Tracers in Ocean and Climate Models* Matthew England CEMAP, School of Mathematics The University of...

Tracers in Ocean Tracers in Ocean and Climate and Climate

Models*Models*

Matthew EnglandMatthew England

CEMAP, School of MathematicsCEMAP, School of Mathematics

The University of New South WalesThe University of New South Wales* See also www.maths.unsw.edu.au/~matthew/publications.html#MR98

Possible due to: GEOSECS, TTO, SAVE, WOCE, …..

Why bother with tracers in models?

•Ocean model “validation”

•Diagnosis of model circulation mechanisms

•Studies of the ocean carbon cycle

•Data assimilation studies

•Paleoceanographic considerations

•Ocean model “validation” (e.g. CFCs, 14C)

•Diagnosis of model circulation mechanisms (e.g. dye/age tracers, 39Ar)

•Studies of the ocean carbon cycle (carbon compounds, oxygen, phosphate, nitrate,…)

•Data assimilation studies (e.g. CFCs, tritium)

•Paleoceanographic considerations (e.g. carbon-13, oxygen-18)

Why bother with tracers in models?

100-1000 year ventilation

10-100 year ventilation

Robust diagnostic:T-S restored to observed in the

interior

Robust diagnosticObserved 14C

Prognostic

Supressed convection and vertical motion

Prognostic experiment:Interior T-S free to evolve

Toggweiler et al. [1989]

Chlorofluorocarbons

Plate 2. Distribution of CFC-12 on isopycnal surfaces corresponding to maximum NADW outflow in 1988 in the Redler and Dengg [1999] simulations. (a) In the 4/3° model, and (b) in the 1/3° model. The color bar indicates CFC concentrations in pmol/kg, with iso pycnal layer depths contoured (meters).

Ajax section in the South Atlantic

CDW

CDW

AAIW

AABW

Forcing functions for tracers ?

sea-ice

CFC

14C

3He

CFC14C

3He

CFC

CFC14C

CFC 14C

14C

CFC

14CCFC

Air-sea gas flux = f (k, ice, )

k = piston velocity ~ wind speed, U2 or U3

= solubility ~ SST, (SSS)

How to compute gas uptake:

•Use model-generated ice, winds, T-S?

•Use observed ice, winds, T-S?

Tracers in coupled climate models:

Both approaches can give an apparently goodtracer simulation but for the wrong reason

(see England and Maier-Reimer 2001 for details)

Case 1:

Spurious convection

Case 2:

Other tracer techniques:

•Age/Dye tracers

•Tracer data assimilation

•Off-line tracer models

(Cox, 1989, England 1995, O’Farrell 2000….)

(Haine 1999, Schlitzer 1996, …)

(Aumont 1998, Sen Gupta & England 2003)

Off-Line Tracer Model

Tracer Conservation Equation

OGCM

Horizontal

Velocity Fields

Source Terms

Mixing Terms

Tracer Concentration

T (x, y, z, t)

Continuity Equation

u , v w

•Interannual

•Seasonal

•Intraseasonal

•Water-mass source regions

•CFCs, 14C, 3He

•Radioactive waste

T, S

•Pollution, etc….

•Eddy statistics

•Isopycnal mixing

•GM (1990)

•Convective ML

•Wind Driven MLT, S, CFCs, 14C,….

PhD project: Alex Sen Gupta

Example: CFC simulations in a ¼ degree model

Integrated CFC content below 2000m

Year = 1980

PhD project: Alex Sen Gupta

Integrated CFC content below 2000m

Year = 2000

PhD project: Alex Sen Gupta

Integrated CFC content below 2000m

Year = 2020

PhD project: Alex Sen Gupta

Integrated CFC content below 2000m

Year = 2040

Advantages: • Independent assessment of model (complements T-S)• Can validate within ocean-only component• Complements GBC and carbon cycle modelling

Drawbacks:•Extra CPU (especially 14C)•Can manipulate outcome to some extent

Uncertainties:•Tracer source functions•Sparse data •How to formalise assessment?

Summary – Tracers in climate models