Predictive hybrid modelling for municipal water and wastewater: A global framework demonstrated in Türkiye

Abstract

This study introduces a novel hybrid empirical-analytical modeling framework for predictive forecasting of municipal water abstraction and wastewater generation. Applied in Türkiye, this framework generates quantitative projections critical for design, simulation, and operational planning of urban water systems. It rigorously assesses the predictive capacities of these advanced models. Forecasts indicate municipal water abstraction in Türkiye will range from 7.2 to 8.0 billion m3 in 2030 and 7.4–8.6 billion m3 in 2040, with wastewater discharge estimated between 5.7 and 7.5 billion m3 and 6.1–8.9 billion m3 for the same years. Projections are supported by robust statistical validation including high coefficient of determination (R2) and low mean absolute percentage error (MAPE) values, and show strong compatibility with previous 2035 forecasts (106 %-116 % match). This highly adaptable and transferable computational framework offers a powerful tool for environmental modelers and systems analysts globally, supporting informed decision-making for sustainable urban water resource management, and advancing cleaner water outcomes and circular economy principles.

Keywords

• Environmental modelling
• Urban water systems
• Wastewater systems
• Forecasting
• Hybrid models
• Predictive analytics