Adrian, Juliane (2017). Interpretation of DC and IP Time-Domain Data Observed on a Copper Deposit in Turkey Using a Newly Developed 2D Finite Element Inversion With Unstructured Meshes. PhD thesis, Universität zu Köln.
|
PDF
DissertationJulianeAdrian2017publicationA.pdf - Published Version Download (186MB) |
Abstract
Within the framework of this thesis, a new two-dimensional inversion algorithm for direct current (DC) resistivity and time-domain induced polarization (IP) data was developed. The algorithm consists of a novel composition of finite element forward calculation and a smoothness constrained local regularization inversion approach including error weighting and a non-linear IP chargeability inversion. All forward and inverse calculations are conducted on unstructured triangular meshes. The advantage of employing an unstructured mesh over the use of structured grids is that complicated surface topography can easily be implemented and complex subsurface structures can be reproduced better. This is among other things due to flexibility and because local mesh refinement is possible without globally increasing the number of elements significantly. An IP perturbation model is implemented that explains the effect of chargeable material by an increase of effective resistivity. Consequently, the inversion is conducted in two steps. The first step is a non-linear DC inversion recovering a resistivity model that corresponds to the effective resistivity and explains the measured DC data within the error bounds. In the second step, a chargeability model is derived by non-linear IP inversion. The employment of unstructured grids pleads for special smoothness constraints because model elements are not arranged in a regular manner and are oriented arbitrarily. Therefore, three types of smoothness matrices are examined that approximate first order derivatives in the triangular mesh. Furthermore, a measure of the depth of investigation is presented that is based on the normalized coverage which is a cumulative sensitivity. A coverage threshold applicable for 2D models is derived by the coverage value of a homogeneous model that corresponds to the depth of investigation according to a rule of thumb. The accuracy of the implemented forward algorithm is verified by modeling of 1D and 2D models and comparison to analytical solutions. By inversion of synthetic data, it is assessed whether the algorithm can reproduce the original models. Reliability is also checked by comparison of the forward and inverse modeling of synthetic data with established algorithms. The new algorithm was successfully applied to DC and time-domain IP field data acquired on a copper ore deposit in Elbistan, Turkey. Due to the state of the ore deposit, which is mainly disseminated, it was expected that the presence of the metallic particles does not produce a prominent resistivity anomaly. Thus, the deposit is difficult to detect with the DC resistivity method alone. However, the assumed sulfide content leads to strong induced polarization effects. These assumptions are confirmed by the resistivity and chargeability models obtained by inversion with the new algorithm. As expected, highly chargeable structures with more than 100 mV/V are recovered that are associated with the ore deposit, while the outline is not clearly distinguishable in the resistivity models. Comparison with inversion results obtained by other algorithms and a near surface resistivity model derived from Radiomagnetotelluric data confirm the resistivity and chargeability results of the new algorithm.
Item Type: | Thesis (PhD thesis) | ||||||||
Translated abstract: |
|
||||||||
Creators: |
|
||||||||
URN: | urn:nbn:de:hbz:38-77980 | ||||||||
Date: | 2017 | ||||||||
Language: | English | ||||||||
Faculty: | Faculty of Mathematics and Natural Sciences | ||||||||
Divisions: | Faculty of Mathematics and Natural Sciences > Department of Geosciences > Institute for Geophysics and Meteorology | ||||||||
Subjects: | Physics Earth sciences |
||||||||
Uncontrolled Keywords: |
|
||||||||
Date of oral exam: | 2 May 2017 | ||||||||
Referee: |
|
||||||||
Refereed: | Yes | ||||||||
URI: | http://kups.ub.uni-koeln.de/id/eprint/7798 |
Downloads
Downloads per month over past year
Export
Actions (login required)
View Item |