The measuring of compound dunes – when height becomes a matter of perspective

Subaqueous bedforms are a fascinating morphological feature that concerns natural scientists and engineers alike. Under certain conditions, the different scales of these natural seafloor patterns merge into compound dunes consisting of large-scale primary and superimposing secondary bedforms. When it comes to the measuring of these composites, however, scholarly opinion varies depending on the investigator’s perspective. Specifically, compound dunes can either be interpreted as a superposition of their respective constituents, whose individual heights are measured independently after mathematical disintegration, or as one coherent bedform with readily measurable extents. Both methodologies, undoubtedly, have fully legitimate scopes of application, but little is written about the actual discrepancy that can result from signal pre-processing or differing geometric height definitions.

We experienced this problem when recently validating a method for the decomposition of compound dunes by comparison with three alternative approaches, of which two relied on detrending the bed elevation profiles before examination, whereas the third approach (similar to the newly proposed one) assessed unfiltered profiles. Although all tools were applied to the same bathymetric raw data, the statistical values of obtained dune dimensions diverged significantly. Even between approaches that generally showed comparable mean dune lengths, the corresponding height values differed by a factor of 2 or so. These results suggest that detrending or band-pass filtering of bed elevation profiles, as it is commonly applied before dune identification, leads to a systematic underestimation of profile amplitudes and thus dune heights. We therefore recommend refraining from these pre-processing steps in all cases where unambiguous absolute heights are needed. Dune identification from unfiltered bed elevation profiles, in return, necessitates that dune dimensions are calculated in consideration of the inherent inclinations. When analyzing the respective behavior of primary and secondary bedforms and their complex interplay, however, mathematical disintegration is the method of choice and, accordingly, dune height remains a matter of perspective.