Geostatistics and Data Analysis
Code
10666
Academic unit
Faculdade de Ciências e Tecnologia
Department
Departamento de Ciências da Terra
Credits
6.0
Teacher in charge
José António de Almeida
Weekly hours
4
Total hours
68
Teaching language
Português
Objectives
On completion of this module, students should be able to understand concepts of data analysis and geostatistics, namely statistical analysis of geological data, and integrate project teams for this subject.
In particular students should be able to:
- Develop processing and interpretation approaches for preliminary statistical data analysis and summarise results;
- For each particular data set (categorical/numerical variables), select the most adequate statistical tools;
- Analyse data redundancy and representativeness;
- Evaluate spatial patterns of correlation between samples;
- Produce estimated maps of a numerical variable and validates results;
- Report and comment results in technical language.
Prerequisites
Elementary knowledge of probability and statistics.
Subject matter
Introduction to data analysis and geostatistics. Types and strategies for data analysis. Data types, variables and preparation of data files. Categorical and numerical variable types. Georeferenced data. Sampling strategies in Earth Sciences. Sample locations and mapping.
Exploratory data analysis. Univariate analysis: summary measurements (first, second and third moment tools and position statistics) and graphical representations. Bivariate analysis: correlation measurements (covariance, Pearson and Spearsman), bi-histograms and graphical representations.
Multivariate analysis: principal components (PCA), correspondence analysis (CA) and classification.
Distribution laws in earth sciences for common data. Random variables. Normal, lognormal and uniform distribution laws.
Regionalized variables theory. Characteristics of regionalized variables. Random functions. Restrictive hypothesis. 2nd order stationary. Intrinsic stationary.
Spatial analysis: calculation of H-scattergrams, spatial covariance and variograms. Variogram surfaces. Modelling of experimental variograms. Calculation of variograms for regular and irregular sample datasets.
Geostatistical kriging estimation: linear estimator properties, solution of the kriging equations system, and kriging variance. Use kriging for the estimation of points, areas and volumes.
Case studies of application in environmental and earth sciences.
Bibliography
[1] Richard A. Johnson & Dean W. Wichern, Applied Multivariate Statistical Analysis, Prentice Hall, 2002, ISBN: 0-13-092553-5 (paperback).
[2] Amílcar Soares. Geoestatistica para as Ciências da Terra e do Ambiente. IST Press, 2000, ISBN: 972-8469-10-1 (paperback).
[3] Edward H. Isaaks, R. Mohan Srivastava. Applied Geostatistics. Oxford University Press, 1989, ISBN: 0-195050134 (paperback).
[4] Pierre Goovaerts. Geostatistics for Natural Resources Evaluation. Oxford University Press, 1997. ISBN: 0-195115384 (hardcover).
Teaching method
The unit is expected to work with theory sessions and practical sessions of 2 hours each. In theory sessions with all students simultaneously, are taught the theoretical background and for specific subjects, demo problems are solved. The practice sessions take place in the computer lab and are occupied with the resolution, in groups of 2/3 students, of 5 problems of data analysis and geostatistics. All information and presentations are available in the Clip web page.
Evaluation method
The evaluation is of continuous type, with an alternative exam after the period of classes. Two(2) tests are scheduled: 1) univariate, bivariate and multivariate analysis, 2) variography and estimation, with a duration of 2 hours in the sessions. Tests have a weighting of 30%+30% of the final grade, a total of 60%. The remaining 40% correspond to the presentation of 5 problems to be solved in the practice sessions and are submited for evaluation until the last session of the unit. All students with practice work with a grade higher than 13 are subject to an oral evaluation, without which will have the grade of the work limited to a maximum of 13.