Well, the way to hold one covariance parameter fixed while the other(s) vary is to use the PARMS statement with the HOLD option.  To get the likelihood profile, you could step through values for the other covariance parameters (also in the PARMS statement) and use the NOITER option.  All of the likelihoods could then be output via ODS, and could then be plotted.   SteveDenham ... View more