Seamount detection and isolation with a modified wavelet transform

The size, shape and number of seamounts, once detected and isolated from other features such as oceanic plateaus or trenches, have the potential to provide valuable constraints on important solid Earth processes, e.g. oceanic volcanism. The variability of seamount size and morphology, however, presents problems for computational approaches to seamount isolation. This paper develops a novel and efficient wavelet-based seamount detection routine ‘Spatial Wavelet Transform (SWT)’; the first use of multiple scales of analysis to directly isolate edifices from bathymetric data. Only weak shape-related criteria are used and no a priori knowledge of the scale and location of the seamounts is required. For a bathymetric profile collected on cruise v3312 SWT matches, to within 25%, the dimensions of five times the number of the features determined by manual inspection than does the best statistically based (e.g. mean, median or mode) sliding window filter. The size–frequency distribution, a key descriptor of seamount populations, is also much better estimated by the SWT method. As such, the SWT represents a step towards the goal of objective and robust quantification and classification of seamounts.