Memory intensive statistical algorithms for multibeam bathymetric data

Abstract A set of algorithms is presented for analyzing and processing the large spatial data sets which are derived from multibeam bathymetry systems. These algorithms are designed to make use of large two-dimensional arrays to enable: (1) the estimation of how many times a particular area has been sampled, (2) the approximation of the bathymetric surface using a weighted running average, (3) the estimation of the data standard deviation continuously over the surface given the assumption of a noisy signal, (4) the breakdown of variance into within-line and between-line variance for situations where multiple ship's survey lines cover the same area and (5) the shoal-biased thinning of data. These algorithms are not restricted to processing bathymetric data, they can be applied to any data which has the property of being (approximately) randomly distributed on a plane with a roughly uniform density.