3D Curve/Surface Fitting

Hello all,

This forum has been critical to my Matlab scripting to date, but I have never posted a question myself.

My question is simply how to measure the goodness of fit between two 3 dimensional curves. I have two surface tracks that I currently manipulate using the optimization toolbox, and at the conclusion of my script I would like to generate a reliable measure of how closely they fit together.

Currently I generate an RMS value from the optimization function, but this value does not seem to reliably reflect how well the two curves match. I have explored the Curve Fitting Tool at length, but it does not seem to provide what I need in this situation.

I could possibly simplify this to a 2D problem without sacrificing much accuracy - because they are surface tracks there is very little (ideally no) variation in the x direction. The link below directs to a 2D plot of the data I need to measure fit for. As you can see the plots have a variety of surface features, but are closely matched. From the optimization toolbox these two tracks have a very high RMS, but given the plots I think there is an error in my RMS calculation. Any ideas on how I could simply measure the fit between these two data sets in either 2D or 3D?