Long-term interpretation of displacement measurements in arch dams remains challenging due to the combined influence of hydrostatic loading, thermal effects, time-dependent behavior, and interactions with the foundation and drainage system. This study presents a data-to-decision analysis applied to a double-curvature arch dam using long-term monitoring data comprising radial displacements from dam-body and foundation pendulums, reservoir water level, air temperature, and complementary measurements, including piezometric pressures, seepage, and selected crack-opening observations. Displacement behavior was first characterized using hydrostatic-season-time (HST) and hydrostatic-season-temperature-time (HSTT) statistical models to establish a baseline response under normal operating conditions. The HSTT formulation provides an improved separation of the thermal effects, resulting in a clearer and more stable residual displacement signal. The residuals were subsequently examined in conjunction with hydraulic indicators through correlation and lag analyses to investigate dam–foundation–drainage interactions and distinguish load-driven responses from hydraulically influenced behavior. Deviations in the residual patterns were interpreted based on the consistency across multiple instruments and their temporal association with reservoir-level variations and drainage performance, rather than isolated sensor responses. This study demonstrates a transparent and implementable workflow that improves the baseline behavior definition by incorporating measured temperature and structural indicators and provides a framework for early anomaly detection through residualhydraulic correlation analysis. The operational validation of the detection sensitivity and false alarm rates remains future work. The proposed approach demonstrates how integrated analysis of displacement, piezometric, and seepage data reduces prediction error by 15% relative to the HST baseline (1.978 mm vs. 2.329 mm Mean Absolute Error) and reveals quantifiable hydraulic lag signatures (r ≈ 0.40 at 7–14 days), suggesting potential for early-warning indicators of drainage system performance changes, subject to operational validation of detection sensitivity and false alarm rates