The program for Regularly Spaced Data (gam) is for gridded secondary data (from seismic or remote sensing) or to check the result of simulation. The directions are specified by unit increments in the grid directions. The output from this program is not in GSLIB data format; it is in a format ready for the variogram plotting program.
In general, the primary variable of interest will not fall on a regular grid. The program for Irregularly Spaced Data (gamv) is commonly used for such data. The program will simultaneously calculate an arbitrary number of variograms in an arbitrary number of directions. The lag distance specification, however, is the same for all directions; therefore, multiple executions of the program are necessary in practice. The output from this program is not in GSLIB data format; it is in a format ready for the variogram plotting program.
A Variogram Map (varmap) is useful for determining directions of continuity. After computing and displaying a variogram map it is common to go back and use the gam or gamv program to calculate the variogram in the directions of greatest and least continuity. The output from this program is in GSLIB data format and is intended to be used in pixelplt.
Most geostatistical simulation is based on the Gaussian model. At times this may be inappropriate and one could check the theoretical indicator variograms (that correspond to the Gaussian model) to the actual indicator variograms. The Indicator Variogram Computation (bigaus) calculates the theoretical indicator variograms, which can be plotted with the variogram plotting program.
The variogram file from model (vmodel) program calculates the variogram values for specific distances and directions to be plotted with experimental values for fitting and checking a variogram model. The user must specify the variogram model - this program just calculates the values for plotting. It is common to set up this program and the vargplt program in a batch file so that the plot can be displayed right away after changing a variogram model parameter.