- journal_title：Vadose Zone Journal
- Contributor：Thomas Wöhling ; Niels Schütze ; Ben Heinrich ; Jirka Šimůnek ; Gregory F. Barkle
- Publisher：Soil Science Society of America
- Date：2009-
- Format：text/html
- Language：en
- Identifier：10.2136/vzj2009.0040
- journal_abbrev：Vadose Zone Journal
- issn：1539-1663
- volume：8
- issue：4
- firstpage：1051
- section：ORIGINAL RESEARCH
Using numerical simulation, we analyzed a novel vadose zone sampling system installed in a vadose zone of volcanic origin in the Lake Taupo catchment, New Zealand. The system comprises 15 Automated Equilibrium Tension Lysimeters (AETLs) installed at five depths. The vacuum in each individual lysimeter is controlled to match the pressure head measured at a reference location in the undisturbed vadose zone. The three-dimensional numerical flow model HYDRUS-3D was used to investigate the impact of flow impediment and shadow effects caused by the installation and operation of the AETLs. The analysis was conducted for steady-state conditions in homogeneous sand, loam, and ignimbrite, as well as in the layered materials present at the experimental site. The hydraulic properties of the different layers were estimated using one-dimensional inverse modeling and global optimization techniques. Horizontal, vertical, and radial two-dimensional sections through the model domain were analyzed for divergence from steady-state conditions. The results suggest that the sampling efficiency of the AETLs is relatively large and the errors caused by the setup and operation of the vadose zone sampler are relatively small. We further utilized a transient HYDRUS-3D model of the vadose zone sampler to simulate transient tensiometric pressure heads using field data. The simulations for the various depths compared favorably with both the measured pressure heads and the simulations of the one-dimensional model. In contrast, the simulated cumulative water fluxes during the evaluation period were larger than the measurements of the corresponding AETLs. These findings highlight the importance of an adequate model structure and information content of the calibration data for the accurate prediction of vadose zone water fluxes.