Numerical test of the laboratory evaporation method using coupled water, vapor and heat flow modelling

Titelangaben

Iden, Sascha C. ; Blöcher, Johanna R. ; Diamantopoulos, Efstathios ; Peters, Andre ; Durner, Wolfgang:
Numerical test of the laboratory evaporation method using coupled water, vapor and heat flow modelling.
In: Journal of Hydrology. Bd. 570 (2019) . - S. 574-583.
ISSN 0022-1694
DOI: https://doi.org/10.1016/j.jhydrol.2018.12.045

Angaben zu Projekten

Abstract

Laboratory evaporation experiments are used to determine soil hydraulic properties (SHP). In most cases, data are evaluated with the simplified evaporation method (SEM). Numerical simulations were used before to quantify the accuracy of the SEM and it was found that the method yields accurate estimates of SHP. However, previous tests neither accounted for heat flow, nor thermal fluxes of liquid water and water vapor, nor temperature effects on the transport properties. Since evaporation experiments are under most circumstances non-isothermal, past studies were therefore oversimplified and likely inaccurate. The objective of this article is to test the accuracy of the SEM using numerical simulations with a coupled model of water, vapor, and heat flow which is based on the Philip-de Vries theory and solves the surface energy balance. The model provides a state-of-the-art description of the fluxes of water, vapor and energy during laboratory evaporation from bare soil. We present simulation results for different soil textures and resistances to vapor flow between soil and air, and analyze the accuracy of the SEM using the simulated data. The resulting average error for the water retention curve is smaller than 0.0025 m3 m−3 and the relative error of hydraulic conductivity ranges from 5 to 15% for sandy loam and clay loam. For sand, the error in conductivity is higher but the structural shape of the conductivity curve is still identified relatively well. Compared to previous analyses of the evaporation method assuming isothermal flow, the average error of the SEM turned out to be only slightly higher.