Untitled Document

We hope that our methodology will prove useful for use with other pelagic species. As a result the model code used in this study will soon be made available for use by the scientific community. Please contact us for further information.

The dispersion of Lagrangian particles is simulated over a global domain using an off-line ocean model. Advection uses monthly averaged horizontal velocity fields from the POCM OGCM. The POCM fields are at 'eddy permitting' resolution (an average of ¼º x ¼º gridboxes). The displacement of a particle over the time step dt (=6 hrs) is simply dx=U x DT, where U is calculated using a linear averaging of velocities from the 3 closest velocity gridpoints (an equivalent relation holds in the y-direction). Where a component of the displacement would result in a particle becoming grounded, that component is ignored. As a result all particles remain active over their lifetime.
Much of the ocean dynamics important to particle dispersion takes place at scales below those resolved by the POCM advection field. In addition some transient eddies simulated by the POCM model would become 'averaged out' when taking monthly averages. (NB each advection field file requires a substantial amount of storage space and it would be impossible to store advection fields at each POCM timestep). As a result it is common practice to simulate the mixing effect of sub-grid scale process by a parameterized mixing term. Here we use a random walk parameterization that in effect simulates a constant horizontal eddy-diffusivity
.
Basic operation allows nmult multiple runs (life span segments) of no_years years (default=2 yrs) , where each Lagrangian particle has a maximum lifespan, max_age (default=1 yr). The difference between these allows initial release of particles over the period of a year (since we have different velocity fields for each month). By setting nmult=1 and increasing no_years & max_age we can see the long term evolution of the Lagrangian particles.


*Migration of Aurelia aurita* in a Lagrangian Ocean Circulation Model**
1. Introduction 2. Model 3. Animated sequence - Japan release - 1000 yr integration
In collaboration with Mike Dawson & Matthew England (UNSW)
1. Introduction
The Aurelia aurita jellyfish is located at numerous sites around the world, including Australia, Japan, Western USA and France. Genetic data provides evidence that the species originates in Japan. Here, high levels of genetic diversity are consistent with a naturally evolved population. Genetic data from the other sites are more indicative of introduced populations.

	Here we investigate the possibility of migration of aurita between the various sites by natural means, traveling across the oceans and around coastlines.
Model Information
We hope that our methodology will prove useful for use with other pelagic species. As a result the model code used in this study will soon be made available for use by the scientific community. Please contact us for further information. The dispersion of Lagrangian particles is simulated over a global domain using an off-line ocean model. Advection uses monthly averaged horizontal velocity fields from the POCM OGCM. The POCM fields are at 'eddy permitting' resolution (an average of ¼º x ¼º gridboxes). The displacement of a particle over the time step dt (=6 hrs) is simply dx=U x DT, where U is calculated using a linear averaging of velocities from the 3 closest velocity gridpoints (an equivalent relation holds in the y-direction). Where a component of the displacement would result in a particle becoming grounded, that component is ignored. As a result all particles remain active over their lifetime. Much of the ocean dynamics important to particle dispersion takes place at scales below those resolved by the POCM advection field. In addition some transient eddies simulated by the POCM model would become 'averaged out' when taking monthly averages. (NB each advection field file requires a substantial amount of storage space and it would be impossible to store advection fields at each POCM timestep). As a result it is common practice to simulate the mixing effect of sub-grid scale process by a parameterized mixing term. Here we use a random walk parameterization that in effect simulates a constant horizontal eddy-diffusivity . Basic operation allows nmult multiple runs (life span segments) of no_years years (default=2 yrs) , where each Lagrangian particle has a maximum lifespan, max_age (default=1 yr). The difference between these allows initial release of particles over the period of a year (since we have different velocity fields for each month). By setting nmult=1 and increasing no_years & max_age we can see the long term evolution of the Lagrangian particles.