Influence of fluid density on the estimation of vertical streambed fluxes and sediment thermal properties

• A density-based heat transport model for vertical streambed fluxes presented. • Fluid density has an obvious influence on the estimation of vertical streambed fluxes (VSFs). • Fluid density has no significant influence on the estimation of effective thermal diffusivity. Freshwater salinization alters the fluid density in the river and affects subsurface flow, solute and heat transport. Studies for quantifying vertical streambed fluxes (VSFs) and the effective thermal diffusivity of streambed sediments using heat as a tracer, considering the influence of fluid density, are limited. This study introduces a coupled conceptual model of subsurface flow-solute-heat transport within a streambed of losing streams, with the numerical solution derived using the finite element method. To investigate the impact of fluid density on parameter estimation, a combination of particle swarm optimization (PSO) and an existing analytical solution is employed to estimate VSFs and effective thermal diffusivity based on temperature–time series across varying salt concentrations in the river water. The accuracy and reliability of the VFLUX2 method are assessed under the influence of the fluid density effect. Results show that fluid density has a more significant influence on the estimation of VSFs than on the effective thermal diffusivity, and the impacts of fluid density on VSFs increase with increasing depth. As river water density increases, the use of traditional analytical models that quantify VSFs could result in an underestimation of VSF, especially when the observation depth is greater than 0.2 m. When using the VFLUX2 method to estimate effective thermal diffusivities of streambed sediments amidst fluid density variations, the VSFs may also be underestimated significantly when temperature measurements are taken at depths greater than 0.2 m. In the lower portion (less than 0.2 m), parameter estimation results indicated that the effect of density effects failed to appear due to the computational errors and short seepage times. Once the fluid density within the streambed stabilizes, even in the presence of higher density than before, the VSFs estimated by VFLUX2 or conventional analytical models remain reliable. This study provides new insights into the use of heat as a tracer to quantify VSFs when subject to fluid density variation.