Thermal dynamics of an intermittent subtropical stream controlled by surface water-groundwater interactions: evidence from distributed temperature sensing and heat-budget modeling

Stream temperature in intermittent subtropical rivers exhibits pronounced spatial and temporal variability driven by surface water-groundwater interactions, yet such variability is often poorly resolved by point-based measurements. This study integrates fiber-optic distributed temperature sensing (FO-DTS), precisely georeferenced using a Real-Time Kinematic (RTK) positioning system, with a physics-based heat-budget model (HFLUX) to investigate thermal dynamics along a 782-m intermittent stream reach in Taiwan. Meter-scale FO-DTS measurements collected over a 3-day monitoring period revealed distinct longitudinal thermal regimes. Model simulations reproduced measured stream temperatures with high accuracy across upstream, midstream, and downstream segments, with root mean square errors (RMSE) ranging from 0.328 to 0.399 °C and normalized RMSE values below 4.1%. Sensitivity analysis indicated a downstream shift in dominant thermal controls. Discharge exerted the strongest influence in the upstream losing reach, whereas groundwater temperature became the primary control in the midstream and downstream segments, where increasing subsurface inflows dampened diel temperature variability. The optimal groundwater temperatures inferred by the model closely matched independent field measurements, allowing differentiation between hyporheic-dominated inflow in the midstream segment and shallow groundwater contributions downstream. These results demonstrate that coupling RTK-referenced FO-DTS with heat-budget modeling provides a robust framework for diagnosing surface water-groundwater interactions and thermal regimes in intermittent streams, particularly in subtropical and drought-prone regions.