A particle tracking model to analyze transport in ocean

Abstract

The ocean circulation is the global movement of water, which occurs in the ocean basins. On the one hand, the ocean currents distribute heat from the equator to the poles and on the other hand, they transport all kind of materials like debris and organisms like nutrients or larvae from their origin to almost anywhere in the planet. Ocean circulation can be classified into density driven deep ocean circulation and wind driven surface ocean circulation. The first one act like a conveyor belt moving warm surface water and forcing an upward movement of cold water. The second one is due to the effect of the global wind pattern and originate oceanic gyres at both hemispheres. The aim of this thesis is to contribute to the knowledge of the drift of particles into different ocean currents using lagrangian simulations. Lagrangian models are a useful tool to analyze pathways in the ocean by using a set of virtual particles, which are transported by the ocean currents. The main variable to run the lagrangian model is the velocity field provided by databases. Here, we will use velocity data from SODA and HYCOM along with other variables like sea surface temperature (SST), radiance, Mississippi and Atchafalaya river discharge, wind, salinity, water temperature and data of the teleconnection patterns NAO and AMO. The changes in the probability of larvae crossing the North Atlantic Ocean over the period 1899-2010 was analyzed in the first study case using lagrangian trajectories of passive tracers. Virtual particles were released in the Strait of Florida where the main driving forcing is the Gulf Stream. An increase in the chances of crossing the Atlantic was obtained considering species with long planktonic larvae duration. In addition, a minimum travel period around 6-7 months was calculated. The increase was related with the intensification of the changes in eddy kinetic energy along the Gulf Stream path. In the second study case, the minimum migration duration of the European eel was analyzed in order to obtain similar results than those from the microstructure of eel otoliths, which are around 7-9 months. The lagrangian simulations were tested under different conditions like spatial and time resolution, depth, release area and also initial distribution. Overall, the results showed faster migration when increasing the resolution of the model and also when decreasing the release depth. In addition, the fastest migration was obtained when the virtual particles were passive tracers horizontally advected by the currents. However, the migration strategy with downstream swimming resulted in faster migrations. The continuous displacement of water masses due to ocean currents contributes to larval dispersion allowing connectivity among populations from different geographic areas. In the third study case, the connectivity among five populations (Red Sea, Madagascar, Maldives, Java and Aceh) of spiny lobster Panulirus penicillatus in the Indian Ocean was analyzed using numerical simulations of passive tracers over the period 1858-2008. Biological parameters of the spiny lobster as release depth and areas, larval duration and spawning season were selected and used in the lagrangian model. The results showed how the Panulirus penicilatus larvae moved among different regions in the Indian Ocean, favoring the genetic connectivity among them. Specially, two significance positive trends were obtained for particles that reached Aceh from Maldives and for particles that reached Java from Aceh. Finally, the sea surface temperature pattern in the Texas-Louisiana shelf as well as the influence of the plume formed by the Mississippi-Atchafalaya (MA) river system on it were investigated. The lagrangian model was one of the methods applied in this study case to analyze the current in the area under scope. Passive particles were released at each river mouth over the period 1993-2015. The results showed a buoyancy lost in the area occupied by the MA river system turbid plume due to the reduction of westward transport of freshwater, the reduction of river discharge and the decrease in the frequency of easterly winds. This buoyancy lost causes a higher vertical mixing favoring the entrainment of colder water, which explain the SST decrease in the area under study. In conclusion, this thesis not only proves the interest of lagrangian models to describe transport of larvae into ocean currents or connectivity among different areas but also proves that they constitute a complementary tool for climate studies.