Modeling the Influence of the Dardanelles Outflow on the Aegean Sea Dynamics

Modeling the influence of the Dardanelles outflow on the Aegean Sea dynamics

V. Kourafalou Univ. of Miami / RSMAS

I. Androulidakis (Univ. of Thessaloniki, Greece) D. Raitsos, K. Tsiaras (HCMR, Greece) A.Wallcraft and O.M. Smedstad (NRL-SSC)

Collaborators: E. Jarosz, C.A. Blain (NRL-SSC) P.M. Poulain (OGS, Italy) S. Besiktepe (NURC, Italy/international) EU-SESAME project (HCMR, coordinator) The Aegean Sea and the Black Sea are coupled through the Bosporous-Marmara Sea-Dardanelles system

SeaWiFS MUMM/SEADAS 4.0 chl-a (mg/m3) 18/04/1999 Analysis by L. Jönsson Observations: 1997-1998 hydrography

Salinity at 5m (psu) / May 1997 Salinity at 5m (psu) / September 1998

40 40

39 39

23 24 25 23 24 25

32 33 34 35 36 37 38 32 33 34 35 36 37 38 Salinity Salinity (May 1997) (Sept. 1998)

Data provided by: H. Kontoyiannis (METROMED) and G. Georgopoulos, V. Zervakis (INTERREG) Observations: The 2002-2003 Drifter study

 4 deployments (March, June, Sept. 2002; Febr. 2003)

 45 drifters

 ~ 10 m drogue depth

 Generally deployed at depths larger than 200 m

Olson et al., JPO 2007 The North Aegean Hybrid Coordinate Ocean Model (NAEG-HYCOM)

160

1470 140 Samothraki 1400 1330 Res.: 1/500 1260 Limnos Basin 0 120 1190 Nested in 1/25 MED-HYCOM 1120 1050 Hmin=2m 100 980 Limnos BSW outflow 910 Athos Basin 840 Monthly mean rivers & outflow 80 770 700 630 Atm. Forcing: 60 560 •NOGAPS 1 deg 490 •SKIRON/POSEIDON 1/10 deg Skyros Basin 420 40 Sporades 350 280 Evia Trough Lesvos 210 20 140 Chios Basin 70 0

20 40 60 80 100 120 140 160 180 200 220 •What is the role of outflow properties, strait dynamics and atmospheric forcing in the development of the Dardanelles plume? •How does the transport and fate of BSW waters vary in seasonal and inter-annual time scales, how is it modified by the complex topography and how does it impact the North Aegean general circulation? •What is the role of BSW on the North Aegean dense water formation? Model simulations– Study Period 2002-2009

43 45 49 53 54 48 34 19 5.2 53 56 59 57 46 3.7 15 57 6.2 64 64 58 45 27 63 67 70 69 59 47 68 72 75 72 6.9 45 72 75 76 7.4 62 73 76 73 67 57 BSW discharge 72 72 65 57 68 65 53 37 63 58 43 26 57 52 40 26 5.5 44 59 65 66 52 51 48 41 36 4.2 46 62 66 60 53 32 45 42 39 38 37 50 54 48 35 13 6.4 37 34 31 28 9.9 17 17 14 7.7 28 24 2.6 19 16 12 12 1.7 7.3 6 4.8 3.7 4.9 3.6 4 3 4.4 2.9 4.9 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198

Dardanelles outflow current parameterization: modified river upper layer inflow over top 25 m and spread over 5 cells (Schiller and Kourafalou, 2009; Besiktepe, 1994) Dardanelles Outflow: variability in buoyancy input

2 layer outflow structure •Maximum: Spring - Summer •Minimum: Autumn - Winter

July 9, 2003

Oct. 1, 2003 Samothraki

Limnos Basin .BSW outflow is the largest lateral buoyancy input in the Eastern Med. .The BSW pathways exhibit variability in many time scales and are largely influenced by the complex topography Limnos BSW outflow .Parameterizations of outflow properties influence basin scale dynamics .Employ new data from ancillary projects (ONR-Poulain, NRL-Jarosz, NURC-Besiktepe, EU-SESAME) BSW pathways: influence of atmospheric forcing and topographic constrains NOGAPS (resolution: 1 degree) SKIRON (resolution: 1/10 degree)

160 160 0.13 0.125 0.055 0.12 140 140 0.115 0.11 0.05 0.105 0.045 120 0.1 120 0.095 0.09 0.04 0.085 100 100 0.08 0.035 0.075 0.07 0.03 80 0.065 80 0.06 0.055 0.025 0.05 60 0.045 60 0.02 0.04 0.035 0.015 0.03 40 40 0.025 0.01 0.02 0.015 0.005 20 0.01 20 0.005

20 40 60 80 100 120 140 160 180 200 220 20 40 60 80 100 120 140 160 180 200 220

Details in high res. wind curl allow the biforcation of the BSW pathways 26 March NAEG-HYCOM SST 160 (NOGAPS) 20

19 140 SST evaluation (3/26/03) 18

120 17 MODEL 16

(w/ NOGAPS) 100 15

14 MODIS SST 80 13 26 March MODIS-SST 60 12

11 40 10 41 20 9

8 40.5 26 March20 NAEG-HYCOM 40 60 80 SST 100 120 140 160 180 200 220 20 40 160 (SKIRON) 19

140 18 39.5 17 120 39 16

100 15 38.5 14 80 13 38 12 22.5 23 23.5 24 24.5 25 25.560 26 26.5 27 11

40 10 MODEL 9 •Atmospheric fluxes (w/ SKIRON) 20 •BSW temperature 8

20 40 60 80 100 120 140 160 180 200 220 BSW pathways: influence on local and basin-wide dynamics

2 April 2003 No cross basin influence

30 May 2003 Strong cross basin influence 160

1470 140 Samothraki 1400 1330 Limnos Basin 1260 120 1190 1120 1050 BSW pathways: cross-basin structure 980 100 Limnos BSW outflow 910 Athos Basin 840 (upper 100m) 80 770 700 630 60 560 490 Skyros Basin 420 40 Sporades 350 280 Evia Trough Lesvos 210 20 140 Chios Basin 70 0

20 40 60 80 100 120 140 160 180 200 220

30 May 2003 Strong cross basin influence BSW pathways: seasonal variability (spring/left panels– fall/right panels)

SALINITY (model) Chl-a (SeaWiFS data)

160 160 APR OCT 39.5 140 140 39 120 120 38.5 100 100 38 80 80 37.5 60 60 37 40 40 36.5 20 20 36

20 40 60 80 100 120 140 160 180 200 220 20 40 60 80 100 120 140 160 180 200 220 160 NOV 35.5 160 MAY

140 35 ) 140

120 34.5 s 120

(

100 34 100

80 33.5 i 80

60 33 l 60

40 32.5 40 S

20 32 20 31.5 20 40 60 80 100 120 140 160 180 200 220 20 40 60 80 100 120 140 160 180 200 220 JUN DEC 160 31 160

140 30.5 140

120 30 120

100 29.5 100

80 29 80

60 28.5 60

40 40 28

20 20 27.5 27 20 40 60 80 100 120 140 160 180 200 220 20 40 60 80 100 120 140 160 180 200 220 160

1470 140 Samothraki 1400 1330 Limnos Basin 1260 120 1190 1120 1050 100 980 Vertical structure in the Limnos deep basin Limnos BSW outflow 910 Athos Basin 840 80 770 (1400 m) 700 630 60 560 490 Skyros Basin 420 40 Sporades 350 280 Evia Trough Lesvos 210 20 140 Chios Basin 70 0

20 40 60 80 100 120 140 160 180 200 220

Feb 09, 2003 June 29, 2003 North Aegean subdomains

2 4 6

7 3

1 8 Generalised Additive Models (GAMs) for Area 3 (Broad Dardanelles plume area) flexible regression technique ability to model nonlinearities using nonparametric smoothers advantage over traditional regression methods (General Linear Models) suitable for scenario evaluation Salinity: maximum Chl- a abundance is reached at 35.5-36, whereas after that there is a decline

SST: the colder the waters the highest Chl-a (BSW and upwelling)

SSH: high chl-a with SALINITY TEMPERATUR SSH high SSH (BSW input) E

Nutrients Chl-a increases as phosphates increase; while Chl-a decreases after a certain amount of nitrates (0.9) (Phosphate limited PHOSPHATES NITRATE environment) S

The models stated that the most important factors influencing the Chlorophyll-a in Area 3, are Salinity, SST, SSH, Phosphates and Nitrates (73%) – MLD not important. Model Mixed Layer Depth vs. observed chl-a in Area 1 (open sea) (2003)

0.28 200

180 0.24 160

) 140 3 0.2 120 MLD MLD (

m) Chl‐a 100 0.16 80 Chl-a (mg/m 60 0.12 40

20 0.08 0 1 2 3 4 5 6 7 8 9 101112 ÌMonthsÞí åò

The NAEG‐HYCOM MLD follows the observed Chl‐a concentrations in the seasonal cycle  In summer, the MLD has small values (stratification) preventing the nutrients ascension to the surface layer while in autumn (MLD increase), the nutrient concentration begins to increase resulting in a relative chl‐a rise Comparison of model trajectories to drifter data (2002-2003)

Kourafalou and Tsiaras (2007) Comparison of model trajectories to drifter data (2002-2003)

40.5 08/03/2002 - 04/05/2002 40.5

40 40 02/04/2003 - 02/04/2003 39.5 39.5

39 39

38.541 22.5 23 23. 5 24 24. 5 25 25.5 26 26.5 27 38.5 22.5 23 23.5 24 24.5 25 25.5 26 26.5 27 41

40.5 03/21/2002 -05/13/2002 40.5

40 40

39.5 39.5 03/08/2002 -05/13/2002 09/05/2002 - 09/27/2002 39 39

38.5 22.5 23 23.5 24 24.5 25 25.5 26 26.5 27 38.5 22.5 23 23.5 24 24.5 25 25.5 26 26.5 27 Drifter data Model trajectory Model trajectory (delayed) Measurements and modeling at the Dardanelles Strait EU-SESAME project Provided by E. Ozsoy, METU

Dardanelles Strait ADCP

Nutrients, chlorophyll, plankton continuous measurements station Ancillary project: NRL-SSC (E. Jarosz)

BARNY Current Speed (cm/s) BARNY 0 120

10 100 20

30 80

40 60 50 Depth (m) 60 40

70 20 80

90 0 09:00 12:00 15:00 August 28, 2008 (UTC Time)

T/C MOORING

Moorings and Current Observations in the Dardanelles Strait (Aegean Exit) Ancillary project: NRL-SSC (C.A. Blain)

ADCIRC TSS Model Domain Black Sea

Marmara Sea Bosporus Strait

t rai St lles ne rda Da Unstructured Finite Element Mesh 38,589 nodes, 73,433 elements 300 m – 12.4 km resolution Aegean Sea

Battlespace Environments Ancillary projects: OGS and NURC (P.M. Poulain and S. Besiktepe) Provided by P.M. Poulain, OGS