Seasonal Atlas of the Western Mediterranean Sea from climatologies and high resolution numerical models of Mercator by FLORE MOUNIER1,2, Karine Béranger1,2, Romain Bourdalle-Badie3, Yann Drillet3 & Laure Siefridt3 1 LODYC, Tour 26, 4e étage, boite 100, 4 Place Jussieu, 75252 Paris Cedex 05, France firstname.lastname@example.org, email@example.com 2 ENSTA, UME, Chemin de la Hunière, 91761 Palaiseau Cedex, France 3 CERFACS, Mercator Project, Gustave Coriolis, Toulouse, France firstname.lastname@example.org, email@example.com,,firstname.lastname@example.org Introduction Results from modelling projects and climatological databases for the Mediterranean Sea are now available. This atlas has been designed as a flexible tool to compare outputs from general circulation models with climatologies derived from observations. It has been implemented for PAM and MED16 present simulations (see part I) but it will be used for further simulations. Basically, in the Western Mediterranean Sea, three major water masses are identified according to their formation sites: - one in the upper layer, the Modified Atlantic Water (MAW) which originates from the Strait of Gibraltar; - another one at mid-depth, the thick layer of the Levantine Intermediate Water (LIW), that propagates from the eastern basin to the western basin; - and finally one at the bottom, which is a reservoir filled by Western Mediterranean Deep Water (WMDW) which are produced during the winter season in the Gulf of Lions. This atlas allows to compare views of seasonal horizontal distributions of properties at significant levels of these water masses : temperature, salinity, velocities, eddy Kinetic Energy. The barotropic stream function is also available. Seasonal vertical distributions of temperature and salinity are shown for 11 preselected sections. Part I provides a description of the four simulations and Part II shortly describes the two climatologies. Part III describes useful information concerning the positions of the vertical sections and the descriptions of the figures. I. Simulations from high resolution models of Mercator A very high resolution model of the Mediterranean Sea has been developed in the context of the MERCATOR project (Bahurel et al., 2002; www.mercator.com.fr). The PAM model (the Atlantic-Mediterranean Prototype, in French 'Prototype Atlantique Méditerranée'; Drillet et al., 2000; Béranger et al., 2001; Siefridt et al., 2002) uses a 1/16°coshorizontal grid mesh (the latitude), and 43 z-levels on the vertical. The horizontal grid is stretched at Gibraltar in order to better fit the coast line (Blanchet and Siefridt, 1998). The topography is based on the 1/30° bathymetry (Smith and Sandwell, 1997) averaged onto the 1/16° grid. The numerical model is an extended version of the primitive equation numerical model OPA (Ocean PArallel, Madec et al., 1997) with the rigid lid approximation. The horizontal diffusion is parameterised by a biharmonic operator. On the vertical, the mixing is parameterised by a turbulent kinetic energy scheme. The model has been forced by daily atmospheric fluxes from ECMWF analyses over the period March 1998 to June 2002. This choice was motivated by the relatively fine resolution of the fluxes allowed from March 1998 at the ECMWF centre (0.5° per 0.5°, TL319 grid). It is know that the resolution of the atmospheric forcings can play a major role to simulate some circulation patterns (Horton et al., 1994; Lascaratos and Hatziapostolou, 2001). The fluxes are applied using the flux correction method (Barnier et al., 1995) with a retroaction coefficient . In the retroaction term, the model sea surface temperature (SST) is relaxed to the climatological SST. Freshwater fluxes (Evaporation, Precipitation, River runoffs) are put as a virtual salt flux, with an added term that relaxes the model sea surface salinity (SSS) toward the climatological SSS. For the Mediterranean Sea, runoffs of 31 rivers (Vörösmarty et al., UNESCO, 1996) are incorporated using an upstream scheme at the river mouth. The Black Sea discharge is modelled using the estimates of Stanev et al. (2000). The Atlantic Ocean initial state was provided by the seasonal climatology of Reynaud et al. (1998) . The simulations PAM-20 and PAM-21 (see below), which have been incorporated in this atlas have been computed at the ECMWF centre (Drillet et al., 2000). The second model MED16 uses the Mediterranean the configuration of the PAM model but on a smaller domain. MED16 has been extracted from the PAM model and the exchanges with the Atlantic Ocean are modelled by an Atlantic box (bufferzone). The simulations MED16-05 and MED16-07 (see below) have been done at the IDRIS centre (Béranger, 2001; Béranger et al., 2002). 1. PAM simulations In the PAM model, the diffusivity coefficients are taken equal to -3.109 m2.s-1 for the tracers and equal to –9.109 m2.s-1 for the velocities. A free slip condition is used and the time step is 900s. Only the major Mediterranean rivers (large runoff) are represented according to MED16. PAM-20 - Duration of the simulation : about 5 years - Initial state in the Mediterranean Sea : MODB5 seasonal climatology - Atmospheric forcing : the PAM model has been forced, during 16 months by the daily fluxes averaged over the period March 1998 to February 2001, and then by interannual fluxes from March 1998 to February 2001 - Retroaction term for SST : the model SST is relaxed to the weekly SST of Reynolds estimated from satellite data, with a constant coefficient equal to –20 W.m-2 - Relaxation term for SSS : the model SSS is relaxed to the MODB5 climatological SSS with a constant coefficient equivalent to =–20 W.m-2 . PAM-21 - Duration of the simulation : about 5 years - Initial state in the Mediterranean Sea : MEDATLAS-II seasonal climatology (that comes from an averaged of monthly means) - Atmospheric forcing : the PAM model has been forced, during 16 months by the daily fluxes averaged over the period March 1998 to February 2001, and then by interannual fluxes from March 1998 to June 2002 - Retroaction term for SST : constant coefficient equal to -40 W.m-2, using Reynolds SST -Relaxation term for SSS : constant coefficient equivalent to -40 W.m-2, using MEDATLAS-II climatological SSS. MED16 simulations 1. In the MED16 model, the diffusivity coefficients are taken equal to -4.109 m2.s-1 for tracers and velocities. A no slip condition is used at the coast. The time step is 600s. MED16-05 - Duration of the simulation : 11 years - Initial state : MODB5 seasonal climatology - Atmospheric forcing : daily air-sea surface fluxes applied in a yearly perpetual mode (period March 1998 to February 1999) - Retroaction term for SST : constant coefficient equal to -40 W.m-2 using Reynolds SST - Relaxation term for SSS : constant coefficient equivalent to -40 W.m-2 using MODB5 climatological SSS. MED16-07 - Duration of the simulation : 13 years - Initial state in the Mediterranean Sea: MEDATLAS-II monthly climatology - From year 1 to year 8 : - Amospheric forcing : daily fluxes applied in a yearly perpetual mode (year 2000). The net heat flux has been corrected by a factor of 1.12 to allow the heat budget to be equal to –7 W.m-2 over the Mediterranean Sea, a value commonly known (Béthoux, 1980) - Retroaction term for SST : the model SST is relaxed to the climatological SST with a coefficient . ranges from –10 in winter to –40 W.m-2 in summer. Spatial means of this retroaction term have been applied over a box of half a degree - Relaxation term for SSS : the model SSS is relaxed to the MEDATLAS-II climatological SSS with a coefficient equivalent to =–40 W.m-2. Spatial means of this relaxation term have been applied over a box of half a degree - From year 9 to year 13 : - Atmospheric forcing: daily fluxes over the period March 1998 to June 2002. No correction of the net heat flux budget was applied (balance around -30 W.m-2). The net evaporation budget is about 700 mm.yr-1, a value in agreement with the compilation of Boukthir and Barnier (2000) - Retroaction term for SST : the model SST is relaxed to the weekly SST of Reynolds (2 and spatial means) from March 1998 to December 2000, and then to the daily SST of Reynolds (2 and spatial means) from January 2001 to May 2002 - Relaxation term for SSS : the model SSS is relaxed to the MEDATLAS-II climatological SSS with a coefficient equivalent to = –40 W.m-2 (spatial means). II. Climatologies Two climatologies have been used for the initial state of the models. The MODB5 seasonal climatology of Brasseur et al. (1996) and the monthly climatology of the MEDATLAS/MEDAR group (2002). These climatologies have been interpolated onto the Mediterranean grid of the PAM-MED16 models. These interpolated fields are showed in this atlas. More informatiosn are available at the following WWW locations : MEDATLAS: http://modb.oce.ulg.ac.be/Medar/medar.html and http://www.ifremer.fr/sismer/program/medar/ MODB http://modb.oce.ulg.ac.be/atlas/atlas.html III. Useful informations 1. Atlas interface Access to the data can be done by two different ways: - Through the html interface by selecting a figure with a click on the mouse, a quick look of a low resolution which size is adjusted user’s screen size - By downloading the figure (jpeg format), a high resolution figure can be obtained. The atlas is organised in two parts given on the main page of the interface : model results and climatologies. For both, figures for the four seasons are proposed (winter = January to March). A selection of sections (Part III.1) is proposed for each parameter. 2. Sections available in the atlas - Horizontal sections are located at surface and 400 meter depth levels. - A full line and a name on the following figure indicate vertical sections. Their detailed positions are given in Tab.1. The vertical fields are characterised by a colour scale that does not change neither with seasons nor with sections. On the other hand, the contour scales change according to season and section to follow easily the major water masses through the whole basin. Begin End Section Longitude Latitude Longitude Latitude MED1 -5.65°E 35.78°N -5.90°E 36.15°N MED3-4 3.09°E 36.08°N 3.09°E 41.78°N MED5 1.55°E 39.10°N 2.41°E 39.55°N MED8 9.16°E 37.10°N 9.16°E 39.16°N MED9 9.47°E 42.98°N 9.47°E 44.42°N MED10 9.59°E 42.43°N 11.16°E 42.43°N MED11 11.09°E 37.05°N 12.53°E 37.99°N MED26 0.28°E 36.04°N 0.28°E 38.73°N MED27 4.34°E 39.84°N 16°E 39.84°N MED28 0.16°E 38.63°N 1.28°E 38.92°N MED29 6.28°E 36.85°N 6.28°E 42.98°N Tab.1: Geographical coordinates of the vertical sections presented in the atlas. 3. 4. 3. Figures Model results: temperature, salinity, velocity, barotropic stream function, and eddy kinetic energy For each parameter, the same section (Tab.1) is plotted for the four simulations, and is indicated at the top of the figure, with the chosen season, the location of the model results inside the figure, and, the name of the chosen parameter. The same palette is used for the four figures. For example, for the potential temperature in winter at the vertical section med1, the title of the figure is : ' Section : med1 / Season: Winter / MED16-05 (top-left), MED16-07 (top-right), PAM-20 (bottom-left), PAM-21 (bottom-right) / Temperature ' The winter temperature of MED16-05 is put at the top-left in the figure, the winter temperature of MED16-07 is put at the top-right in the figure, the winter temperature of PAM-20 is put at the bottom- left in the figure, and the winter temperature of PAM-21 is put at the bottom-right in the figure. Climatological fields: temperature and salinity Temperature and salinity of the two climatologies have been plotted on the same page. Figures are organised as follows : At the top of the page, the name of the section is indicated with the chosen season, the location of the climatological results inside the figure, and, the names of the chosen parameters. One common palette is used per parameter. For example, in winter at the vertical section med1, the title is : Section : med1 / Season: Winter / MEDATLAS-II (left), MODB5 (right), Temperature (top), salinity (bottom) The winter temperature of MEDATLAS-II is put at the top-left in the figure, the winter temperature of MODB5 is put at the top-right in the figure, the winter salinity of MEDATLAS-II is put at the bottom-left in the figure, and the winter salinity of MODB5 is put at the bottom-right in the figure. Useful information Horizontal sections TEMPERATURE Isotherms every 0.05°C VELOCITY One vector of five is plotted (m.s-1) SALINITY Isohalines every 0.1 psu EDDY KINETIC ENERGY Isoclines every 0.5 cm2.s-2 BAROTROPIC STREAM FUNCTION Isoclines every 1.5 Sv Vertical sections Annual EDDY KINETIC ENERGY Section Isoclines in cm2.s-2 every 0.5 up to 5 MED1 then every 5 every 1 up to 10 MED3-4 then every 2.5 every 1 up to 10 MED5 then every 2.5 MED8 every 2 MED9 every 0.5 MED10 every 0.5 MED11 every 1 every 0.5 up to 5 MED26 then every 2 MED27 every 2 MED28 every 1 every 1 up to 10 MED29 then every 5 Winter & Autumn TEMPERATURE SALINITY Section Isotherms in ° C Isohalines in psu from 36.3 to 37.5 by 0.2 from 12.7 to 13.6 by 0.05 MED1 then every 0.05 until 38.2 then every 0.4 then every 0.025 from 12.4 to 13 by 0.05 from 36.3 to 38.2 by 0.1 MED3-4 then every 0.1 until 13.3 then every 0.025 then every 0.5 from 13 to 13.3 by 0.025 from 36.3 to 37.5 by 0.2 MED5 then every 0.1 until 13.5 then every 0.05 until 38.2 then every 0.2 then every 0.025 from 12.4 to 14 by 0.1 from 37 to 38.6 by 0.1 MED8 then every 0.4 then every 0.05 12.7 from 37 to 37.9 by 0.2 MED9 from 12.8 to 14.3 by 0.05 then every 0.05 then every 0.4 12.7 from 37.3 to 37.9 by MED10 from 12.8 to 14.3 by 0.05 0.05 then every 0.1 then every 0.4 12.7 13.2 from 37.3 to 38.6 by 0.1 MED11 from 13.7 to 15 by 0.05 then every 0.025 then every 0.25 from 12.7 to 13.3 by 0.05 from 36.3 to 37.5 by 0.2 MED26 then every 0.1 until 14 then every 0.05 until 38.2 then every 0.4 then every 0.025 from 37 to 38.4 by 0.2 from 12.7 to 13.6 by 0.05 MED27 then every 0.05 until 38.6 then every 0.2 then every 0.025 from 12.7 to 13.6 by 0.05 from 37 to 37.5 by 0.1 MED28 then every 0.4 from 38.2 to 38.7 by 0.025 from 12.7 to 13.6 by 0.05 from 37 to 38.4 by 0.1 MED29 then every 0.4 then every 0.02 Spring & Summer TEMPERATURE SALINITY section Isotherms in °C Isohalines in psu from 12.8 to 13.9 by 0.1 from 35.5 to 38 by 0.5 MED1 then every 0.2 until 15 then every 0.1 then every 0.4 from 12.4 to 13 by 0.05 from 36.2 to 38.3 by 0.2 MED3-4 then every 0.1 until 13.3 then every 0.025 until 38.7 then every 0.5 then every 0.1 from 12.9 to 13.3 by 0.05 from 37 to 38.1 by 0.5 MED5 then every 0.5 then every 0.025 from 38.2 from 37 to 38.4 by 0.2 from 12.4 to 14 by 0.1 MED8 then every 0.05 until 38.5 then every 0.4 then every 0.025 12.7 MED9 from 12.8 to 14 by 0.1 every 0.04 from 37.3 then every 0.4 from 13 to 13.6 by 0.1 from 37.3 to 38.5 by 0.1 MED10 then every 0.1 then every 0.025 from 13.7 to 15 by 0.05 from 37.3 to 37.9 by 0.1 MED11 then every 0.1 then every 0.2 38.2 to 38.4 by 0.1 MED26 every 0.05 from 12.7 then 38.425 then every 0.01 38.4 38.45 MED27 every 0.1 from 12.7 38.525 then every 0.05 from 12.7 to 13.2 by 0.05 from 37 to 38.225 by 0.2 MED28 then every 0.4 then every 0.05 from 12.7 to 13.6 by 0.1 from 37 to 38.422 by 0.2 MED29 then every 0.4 then every 0.02 References Bahurel P, De Mey P, Le Provost C, Le Traon P-Y, Mercator project (2002). GODAE Prototype system with applications. Example of the Mercator system. European Geophysical Society XXVII General Assembly, Nice, France, April 2002. Barnier B, Siefridt L, Marchesiello P (1995). Thermal forcing for a global ocean circulation model using a three year climatology of ECMWF analyses. Journal of Marine Systems, 363-380. Béranger K (2001). Modélisation aux équations primitives à très haute résolution de la circulation générale de la Méditerranée. Technical report (SHOM), LODYC, Paris, France. Béranger K, Testor P, Mortier L, Gascard J-C, Crépon M, Siefridt L, Drillet Y (2001). Modélisation haute résolution de la Mer Méditerranée : le bassin occidental. The 36th CIEMS Congress, Monaco, pp53. Béranger K, Mortier L, Crépon M (2002). Seasonal transport variability through Gibraltar, Sicily and Corsica straits. 2nd Meeting on the Physical Oceanography of Sea Straits, Villefranche, France, 77-80. Béthoux JP (1980). Mean water fluxes across sections in the Mediterranean Sea, evaluated on the basis of water and salt budgets and of observed salinities, Oceanologica Acta 3, 79-88. Blanchet I, Siefridt L (1998). Achieving the grid and bathymetry construction for the MERCATOR prototype. Technical report of MERCATOR project, CERFACS, Toulouse, France. Boukthir M, Barnier B (2000). Seasonal and inter-annual variations in the surface freshwater flux in the Mediterranean Sea from the ECMWF re-analysis project. Journal of Marine Systems 24, 343-354. Brasseur P, Beckers J-M, Brankart J-M, Schoenauen R (1996). Seasonal temperature and salinity fields in the Mediterranean Sea: Climatological analyses of an historical data set. Deep-Sea Research, 43(2):159-192. Drillet Y, Béranger K, Brémond M, Gaillard F, Le Provost C, Siefridt L, Theetten S (2000). Rapport de projet MERCATOR, Expérimentation PAM, CERFACS, Toulouse, France. Horton C, Kerling J, Athey G, Schmitz J, Clifford M (1994). Airborne expendable bathythermograph surveys of the eastern Mediterranean. Journal of Geophysical Research 99(C5), 9891-9905. Lascaratos A, Hatziapostolou E (2001). A numerical study of the cause of the Eastern Mediterranean transport: The role of the northern Aegean and the Black Sea waters. The 36th CIEMS Congress, Monaco, pp73. Madec G, Delecluse P, Imbard M, Levy C (1997). OPA, release 8, Ocean General Circulation reference manual. LODYC/IPSL, France, internal report 96/xx, February 1997. MEDAR/MEDATLAS Group (2002). MEDAR/MEDATLAS 2002 Database. Cruise inventory, observed and analysed data of temperature and bio-chemical parameters (4 CDrom, in print). Reynaud T, LeGrand P, Mercier H, Barnier B (1998). A new analysis of hydrographic data in the Atlantic and its application to an inverse modelling study. International WOCE Newsletter 32, 29-31. Siefridt L, Drillet Y, Bourdalle-Badie R, Béranger K, Talatier C, Greiner E (2002). Mise en oeuvre du modèle Mercator à haute résolution sur l’Atlantique Nord et la Méditerranée. Newsletter Mercator, Letter number 5. Smith WHF, Sandwell DT (1997). Global sea floor topography from satellite altimetry and ship deph soundings. Science 277, pp1956-1962. Stanev E, Le Traon P-Y, Peneva EL (2000). Sea level variations and their dependency on meteorological and hydrological forcing : Analysis of altimeter and surface data for the Black Sea. Journal of Geophysical Research 76(24), 5877-5892. Vörösmarty CJ, Fekete BM, Tucker BA (1996). Global river discharge database. RivDIS, vol. 0 to 7, International Hydrological Programme, Global Hydrological Archive and Analysis Systems, United Nations Educational Scientific and Cultural Organization, Paris, France.