See more from this Session: Complexity - Linked Nonlinear Processes

Enhanced water vapor diffusion under temperature gradients has been proposed as a mechanism to explain the discrepancies between measured and predicted water fluxes in soils. Due to the difficulties in measuring soil vapor diffusion directly, modeling approaches have been used to estimate the vapor enhancement factor (η) by matching theory to measurements. In the method proposed by Hiraiwa & Kasubuchi (2000), soil thermal conductivity associated with conduction heat transfer (λ_{c}) is assumed to be equal to the apparent soil thermal conductivity (λ) measured at a low temperature, and η is significantly underestimated. In this study, an improved approach for estimating η is used in which λ_{c} is taken as the apparent soil thermal conductivity associated with infinite atmospheric pressure. The λ at infinite atmospheric pressure is estimated by extrapolating λ measurements made at finite air pressures. By subtracting λ_{c} from measured λ values at a given atmospheric pressure, the contribution of thermal vapor diffusion to heat transfer (λ_{v}) is obtained and then used to estimate η. On a Lysimeter sand, λ_{v} accounts for 4 to 25%, 8 to 29%, and 13 to 35% of λ at 3.5°C, 22.5°C and 32.5°C, respectively, at soil water contents greater than 0.02 m^{3} m^{-3}. Thus, the latent heat transfer due to vapor diffusion is important even at temperatures as low as 3.5°C. The agreement between predictions from the new method and selected literature values suggests that the improved approach is able to provide accurate estimations of η.

The results from this study show that the magnitude of latent heat transfer due to thermal vapor diffusion is strongly soil texture dependent. Thus, it is important to estimate η on specific soils rather than assuming η from literature values.

See more from this Session: Complexity - Linked Nonlinear Processes