A Lagrangian particle random walk model for simulating a deep-sea hydrothermal plume with both buoyant and non-buoyant features

Abstract

This paper presents a computational model of simulating a deep-sea hydrothermal plume based on a Lagrangian particle random walk algorithm. This model achieves the efficient process to calculate a numerical plume developed in a fluid-advected environment with the characteristics such as significant filament intermittency and significant plume meander due to flow variation with both time and location. Especially, this model addresses both non-buoyant and buoyant features of a deep-sea hydrothermal plume in three dimensions, which significantly challenge a strategy for tracing the deep-sea hydrothermal plume and localizing its source. This paper also systematically discusses stochastic initial and boundary conditions that are critical to generate a proper numerical plume. The developed model is a powerful tool to evaluate and optimize strategies for the tracking of a deep-sea hydrothermal plume via an autonomous underwater vehicle (AUV).