Download the full-sized PDF of Geostatistics and Clustering for Geochemical Data AnalysisDownload the full-sized PDF



Permanent link (DOI):


Export to: EndNote  |  Zotero  |  Mendeley


This file is in the following communities:

Graduate Studies and Research, Faculty of


This file is in the following collections:

Theses and Dissertations

Geostatistics and Clustering for Geochemical Data Analysis Open Access


Other title
geochemical analysis
Type of item
Degree grantor
University of Alberta
Author or creator
Prades, Carlos
Supervisor and department
Deutsch, Clayton (Civil and Environmental Engineering)
Examining committee member and department
Askari-Nasab, Hooman (Civil and Environmental Engineering)
Boisvert, Jeff (Civil and Environmental Engineering)
Department of Civil and Environmental Engineering
Mining Engineering
Date accepted
Graduation date
2018-06:Spring 2018
Master of Science
Degree level
This thesis addresses challenges in geostatistical analyses of multivariate geochemical data that commonly contain complexities that have a significant influence on geostatistical modeling and cluster analysis. For geostatistical modeling, the effect of the most common despiking methods is investigated and their problems documented. It is shown that both local average despiking and random despiking lead to bias in the observed variogram and predicted uncertainty. A new despiking method is proposed and implemented to improve variography when the variable has a significant spike. The developed approach combines a random despiking component and a local average despiking component. Cluster analysis can be applied for mineral exploration purposes. It can be used to find large structures in the data and also to detect multivariate anomalous samples. Data transformations are shown to have a significant impact on clustering results. Guidance and recommendations on appropriate data transformations for improving cluster analysis performance are provided. Three different methods are developed for identifying multivariate anomalies with cluster and spatial analysis. The first method uses different combinations of clustering and data transformations for finding small anomalous clusters. The second uses different clustering outputs for identifying samples that do not clearly belong to any cluster. The third recognizes samples that are spatially anomalous. Each of these multivariate methods detects anomalies from a different point of view. A combination of these detection methods is recommended. The goal is to obtain more stable and reliable results. Its application in stream silt samples from the Northwest Territories shows that the proposed multivariate anomaly detection methods are capable of identifying several showings (known mineral deposits). Some of these showings are not detected from the histograms of different elements; this supports and motivates the use of multivariate anomaly detection methods for mineral deposit exploration.
This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for the purpose of private, scholarly or scientific research. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.
Citation for previous publication

File Details

Date Uploaded
Date Modified
Audit Status
Audits have not yet been run on this file.
File format: pdf (PDF/A)
Mime type: application/pdf
File size: 9916822
Last modified: 2017:11:08 17:28:12-07:00
Filename: Prades_Carlos_201709_Msc.pdf
Original checksum: ef2c67fff0df39f65528da58e8688f46
Activity of users you follow
User Activity Date