Oberdörster, Christoph: Hydrological Characterization of a Forest Soil Using Electrical Resistivity Tomography. - Bonn, 2010. - Dissertation, Rheinische Friedrich-Wilhelms-Universität Bonn.
Online-Ausgabe in bonndoc: https://nbn-resolving.org/urn:nbn:de:hbz:5N-22100
@phdthesis{handle:20.500.11811/4217,
urn: https://nbn-resolving.org/urn:nbn:de:hbz:5N-22100,
author = {{Christoph Oberdörster}},
title = {Hydrological Characterization of a Forest Soil Using Electrical Resistivity Tomography},
school = {Rheinische Friedrich-Wilhelms-Universität Bonn},
year = 2010,
month = jul,

note = {An explicit knowledge of soil properties is required in agronomy, nature conservation, and hydrology to characterize water storage and water flow processes, even more in the context of climate change. Electrical resistivity tomography (ERT) has become a more frequently used method in soil science and hydrogeology to obtain this information since the bulk soil electrical conductivity, σb, derived from ERT is directly linked to the soil water content, θ.
In this work, a field plot (10 m x 10 m) which was located in a forest on the premises of the Forschungszentrum Jülich (Jülich, Germany) was equipped with 36 boreholes to investigate the soil hydraulic properties of a forest stand by means of ERT.
First, the impact of the ERT data errors on σb was analyzed. A synthetic experiment was performed to clarify whether there is a significant difference between inverted ERT data sets once produced from a water saturated soil profile, and once from a drier profile. The related ERT data were noised in the framework of a Monte Carlo approach by means of authentic error distributions derived from field measurements. Different error models were used within the consecutive inversion process. It became obvious that data errors propagated ruthlessly into the final model, leading occasionally to an overlap of resulting σb distributions related to dry and wet soil conditions, respectively. The results of this study suggested to evaluate data errors precisely. If possible, data errors should be detected in dependence of the corresponding measurement geometry.
Additionally, a long-term study was performed in the field to monitor changes in soil water content by means of ERT. A period of dewatering was chosen to calibrate the relationship between σb obtained from ERT and θ derived from TDR. This petrophysical relationship was used to derive water contents in an ERT image plane for a period of nine months. The plausibility of the imaged spatial distributions of soil water content changes could be verified by different independent measurements (e.g., by TDR). The agreement with those measurement techniques as well as the plausibility of spatial soil water changes caused by root water uptake of the trees demonstrated the additional benefit when a median filter was applied to noisy time-lapse inversion results.
Finally, a saline tracer experiment was performed in order to investigate the transport behavior of the soil. To parameterize solute transport processes, the convection-dispersion equation (CDE) and the mobile-immobile model (MIM) were fitted to ERT and TDR data. Although σb derived from ERT was lower than TDR measurements in almost all depths, estimated pore water velocities of the CDE model were very similar. Early peak arrival times at lower depths and long tailings of the breakthrough curves (BTCs) clearly indicated preferential flow phenomena which could not be described with an appropriate parameterization using classical transport approaches such as the CDE. Also the adaption of the MIM model did not lead to more reasonable solute transport parameters. However, typical features of preferential transport could be detected and the spatial variability of the preferential transport process could be imaged by ERT.},

url = {https://hdl.handle.net/20.500.11811/4217}
}

The following license files are associated with this item:

InCopyright