It Is difficult to prove a direct relationship between statistical 

 relief and combined process and material, but an Interesting Insight can 

 be gained from a study of a highly variable environment, sedimentary 

 micro topography. In 1981, Mark Wlmbush of the University of Rhode 

 Island deployed stereo camera on a structure on the upper continental 

 rise northeast of Cape Hatteras. Stereo-pair bottom photographs were 

 taken at an Interval of twenty-seven days. Time-lapse photography 

 showed that between the dates of these stereo-pair photographs, the fine 

 scale sea-floor relief beneath the structure was altered both by biolog- 

 ical activity and episodic bottom current events. 



Two mlcrorellef maps were generated from the stereo-pair Images and 

 these are Illustrated In Figures 5-3 and 5-4. Transects of heights were 

 taken at 0.5 cm Intervals (labelled A, B, C, D) across the surface. 

 Amplitude spectra all showed the power law form found In spectra at 

 lower frequencies, and In addition the spectral parameters showed no 

 significant differences In spite of the gross change In the surface (see 

 Figure 5-5). This Implies that the two surfaces merely represent two 

 realizations of the same statistical process. In terms of frequency 

 domain analysis, only the phase spectrum Is altered by the redistribu- 

 tion of features, not the amplitude spectrum. This simple experiment 

 does not tmequlvocally prove a causal relationship between statistical 

 relief and process, but It does provide an encouraging result. 



The Phase Spectrum 



To reconstruct a profile or surface from Its frequency domain repre- 

 sentation. It Is not sufficient to model only the amplitude of each com- 



56 



