Application Of The Thermo-TDR Technique For Determining Water Flux Density And It's Relationship With Thermal Dispersion And Solute Dispersion In Saturated Soils

Soil water flux density is an important parameter in studying runoff, infiltration, and chemical transport. Thermal dispersion and solute dispersion, resulting from the discrepancies of pore water velocity, are important parameters in studying water, heat, and solute transport in porous media. In this study, heat and Cl^- were used as tracers to study heat and solute transport through three packed media: glass beads, a sandy loam soil and a sandy clay loam soil, under steady state and saturated flow conditions. The Thermo-TDR probe was used to measure temperature change as a function of time and solute breakthrough curves, and the performance of the heat pulse technique in measuring water flux was investigated experimentally, and the reason that might account for the discrepancy between heat-pulse estimated water flux densities and actual measurements were investigated. An amplification factor is introduced to correct the influence of wall flow. Analytical heat and solute transport models were used to estimate the thermal dispersion coefficient (λ_d), solute dispersion coefficient (D) from the observed temperature and solute breakthrough curves. The relationship of λ_d, D with water flux to different media was examined. And the magnitude of thermal dispersivity (β) and solute dispersivity (κ) was compared at different water flux. The major findings are:(1) Three different methods (the maximum dimensionless temperature difference, temperature ratio between downstream and upstream sensor, and the time corresponding to the maximum temperature change) are used to estimate the water flux from the temperature change by time data and the stationary thermal properties of the soils. There is a linear relationship (r²>0.98) between water fluxes estimated by the maximum dimensionless temperature difference and the ratio method and water flux determined from the columns outflow. The water fluxes estimated by the maximum dimensionless temperature difference and the ratio method are lower than that determined from the outflow measurements. The simple temperature ratio method gives more accurate water flux estimates than the maximum dimensionless temperature difference method. However, our attempt trying to relate water flux to the time corresponding to the maximum temperature change is not successful.(2) A novel finding from this study is that in the packed columns, wall flow is responsible for the deviations between water flux estimates from heat pulse data and water flux obtained from outflow, and the magnitude of wall flow is largely determined by soil texture. An amplification factor (sandy loam, 1.12, sandy clay loam, 1.24) is introduced to correct the influence of wall flow, which reduces the errors of the heat pulse measurements to within 5%. Further more, we demonstrate that the "reduced convection" assumption of Ochsner et al. (2005) can be explained by the influence of wall flow on water flux density. Including a thermal dispersion term in the conduction-convection heat transfer equation does not improve the model significantly.(3) Statistical examination reveals that the lower detection limit of the heat pulse method is soil texture dependent, i.e., 6.630 × 10^-7 on the glass beads, 1.010 x 10"6 on the sandy loam, and 1.657 × 10^-6 m s^-1 on the sandy clay loam.(4) The thermal dispersion coefficient (\) is found to be a function of the soil texture and water flux density. >y increases with /. A clear linear relationship between the solute dispersion coefficient and the water flux is also observed.(5) There is a power function relationship between thermal dispersivity and water flux density. When water flux is less than 0.7 x 10"5 m s"1, thermal dispersivity decreases significantly as water flux increased. However when water flux is greater than 0.7 x 10"5 m s'\ thermal dispersivity does not chang as water flux increased. The thermal dispersivity of different soil texture is: sandy clay loam (0.074⁰.084cm) > sandy loam (0.042-0.05 lcm) > glass beads (0.038⁰.047cm). On the other hand, the solute dispersivity dose not change with water flux but depends on soil texture (sandy clay loam, 0.2⁰.3cm > sandy loam, 0.1⁰.2cm > glass beads, 0.04⁰.05cm).(6) The magnitude of the thermal dispersivity and solute dispersivity depends on the range of water flux density. In the lower water flux range (J< 0.7 x 10"5 m s"1), thermal dispersivity is significantly larger than the solute dispersivity. However when the water flux density is greater than 0.7 x 10"5 m s'1, thermal dispersivity and solute dispersivity do not vary with the water flux. At this time, thermal dipersivity and solute dispersivity are at the same magnitude. But the absolute value of solute dispersivity is greater than the thermal dispersivity.