Integrated statistical and feature based time series analysis of natural radionuclides and physicochemical parameters in surface water and groundwater of a hydrogeological system

Rn-222 is a well-established tracer for groundwater-surface water interactions, but long-term analyses of radiological parameters are scarce. Time series analysis of radiological parameters in waters is a recent, case-specific approach. This study applies a feature-based time series analysis to radiological and physicochemical data from groundwater (GW) and surface water (SW) within a river and two tributaries spanning contrasting lithologies. Over one hydrological year, we monitored Rn-222, gross alpha and beta radiation, pH, electrical conductivity, and water temperature. Groundwaters showed significant spatial differences (p < 0.05) across sampling sites and lithologies for all variables, while in surface waters only gross alpha, pH, and EC varied significantly. Correlations were site-specific, with groundwater showing stronger internal relationships among variables than surface water. Air temperature correlated more consistently with water parameters than precipitation. Nevertheless, Rn-222 was found to be inversely correlated with discharge (in groundwaters) and positively correlated with air temperature. Seasonal-trend decomposition (STL) revealed long-term patterns, seasonal signals, and noise, allowing hierarchical clustering of sampling sites by variable. The results clarified the role of geological setting and climatic factors in shaping water dynamics. Seasonal variations were observed across multiple parameters, with Rn-222 concentrations exhibiting delayed responses to precipitation events, reflecting heterogeneous recharge pathways and subsurface residence times. Hierarchical clustering distinguished deep, lithologically confined aquifers from more responsive, shallow systems, offering insights into the groundwater discharge contributions to surface water. Notably, surface waters of the tributaries were more responsive to groundwater inputs, with a clear seasonal signal attributed to changes in flow and weather conditions. The study underscores the importance of integrating time series analysis and Hierarchical Clustering into routine water quality analyses to better understand the spatiotemporal dynamics of groundwater and surface water interactions in geologically complex environments. Findings provide insights for water resource management and monitoring, with applications in comparable hydrogeological regions.