I'd like to construct a 1D spatial spectrogram for nonuniformly sampled, nonstationary (xy-elevation) data with a wide range of sampling rates (~1-10000 pt/m).
One way I think I could do this without running into problems due to the large number of NaNs in the data is to use Matlab's Lomb-Scargle periodogram function (plomb) in overlapping windows.
Are there any reasons why this might be a poor approach, or any better ways to do this instead?
