! Define your variables to be correlated let p = sst let q = wind ! (or what ever the variable names are) go variance contour correl ! (contours the correlations) This will give simultaneous correlations. To get lag and leads use the transformation @shf when defining p or q. For example: Let q = wind[L=@shf:5] The 6th time step in the original time series of q now becomes the first etc. Thus p (sst) leads q (wind) by 5 when you do the correlation. You can use negative numbers also.
So, Simple. 🙂