1D interpolation of 3D matrix

Hi

I have a question regarding interpolation of 3d matrices.

So lets say I have A = [X Y Z] =

A(:,:,1) = [1, 1, V; 
            1, 7, V;
            NaN,NaN,NaN
            NaN,NaN,NaN]; 
A(:,:,2) = [2, 1, V;
            2, 3, V;
            2, 5, V;  
            2, 7, V];
A(:,:,3) = [3, 1, V; 
            3, 4, V;
            3, 7, V;
            NaN,NaN,NaN];

• V is in this case just a measured value like temperature or velocity that changes in each point.
• X and Y is a coordinate position and each matrix is fixed in X and has same length in Y but different resolution.

So I want to interpolate each matrix to have the same resolution. Most preferrably I only want the interpolation to be a linear interpolation of V considering the Y position and not X position.

How do I go about this? I would know what to do if I had a 2d matrix and only one value of V in each matrix, however I have stored 20 different values of V. Also for my problem I have a set of 21 and not 3 different matrixes where 3 different resolutions are used. Do I need to separate the parts of the matrix with different resolution or can I interpolate with everything stored in the same 3D matrix?

Thanks in advance!