Objective: The lab's main objective is examining how the several software packages (Surfer, ArcGIS (Geostatistical Analyst), and R (geoR?) present variograms by examining a couple of our data sets.
Remember: You also have documentation for the Geostatistical Analyst (ESRI) on CD and another version via the interactive tutorial within the extension. Surfer has some useful on-line documentation. For R you have the D&R text and Bivand, Pebesma, and Gomez-Rubio (2008) chapter 8. And the web is out there...
Data: Based on feedback from the first lab, let's hold the data to just two variables this time:
Task: Load each of the datasets into each of the software packages and examine the access to variograms each makes available. Look for how to control "lag width", "number of lags", "maximum lag distance", "direction", "angle tolerance", and which variogram you are seeing. Can you 'scale' the variograms to help compare the two different spatial processes? In short, we are continuing with step A.7 and toward A.8 from the "Aide-memoire for Spatial Analysis", Appendix A from the W&O text.
Surfer. Grid -> Variogram -> New Variogram. Right-click the variogram and select "Properties".
Arc. Be sure your extensons are licensed, available and on. If you need to increase the memory space for more than the default number of point pairs, running...
and, in the Geostatistical tab, resetting the number of points, and then restarting Arc should do it. Expand the variogram panel. See what's different in the "wizard". Note the linkage of displays.
R. I'm still new on this but try the following, being aware that you probably have named data differently from what I have, and that you need to remember to set your working directory.
require(geoR) library(lattice) maas <- read.geodata(maas, header=TRUE, coords.col=1:2, data.col=3:11) plot.geodata(maas, col.data="Cu") ??variogram help(plot.variogram) help(variog) myvg <- variog(maas, data=maas$data[,"Cu"]) myvg plot.variogram(myvg) myvg2 <- variog(maas, data=maas$data[,"Cu"], option="cloud") plot.variogram(myvg2) myvg3 <- variog(maas, data=maas$data[,"Cu"], direction=pi/3, tolerance=pi/16) plot.variogram(myvg3) myvg3
Deliverable: A brief (maximum 5 pages) report addressing the following points:
Due: in two weeks.