Heat Budget Analysis of a Braided River in Taiwan Using the HFLUX Model

Temperature is regarded as an effective natural tracer for analyzing river flow sources through heat budgets. This study investigates the thermal dynamics of the Ai-Liao River, Taiwan, during a dry-season period (December 6–9, 2022). A fiber-optic distributed temperature sensor (FO-DTS) was deployed along a 782-meter river section, and the HFLUX model was employed to analyze the river’s heat budget. Additionally, thermal imagery obtained via drone-assisted surveys was used to identify potential groundwater inflow locations. FO-DTS data revealed high spatial temperature variability, dividing the study area into three segments. In the upper segment, daily temperature differences (DTD) increased downstream. In the middle segment, DTD decreased downstream, while in the lower segment, DTD remained stable. The HFLUX model simulations yielded RMSE values of 0.42°C, 0.33°C, and 0.30°C for the respective segments. Results indicated that the upper segment exhibited high sensitivity to heat budget changes due to low flow. In contrast, the middle segment demonstrated increased groundwater energy contributions, with an average of −83.3 W/m² over three days, moderating DTD. Thermal imagery captured "tongue-shaped" inflow patterns along the middle riverbanks, indicating significant water source inputs. In the lower segment, increased flow stabilized DTD. The integrated analysis from observations, modeling, and thermal imagery suggests that modeled groundwater inflows predominantly enter the river as hyporheic flows.