies and data-driven decision-making. By leveraging a decentralized intelligent framework for global water governance, blockchain-based security protocols can be integrated with AI algorithms trained on real-time water sensing data. In the realm of remote water distribution, this technological synergy offers substantial benefits in optimizing water availability, improving efficiency, and identifying scarcity trends. Such advancements facilitate the equitable management of multi-source water resources, particularly in the face of climate change (Lin et al., 2018).

The objective of sustainable water resources management is to maintain the ecological, economic and hydrological integrity of the current society and to ensure the viability of future prospects. The application of artificial intelligence methods in urban water resource planning is largely due to their exceptional capacity for reasoning, adaptability, modelling and forecasting of water demand (Xiang et al., 2021).

An advanced artificial neural network (ANN) framework, integrating a constriction coefficient-based particle swarm optimization and gravitational chaotic search algorithm (CPSOCGSA), was employed to predict monthly salinity levels. The model was developed and validated using historical monthly data on total dissolved solids (TDS) and electrical conductivity (EC) of the Euphrates River at Musayyib and Babylon, alongside relevant climatic variables from 2010 to 2019. To evaluate its performance, the CPSOCGSA-ANN approach was compared with alternative optimization techniques, including the slime mold algorithm (SMA-ANN), particle swarm optimization (PSO-ANN), and the multi-verse optimizer (MVO-ANN). Results demonstrated that data preprocessing significantly enhanced data quality and facilitated the identification of the optimal forecasting scenario. Moreover, the CPSOCGSA-ANN model outperformed the comparative methods across multiple statistical performance metrics. The proposed approach demonstrated high predictive accuracy in modeling TDS and EC time series, yielding R² values of 0.99 and 0.97, respectively, and scatter index (SI) values of 0.003 for both parameters (Khudhair et al., 2022).

An advanced intelligent management system has been developed to address water scarcity in Palestine, focusing on four key aspects: water scarcity assessment, protection of traditional water resources, implementation of rainwater harvesting (RWH), and the application of dynamic intelligent water management strategies. The smart RWH system has demonstrated the capability to fulfill approximately 41% of household water demands. Moreover, the smart dual water system has exhibited greater reliability in meeting water needs compared to the standalone smart RWH system. Through the integration of dynamic management strategies and an innovative supply framework, the smart dual system has proven its effectiveness in mitigating water scarcity comprehensively across all elevation levels within the study area. Additionally, the use of a smaller storage tank has been identified as a critical factor in optimizing system performance, with potential implications for the social and economic development of the region (Judeh, 2022).

In the United States, increasing urban populations, rising living standards, limited expansion of new water sources, and the depletion of existing supplies have created a growing imbalance where the demand for treated municipal water surpasses available resources. The amount of water used for irrigating residential urban landscapes is influenced by various factors, including landscape type, management practices, and regional geographic conditions. In general, however, landscape irrigation can account for a significant proportion of household water use, with figures ranging from 40% to 70%. It is therefore recommended that the primary focus of water conservation efforts be the efficient use of irrigation water in urban landscapes. Furthermore, plants in typical residential landscapes are frequently irrigated with greater quantities of water than is necessary to sustain essential ecosystem functions, including carbon regulation, climate control, and the maintenance of aesthetic appeal (Hilaire et al., 2008).

Effective water demand management is essential for water utilities, as they play a critical role in delivering safe drinking water from sources to end users through distribution networks. To ensure both current and future system performance, utilities must make informed decisions regarding water distribution. In this context, data-driven approaches, including artificial intelligence models, can be utilized to predict water demand, enabling the development of advanced tools to enhance overall water management (Zanfei et al., 2023).

Jafar et al. (2023) studied the surface water quality of Lake Seine in the Latakia Governorate using multiple linear regression (MLR) and machine learning (ML) models to predict the Water Quality Index (WQI). Their research analyzed water samples collected from a key drinking water source in Latakia, Syria, during 2021–2022, assessing water quality through statistical performance metrics such as the coefficient of determination (R²) and root mean square error (RMSE). The findings demonstrated that the MLR model, along with three ML techniques—linear regression (LR), least angle regression (LAR), and Bayesian ridge (BR)—exhibited strong predictive accuracy in estimating WQI. The MLR model achieved an R² of 0.999 with an RMSE of 0.149, while the ML models demonstrated near-perfect prediction accuracy, attaining an R² of 1.0 and an RMSE close to zero. These results highlight the effectiveness of both statistical and ML-based approaches in water quality prediction, offering a reliable framework for enhancing water management strategies and ensuring sustainable water resource planning.

Chamizo-Checa et al. (2020) conducted a comprehensive study on the Mezquital Valley in Mexico, utilizing the Water Evaluation and Planning System (WEAP) to model water balance scenarios from 2005 to 2050. Their research assessed the impacts of climate change on water availability, considering both steady-state and transient scenarios influenced by projected climate perturbations. The study highlighted the region’s reliance on untreated wastewater from Mexico City for agricultural irrigation and examined the implications of implementing wastewater treatment and advanced irrigation techniques. Findings indicated that while measures like drip irrigation could reduce agricultural water demand by 42%, they might not suffice to meet downstream hydroelectric requirements, especially with anticipated reductions in imported wastewater. This underscores the necessity for integrated water resource management strategies to address future water scarcity challenges in the region.​

Krishnan et al. (2022) highlight the growing urgency of effective water management amid the escalating global water crisis, emphasizing the need for advanced harvesting, recycling, and distribution strategies. As population density increases, intelligent water management systems become essential for optimizing resource utilization, conserving water, and maintaining quality standards. Their study explores key advancements in wastewater recycling, rainwater harvesting, and irrigation management through the application of Artificial Intelligence (AI) and Deep Learning (DL) within an Internet of Things (IoT) framework. Given the complexity of water management data, adaptable AI-driven models are necessary to provide integrated solutions across various sectors. Through case studies and statistical analyses, the research presents a structured framework for enhancing decision-making in water resource management, demonstrating the transformative potential of AI and IoT in addressing global water challenges.

Fu et al. (2022) explored the transformative impact of deep learning techniques on urban water system planning and management, emphasizing their potential to revolutionize economies, environments, and societies worldwide. While deep learning has been applied across various aspects of water management, a comprehensive review of its current applications and future potential remains limited. To address this gap, the study examines key areas where deep learning is being utilized, including water demand forecasting, leakage detection, contamination assessment, sewer defect evaluation, wastewater system state prediction, asset monitoring, and urban flood modeling. Despite its promise, deep learning applications in urban water management are still in their early stages, with most studies relying on benchmark networks, synthetic datasets, or pilot systems, and only limited real-world implementation. Among these applications, leakage detection has shown the most progress in practical deployment. The study further identifies five critical research challenges that must be addressed to advance deep learning in this field: data privacy, algorithm development, explainability and trustworthiness, multi-agent systems, and digital twins. Overcoming these challenges is expected to drive greater intelligence and autonomy in urban water systems, accelerating innovation and digital transformation within the global water sector. By highlighting these key areas, the study encourages further research and development, leveraging deep learning to enhance sustainable water management and optimize resource efficiency.

Vallejo-Gómez et al. (2023) conducted a systematic review focusing on smart irrigation systems that integrate artificial intelligence (AI) and Internet of Things (IoT) technologies in both urban and rural agricultural settings. The study analyzed 170 articles, identifying key technologies such as fuzzy logic, neural networks, and machine learning as pivotal in enhancing irrigation efficiency. These technologies enable real-time monitoring and precise control of irrigation processes, optimizing water usage and improving crop yields. The review highlights the potential of AI and IoT in transforming traditional irrigation practices, emphasizing their role in promoting sustainable agriculture through efficient resource management.​

Jung et al. (2021) developed an artificial neural network (ANN)-based model to predict surface water quality in the Nam River watershed, South Korea. By integrating meteorological data and water quality parameters, the study aimed to enhance the accuracy of water quality forecasts. The ANN model demonstrated high predictive performance, with coefficients of determination (R²) ranging from 0.810 to 0.929 for dissolved oxygen (DO) and 0.671 to 0.863 for biochemical oxygen demand (BOD₅). The results underscore the efficacy of ANN models in capturing the nonlinear relationships between meteorological factors and water quality indicators, offering a reliable approach for water quality management and decision-making processes.