Deployment Strategies for Representative Surveys using Passive Drifting Seafloor Imaging Floats

This paper develops a method to simulate the trajectories of Lagrangian imaging floats and evaluate the effectiveness of different deployment strategies when planning sparse sampling surveys of seafloor regions in the order of 1-25 km
²
over periods of several weeks or months. The method combines time-varying current depth profiles with bathymetry to simulate the drift trajectories of platforms that keep a constant altitude off the seafloor. To achieve better targeting of observations, time-reverse simulations are used to determine surface deployment locations that are predicted to pass through specific points on the seafloor. The reward function used to evaluate each deployment plan considers both the total area and distribution of the predicted observations in a region of interest. We assess different observation strategies, comparing stratified and randomised distributions, and investigating the impact of different dive durations, numbers of platforms and seasonal sensitivity on the reward score for four real-world deployment scenarios. The results show that for a given survey duration there exists an optimum dive duration for different depths and sizes of region for efficient representation, and that the sensitivity of the reward to operational choices is higher for smaller survey regions.