The tropical Atlantic circulation:a comparision between ORCA025,

ORCA05 and FLAME 1/12°

Drakkar-Meeting Grenoble, 25.1.-26.1.2007

Sabine Hüttl

1) How realistic is the mean tropical Atlantic circulation in ORCA025 compared to observations, ORCA05 and FLAME?

zonal section along equator meridional section at 23°W section along 5°S

2) How fits the variability of the interannually forced runs into the results from the ORCA05 runs and similar experiments within the FLAME framework ?

Questions to answer:

23°W

0°N5°S

Section along the equator

FLAME 1/12° ORCA05

ORCA025

0.70.8

0°N

EUC core depths more realistic in ORCA than in FLAME (see Schott et al. 2003 and 2005, Brandt et al. 2006)

similar velocities and strength, stronger eastward vertical inclination of EUC in ORCA

Meridional section at 23°W

observations by Brandt et al. 2006,11 sections

EUC:13.8 Sv

SEUC:7.0 Sv

EUC: 12.13 SvSEUC: 2.61 Sv NEUC: 2.43 Sv

Meridional section at 23°W

EUC: 12.13 SvSEUC: 2.61 Sv NEUC: 2.43 Sv

EUC: 18.44 SvSEUC: 0.00 Sv NEUC: 0.00 Sv

Meridional section at 23°W

EUC: 19.89 SvSEUC: 1.94 Sv NEUC: 5.71 Sv

EUC: 12.13 SvSEUC: 2.61 Sv NEUC: 2.43 Sv

Mean annual cycle of EUC transport 23°W

G70KAB001KAB002ORCA05FLAME 1/12°

transport maximum in autumn (response to zonal wind, see Arhan et al. 2006) in all models, second maximum in spring: not in ORCA05, in G70 only weak max. FLAME (and OCCAM ¼°) maxima occur one month earlier

amplitude ann. cycle

8 Sv10 Sv10 Sv14 Sv8 Sv

ORCA05 shows too strong annual cycle ORCA025 and FLAME in good accordance to results from OCCAM ¼° (see Hazeleger et al. 2003)

mean

16.91 Sv19.89 Sv19.89 Sv18.44 Sv12.13 Svtr

ansp

ort

in S

v

Interannual EUC transport variability 23°W

G70KAB001KAB002ORCA05FLAME 1/12°

Most observational and model studies agree, that interannual EUC variability is mainly wind-driven by local surface Ekman divergence. This implies:

differences between different experiments with different surface forcing/parameters:

KAB001 and KAB002 differ in amplitude (different strength of surface restoring) KAB001/002 and G70 show different behaviour

between variability of the ORCA025-runs, ORCA05 and FLAME 1/12° even the sign is unclear !

interannual anomalies EUC 23°W in Sv

tran

spor

t in

Sv

Western boundary regime 5°S

observations by Schott et al., 2005

NBC: 26.5 SvDWBC: 25.5 Sv

ORCA025KAB001

NBC: 28.65 SvDWBC: 24.01 Sv

NBC

DWBC

Shallow western boundary current 5°S

FLAME 1/12° ORCA05ORCA025KAB001

core next to western boundary is realistic only but respresented only in FLAME ! mean northward NBC from measurements by Schott et al., 2005: 26.5 Sv differences between ORCA025-runs? (two-core structure around 200m in KAB001 looks strange…)

NBC:29.17 Sv

NBC:21.8 Sv

NBC:28.65 Sv

Shallow western boundary current at 5°S

KAB001/002 G70

NBC:28.65 Sv

NBC:30.61 Sv

both currents are too strong and show high velocities only near the surface G70 shows stronger current core velocities (around 1.4 m/s) but core still located at the shore

Annual cycle & interannual variability NBC 5°S

G70KAB001ORCA05FLAME 1/12°

only KAB001 & ORCA05 show similar annual cycle (offset)

G70 and FLAME 1/12° totally different structures !

interannual variability of KAB001 and FLAME 1/12° shows some similarity for the last 10 years

ORCA05 and G70 do their own…

tran

spor

t in

Sv

NBC transport anomaly at 5°S in Sv

Conclusions1) Mean tropical Atlantic circulation in ORCA025, ORCA05 and FLAME

EUC structure and velocities quite realistic in all ORCA runs western boundary current (NBC) in all ORCA runs not seperated from the coast as observed ! higher resolution ? mean transports of EUC and NBC in most ORCA runs higher than observations, small-scale zonal currents only in ORCA025 and FLAME, but transports too low

2) Annual cycle and interannual variability

EUC phases of the annual cycle of the EUC quite robust between different ORCA025- runs and FLAME but amplitudes differ, ORCA05 not able to reproduce observed EUC annual cycle interannual variability depends on strength of surface restoring and wind-forcing NBC annual cycles between KAB001, G70 and FLAME not comparable ! interannual variability also different…