Water masses, mass transport and variability of the Canary current in autumn

1 Water Masses, Mass Transport and Variability of the Canary Current in Autumn CHRISTINA ANN MOCK Facultad de Ciencias del Mar, Universidad de Las Palmas de Gran Canaria, Las Palmas, Spain 2 Abstract CTD casts from four separate autumn cruises in 1997, 2003, 2006, and 2009 sampled the Canary Basin in order to describe the autumn characteristics of the Canary Current. To calculate geostrophic velocities and transport, a reference level of no motion of ƴn = 28.072 kg m-3 (approximately 3000 m depth) was chosen to integrate the thermal wind equation, as supported from LADCP data used from the ORCA 2009 campaign. In autumn, the average section shows a Canary Current flowing to the south at a rate of -3.6 Sv (1 Sv = 109 kg s-1) in the surface layer (0-700 m) at the latitude of the Canary Islands, from the African coast to 18ºW. This net flow is divided into a northward component of 1.1 Sv in the Lanzarote Passage (LP) and a southward component to the west of Lanzarote Island of -4.7 Sv transporting North Atlantic Central Water (NACW). In the intermediate layers (700 – 1600 m depth), Antarctic Intermediate Water (AAIW) is transported northward in the LP with 0.7 Sv. West of Lanzarote Island, the intermediate layer transport is nearly negligible although there are high energetic eddies with warm, salty Mediterranean Water (MW) presumably flowing toward the south. Deep water layers have negligible transport fluctuations, consisting of North Atlantic Deep Water (NADW) properties. While the autumn tendency is flow to the north at the surface layers in the LP, fluctuations exist from 2.2 to -1.9 Sv. However, in the intermediate layers the transport is rather consistent with variations from 0.1 to 0.7 Sv to the north, as all show AAIW features in each of the stations sampled in the LP. West of Lanzarote Island, all four cruises show a surface Canary Current component to the south ranging from -4.1 to - 5.0 Sv, with the exception of 2009 where the transport is -1.5 Sv. Key words: water masses, mass transport, Canary Current, annual variation 3 1. Introduction The eastern dynamic boundary of the North Atlantic Subtropical Gyre is defined by the Canary Current which is primarily fed by the easterly branch of the Azores Current and flows southwestward through the Canary Islands and along the Northwestern coast of Africa, turning westward into the North Equatorial Current (Hernández-Guerra et al., 2005; Machín et al., 2006). This southward flow transports surface North Atlantic Central Waters (NACW) across the Canary Islands. At first, studies of the Canary Current were made using historical data from the general North Atlantic area (Stramma, 1984; Stramma and Siedler, 1988; Stramma and Isemer, 1988) due to the lack of direct sampling in the Canary Island Basin. In the late 1990’s, the European Union funded CANIGO (Canary Islands Azores Gibraltar Observations) Project collected data north of the Canary Islands (Parrilla et al., 2002). Four cruises were conducted between Madeira and Canary Islands for each season (Knoll et al., 2002). Machín et al. (2006) applied an inverse box model to examine the seasonal characteristics of the Canary Current. Meanwhile, a more consistent interest continued to research the Canary Basin through the CORICA, RAPROCAN and ORCA projects. The Canary Current became more defined as the southward transport of surface North Atlantic Central Waters (NACW) across the Canary Islands to approximately 700 m depth (Hernández-Guerra et al., 2001, 2005; Machín et al., 2006). The intermediate waters showed evidence of Mediterranean Water (MW) spreading from the Straits of Gibraltar with saltier, warmer cores crossing the Canary Island Basin. Deep waters consisting of North Atlantic Deep Water (NADW) have shown no predominant direction flow. Likewise, special interest was also taken in the channel between the African coast and Lanzarote Island, the Lanzarote Passage (LP). This area is a shallower region with a 4 maximum depth of 1300 m. Whereas the Canary Current shows a general mean southward transport, significant seasonal transport variations in the LP were noted (Hernández-Guerra et al., 2002, 2003; Machín et al., 2006, 2010; Fraile-Nuez et al., 2010). Current meter moorings were deployed in the late 1990’s in the LP (Knoll et al., 2002; Hernández-Guerra et al., 2001, 2002, 2003). Hernández-Guerra et al. (2003) calculated northward transport of NACW in the LP from these moorings at the surface layer in the autumn and recognized the presence of AAIW in the intermediate levels. Fraile-Nuez et al. (2010) and Machín et al. (2010) used nine years of data to further analyze the mass transport variation in the LP. The intention of this study is to focus on the Canary Current in the autumn season defining its water masses and calculating the mass transport in this region. Section 2 describes the campaigns used to collect the hydrographic data used in this study and the methods used to calculate geostrophic velocities and mass transport and how the reference level of no motion was determined. Potential temperature/salinity diagrams and vertical sections of potential temperature, salinity and neutral density are used in conjunction with geostrophic velocity and accumulated mass transport diagrams to analyze the net mass transport of the Canary Current in autumn in section 3. The interannual variability is presented in section 4 and the final results are discussed in section 5. 2. Data and Methodology This study makes use of four separate cruises that took place during the autumn months in the Canary Island Region. This area encompasses the Canary Basin from 24ºW 5 eastward along the 29.5ºN parallel until North of Gran Canary Island, then southeast through the islands of Lanzarote and Fuerteventura to the African continental coast, referred to as the Lanzarote Passage (Fig. 1). The mean autumn ensemble was calculated from data where two or more cruise Conductivity-Temperature-Depth (CTD) stations coincided. In this study a total of 32 stations were used to calculate the autumn average, eight of these within the LP. Table 1 shows the corresponding dates and number of CTD stations from each cruise considered in our area of study. 32 of the stations were duplicated at least once in order to provide us means to conduct a study generalizing the characteristics of the Canary Current in the autumn season. CANIGO is chronologically the first campaign in this study which took place in September 1997 and provides us the most stations in the LP, 12 total. Its stations only extend westward to 18ºW. CORICA is the next fall campaign in which data is taken in 2003. This campaign provides us with 7 stations in the LP, and 24 stations west of Lanzarote, even though in this campaign continues further westward. The RAPROCAN campaign in fall of 2006 provides us data from all 28 of its stations, with 5 in the LP. The ORCA cruise of 2009 is the only campaign of this study to occur during November. It only provides us 2 data stations in the LP. Although 18 data stations are available westward of the LP, only 10 correspond with positions from previous cruises and are used for the ensemble data set. From here out, each cruise will be referred to by its corresponding year and the mean data of the four cruises will be referred to as the average or ensemble. Generally reflected in the four cruises, data stations are closer together in the LP, as this area samples the continental slope. Stations are closely spaced in order to minimize interpolation errors when calculating mass transport due to the bottom triangle formed by steeper slopes. 6 Each CTD cast collected temperature and salinity data from dual sensors and water samples from a 24-bottle rosette, with 10 liter bottles. Temperature and pressure sensors were calibrated in accordance with WOCE standards at the SeaBird laboratory. Salinity measurements were calibrated onboard by analyzing water samples with a Guildline AUTOSAL 8400B salinometer with accuracy better than 0.002 for single samples (salinity values are noted in the Practical Salinity Scale). Conductivity measurements were calibrated with the application of slope correction. In all four campaigns, potential temperature, θ, and salinity, S, were obtained from CTD measurements at each station with values gridded to a vertical resolution of 2 dbar. Neutral density, ƴn, was computed according to Jackett and McDougall (1997). On the ORCA 2009 cruise, the Lowered Acoustic Doppler Current Profiler (LADCP) system employed a two Teledyne/RDI Workhorse (WH) 300 kHz, which were mounted on the rosette and deployed at each CTD cast. Both narrow band units were run in master/slave mode, one upward looking (slave) and one downward looking (master) with a shared battery pack. LADCP data was processed with the Visbeck software, developed at Colombia University (Fischer and Visbeck, 1993). A GPS reference is applied to calculate the absolute velocity. In order to compare the LADCP velocities with geostrophic velocities, the barotropic tidal component was subtracted from each LADCP profile. Using the OSU (Oregon State University) TOPEX/POSEIDON global tidal model (TPXO), the tidal component was predicted for 2009, using the bottom track time, halfway through CTD cast, as the time for the tidal prediction. (Comas-Rodríguez et al., 2010). Geostrophic velocities are calculated using the thermal wind equation. In order to correctly integrate this equation, a reference level with a known velocity is necessary. Usually the reference level is chosen at the level of no motion, or zero velocity that is 7 unknown. In order to determine which reference levels of no motion could be considered more accurate for our study, we use the LADCP data from 2009 cruise, the only one with these measurements. Geostrophic velocities are computed with different levels and compared with the LADCP velocities following a similar procedure as described in Comas-Rodríguez et al. (2010). Initially, a neutral density of 28.072 kg m-3 for the reference level of no motion was chosen as per Hernández-Guerra et al. (2005). This level is located at approximately 3000 m depth. This reference level also is supported in the study of the North Atlantic carried out by Ganachaud (2003). However, when the sea floor is shallower than the reference isoneutral, the sea bottom becomes the reference level. Likewise, calculations were made with reference level values of ƴn = 27.38 kg m-3 (approximately 700m) for shallow-water stations and ƴn = 27.922 kg m-3 (approximately 1600m) for deep-ocean stations in accordance with Machín et al. (2006). Figure 2 shows the results considering both reference layers. Figure 2a shows the θ/S diagram for the 2009 cruise where Antarctic Intermediate Water (AAIW) is clearly seen in the LP (Hernández-Guerra et al., 2001, 2003, 2005; Machín et al., 2006, 2010; Fraile-Nuez et al., 2010). This AAIW is transported by a north flow due to the stretching of intermediate water strata as stated by Machín et al. (2010). LADCP transport in Figure 2c shows this northward flow at intermediate layers in the LP as well as the geostrophic transport referenced to 28.072 ϓn. As previously mentioned, in the case of the LP, where the maximum ƴn < 27.82, the reference level is the sea bottom. In contrast, the geostrophic transport referenced to 27.38 ϓn shows a southward transport. It is well known that the surface transport in the LP in autumn is to the north (Machín et al., 2006, 2010; Fraile-Nuez et al., 2010). Figure 2b shows that the LADCP transport represents this northward flow in the LP. As in intermediate layers, the 8 northward geostrophic transport in the surface layer is obtained using a reference layer at 28.072 ϓn and not 27.38 ϓn. Westward of the LP, geostrophic transport from both reference layers shows a similar behavior and roughly agree with the LADCP transport. Although the overall Canary Current transport visually looks similar for both sets of reference layer values, the 28.072 ϓn is chosen because it correctly reflects the northern transport in the LP. Overall, the greatest correspondence is seen with the 28.072 ϓn reference layer. 3. Autumn Average Transport As mentioned before, the data from stations where 2 or more CTD casts from different cruises coincided were combined to create the average autumn ensemble, consisting of 32 stations. Fig. 3 displays the θ/S diagram which demonstrates four notable vertical regions; two with wide variability (surface and intermediate waters) and two with close clustering (NACW and NADW). Superimposed are σθ isolines which help define different water layers (surface, intermediate and deep; σθ = 26.5, 27.3, and 27.9, respectively). The shallowest layer above the seasonal thermocline (approximately 70 m) is represented by a scattering of θ/S values due to seasonal heating, evaporation and lowsalinity water advected by upwelling filaments generated off the northwest African coast during autumn (Van Camp et al., 1991; Hernández-Guerra et al., 1993; Nykjaer and Van Camp, 1994; Barton et al., 1998; Pacheco and Hernández-Guerra, 1999). NACW is seen from the seasonal thermocline to approximately 700 m depth, which corresponds to approximately σθ = 27.30 (Harvey 1982). The intermediate stratum (27.38 < σθ < 27.922, roughly 700 -1600 m depth) contains two different water masses: a relatively fresh (S < 35.4) AAIW and a warmer and saltier (S > 35.5) MW. All 8 stations in the LP contain 9 AAIW properties in the intermediate water levels as seen in the vertical sections of potential temperature and salinity reflected by the salinity minimum and cooler temperatures (Figures 4 and 5). Likewise, MW is defined by salinity maximum and warmer temperatures, found at stations 18 and 24. A mixture between AAIW and MW is observed in the rest of the sections. A thick deep branch from approximately 1600 m depth to the ocean bottom is occupied by NADW. In Fig. 4, temperatures range from less than 2ºC in deep waters to greater than 23ºC at the surface. Surface temperatures are slightly cooler along the African coast and gradually get warmer further into the open ocean. As the surface temperature increases further away from the African coast, similarly there is an increase in salinity as seen in Fig. 5. Average salinity values for autumn along this transect are slightly higher than 37 at the surface and lower than 34.89 in deep waters. The cooler and less saline water in the surface layer close to the African coast is due to the advection of upwelled waters from the African coast to offshore as already mentioned. In general, salinity decreases with depth, except in intermediate waters near 800 – 1300 m depth where a subsurface salinity maximum is present. This confirms the presence of the aforementioned MW in the θ/S diagram corresponding to salinity maximums at stations 18 and 24 (approximately 17º and 19.3ºW) and warm temperature dips. The presence of AAIW in the θ/S diagram is clearly supported by the pronounced salinity minimum pocket of salinity < 35.3 with corresponding cooler temperatures (θ < 9ºC) in the LP. The average neutral density (ƴn), in Fig. 6, is calculated from the autumn average values for potential temperature and salinity at each station with values of kg m-3. These isoneutral contours reflect the natural separation of the water column, dividing the ocean into various layers to distinguish different water masses. As sited in Ganachaud (2003), 16 Acknowledgments. I wish to acknowledge the support and encouragement from my tutor Alonso Hernández-Guerra during the preparation for my Master’s Thesis. Special appreciation to Eugenio Fraile-Nuez, Isis Comas-Rodríguez and M. Dolores PérezHernández for their invaluable MATLAB help in instruction and programming. The Centre ERS d’Archivage et de Traitement provided the data base from which the wind stress field was obtained. I thank the Ministerio de Educación y Ciencias de España for the scholarship received to study my Masters of Science in the University of Las Palmas. This study obtained data from the following projects: CANIGO (MAS-CT96-0060) of the European Union; CORICA (REN2001-2649) and ORCA (CTM2005-04701-C02-01) funded by the Spanish Ministry of Science and Technology and Feder; and RAPROCAN funded by the Instituto Español de Oceanografía. 17 References Barton, E., J. Arístegui, P. Tett, M. Cantón, J. García-Braun, S. Hernández-León, L. Nykjaer, C. Almeida, J. Almunia, S. Ballesteros, G. Basterretxea, J. Escánez, L. García-Weill, A. Hernández-Guerra, F. López-Laatzen, R. Molina, M. F. Montero, E. Navarro-Pérez, J. M. Rodríguez, K. van Lenning, H. Vélez, and K. Wild, 1998: The transition zone of the Canary Current upwelling region. Prog. Oceanogr., 41, 455– 504. Comas-Rodríguez, I., A. Hernández-Guerra and E. L. McDonagh, 2010: Referencing geostrophic velocities using ADCP data at 24.5ºN (North Atlantic). Scien. Mar, 74, 331-338. Fischer, J. and M. Visbeck, 1993: Deep velocity profiling with self-contained ADCPs. J. Atmos. Ocean. Tech., 10, 764-773. Fraile-Nuez, E., Machín, F., Vélez-Belchí, P., López-Laatzen, F., Borges, R., BenítezBarrios, V., and A. Hernández-Guerra, 2010: Nine years of mass transport data in the Eastern boundary of the North Atlantic Subtropical Gyre. J. Geophys. Res., 115, C09009, doi:10.1029/2010JC006161. Ganachaud, A., 2003: Error budget of inverse box models: The North Atlantic. J. Atmos. Ocean. Tech., 20, 1641–1655. Harvey, J., 1982: ϴ-S relationships and water masses in the eastern North Atlantic. Deep Sea Res., 29, 1021– 1033. Hernández-Guerra, A., J. Arístegui, M. Cantón, and L. Nykjaer, 1993: Phytoplankton pigment patterns in the Canary Islands area as determined by using CZCS data. Int. J. Remote Sens., 14, 1431–1437. 18 Hernández-Guerra, A., F. López-Laatzen, F. Machín, D. de Armas, and J. L. Pelegrí, 2001: Water masses, circulation and transport in the Eastern Boundary Current of the North Atlantic Subtropical Gyre. Scien. Mar., 65 (Suppl. 1), 177-186. Hernández-Guerra, A., F. Machín, A. Antoranz, J. Cisneros-Aguirre, C. Gordo, A. Marrero-Díaz, A. Martínez, A. W. Ratsimandresy, A. Rodríguez-Santana, P. Sangra, F. López-Laatzen, G. Parrilla, and J. L. Pelegrí, 2002: Temporal variability of mass transport in the Canary Current. Deep-Sea Res. I, 49, 3415-3426. Hernández-Guerra, A., E. Fraile-Nuez, R. Borges, F. López-Laatzen, P. Vélez-Belchí, G. Parrilla, and T. J. MuÄller, 2003: Transport variability in the Lanzarote passage (eastern boundary current of the North Atlantic Subtropical Gyre). Deep-Sea Res. I, 50, 189-200. Hernández-Guerra, A., E. Fraile-Nuez, F. López-Laatzen, A. Martínez, G. Parrilla, and P. Vélez-Belchí, 2005: Canary Current and North Equatorial Current from an inverse box model. J. Geophys. Res., 110, C12019, doi:10.1029/2005JC003032. Jackett, D. and T. J. McDougall, 1997: A neutral density variable for the world's ocean. J. Phys. Oceanogr., 27, 237-263. Knoll, M., A. Hernández-Guerra, B. Lenz, F. López-Laatzen, F. Machín, T. Müller, and G. Siedler, 2002: The Eastern Boundary Current system between the Canary Islands and the African Coast. Deep Sea Res. Part II, 49, 3427–3440. Machín, F., A. Hernández-Guerra, and J. L. Pelegrí, 2006: Mass fluxes in the Canary Basin. Prog. Oceanog., 70, 416-447. Machín, F., J. L. Pelegrí, E. Fraile-Nuez, P. Vélez-Belchí, F. López-Laatzen, and A. Hernández-Guerra, 2010: Seasonal flow reversals of Intermediate Waters in the Canary Current system east of the Canary Island. J. Phys. Oceanogr., 40, 1902-1909. 19 Mason, E., F. Colas, J. Molemaker, A. F. Shchepetkin, C. Troupin, J. C. McWilliams, and P. Sangrà, 2011: Seasonal variability of the Canary Current: A numerical study. J. Geophys. Res., doi:10.1029/2010JC006665, in press. Pacheco, M., and A. Hernández-Guerra, 1999: Seasonal variability of recurrent phytoplankton pigment patterns in the Canary Islands area. Int. J. Remote Sens., 7, 1405–1418. Parrilla, G., S. Neuer, P.Y. Le Traon, and E. Fernández-Suárez, 2002: Topical studies in oceanography: Canary Islands Azores Gibraltar Observations (CANIGO). Volume 1: Studies in the northern Canary Islands basin. Deep-Sea Res. II, 49, 3409–3413. Stramma, L., 1984: Geostrophic transport in the warm water sphere of the eastern subtropical North Atlantic. J. Mar. Res., 42, 537–558. Stramma, L., and H. Isemer, 1988: Seasonal variability of meridional temperature fluxes in the eastern North Atlantic Ocean. J. Mar. Res., 46, 281–299. Stramma, L., and G. Siedler, 1988: Seasonal changes in the North Atlantic Subtropical Gyre. J. Geophys. Res., 93, 8111–8118. Van Camp, L., L. Nykjaer, E. Mittelstaedt, and P. Schlittenhardt, 1991: Upwelling and boundary circulation off northwest Africa as depicted by infrared and visible satellite observations. Prog. Oceanogr., 26, 357– 402. 20 List of Tables 1 Cruise, dates and number of stations used in this study. Ensemble refers to the stations used to calculate the average mean, where 2 or more stations from different campaigns coincide. LP stands for the Lanzarote Passage. 2 Layer definitions and approximate equivalences with water masses. 21 Table 1. Cruise, dates and number of stations used in this study. Ensemble refers to the stations used to calculate the average mean, where 2 or more stations from different campaigns coincide. LP stands for the Lanzarote Passage. Table 2. Layer definitions and approximate equivalences with water masses. Cruise Dates Number of Stations Total LP CANIGO Poseidon 233 7-18 September 1997 24 12 CORICA 7-29 September 2003 36 7 RAPROCAN 0906 6-16 September 2003 28 5 ORCA 3–8 November 2009 21 2 Ensemble 32 8 Layer Lower Interface ƴn kg m-3 Water Mass 1 26.44 surface water 2 26.85 NACW 3 27.162 NACW 4 27.38 NACW 5 27.62 MW/AAIW 6 27.82 MW/AAIW 7 27.922 MW/AAIW 8 27.975 UNADW 9 28.008 MNADW 10 28.044 MNADW 11 28.072 MNADW 12 28.0986 LNADW 13 28.11 LNADW 14 bottom diluted AABW 22 List of Figures 1 Area of study, Canary Island Basin. Symbols mark where two (star), three (triangle) or four (circle) CTD stations from different cruises coincide in order to calculate the Autumn Average. For reference, isobaths are shown in meters (Smith and Sandwell, 1997). 2 Data from ORCA 2009 cruise. (a) Potential temperature (θºC) / Salinity diagram. Superimposed are σθ isolines which help define different water layers (surface, intermediate and deep; σθ= 26.5, 27.3, and 27.9, respectively). Accumulated mass transport stream functions for (b) surface layer and (c) intermediate waters. Red line indicates reference level of no motion ƴn =27.38 kg m-3 (~700m) for shallowwater stations and ƴn = 27.922 kg m-3 (~ 1600m) for deep-ocean stations. Blue line represents reference level ƴn = 28.072 kg m-3, or sea bottom where floor is shallower than isoneutral. Dashed green line represents transport using LADCP data. Positive/negative values reflect northward/southward transport. Ekman transport has been applied to the most surface layer. 3 θ/S diagram using all the data from the four cruises to build an average autumn ensemble. Superimposed isolines, σθ, as in Fig. 2a. 4 Autumn average potential temperature vertical section. Bathymetry comes from the Sandwell-Smith (1997) database. Note that the vertical scale changes at 1000 m depth. Top axis indicates autumn average station numbers. 5 Same as Fig. 4 but for the autumn average salinity. 6 Same as Fig. 4 but for the autumn average neutral density. 7 Vertical section of geostrophic velocity for the autumn average ensemble. Dashed lines correspond to negative (southward) velocities (cm s-1). 23 8 Autumn average accumulated mass transport stream functions for different groups of layers corresponding to the surface (layers 1-4, ƴn < 27.38), intermediate (layers 5-7, 27.38 < ƴn < 27.922), and deep (layers 8-14, ƴn > 27.922) waters. The integration begins at the coast of Africa. Positive/negative accumulated mass transport reflects northward/southward flow, respectively. The gap in the stream function is at Lanzarote Island. Ekman transport has been applied to the surface layer. 9 CTD stations map for each of the 4 campaigns used in this study. Triangles represent presence of MW, stars represent presence of AAIW. Bathymetry comes from the Sandwell-Smith (1997) database. 10 θ/S diagram for each of the four cruises. Superimposed isolines, σθ, as in Fig. 2a. 11 Overall integrated mass transport stream functions for each of the cruises at (top) surface (layers 1-4, ƴn < 27.38), (middle) intermediate (layers 5-7, 27.38 < ƴn < 27.922), and (bottom) deep (layers 8-14, ƴn > 27.922) waters. The integrations begin at the coast of Africa. Positive/negative accumulate mass transport reflects northward/southward flow, respectively. The gap in the stream function is at Lanzarote Island. Ekman transport has been applied to the most surface layer. 24 Fig. 1. Area of study, Canary Island Basin. Symbols mark where two (star), three (triangle) or four (circle) CTD stations from different cruises coincide in order to calculate the Autumn Average. For reference, isobaths are shown in meters (Smith and Sandwell, 1997). 25 Fig. 2. Data from ORCA 2009 cruise. (a) Potential temperature (θºC) / Salinity diagram. Superimposed are σθ isolines which help define different water layers (surface, intermediate and deep; σθ= 26.5, 27.3, and 27.9, respectively). Accumulated mass transport stream functions for (b) surface layer and (c) intermediate waters. Red line indicates reference level of no motion ƴn =27.38 kg m-3 (~700m) for shallow-water stations and ƴn = 27.922 kg m-3 (~ 1600m) for deepocean stations. Blue line represents reference level ƴn = 28.072 kg m-3, or sea bottom where floor is shallower than isoneutral. Dashed green line represents transport using LADCP data. Positive/negative values reflect northward/southward transport. Ekman transport has been applied to the most surface layer. 2009 Surface Layar (y < 27.38) e ' ~b~--------------------' [ • ¡;¡; Intermediate Layer (27.38 < Y < 27.922) e "