What you are saying that you do not think that two different
parameters/variables/sets of measurements are *not* related to each
And you want to plot them on the same graph.
So do that; now if ther are a large number of samples for each
parameter, then use time as the dependent variable: plot each point so:
first sample of X on X axis and first sample of Y on Y axis; second
samples of X and Y in same manner, etc.
You have preserved ordering of the samples, and one set could have
been measured last year and the other set taken recently.
Now if they are truly in-related, you will have a scatter-plot tha
approaches a uniform greyness (black dots, white background) where
limits on one axis cannot guarantee a "reasonable" percentage of points
in the other axis, other than limits beyond three (or more) sigma of
well-behaved (ie gaussian) processes - which is meaningless.