Abstract: To elucidate the coupled mechanisms of seepage, heat transfer, and mass transfer in rough rock fractures, a three-dimensional numerical model was developed based on the lattice Boltzmann method (LBM). The model employs three distinct distribution functions to capture the evolution of the velocity, temperature, and concentration fields, explicitly accounting for the dependence of fluid kinematic viscosity, thermal diffusivity, and solute diffusivity on temperature. Model accuracy was verified against two benchmark cases. The model was subsequently applied to investigate the effects of rock temperature, pressure gradient, and fracture roughness on solute transport within fractures. A comprehensive sensitivity analysis was performed to evaluate the combined influence of five key parameters: fracture aperture, pressure gradient, rock temperature, fracture roughness, and initial solute diffusion coefficient. The results reveal that increasing rock temperature reduces the kinematic viscosity of the fluid while enhancing the solute diffusion coefficient, thereby markedly accelerating solute transport. Specifically, when rock temperature rises from 20°C to 70°C, kinematic viscosity decreases to 48% of its initial value, whereas the solute diffusion coefficient increases by a factor of 5.17 relative to its initial value. Higher seepage velocity further promotes solute transport through the combined effects of advection and thermal enhancement; however, increased flow velocity partially offsets the temperature-induced acceleration. Increasing fracture roughness leads to reduced fluid velocity and elevated temperature within the fracture. The deceleration of flow exerts a substantially greater inhibitory effect on solute transport than the enhancement of diffusion arising from temperature elevation. Consequently, greater fracture roughness results in slower solute transport. Among all investigated factors, fracture aperture exhibits the highest sensitivity, whereas the initial solute diffusion coefficient shows the lowest sensitivity.