Modeling and Assessment of a Hybrid Pyramid Solar Still-Humidification-Dehumidification Desalination System for Sustainable Freshwater Production in SWGHs: A Case Study

Demand and the urgent need for energy conservation and sustainable resource management. Addressing this challenge requires desalination systems that are both energy-efficient and adaptable to climate variability. This study proposes a hybrid seawater greenhouse (SWGH) system that integrates humidification–dehumidification (HDH) desalination with a pyramid solar still (PSS) through thermal coupling to enhance overall energy utilization and freshwater productivity. To enable long-term performance assessment, convolutional neural network (CNN), gated recurrent unit (GRU), and hybrid CNN–GRU deep learning models were developed to predict freshwater production using historical and projected climatic data, including global horizontal irradiation (GHI) and air temperature (T2M), over the period 1985–2034. The CNN–GRU model demonstrated superior predictive performance and was coupled with thermodynamic simulations of the hybrid system to estimate long-term freshwater yield. Results indicate an annual cumulative freshwater production of approximately 4250-4400 L/m²·year. The HDH subsystem, with a surface area of 300 m², contributed 1450–1500 L/m²·year, while the 150 m² solar still unit produced 2800–2900 L/m²·year annually. These findings confirm that thermal integration between the desalination units significantly improves overall freshwater production while reducing thermal losses. The novelty of this work lies in the combined application of a CNN–GRU predictive framework with a thermally integrated hybrid solar desalination system, enabling reliable long-term forecasting under climate-sensitive conditions. The proposed approach provides a scalable, energy-efficient solution for sustainable freshwater generation and offers a data-driven decision-support tool for adaptive water management in arid coastal environments.