Assessing and Forecasting Groundwater Resources in the Context of Climate Change Using AI Techniques for the Industry Zones in Tiruppur, India

Groundwater resources in semi-arid and industrial regions are increasingly threatened by unsustainable extraction, groundwater contamination, and climate-induced variability. Tiruppur District in Tamil Nadu, India, represents a critical case where rapid industrial growth, intensive agricultural activity, and changing climatic patterns have resulted in severe groundwater stress. This study proposes a hybrid artificial intelligence–based framework for the assessment and forecasting of groundwater levels and quality under climate change conditions. The framework integrates multi-source datasets comprising historical groundwater level records (1994–2024), groundwater quality parameters, and meteorological data. To address the non-stationary nature of hydrological time series, Improved Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (ICEEMDAN) and Variational Mode Decomposition (VMD) are employed prior to model training. Deep learning models, including Slime Mould Algorithm–optimized Long Short-Term Memory (SMA–LSTM) networks and CNN–LSTM hybrids, are developed to capture temporal and spatial dependencies. An Adaptive Weighting Model is used to ensemble predictions and improve robustness. Model performance is evaluated using Mean Absolute Error, Root Mean Square Error, Coefficient of Determination, and Nash–Sutcliffe Efficiency. The proposed ensemble framework demonstrates superior predictive accuracy, achieving an R² value of 0.948 and an NSE of 0.938. The results confirm the effectiveness of hybrid deep learning approaches for climate-resilient groundwater management and highlight their scalability to other water-stressed regions.