Sensitivity analysis of determining seepage velocity using temperature records

Vertical velocities of groundwater flow can be determined by temperature time series measurements at different groundwater depths. The most typical analytic model is the Hatch model, which uses the characteristic of temperature damping or the phase lag of temperature. A local sensitivity analysis is conducted on the parameter of Hatch model in an effort to identify the most important impact factors of the model, and subsequently, the effect of each parameter on the modeling result. The global sensitivity method is then applied to analyze the interaction among parameters and their effects on model simulations. Results show that the distance between the two measuring points (Δz) is the most influential factor on the model simulation, and followed by the heat capacity of solids (ρ_sc_s), the effective porosity (n_e). The model simulation is least sensitive to the variations of the heat conduction coefficient (λ₀) and the degree of heat dispersion (β), and a negative correlation is found between vertical velocities of groundwater flow and the two parameters λ₀ and β. While for other parameters, positive correlations exist. As a result, a special attention should be paid on measurement of the distance between the two measuring points, while empirical values can be assigned to the two parameters λ₀ and β.