Is it possible to use surface curve fit generated fittedmodel to normalize other set of data

Hello,

I need to do some normalization of my data. The set of data I have has references and actual samples. I used references (4X2 data points) to generate surface curve fit and fittedmodel, and I was wondering whether there is a way to normalize the sample data with that model. It is 3D data, with X, Y, and Z, and I only need to normalize the Z. For the sample, Y is the same as the references, but X is little different (might need to do some prediction as well??). Also, the value for the Z is bigger for the reference but smaller for the samples. I just need to use either the model, equation, or magnitude for the normalization.

Thanks,