Characterizing and mapping the seafloor and its features requires collecting and analyzing datasets of varied types and scales. Remotely-sensed data can increase our understanding of seafloor processes by providing insight into seabed geomorphology, substrate characteristics, etc., over large areas. Collecting direct observations or physical samples of the seafloor, also known as ground-truth data, holds the key to interpreting and validating remote sensing data. In this thesis, the case is made that by concurrently analyzing some or all of these datasets, knowledge about seafloor processes is gained more efficiently and to demonstrate this a suite of software tools has been developed to support seafloor habitat mapping. 

In Paper I, novel ways to use Digital Bathymetric Models (DBMs) and their derivatives to describe bedforms are proposed. In particular, a spatial analysis procedure for the quantitative characterization of large, straight, isolated (LSI) bedforms found in the Great South Channel (GSC) is presented. The procedure is objective and repeatable. This study led to an increased understanding of the large, straight, isolated bedforms of the GSC, including the geological processes associated with them. It was also discovered that these bedforms do not migrate, contrary to previous thought. This discovery prompted the study of the role of these bedforms as habitat, which is described in Paper II, where the outputs from the previous analysis are complemented with co-registered seafloor images. 

Paper III, lastly, directly addresses the need for a unified software platform for collecting and concurrently analyzing such diverse datasets. In this paper, I present Groundtruther, a toolbox for linking marine datasets in space and time, thus enabling the user to interact with multiple datasets simultaneously, with the aim of obtaining a synoptic view of a particular portion of the seafloor. The contributions from this thesis feed directly into spatial planning and ecosystem-based management.