ABSTRACT
Harmonic and spectral analysis techniques are nothing more than a set of procedures based upon a particular transformation of data. Common to any transformation of this type the purpose is to rearrange the original information in such a way that it can be more easily manipulated or more easily interpreted. In its simplest and basic form the Fourier transformation of a two-dimensional field involves the fitting by least squares, of a set of parallel sinusoidal waves of varying wavelength and orientation. Many of the calculations performed on spatial data such as the application of a weighting function and the comparison of patterns for identification involve a convolution. As indicated in the discussion of the formulae for differentiation the results are very dependent upon the higher frequencies. This usually means that, for undulating relief, nearly every grid point will have a slope which is significantly different from the adjacent ones.
