Deciphering the dependencies of piezometric signals on hydroclimatic and anthropogenic forcings using time-series models

Understanding how groundwater levels respond to hydroclimatic forcings and human activities is essential for sustainable groundwater management yet remains challenging in many regions due to incomplete pumping records, heterogeneous datasets, and non-stationary system behavior. In this study, we explore a hybrid, data-driven framework to disentangle climate-driven and anthropogenic influences on long-term groundwater head time series using monitored piezometric networks. We first apply transfer function–noise (TFN) modelling, implemented through the Pastas framework, to simulate groundwater head dynamics as a response to observed hydroclimatic forcings, including precipitation, evapotranspiration, and river stage. The resulting model residuals exhibit structured, behaviours that cannot be attributed to random noise alone, suggesting the presence of missing processes or stresses not explicitly represented in the model. The results show that the Pastas model is able to optimize the parameters of response functions of the recharge while separating out other drivers of the groundwater head. We then analyse residual pattern, temporal shifts, and spatial coherence across multiple wells to assess the residual patterns and classify them between unresolved natural processes and anthropogenic stresses. To this end, we deploy unsupervised anomaly detection algorithms (Isolation Forest) on these residuals to automatically classify the temporal schedule of pumping events without prior labelling. This work demonstrates how interpretable time-series models and data-driven learning can be combined to reduce uncertainty, improve process understanding, and extract management-relevant information from groundwater monitoring data under data-scarce conditions.