Top Banner
The Eastern Mediterranean Transient studied with Lagrangian diagnostics applied to a Mediterranean GCM forced by satellite SST and ECMWF wind stress for the years 1988-1993. Volfango Rupolo a , Salvatore Marullo a and Daniele Iudicone b a C.R. Casaccia ENEA Rome, Italy b, Istituto di Fisica dell'Atmosfera-CNR, Rome, Italy ( subm. to JGR) http://clima.casaccia.enea.it/staff/rupolo After a long spin up phase (‘climatological’ field), the model is forced using daily ECMWF wind stress and relaxing temperature to daily satellite SST for the years from 1988 to 1993. (salinity continues to be restored to climatological values) The results are analysed computing: Water mass formation rate Transport in the straits (CretanArcs and Otranto Strait) and in the EM using Lagrangian diagnostics Upwelling (transport and location) of the
39

After a long spin up phase (‘climatological’ field), the model is forced using daily ECMWF wind stress and relaxing temperature to

Feb 25, 2016

Download

Documents

ita

The Eastern Mediterranean Transient studied with Lagrangian diagnostics applied to a Mediterranean GCM forced by satellite SST and ECMWF wind stress for the years 1988-1993. Volfango Rupolo a , Salvatore Marullo a and Daniele Iudicone b - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

The Eastern Mediterranean Transient studied with Lagrangian diagnostics applied to a Mediterranean GCM forced by satellite SST

and ECMWF wind stress for the years 1988-1993.Volfango Rupoloa, Salvatore Marulloa and Daniele Iudiconeb

aC.R. Casaccia ENEA Rome, Italyb, Istituto di Fisica dell'Atmosfera-CNR, Rome, Italy (subm. to JGR) http://clima.casaccia.enea.it/staff/rupolo

After a long spin up phase (‘climatological’ field), the model is forced using daily ECMWF wind stress and relaxing temperature to daily satellite SST for the years from 1988 to 1993. (salinity continues to be restored to climatological values)

The results are analysed computing:

Water mass formation rate

Transport in the straits (CretanArcs and Otranto Strait) and in the EM using Lagrangian diagnostics

Upwelling (transport and location) of the ‘uplifted water’

Page 2: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Water mass formation analysis:

From 1989 to 1991, the Adriatic produce less and lessdense water while the Aegean continues its ‘normal’ activity

In 1992 and especially 1993 all basins enhances activity but the densest water is produced in the Aegean

Geographical distribution of formation of water denser than 29.10, units are in 102m3sec-1Km-2

Page 3: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Dashed line: density at the Bottom level in the Otranto strait

Full line: density at the silllevel in the Cretan See (WCA)

No tracer drift.

Interannual Variability is introduced after the 100th year of the spin up. Otranto overflow becomes lighter during winter 1992 and1993

Density at the Antickithera sill depth and at the bottom of Otranto Straits

‘Standard run’Yearly mean

Years from 1988 to 1993Monthly mean

Page 4: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Yearly mean salinity vertical profiles:the decrease of inflow ofsalty intermediate water in the Adriatic precedes the change in the stratification in the Ionian (fully development of the EMT occurring in 1992 and 1993)

(‘+’=1989, ‘*’=1990, ‘’=1991, ‘Δ’=1992 and ‘’=1993Full line without symbols= ‘climatologic year’)

Page 5: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Density vertical section at 35° N, yearly mean values

‘climatologic’year 1993

The general mechanismof the transient is reproduced

Page 6: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Salinity vertical section at 37.5° N, yearly mean values

‘climatologic’year 1993

Dashed area = S > 38.85 PSU

The general mechanismof the transient is reproduced

Page 7: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Left =1993, rigth=climatology

In the climatology ADW is deeperand moves ciclonically in the Ionian.

In 1993 the Aegean overflow is deeper and spreads in the Ionian developing energetic Coherent structures characterized by velocity O(10) cm/sec and L 100-200 Km

Particles initially sedded uniformly in the horizontal (300 <z 600 m)

Page 8: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

1993: about 60 000 particles are released at intermediateDepth. Colours indicate depth (blu 200 m, yellow 2700 m.)

climatology

Page 9: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Lagangian Transport in the cretan Arcs.

W

E

Sv

Sv

1014 m3 Integrated transport of overflowDeeper than 600 m.

From 1988 to 1993 about 1.2 1014 m3 flow deeper than 600 m.From the Cretan Arcs (roughly the halfof he estimate of Roether et al.; 1996)

Considering only water flowing deeper than 1200 m in this period the Aegeanis 4.5 times more active than Adriatic

About the 80 % of this overflow occursduring 1992 and 1993.

Strait ofOtranto Antikithera

StraitKassosStrait

O1

O2

A1 A2

A£ A4

1988 ----------1990----------1992---1993

Page 10: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Strait ofOtranto Antikithera

StraitKassosStrait

O1

O2

A1 A2

A£ A4

1992

1993Deep overflow (> 1500 m) overthe Antikithera Strait during1992 and 1993

End of February

Page 11: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

To compute sub basins lagrangian transport estimate we considerThree –four years long – velocity field:

1) Four iterated years representative of the ‘climatologic’ – or pre EMT situation. (Clim4)

2) Years from 1988 to 1991 (F4, representative in our simulation of the pre conditioning phase)

3) 1993 followed by three years inwhich the model is forcedwith standard forcing (S4, useful to follow the propagation ofa single event of deep Aegean overflow)

Subbasins transport estimates

Page 12: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

A3

SIC

A1

O1

0.11 2.00.03 0.80.11 0.4

0.01 2.80.12 1.70.30 1.6

0.04 2.90.48 1.70.40 1.7

Green= ‘climatology’(Clim4)

Blu= preconditioningPhase (F4)

Red = Fully developpedEMT

In the pink boxes are indicated the mean arrival times

Intermediate Water: Only particles starting from A1 reaching ending sections in less than 4 years and between 200 and 600 m. are considered.

Note the transport values in the preconditioning phase

Recirculation A1A3persists in S4

Page 13: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

O1

I37

A1

<0.01 <0.01 – 0.26 < 0.01 – 0.42 0.15 0.8

0.33 0.06 2.80.10 < 0.01 –0.25 0.05 1.1

Green= ‘climatology’(Clim4)

Blu= preconditioningPhase (F4)

Red = Fully developpedEMT

Deep Water: Only particles starting from A1and O1 and reaching ending section I37 deeper than 600 m. and 1500m in less than 4 years

are considered. Mean arrival times and second row

In the pink boxes are indicated the mean arrival timesfor particles reaching I37 deeper than 1500 m.

Page 14: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Vertical section of particles coming from the Adriatic (red)And the Aegean (black))

clim4Clim4 F4 S4

Page 15: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Kinetic energy in the deep (> 700 m) Ionian

S4: 1993 + 3 years forced with climatological forcing

The maximum is reachedafter 50 days from the insetof the deep Aegena flow

The succesive e-folding timeis 150 days

Page 16: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Upwelling through the nutricline: particles are released (homogeneously) at 160 m. of depth and they are integrated till they reach the depth of 30, 15 and 5 m.

From 160 to 5: 0.02 Sv

From 160 to 15: 0.04 Sv

From 160 to 300: 0.07 Sv

1993:Fully developpedEMT

standard year

From 160 to 5: 0.01 Sv

From 160 to 15: 0.02 Sv

From 160 to 300: 0.03 Sv

Page 17: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

standard year

From 160 to 5:

From 160 to 15:

From 160 to 30:

1993

From 160 to 5:

From 160 to 15:

From 160 to 30:

Time behavior of flux through the ‘nutricline’Red = flux at the starting section, black= flux at the ending section

Page 18: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

SummaryRelaxing model SST to satellite SST from 1988 to 1993, the general mechanism of the EMT is reproduced

Lagrangian diagnostics make easier the analysis of the developmentof the EMT as it is represented by the model and allows quantitative estimates, in particular:

i) In a preconditioning phase IW inflow in the Adriatic (Aegean) decreases (increases). Probably wind induced (Samuel et al, Demirov and Pinardi)

ii) The EMT fully develops during 1992 and 1993, the overflow from the Aegean is concentrated during two events (O(months)). The total flow over the Cretan Arcs is 1.2 1014 m3 (roughly the half of he estimate of Roether et al.; 1996)

iii) Relaxation toward pre-EMT situatiuon (qualitative behavior and estimate of characteristic time)

Moreover: Quantitative estimates of vertical transport before and after the EMT (up lifted water), Time statistics

http://clima.casaccia.enea.it/staff/rupolo

Page 19: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Water mass formation rate

),,(),,(),,(),,( tyxtyxQtyxStyxH

Cdxdy

Area pF()=

T'1

year

dt1

Adriatic

Aegean

[F()]=[Sv]

Page 20: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Transport section to section:

Quantitative analysis

Water path

Tracers analysis

Time statistics

Particles are released from section 1 with some criterium (e.g. u>0, S<So)

Each particle is integrated till up it reaches 2, 3 or recirculates in 1

The transport function relative to the transport ‘section to section’ is computed summing the contribution of each particle

Note that the transport from 1 to 3 is different with or without 2

Lagrangian diagnostics in a OGCMLagrangian diagnostics in a OGCMOff line integrationOff line integration

1

3

2

1

Page 21: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

In a first phase (89-91) salinity decreases at intermediate depth both at Otranto Strait and in the Adriatic, indicating a less inflow of LIW (preconditioning). In this phase (not shown) density slowly increases in the Aegean below the sill of the Cretan Arc

Only during 1992 and 1993 Ageean water replaces ADW.

Note the wrong depth of ADW

Salinity vertical profiles

(‘+’=1989, ‘*’=1990, ‘’=1991, ‘Δ’=1992 and ‘’=1993Full line without symbols= ‘climatic year’)

Page 22: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

  Climatic year

1988 1989 1990 1991 1992 1993

Flux (Sv)T S

0.31 (0.14)13.3938.6329.13

0.12 (0.06)13.3538.6229.13

0.002 (0.0006

)13.3638.6229.13

0.02 (0.003

)13.3538.6229.13

0.02 (0.009)13.4138.6429.13

0.03 (0.01

)13.5338.6929.15

0.04 (0.01)13.4838.6629.14

Depth in the end section

Climatic

year

1988 1989 1990 1991 1992 1993

0< z < 200 mFlux (Sv)TS

 0.37

16.5638.7528.46

 0.36

16.7638.7628.42

 

 0.56

16.3038.7828.56

 0.56

16.6838.7828.46

 0.37

16.8938.7328.37

 0.81

15.87

38.74

28.62

 0.68

16.1238.77 28.58

200< z < 600 mFlux (Sv)TS

 0.76

14.6438.8829.05

 

 0.75

14.69

38.88

29.04

 0.90

14.5438.8729.06

 1.00

14.5738.8729.05

 0.78

14.63

38.85

29.03

 1.01

14.21

38.82

29.10

 0.81

14.0   38.8029.13

600< z < 1200 mFlux (Sv)TS

 0.20

14.1638.8129.10

 0.16

14.18

38.81

29.10 

 0.41

13.9638.7829.12

 0.35

14.0938.8029.11

 0.16

14.17

38.81

29.10

 0.53

13.93

38.78

29.13

 0.62

13.79 38.78 29.15

1200< z < 1500 mFlux (Sv)TS

 < 0.01

 < 0.01

 0.05

13.5538.6929.14

 

 < 0.01

 <

0.01

 0.13

13.68

38.73

29.14

 0.23

13.71 38.77 29.16

1500< z < 2000 mFlux (Sv)TS

 < 0.01

 < 0.01

 0.02

13.6038.7029.14

 

 < 0.01

 <

0.01

 0.11

13.6338.7229.15

 0.27

13-7038.7729.17

2000< z < 3000 mFlux (Sv)TS

 < 0.01

 < 0.01

 < 0.01

 < 0.01

 <

0.01

 0.07

13.6138.7429.16

 0.33

13.7038.7829.18

Otranto Strait

Antikithera Strait

Fully developped EMT

Preconditioning phase

Page 23: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

O1

A1

L25

<0.01 <0.01 – 0.08 < 0.01 – 0.13 0.02 2.8

<0.01 <0.01 – 0.01 < 0.01 – 0.05 <0.01 2.8

Green= ‘climatology’(Clim4)

Blu= preconditioningPhase (F4)

Red = Fully developpedEMT

Deep Water: Only particles starting from A1and O1 and reaching ending section L25 deeper than 600 m. and 1500m in less than 4 years

are considered. Mean arrival times and second row

In the pink boxes are indicated the mean arrival timesfor particles reaching I37 deeper than 1500 m.

Page 24: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

S4: 1993 + 3 years forcedWith climatological forcing

Distribution of arrival timesOf particles starting from the westernCretan Arc and reaching endingSections I37 and L25

Page 25: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Upwelling from deep layers: particles are released (homogeneously) at 1250, 850 and 620 m. of depth and they are integrated till they reach the depth of 420m

From 620 to 420: 1.30 Sv

From 850 to 420: 0.54 Sv

From 1250 to 420: 0.21 Sv

standard year

From 620 to 420: 0.33 Sv

From 850 to 420: 0.11 Sv

From 1250 to 420: 0.04 Sv

1993:Fully developpedEMT

Page 26: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Black point: initial conditions in the

lower surface

Red points: final conditions in the

upper surface

Page 27: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

  ‘Climaticyear’

1988 1989 1990 1991 1992 1993

 Adriatic (Reg. 1, in Fig 1)

13.1038.5429.12

13.6138.5028.98

13.4938.4929.00

13.7438.4528.92

13.4038.4729.00

13.0638.4929.09

12.7538.4829.15

 South Aegean(Reg. 2 in Fig1)

14.6838.8929.05

14.8938.8929.00

14.4038.8729.09

14.4638.8729.08

14.6738.8729.04

14.2238.8629.13

13.7938.8329.19

Δρ(Aegean –Adriatic)

 -0.07

 0.02

 0.09

 0.16

 0.04

 0.04

 0.04

Table. 1 Mean Hydrological values of the 20 days characterised by having the densest surface water in reg. 1 and reg. 2 of fig.1. Note that the surface water is denser in the Adriatic in the ‘climatic’ year (spin up of the model). The low density values in the Adriatic in the 1990 are due to an anomalous warming in the late winter.

Model SST close to satellite SST

Cold winters1992 and 1993

Page 28: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Cross isopycnal surface flux F()

),,(),,(),,(),,(1)(

1

tyxtyxQtyxStyxHC

dxdydtT

FArea pyear

F() ([F()]=[Sv]) is the contribution of air-sea fluxes to the mass flux across the isopycnal surface =, normalised on the entire year.

The water mass formation induced by air sea interactions in the density range 1 < < 2, is given by the difference between the flux entering 1 an the flux leaving 2, i.e. it is equal to the difference F(2)-F(1)..

(T is the time interval of integration, H(x,y,t) and Q(x,y,t) are the heat and freshwater surface fluxes, C p is the specific heat capacity of water, S(x,y,t) and (x,y,t) are the surface salinity and density and and are the derivatives of density with respect to temperature and salinity.)

A water mass body can cross a given isopycnal surface fixed in space or can simply change its density, as a result of air-sea interactions, but remain in place. In this latter case the cross isopycnal flux it is not across any physical surface

Page 29: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Surface cross isopycnal flux

All basin

Aegean Adriatic

Levantine

- Net decrease of ADW production during 89-91

- Net increase of DW production in all the basin (specially in the EM) during 1992 and 1993 (about 2 Sv).

- In these years the Adriatic returns to its typical production rate but the Aegean Sea produce more and denser water

Page 30: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

F(,x,y), geographical distribution of the flow

crossing the isopycnal 29.16

During 1993 the shift from Adriatic and Levantine region to Aegean as predominant Source of DW is observed.

Same color scale

Only heat forcing (relaxation to satellite SST, H=C(T-T*) )

Page 31: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Vertical profiles of salinity difference between1993 and the ‘climatic’ year

The general mechanism of theEMT is reproduced, even with biasin salinity values and ADW depth

Page 32: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

(‘+’=1989, ‘*’=1990, ‘’=1991, ‘Δ’=1992 and ‘’=1993Full line without symbols= ‘climatic year’)

Salinity vertical profiles

In a first phase (89-91) salinity decreases at intermediate depth both at Otranto Strait and in the Adriatic, indicating a less inflow of LIW (preconditioning). In this phase (not shown) density slowly increases in the Aegean below the sill of the Cretan Arc

Only during 1992 and 1993 Ageean water replaces ADW.

Note the wrong depth of ADW

Page 33: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

O1

O2

I37

A1 A3

A4

L25

SIC A2

1

2

3

4

Strait ofOtranto

AntikitheraStrait

KassosStrait

AegeanSea

AdriaticSea

IonianBasin

LevantineBasinCretan Passage

Strait ofSicily

Fig. 1

Page 34: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

O1

O2

A1 A3

A4

A2

Strait ofOtranto

AntikitheraStrait

KassosStrait

AegeanSea

AdriaticSea

IonianBasin

Page 35: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Climatic year: about 60 000 particles are released at intermediateDepth. Colours indicate depth (blu 200 m, yellow 1700 m.)

Page 36: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

Tab.4 First row: Total, lost flux and meanders in the starting section A1. Following rows: Total mean flow connecting the western Cretan Arc to the Otranto Strait, Eastern Cretan Arc and Western Ionian (see Fig.1). No selecting criterion is applied at the ending section. 

  Clim4 F4 S4

Total flux [Sv]Lost flux [Sv]‘Meanders’ [Sv]

1.300.920.37

1.570.420.35

2.250.600.28

A1 O1 Flux [Sv] 0.27 0.14 0.35

A1 A3 Flux [Sv] 0.06 0.35 0.72

A1 SIC Flux [Sv] 0.05 0.66 0.58

Page 37: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

  Clim4 F4 S4

I37 O1600 < z < 1500 [Sv]z > 1500 [Sv]mean arrival time, z> 1500 [years]

 0.330.062.8

 0.10

< 0.01

 0.250.051.1

I37 A1 600 < z < 1500 [Sv]z > 1500 [Sv]mean arrival time, z> 1500 [years]

 0.14

< 0.01--

 0.26

< 0.01--

 0.420.150.8

L25 OI 600 < z < 1500 [Sv]z > 1500 [Sv]mean arrival time

 < 0.01< 0.01

--

 0.01

< 0.01--

 0.05

< 0.01--

L25 A1 600 < z < 1500 [Sv]z > 1500 [Sv]mean arrival time, z> 1500 [years]

 0.01

< 0.01--

 0.08

< 0.01--

 0.130.022.8

Tab.6: Deep flow connections between Otranto Strait and Western Cretan Arc (O1 and A1) with a zonal section west of Crete in the Ionian basin (I37) and a meridional section east of the Cretan passage (L25, see Fig. 1). In particular are shown for each experiment the flow relative to particles reaching I37 an L25 with a depth z 600 < z < 1500. For particles that reach the ending section at a depthdeeper than 1500 m. we show both flux and mean arrival time. Mean characteristics are shown only if the flow connection is greater than 0.01 Sv.

Page 38: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to

S4 Aegean A1 Ionian I37 Levantine L25 I37 A1 600 < z < 1500 T [°C] S [PSU] mean depth [m]

14.16 38.85 520

13.70 38.73 920

13.75 38.75 920

I37 A1 z > 1500 T [°C] S [PSU] mean depth [m]

13.95 38.86 570

13.65 38.75 2040

13.67 38.75 1750

Tab.7: Hydrological characteristics and mean depth of particles that starting from the Western Cretan Arc (A1) reach the zonal section west of Crete in the Ionian basin (I37) and the meridional section east of the Cretan passage (L25, see Fig. 1).

Page 39: After a long spin up phase (‘climatological’ field),  the model is  forced using daily ECMWF wind stress and relaxing temperature to