Office:Math East 140 (the building just to the east of the Mathematics Building)
621-6859
http://www.u.arizona.edu/~donaldm
Although I am no longer teaching the Geostatistics course I still have an office on campus but my hours are irregular. I do respond to email inquiries Accessing the R software
Binaries are downloadable for Windows, Linux and Unix
Shortlist of most useful R commands
Overview of R and available packages
List of Geostatistics packages in R
Within R (for packages that have been installed), you can bring up a summary by the command "help(package='name of package')"
Tutorial for gstat using Meuse data set
Introduction to geoR with examples
Documentation for Random Fields
VR actually includes several packages, in particular "Spatial"
Two other useful packages
Documentation for ForeignThis package allows you read in files produced by other programs
Documentation for ShapefilesThis package allows you to read into or write Shapefiles (i.e. for ArcGis
The R Language
R Language Definition
R has links to use with GRASS, MySQL, OCTAVE/MATLAB
Matlab "package" in R
Using GRASS with R
GRASS interface in R
GGOBI
Getting ggobi
Binaries are downloadable for Windows, Linux and OSX
Examples and Tutorial
Some tutorials for R
The R language
More documentation on R
More R documentation
Although there are many sites on the internet that can useful, there is one general site that will be of particular interest
AI-GEOSTATS
The software to be used in class demonstrations included
(1) GEOEAS, see the software listings under www.ai-geostats.org
GEOEAS was released into the public domain by the US-EPA. It consists of multiple separate DOS programs (but can be run in DOS windows under MS-Windows). Each component has a friendly window interface.
(2) VARIOWIN, see the software listings under www.ai-geostats.org
This package was originally commercial and packaged with a book of the same name, after the printing ran out the author released the code into the public domain. The book (user's manual) is also available as a pdf file. It is a "windows" program and essentially replaces two components of GEOEAS, namely PREVAR and VARIO. Both of which have severe data file size limitations due to the DOS format.
Caution: The pair comparison files generated by PREVAR (GEOEAS) and PREVAR (VARIOWIN) are not compatible
Also see
SAS (commercial)
SAS and GIS
PROC Sim2D
PROC Sim2D
GEOSTATISTICAL ANALYST (Commercial, works with ARCGIS)
hFact Sheet on Geostatistical Analyst
GIS
S-PLUS (Commercial)
There is an add-on for S-Plus called S+SpatialStats
Both R and S-Plus are based on the S language (from Bell Labs)
A Note about geostatistics software conventions
Three sets of class notes are accessible on this website. A current set (which are revisions of the "Old Classnotes". In addition there are the "Old Classnotes" and then the "Oldold Classnotes"
SYLLABUS Spring 2007
Class Notes Spring 2007
Old Classnotes
Assignments and Project description Spring 2007
SOFTWARE
Prequisites: basic probability/statistics course, matrix theory and computer
experience(DOS, Windows on PC compatibles, UNIX on workstations) No programming was required
for the course.
Geostatistics is the name commonly associated with both the techniques
utilized and the problems/objectives arising out of applications in the earth sciences
where earth sciences is interpreted in a broad sense (including but not limited to
geosciences, hydrology, soil sciences, mining engineering, environmental monitoring
and assessment, atmospheric sciences). In general the topics are oriented to the
analysis of spatially located data and in particular the estimation of values of the
variable or variables of interest (for example ore grades, pollutant concentrations,
hydrologic parameters) at an unsampled point or the average over an area or
volume using data at a discrete number of (possibly) irregularly spaced points.
The course emphasized applying the methods and software for the analysis
of a specific data set as well as becoming familiar with the literature.
The following is a general description of the course, for more detailed information pertaining to a specific semester see the syllabus listed above.
OUTLINE OF CONTENT OF THE COURSE
Introduction
Review of probability/statistics
Overview of problems/objectives
Review of matrix theory
Random Functions
Puntual Kriging
Estimator
Equations
Properties
Use of software
Variograms
estimation/modeling
valid models
problems/difficulties
cross-validation
Block kriging
averaged variograms
regularized variograms
sample support
dispersion variance
Universal kriging
drift
variogram estimation problems
modification of equations
Space-time modeling
product-sum models
fitting space time models
Intrinsic Random Functions
Generalized increments
Generalized covariances
equations, software
Non-linear transformations
log-normal kriging
bias correction
drift problems
variogram modeling
indicators
Comparison with other techniques
Multivariate methods
cross-variograms and cross-covariances
linear coregionalization model
cokriging estimator
cokriging equations
Simulation
L-U
sequential gaussian
simulated annealing
Bayesian Methods
TEXTBOOK
There was no prescribed text, classnotes were be provided. However, students might find one or more the following useful (none will cover all the material nor will any one of these be completely covered)
Mining Geostatistics, A.G. Journel and Ch. J. Huijbregts, Academic Press, 1978
Geostatistics: Modeling Spatial Uncertainty, Jean-Paul Chiles and Pierre Delfiner, J. Wiley, 1999
The Theory of Regionalized Variables and its applications, G. Matheron, Ecole des Mines, Paris, 1971
An Introduction to Applied Geostatistics, Edward H. Isaaks and R. Mohan Srivastava, Oxford University Press 1989
Geostatistics for Natural Resources Evaluation, Pierre Goovaerts, Oxford University Press, 1997
GSLIB: Geostatistical Software Library and User's Guide, Clayton V. Deutsch, Andre G. Journel, Oxford University Press, 1992
Multivariate Geostatistics, Hans Wackernagel, Springer 1995
Geostatistics and Petroleum Geology, M.E. Hohn
Statistics for Spatial Data, Noel Cressie, J. Wiley, 1993
Statistical Methods for Spatial Data Analysis, Oliver Schabenberger and Carol A. Gotway, Chapman & Hall/CRC 2005
Interpolation of Spatial Data: Some theory for kriging, Michael Stein, Springer 1999
An Introduction to Model-based Geostatistics, P.J. Diggle, P.J. Ribiero, Jr and O.F. Christensen (in Spatial Statistics and Computational Methods, J. Mxller (ed), Springer 2003
CONFERENCE PROCEEDINGS
There have been a series of geostatistics conferences and most have resulted in proceedings. These are an additional important set of reference materials
Advanced Geostatistics in the Mining Industry, (1976) D. Reidel Publishing
Geostatistics for Natural Resource Characterization, G. Verly et al (eds) (1984) D. Reidel Publishing
Quantitative Analysis of Mineral and Energy Resources, C.F. Chung et al (eds) (1988), D. Reidel Publishing
Geostatistics (Vols 1 & 2), M. Armstrong (ed) (1989) Kluwer academic press
Geostatistics Troia '92, A. Soares (ed), (1993) Kluwer academic press
Geostatistics Wollongong '96, E. Baafi and N. Schofield (eds), (1997) Kluwer acdemic press
Geostatistics for the Next Century, R. Dimitrakopoulos (ed), (1994) Kluwer academic press
geoENV I-Geostatistics for Environmental Applications, A. Soares et al (ed) (1997) Kluwer academic press
Geostatistical Simulations, M. Armstrong and P. Dowd (eds) (1994) Kluwer academic press
geoENV II-Geostatistics for Environmental Applications, J. Gomez-Hernandez et al (eds) (1999) Kluwer academic press
geoENV III-Geostatistics for Environmental Applications, P. Monestiez et al (eds)(2001) Kluwer academic press
geoENV IV-Geostatistics for Environmental Applications, X. Sanchez-Villa et al (eds) (2005) Kluwer academic press
Geostatistics RIO 2000, M. Armstrong et al (eds), (2001) Kluwer academic press
Geostatistics Banff 2004, O. Leuangthong and C. Deutsch (2005) Kluwer academic press
SOFTWARE
Geostatistics is computationally intensive hence software is essential. The emphasis will be on software that is free or is readily available to students at the university.
The primary software used for class demonstrations was R