Another Place for Physical Oceanography: Quantifying Connectivity in the Coastal Ocean

Another Place for Physical Oceanography: Quantifying Connectivity in the Coastal Ocean

Mitarai, S., Siegel, D.A., Watson, J.R., Dong, C. & McWilliams J.C.

Will be submitted to JGR-Oceans

• Master piece by Sverdrup (1995)

Physical Oceanography for Biology

a) Sverdrup (1955), b) Falcowski et al (1998)

Sverdrup predictionSverdrup prediction

Modern viewModern view

Gyre circulation

Primary production

& fish production

“The physics of the motion of the ocean

is essential for the biology of the ocean”

“Rational explanation of why the ocean

and its contents are the way they are”

Another Place for Physical Oceanography

• Dispersal of fish larvae & its consequences

Caselle & Warner (1996)

Spatial & temporal variations Retention around islands

Species invasion across border?

Vermeij (1994)

Swearer et al (1991)

Another Place for Physical Oceanography

• Spread of pollutants & ecosystem response

e.g., Exxon Valdez Oil Spill (1989)

Peterson et al, Science (2003)

• Following Taylor’s theory (1921)

Lagrangian PDF Modeling

Poulain and Niiler (1989)

Sample drifter trajectoriesRegional eddy diffusivity

Swenson & Niiler (1994)

Question: how useful are they?

• Missing links between applications & PDF methods

1. Release-position dependence

2. Inter-annual & seasonal variability

3. Expectation (long) vs. instantaneous (short) views

4. Coastal connectivity

Circulation Simulations

• Promising, but physics has not been assessed well

Lagrangian PDFs?

e.g., can physical oceanography help connectivity studies SoCal Bight?

Cowen et al., Science (2006)

Simulated connectivity in Caribbean

Goal of This Paper

• Assess Lagrangian PDFs of circulation simulations

Using simulations of Dong & McWilliams (2007)

• Discuss if obtained info may help marine biology

e.g., fish population connectivity

e.g., for designing drifter experiment / gene connectivity

Circulation Simulations

• Dong & McWilliams (2007)

Eulerian fields have been validated using available data set

6-hourly mean flow fields are generated from 1996--2000

• Release a large number of Lagrangian particles

From nearshore (= within 10 km from coast)

• Assess Lagrangian PDFs

Conditioned upon release position & time

Lagrangian Particle Tracking

• Nearshore waters are delineated into 137 sites

Cover most of waters 100 m of shallower

Sites & Bathymetry

• A turbulent dispersal problem

Eddy-driven; dispersal patterns change depending on release times

Sample Particle Trajectories

• Spread out in 30 days

Nearly isotropic (no strong directionality)

A Sample Lagrangian PDF

Make white circles biggerMake white circles bigger

• Shows good agreements with drifter observation

Lagrangian PDF in the previous page can be reproduced

Lagrangian Time & Length Scales

[ add cases with different release times & locations? ]

• Some Lagrangian PDFs show non-diffusive patterns

Simple Gaussian scaling won’t reproduce these

Release-point Variability

Poleward transport

Eddy retentionMake white circles bigger

San Diego -> OceansidePalos Verdes

Should I show eddy motion vector plots?

Make white circles bigger

San Diego -> OceansidePalos Verdes

Should I show eddy motion vector plots?

Seasonal Variability

Clear poleward transport

Stormy poleward transport (due to stormy wind)

No poleward transport (due to strong equatorward wind)

Rather consistent

Degree of eddy retention varies, though

Inter-annual Variability

Strongest poleward transport(El Nino)

No signals of poleward transport(La Nina)

Degree of eddy retention is rather

consistent

Example: Site Connectivity

• Local connection to larval pool depending on PLDs

“Larval pool” for longer PLDs

“Self settlement” for shorter PLDs

Along-shore transport

Eddy retention

Summary

• Three distinctive dispersal patterns

Poleward transport, eddy-retention & isotropic spread

• Strong seasonality in poleward transport

Weak seasonality in eddy-transport

• Strong inter-annual variability in poleward transport

Weak inter-annual variability in eddy-retention

• Connectivity as a function of transport time

Local connectivity ~ 10 days

Uniform connectivity ~ 30 days

Help MPA Design

• Which site is an exporter?

Habitats along mainland for longer PLD

• Higher settlement expected for Northern Islands

Because of poleward transport + eddy retention

Expected Settlement

Species Invasion through Borders

• Lagrangian PDFs clearly show the sign

From mexico along the mainland in El Nino years

But, not to N. Channel Islands

Genetic Structures?

• May be hard to find them for a longer PLD

Because all sites can be well connected for PLD > 60 days

• More structures expected for a shorter PLD

More self-seeding for PLD < 10 days

Caveats of PDF Methods

• Stochasticity will have important consequences

e.g., for dose-response, species coexistence

Acknowledgement

• Blah, blah, blah,...

Appendix: SGS problem• PDF methods for LES (Pope, 1998)

Filtered density function transport equation (exact form)

Fokker-Planck equationwith models for unclosed terms

Equivalent particle system

[This would be a separate paper...]