The Azores Current System and the Canary Current from CTD and ADCP data

TESIS DOCTORAL TESIS DOCTORAL MAYO 2011 MAYO 2011 The Azores Current System and the Canary Current from CTD and ADCP data D. SALVADOR GALV ´ AN HERRERA, SECRETARIO DEL DEPARTAMENTO DE F´ ISICA DE LA UNIVERSIDAD DE LAS PALMAS DE GRAN CANARIA, CERTIFICA, Que el Consejo de Doctores del Departamento en su sesi´on de fecha 13 de mayo de 2011 tom´o el acuerdo de dar el consentimiento para su tramitaci´on, a la tesis doctoral titulada “The Azores Current System and the Canary Current from CTD and ADCP data” presentada por la doctoranda Da. Isis Comas Rodr´ıguez y dirigida por el Doctor D. Alonso Hern´andez Guerra. Y para que as´ı conste, a efectos de lo previsto en el Arto73.2 del Reglamento de Estudios de Doctorado de esta Universidad, firmo la presente en Las Palmas de Gran Canaria, a diez y seis de mayo de dos mil once. PROGRAMA DE DOCTORADO EN OCEANOGRAF´ IA Departamento de F´ısica The Azores Current System and the Canary Current from CTD and ADCP data (La Corriente de Azores y la Corriente de Canarias mediante datos de CTD y ADCP) Tesis doctoral presentada por Isis Comas Rodr ´ ıguez para la obtenci´on del t´ıtulo de Doctor por la Universidad de Las Palmas de Gran Canaria Dirigida por Dr. Alonso Hern´ andez Guerra Dr. Alonso Hern´andez Guerra Da. Isis Comas Rodr´ıguez LAS PALMAS DE GRAN CANARIA MAYO 2011 A mi familia Agradecimientos Acknowledgements Dedico mi tesis doctoral a mi familia, en el sentido m´as amplio de la palabra. A todas y cada una de esas personas que considero “m´ıas”, unidas a m´ı y al trabajo que he desempe˜nado estos a˜nos por lazos de muchos tipos, pero todos fundamentales para llegar a escribir estas l´ıneas. Deseo dar las GRACIAS... ... por la beca FPU (Formaci´on de Profesorado Universitario), del Ministerio de Educaci´on y Ciencia, que he disfrutado durante los cuatro a˜nos de realizaci´on de mi tesis. Sin ese apoyo financiero, no habr´ıa podido hacer el doctorado. Las estancias que he realizado en el extranjero no habr´ıan sido posibles sin las convocatorias a las que me acog´ı en el marco de esta beca predoctoral. ... al Departamento de F´ısica de la ULPGC (Universidad de Las Palmas de Gran Canaria) por permitirme la adscripci´on como PIF (Personal Investigador en Formaci´on) y por el apoyo log´ıstico y humano que siempre me han brindado. La asistencia a congresos internacionales ha sido posible adem´as gracias al programa PFPI (Programa de Formaci´on de Personal Investigador) de la ULPGC. ... al grupo de investigaci´on “Oceanograf´ıa F´ısica y Oceanograf´ıa por Sat´elite”. El laboratorio F-214 ha sido mi segundo hogar durante este periodo. La Facultad de Ciencias del Mar ha sido mi escuela, brind´andome adem´as la oportunidad de formarme como docente durante estos dos ´ultimos a˜nos. De aqu´ı me llevo estupendos v amigos y compa˜neros de fatiga que me han ayudado y acompa˜nado durante largas horas de trabajo (Francis, Eugenio, Vero, Lola, David, Juanma, Tina, Johan, Enrique...) o simplemente con una sonrisa de ´animo por los pasillos entre Departamentos (Iv´an, Inma, Minerva, Mayiya, Sheila, Anna...). ... al Personal de Administraci´on y Servicios que, de una manera u otra, han facilitado esta andadura. El “co˜nazo” que damos en Conserjer´ıa, los tr´amites en la Administraci´on y cualquier consulta en la Biblioteca, todo ello acompa˜nado de una sonrisa y de muchas ganas de ayudar. No quiero olvidarme tampoco del resto de personas que, trabajando en la ULPGC y dedic´andome un rato m´as del que era necesario, me han ayudado con los inacabables tr´amites necesarios para llegar hasta aqu´ı, especialmente al personal de Tercer Ciclo y de GRAI (Gesti´on de Recursos y Ayudas a la Investigaci´on). A algunos de ellos (Lola, Alejandro...) les debo, como poco, una caja de bombones :) ... a los proyectos ORCA (CTM2005-04701 y CTM2008-04510) y MOC (CTM200806438), en los que he participado y han apoyado adem´as econ´omicamente mi formaci´on, ambos financiados por el Ministerio de Ciencia y por fondos Feder. Al proyecto Malaspina (Consolider CSD2008-00077), del que formo parte en la actualidad. Un lugar muy especial guardo tambi´en para el proyecto RAPROCAN del IEO (Instituto Espa˜nol de Oceanograf´ıa), que me ha permitido disfrutar de las campa˜nas oceanogr´aficas desde que era estudiante colaboradora en los ´ultimos a˜nos de carrera. Hablando de barcos... ... a la tripulaci´on del B/O Cornide de Saavedra y del BIO Hesp´erides, por contribuir a que la experiencia de embarcarme me haya enganchado hasta acumular un total de siete campa˜nas en los ´ultimos a˜nos. Y a todos mis compa˜neros de profesi´on que han estado embarcados conmigo y han hecho que resonaran las risas durante las largas guardias nocturnas (Mireya, Carolina, Yeray, Amelia, Fran, Marta, peque˜no Timmy...). vi ... con may´usculas a mi director de tesis, Alonso. Por haberme dejado colaborar con ´el desde cuarto de carrera (¡tras insitirle tres veces!) y haberme ense˜nado mucho en la oceanograf´ıa, pero tambi´en en la manera de hacer ciencia, durante los ´ultimos siete a˜nos. Por su paciencia y tiempo dedicado a que esta tesis sea lo que es. Por las campa˜nas, los congresos y el apoyo que me ha brindado, y porque estos a˜nos han permitido que nos entendamos cada vez un poquito m´as. ... a todos los investigadores que me han ayudado, colaborando en mi formaci´on y con esas charlas sobre el futuro que han esclarecido m´as de una vez en qu´e direcci´on seguir (Pedro V´elez, Antonio Mart´ınez, Federico L´opez...). Tambi´en me gustar´ıa mencionar a aquellas personas que hicieron posibles mis estancias en Southampton: I would like to thank Elaine McDonagh and Stuart Cunningham for hosting my two three-month visits to NOCS (National Oceanography Centre, Southampton) and providing the chance to work with them. Thanks both to the institution and its ‘Ocean Observation and Climate’ research group, with special mention to Brian King, who helped me during my stay. They also provided the cruise D279 data (shown in Chapter 2 of this thesis), which was supported by NERC (Natural and Environmental Research Council, United Kingdom) as part of the Core Strategic Research Programme ‘Ocean Variability and Climate’ at NOCS. Also to Alberto Naveira Garabato for sharing impressions with me and a big Thank You to all those people who helped me keep warm memories of the months I spent abroad. ...a mi “gente”. A mi familia al completo, de sangre o no, por el apoyo y los ´animos incondicionales durante mi carrera (a mi madre, mi hermano, mis abuelos, mis t´ıos y primos, Yoly y Pepe, Cintia y Alberto, Saulo y Carlos...). A todo el que me ha hecho sentir querida y en familia. Dedico esta tesis de manera muy especial a mi padre. Como se suele decir, “no le cabr´ıa el culo en la silla de orgullo” si estuviese presenciando este momento. Esa persona que siempre te ve guapa y cree que todo lo haces bien... y en el fondo sabes que es porque el amor lo ciega. Yo tambi´en te quiero, pap´a, mucho :) vii ...a mi “otra gente”, aquellos que con su amistad se han alegrado conmigo por todo lo bueno y han sabido levantarme en los momentos no tan buenos. Hay amigos que llevan a mi lado desde la adolescencia (Varsha...). Algunos entraron en mi vida en cuarto de carrera, cuando empec´e en la investigaci´on (M´onica, Borja...). Otros llevaban ah´ı desde que entramos en Ciencias del Mar, pero pasaron muchas experiencias con la “delegada de clase” :) antes de que nos uniera la amistad (´ Angela, Aridane...). Alguna incluso dej´o este mundillo, pero eso no ha hecho que deje de estar aqu´ı cerquita (Idaira...). Gracias por todo el apoyo que me han brindado y por las experiencias que, a d´ıa de hoy, hacen que sea la persona que soy. ...a Rayco, esa persona que me aguanta (¡pobrecito!) cuando acaba el d´ıa y con la que desahogo los a veces sinsabores de esta profesi´on que tanto me gusta. Su apoyo para la realizaci´on de esta tesis ha sido fundamental y ver el mundo a trav´es de sus ojos ha hecho que, desde hace m´as de ocho a˜nos ya, estar a su lado me haya enriquecido en todos los aspectos. Gracias por tu paciencia, generosidad y sencillez. Gracias por ser familia, amigo y mucho m´as. ...a todo aqu´el que, de un modo u otro, ha contribuido a que haga realidad el sue˜no de ser ocean´ografa. Incluso a aquellos que (disc´ulpenme), me haya dejado en el tintero... viii Resumen Debido al creciente inter´es en el estudio de la circulaci´on global y su papel en la regulaci´on del cambio clim´atico, la frontera occidental del giro subtropical de Atl´antico Norte (NASG, North Atlantic subtropical gyre, en ingl´es) ha sido extensamente estudiada a lo largo de los a˜nos a causa del su intenso transporte de masa y calor hacia el norte. No obstante, pocos trabajos ha suscitado la frontera este hasta que recientemente se ha reconocido el papel esencial que juegan las corrientes en la Cuenca de Canarias. En la actualidad, estos estudios han contribuido a mejorar nuestro entendimiento y cuantificaci´on del NASG. En esta tesis, se propone y aplica una metodolog´ıa para evaluar el Sistema de Corrientes de Azores, as´ı como el transporte promedio de masa y la variabilidad de la Corriente de Canarias, ambos como parte fundamental del NASG. Mediante el uso de Perfiladores de Corriente Ac´usticos Doppler (ADCP, Acoustic Doppler Current Profiler, en ingl´es) para calcular velocidades en la capa de referencia, este procedimiento proporciona un m´etodo para las estimaciones de c´alculos geostr´oficos absolutos. La metodolog´ıa sugerida es aplicada a diversos conjuntos de datos con el fin de mejorar las estimas de transporte en las regiones de Canarias y Azores. Se ha realizado una secci´on hidrogr´afica zonal al norte de las Islas Canarias en el invierno de diferentes a˜nos. Un transporte de masa promedio ha sido obtenido de estas medidas mediante la elaboraci´on de una secci´on promedio, en la cual se aprecia una reducci´on del campo de remolinos de la regi´on. Los resultados muestran un flujo caracter´ıstico hacia el sur a trav´es de la termoclina, de una intensidad de 3.5 Sv, lo cual se corresponde con la magnitud de la Corriente de Canarias. El procedimiento de referenciado mediante ix The satellite image on the cover is owned by ‘The Visible Earth’ (http://visibleearth.nasa.gov/), which is part of the EOS Project Science Office located at NASA Goddard Space Flight Center. xvi Contents Resumen /Summary ix Presentaci´on de la tesis /Thesis preview xiii 1 GENERAL INTRODUCTION 1 1.1 Oceanographic setting . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 The Canary and Azores Currents as part of the NASG . . . . . . . . 3 1.3 Geostrophic calculations . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.4 Aim and thesis outline . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2 METHODOLOGY 11 Referencing geostrophic velocities using ADCP data at 24.5◦N (North Atlantic) 11 2.1 Introduction................................ 13 2.2 Material and methods . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.2.1 Data acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.2.2 LADCP data processing . . . . . . . . . . . . . . . . . . . . . 15 2.2.3 Tidal barotropic component correction . . . . . . . . . . . . . 19 2.2.4 Reference velocity field . . . . . . . . . . . . . . . . . . . . . . 19 2.3 Results................................... 20 2.4 Discussion................................. 24 3 THE CANARY CURRENT 25 xvii Mean mass transport and variability of the Canary Current 25 3.1 Introduction................................ 27 3.2 Data.................................... 29 3.3 Methodology ............................... 34 3.3.1 Average transect ADCP-referencing . . . . . . . . . . . . . . . 38 3.3.2 Yearly transects ADCP-referencing . . . . . . . . . . . . . . . 43 3.4 Mean mass transport and variability . . . . . . . . . . . . . . . . . . 45 3.4.1 Average section mass transport . . . . . . . . . . . . . . . . . 45 3.4.2 Variability between the 1997, 2006, 2008 and 2009 cruises . . . 50 3.5 LADCP vs. inverse model solution . . . . . . . . . . . . . . . . . . . 58 3.6 Discussion and conclusions . . . . . . . . . . . . . . . . . . . . . . . . 60 4 THE AZORES CURRENT SYSTEM 63 The Azores Current System from hydrographic data 63 4.1 Introduction................................ 65 4.2 Dataandmethods ............................ 67 4.3 Results................................... 69 4.3.1 Water mass distribution . . . . . . . . . . . . . . . . . . . . . 69 4.3.2 The Azores Current and Countercurrent System . . . . . . . . 71 4.4 Discussion and conclusions . . . . . . . . . . . . . . . . . . . . . . . . 77 5 DISCUSSION 81 5.1 General discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 5.2 Conclusions ................................ 85 5.3 Furtherresearch.............................. 87 A Resumen en espa˜nol /Spanish summary 89 A.1 INTRODUCCI´ ONGENERAL...................... 89 A.1.1 Localizaci´on oceanogr´afica . . . . . . . . . . . . . . . . . . . . 89 A.1.2 Las Corrientes de Canarias y Azores como parte del giro subtropical del Atl´antico Norte . . . . . . . . . . . . . . . . . . . 92 xviii A.1.3 C´alculos geostr´oficos . . . . . . . . . . . . . . . . . . . . . . . 95 A.1.4 Objetivos y estructura de la tesis . . . . . . . . . . . . . . . . 97 A.2 METODOLOG´ IA............................. 99 A.2.1 Adquisici´on de datos . . . . . . . . . . . . . . . . . . . . . . . 101 A.2.2 Procesado de datos de LADCP . . . . . . . . . . . . . . . . . 103 A.2.3 Correcci´on de la componente barotr´opica de la marea . . . . . 107 A.2.4 Velocidades en el nivel de referencia . . . . . . . . . . . . . . . 108 A.3 RESULTADOS .............................. 113 A.3.1 La Corriente de Canarias . . . . . . . . . . . . . . . . . . . . . 113 A.3.2 El Sistema de Corrientes de Azores . . . . . . . . . . . . . . . 128 A.4 DISCUSI´ ON................................ 139 A.4.1 Discusi´on general . . . . . . . . . . . . . . . . . . . . . . . . . 139 A.4.2 Conclusiones ........................... 143 A.4.3 Trabajos futuros . . . . . . . . . . . . . . . . . . . . . . . . . 145 xix List of Figures 1.1 Geographical location of the area of study. For reference, the main isobaths are shown following the Smith-Sandwell database [Smith and Sandwell,1997]............................... 2 1.2 Surface currents of the Atlantic Ocean from Tomczak and Godfrey [1994]. 4 1.3 Surface currents of the North Atlantic Ocean from Schmitz [1996]. . . 5 2.1 Station positions for the transatlantic hydrographic section occupied during cruise D279. 125 full depth stations were used along a nominal latitude of 24.5◦N. ............................ 15 2.2 LADCP data processed using the Visbeck method at Station 23 (26.5◦N, 75.9◦W): a) eastwards velocity component, b) northwards velocity component. BB and WH master represent data processed individually, while WH master/slave corresponds to the combined data of the downand up-looking WH running in master/slave mode. Note that the vertical scale is different for each depth range. Mean profiles are the average of the upcast and downcast. . . . . . . . . . . . . . . . . . . . 16 2.3 Statistical analysis carried out for the three different instruments deployed during the survey: a) mean and standard deviation of the differences between the BB and WH master bottom track measurements; b) mean and standard deviation of the differences between each LADCP full depth profile and their bottom track record in the depth range near the sea bottom; c) mean and standard deviation between each LADCP full depth profile and the SADCP profile in a depth range near the sea surface. .................................. 17 xxi 2.4 Tidal barotropic component calculated from the OSU TPXO tide prediction model. This velocity is subtracted from the velocity measurements of the LADCP and SADCP profiles. . . . . . . . . . . . . . . . 20 2.5 Comparison between the initial geostrophic profile and the ADCPreferenced for station pair 22-23 (located at 26.5◦N, at 76.1◦W and 75.9◦W respectively). The dashed line represents the initial calculation while the solid one is corrected using ADCP data. Stars and circles correspond to bottom track records of Stations 22 and 23 respectively. Triangles show the SADCP velocity calculated as the mean of the measurements taken during each cast. Note that the vertical scale is different for each depth range. . . . . . . . . . . . . . . . . . . 21 2.6 a) velocity obtained from the depth-averaged offset of each geostrophic profile, considered as the reference velocity providing the new corrected calculations; b) final velocity estimated at the sea bottom before (gray line) and after (black line) the referencing ADCP data contribution was taken into account. . . . . . . . . . . . . . . . . . . . . . . . . . . 22 2.7 Velocity fields (cm s−1) contoured a) before and b) after the correction that applied ADCP referencing. . . . . . . . . . . . . . . . . . . . . . 23 3.1 Location of the CTD stations carried out in January 1997 (grey dots) during the CANIGO project. Stations occupied in February 2006, 2008 and 2009 (black dots), in the framework of the RAPROCAN project, are also shown. For reference, 200, 1000, 2000, 3000 and 4000m isobaths are shown [Smith and Sandwell, 1997]. . . . . . . . . . . . 29 3.2 Mean Θ-S diagram. Stations located in the Lanzarote Passage (stations 1-5) are plotted in grey. . . . . . . . . . . . . . . . . . . . . . . . . . . 32 3.3 Mean potential temperature (◦C) vertical section. Note that the vertical scale is different for each depth range. . . . . . . . . . . . . . . . 33 3.4 Mean salinity vertical section. Note that the vertical scale is different foreachdepthrange............................ 35 xxii 3.5 Mean neutral density (kg m−3) vertical section. Note that the vertical scale is different for each depth range. . . . . . . . . . . . . . . . . . . 36 3.6 Mean geostrophic velocity (cm s−1) vertical section. Isolines are drawn as follows: every 0.1 cm s−1from 0 to 0.5 and every 0.5 cm s−1from 0.5 to 20 (positive range of velocities); every 1 cm s−1from -20 to -6, every 0.5 cm s−1from -6 to -0.5 and every 0.1 cm s−1from -0.5 to 0 (negative range of velocities). Positive/negative values stand for northward/southward velocities. . . . . . . . . . . . . . . . . . . . . . 39 3.7 ADCP-referencing process summary for the averaged data. a) Comparison between the initial and the ADCP-referenced geostrophic profiles for station pair 2-3 (located in the Lanzarote Passage). b) Comparison between the initial and the ADCP-referenced geostrophic profiles for station pair 8-9. c) Mean tidal barotropic component calculated from the OSU TPXO prediction model. d) Mean velocity at the reference level for the average cruise. The stars mark the station pairs corrected with bottom-track data. The dots represent station pairs where the LADCP profile is used. . . . . . . . . . . . . . . . . . . . . . . . . . . 40 3.8 Mean ADCP-referenced geostrophic velocity (cm s−1) vertical section. Isolines are drawn as follows: every 0.1 cm s−1from 0 to 0.5 and every 0.5 cm s−1from 0.5 to 20 (positive range of velocities); every 1 cm s−1 from -20 to -6, every 0.5 cm s−1from -6 to -0.5 and every 0.1 cm s−1 from -0.5 to 0 (negative range of velocities). Positive/negative values stand for northward/southward velocities. . . . . . . . . . . . . . . . 42 3.9 Initial geostrophic velocity fields (cm s−1) for the four winter cruises. Isolines are drawn every 1 cm s−1. Positive/negative values stand for northward/southward velocities. . . . . . . . . . . . . . . . . . . . . . 44 3.10 Comparison between the geostrophic and the ADCP-referenced profiles for station pair 8-9 (located in the main section at 29◦N, 14.7◦W and 15.1◦W, respectively) for the four winter cruises. . . . . . . . . . . . . 45 xxiii 3.11 Mean tidal barotropic component calculated from the OSU TPXO prediction model. This velocity is subtracted from the velocity measurements obtained through the ADCPs. . . . . . . . . . . . . . . . . . . 46 3.12 Mean velocity at the reference level obtained from the depth-averaged offset of each geostrophic profile during each of the four winter cruises. Stars mark the station pairs corrected with bottom-track data. Dots represent station pairs where the LADCP or SADCP profile is used. Errorbars are shown, calculated from the available error estimates. . . 47 3.13 ADCP-referenced geostrophic velocity fields (cm s−1) for the four winter cruises. Isolines are drawn every 1 cm s−1. Positive/negative values stand for northward/southward velocities. . . . . . . . . . . . . . . . 48 3.14 Mean accumulated mass transport (Sv). . . . . . . . . . . . . . . . . 49 3.15 Accumulated mass transport (Sv) in the surface layer (γn<27.38 kg m−3) for the four winter cruises. . . . . . . . . . . . . . . . . . . . . . 53 3.16 Accumulated mass transport (Sv) in the intermediate layer (27.38< γn<27.922 kg m−3) for the four winter cruises. . . . . . . . . . . . . 55 3.17 Accumulated mass transport (Sv) in the deep layer (γn>27.922 kg m−3) for the four winter cruises. . . . . . . . . . . . . . . . . . . . . . 56 3.18 Accumulated mass transport (Sv) in the surface, intermediate and deep layers (1997). Dashed lines stand for the stream function estimated by the inverse model and solid lines represent the estimates calculated through our ADPC-referencing procedure. . . . . . . . . . . . . . . . 59 4.1 a) Location of the CTD stations occupied at the Azores Current System during the ORCA cruise in the fall of 2010. For reference, 200-, 1000-, 2000-, 3000-, 4000and 5000-m isobaths are shown [Smith and Sandwell, 1997]. b) Θ-S diagram. NACW, SAIW, MW (dark gray, stations 41-43), NADW and LSW (light gray, stations 39-43) are shown. 68 4.2 Vertical sections for a) Θ (◦C), b) salinity and c) γn(kg m−3). Note the different vertical scales in plots a) and b). The black dots mark the location of the AzF as suggested by P´erez et al. [2003]. . . . . . . 70 xxiv 4.3 a) Full-depth averaged and b) mean surface velocities from the SADCP. For reference, locations of the CTD stations and main isobaths are shown. 72 4.4 Reference level velocities (cm s−1) obtained from LADCP bottom-track (stars) and full-depth (dots) profiles. . . . . . . . . . . . . . . . . . . 73 4.5 Accumulated geostrophic mass transport (Sv) for thermocline (red), intermediate (green) and deep (blue) layers. Dashed lines show the geostrophic mass transport, solid lines show the LADCP-referenced stream functions. Positive/negative values stand for eastward/westward flows..................................... 75 4.6 SADCP averaged velocities (cm s−1) corresponding to the a) AzCC and b) AzC domains. Geostrophic mass transport (Sv) per layer for the c) AzCC and d) AzC. Dashed lines show the geostrophic mass transport, and solid lines stand for the LADCP-referenced mass transport. . . . 76 4.7 Surface layer data at 24.5◦W. Gray line represents temperature (◦C) recorded by the thermosalinometer. Black lines correspond to the zonal velocity component (cm s−1) measured by the SADCP (solid) and altimeter (dashed). Note the different y-scales. . . . . . . . . . . . . . . 77 4.8 Vertically integrated mass transport per station pair for a) thermocline, b) intermediate and c) deep layers. The thick black line marks the vessel path. At the surface layer, arrows represent the satellite absolute geostrophic velocity. . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 5.1 Geographical location of the area of study including the sections occupied during the ORCA (red dots) and RAPROCAN (black dots) cruises.................................... 84 A.1 Localizaci´on geogr´afica del ´area de estudio. Las principales is´obaras se muestran como referencia seg´un la base de datos de Smith-Sandwell [Smith and Sandwell,1997]. ....................... 91 A.2 Las corrientes superficiales del Oc´eano Atl´antico seg´un Tomczak and Godfrey [1994]. .............................. 92 A.3 Las corrientes superficiales del Oc´eano Atl´antico seg´un Schmitz [1996]. 93 xxv A.4 Posici´on de las estaciones durante la secci´on hidrogr´afica transatl´antica llevada a cabo durante la campa˜na D279. Se realizaron 125 estaciones a lo largo de la latitud de 24.5◦N. .................... 101 A.5 Datos de LADCP procesados mediante el m´etodo Visbeck en la estaci´on 23 (26.5◦N, 75.9◦W): a) componente zonal de la velocidad, b) componente meridional de la velocidad. BB y WH “master” (maestro) representan los datos procesados de manera individual, mientras que WH “master/slave” (maestro/esclavo) corresponde con los datos combinados de los cabezales orientados en sentidos inversos del descenso de la roseta. Debe apreciarse que la escala vertical de los ejes es diferente para los distintos rangos de profundidades. Los perfiles medios (mean, en ingl´es) son el promedio entre el recorrido de descenso y de ascenso delaroseta................................. 104 A.6 An´alisis estad´ıstico para los diferentes instrumentos empleados: a) media y desviaci´on est´andar de la diferencia entre las medidas de “bottomtrack” del BB y del WH “master”; b) media y desviaci´on est´andar de las diferencias entre cada perfil de LADCP y su “bottom-track”; c) media y desviaci´on est´andar de cada perfil de LADCP y las medidas de SADCP en la zona superficial de la columna de agua. . . . . . . . 105 A.7 Componente barotr´opica de la marea calculada del modelo de predicci´on OSU TPXO. Esta velocidad ser´a eliminada de las medidas de velocidad obtenidas por el LADCP y el SADCP. . . . . . . . . . . . . 107 xxvi ORCA Origen y variabilidad de la Corriente de Canarias / Origins and variability of the Canary Current OSU Oregon State University OTIS OSU Tidal Inversion Software RAPID Rapid Climate Change, programme of the NERC (United Kingdom) RAPROCAN Radial Profunda de Canarias / Canary Deep Hydrographic Section RDI (Teledyne) RD Instruments RRS Royal Research Ship SADCP Ship-mounted Acoustic Doppler Current Profiler SAIW Sub-Arctic Intermediate Water Ssalto Segment Sol Multimissions d’Altim´etrie, d’Orbitographie et de Localisation Pr´ecise / Ground Segment Multimission Altimetry, Orbit Determination and Precise Location Sv Sverdrup (106m3s−1∼109kg s−1) TPXO TOPEX/POSEIDON Global Tidal Model ULPGC Universidad de Las Palmas de Gran Canaria / University of Las Palmas de Gran Canaria WH Workhorse (RDI ADCP) xxxv Chapter 1 GENERAL INTRODUCTION 1.1 Oceanographic setting The North Atlantic subtropical gyre (NASG) spans from 15◦to 45◦N and is bounded to the west by the American continent. To the east, it encounters Europe, the Strait of Gibraltar (connecting the Atlantic with the Mediterranean Sea) and Africa. Its importance in the North Atlantic circulation has awakened our interest for decades, leading to studies of both boundaries. The NASG eastern margin is mostly covered by the Canary Basin, which extends between 20◦to 40◦N, a total area of approximately 7 million km2located from 10◦to 40◦W. Due to its large dimensions, different water masses can be found in the Canary Basin. A water mass is constituted by a water volume that can be identified by its area of formation. It is known by its characteristic temperature and salinity values. The water masses are usually formed by the air-sea interaction where they sink. They stabilize at a depth range determined by their density and then flow along isopycnals. In the Canary Basin, the following water masses can be found: North Atlantic Central Water (NACW) is a central water mass located between the surface and 600-800 m depth. It can be easily identified in Θ-S diagrams by a well 1 2CHAPTER 1. GENERAL INTRODUCTION defined line between 26.5 and 27.3 kg m−3(σθ). Below the central waters, the intermediate waters are located. In some cases, these are formed at higher latitudes and therefore characterized by higher densities. In the 700-900 m depth range, Antarctic Intermediate Water (AAIW) can be located, known for being relatively cooler and fresher that the surrounding waters. In contrast, Mediterranean Water (MW), also flowing at intermediate depths, is easily identified by its high temperature and salinity values. At approximately 800-1400 m, isolated MW cores can be found, known as Meddies (Mediterranean eddies). Below 1500 m depth, North Atlantic Deep Water (NADW) occupies the largest water mass volume in the Atlantic Ocean, spreading above the 4500 m depth. NADW also comprises the Labrador Sea Water (LSW), formed due to deep winter convection in the Labrador Sea. In this basin, diluted Antarctic Bottom Water (AABW) is also present in the deepest layer. 28oW 24oW 20oW 16oW 12oW 8oW 28oN 30oN 32oN 34oN 36oN 38oN Azores Canary Islands Madeira Figure 1.1: Geographical location of the area of study. For reference, the main isobaths are shown following the Smith-Sandwell database [Smith and Sandwell, 1997]. 1.2. THE CANARY AND AZORES CURRENTS AS PART OF THE NASG 3 Located at the Canary Basin eastern margin, approximately 100 km off the northwest coast of mainland Africa, are the Canary Islands. This is a Spanish archipelago situated between 27◦and 29◦N, and from 18◦to 13◦W. Near the northern end of the Canary Basin, the Azores archipelago is located about 1500 km west of the Portuguese coast, extending along a west-northwest to east-southeast orientation in the range of latitudes from 36.5◦to 40◦N, and between 31.5◦and 24.5◦W. The area of study covered by this thesis extends over the Canary Basin, from the surface current that flows across the Canary Islands up to the current system located south of the Azores archipelago (Fig. 1.1). 1.2 The Canary and Azores Currents as part of the NASG The prevailing winds north of the tropics consist of the trades (easterlies) and the mid-latitude westerlies. These create Ekman convergence in the mid-latitude North Atlantic, where a high pressure zone is located, driving the subtropical gyre. Circulation around a high pressure is clockwise in the northern hemisphere due to the Coriolis effect. At the NASG western boundary, the Gulf Stream flows polewards along the American coast before detaching and dividing into several branches. The most important are the North Atlantic Current, which circulates to the north, and the Azores Current, that bends to the east forming the northern margin of the NASG. On the gyre’s eastern margin, the Azores Current diverges into several southwardflowing branches, feeding the Canary Current when flowing southwards through the permanent thermocline across the Canary Islands, in line with the African coast. These branches feed the North Equatorial Current, the southern side of the subtropical gyre, that closes the gyre. The existence of the Canary Current was found in the XV century by sailors who navigated taking advantage of the wind forcing and the oceanic surface circulation. This current was first mapped when the Challenger (1872-1876) and Meteor (1930’s) 4CHAPTER 1. GENERAL INTRODUCTION Figure 1.2: Surface currents of the Atlantic Ocean from Tomczak and Godfrey [1994]. expeditions took place. The classic picture of the surface circulation in the region assumed that the North Atlantic Current divided into several branches. One of these, known as the Portugal Current, was believed to feed the Canary Current (Fig. 1.2). The Azores Current was not considered until several oceanographic cruises were carried out south of the Azores archipelago by K¨ase and Siedler [1982]. They suggested that this current was responsible for feeding the Canary Current, a hypothesis later supported by Stramma [1984]; Stramma and Siedler [1988a]; Stramma and Isemer [1988]. This new surface circulation scheme omits the Portugal Current as a cause for the southward flow across the Canary archipelago, presenting the current distribution 1.3. GEOSTROPHIC CALCULATIONS 5 shown in Fig. 1.3. Figure 1.3: Surface currents of the North Atlantic Ocean from Schmitz [1996]. 1.3 Geostrophic calculations In order to quantify the flow in the region of study, velocity estimates are needed. To calculate the vertical shear of geostrophic velocities in the ocean, the thermal wind equations are applied: ∂v ∂z =−g ρof ∂ρ ∂x (1.1) ∂u ∂z =g ρof ∂ρ ∂y (1.2) where uand vare the zonal and meridional velocity components (m s−1), gis the gravitational acceleration (m s−2), ρois the ocean mean density (1026 kg m−3), ρis 6CHAPTER 1. GENERAL INTRODUCTION the sea water density (kg m−3) and fis the Coriolis parameter (s−1) given by 2Ω sin θ, where Ω is the Earth rotational angular velocity (7.29 10−5s−1) and θis the latitude. The thermal wind equations are essential in physical oceanography. Traditionally, temperature and salinity in-situ measurements acquired during a cruise are used to compute density estimates. Consequently, these equations can be applied to hydrographic data in order to estimate geostrophic flows in the oceans. Integrating equation (A.1) over a zonal section, it is found that the geostrophic velocity is defined by the addition of two components. The first component varies vertically (baroclinic velocity), whereas the second depends only upon the chosen reference level: vg(z) = −g ρofZz zo ∂ρ ∂z dz +b(1.3) where bare called reference level velocities. The first component can be calculated using density estimates. However, the estimation of velocities at the reference level is a classical problem in physical oceanography. Having determined the geostrophic velocities, geostrophic transport estimates can also be calculated. To that end, the water column is divided into cells, and each velocity estimate is multiplied by the density and area occupied by a water parcel. Not knowing the value of the reference level velocity (initially assumed to be zero) produces an imbalance in property transports, such as mass, salt, heat or nutrients. Several methods have been studied in order to estimate the velocity in the reference layer. In the first place, the formulation of inverse models has been widely studied. This method adds a new consideration to the thermal wind equations, which is to take into account the mass conservation principle. In these models, the minimum reference velocity that balances the mass transport is considered the final velocity in the reference layer. Today, new methodologies are being proposed to compute better estimates of the reference velocity. Unlike inverse models, these consider any reference velocity (not necessarily the minimum) by comparing the geostrophic estimates to in-situ velocity measurements. The latest is the case proposed in this thesis, in 1.3. GEOSTROPHIC CALCULATIONS 7 which Acoustic Doppler Current Profiler (ADCP) data are used to calculate absolute geostrophic velocities and transports. 8CHAPTER 1. GENERAL INTRODUCTION 1.4 Aim and thesis outline The aim of this thesis is to address several questions: 1. What do we already know about the surface circulation in the Canary Basin? 2. How do we achieve a better estimate of mass transports? Can we improve the methodology to calculate absolute velocity fields by introducing reference level velocities in the geostrophic calculations? 3. Is it possible to get a mean circulation based on a periodically repeated survey in this area? 4. Has the Canary Current suffered changes during the last decade? 5. How does the Azores Current contribute to the circulation in the region? A review of previous studies regarding the Azores and the Canary Currents is carried out to answer the first question (see Chapters 1, 3 and 4). Most of these studies deal with mass transport estimates based on geostrophic calculations. Also the source feeding the flow across the Canaries is discussed. To address the second question, data from a transatlantic hydrographic cruise are used in order to establish a valid procedure to estimate absolute mass transports. To that end, reference level velocities calculated through different combinations of Acoustic Doppler Current Profiler (ADCP) data are compared and a methodology is suggested to be followed in this thesis (Chapter 2). In order to answer the third question, a hydrographic section carried out on a biannual basis in the framework of the RAPROCAN project is averaged. In this thesis (Chapter 3), the mean mass transport of the Canary Current is studied considering 1.4. AIM AND THESIS OUTLINE 9 the four transects carried out in the winter seasons. The flow estimates obtained from summer cruise data will be assessed in future studies. To study changes in Canary Current transports, the RAPROCAN hydrographic section repetitions are analyzed considering winter datasets from 1997, 2006, 2008 and 2009 (Chapter 3). Variability between the measurements carried out in other seasons are evaluated in further research contributions. The fifth question is answered through the study of the Azores Current System net eastward transport using hydrographic data acquired during the ORCA cruise (Chapter 4). This flow’s behaviour in the Canary Basin motivates this research line for future studies to confirm the hypothesis that the Azores Current feeds the Canary Current. 2.2. MATERIAL AND METHODS 17 used as representative of the main transect; however, a statistical study is needed in order to quantify these differences and determine which instrument performed better during the cruise. Close to zero differences between the BB and the downlooking WH bottom track measurements at each station would imply better data quality (Fig. 2.3a). −20 0 20 a) −10 0 10 −20 0 20 Velocity (cm/s) b) −10 0 10 −81 −79 −20 0 20 c) −70 −60 −50 −40 −30 −20 −10 −10 0 10 Longitude (°E) BB WH master WH master/slave Figure 2.3: Statistical analysis carried out for the three different instruments deployed during the survey: a) mean and standard deviation of the differences between the BB and WH master bottom track measurements; b) mean and standard deviation of the differences between each LADCP full depth profile and their bottom track record in the depth range near the sea bottom; c) mean and standard deviation between each LADCP full depth profile and the SADCP profile in a depth range near the sea surface. Each full depth profile was compared to its instrument bottom-track, and the 18 CHAPTER 2. METHODOLOGY downcast, upcast and the mean obtained from both were considered separately. BB, WH master and the WH master/slave package were processed independently. The measurements near the bottom were compared to the bottom-track data in the matching depth range (Fig. 2.3b). We assume that LADCP measurements that are close to the bottom-track data are of better quality, that is, they have closer to zero difference values. SADCP data were also compared to the profile values near the surface in the coincident depths (Fig. 2.3c). Means and standard deviations to each offset were obtained. The instrument that performed better is the one whose mean differences and deviation are the closest to zero. To determine these differences, the western boundary and the ocean interior were considered separately due to their different dynamics. Two values are shown below, the first one corresponds to the western boundary and the second one to the ocean interior. Averaging each instrument’s performance during the cruise, the difference obtained between the BB and WH master bottom track measurements is 8.8 ±0.1 / 1.0 ±0.2 cm s−1(Fig. 2.3a). However, the mean differences between each instrument and its bottom track are -12.5 ±0.1 / -0.2 ±0.2 cm s−1for the BB, -3.1 ±0.1 / -0.1 ±0.1 cm s−1for the WH master processed individually, and -2.6 ±0.0 / -0.7 ±0.1 cm s−1for the WH master/slave combination (Fig. 2.3b). Finally, the mean differences calculated for each LADCP in comparison with the SADCP data are -3.7 ±0.1 / 1.3 ±0.3 cm s−1for the BB, 5.9 ±0.2 / 0.4 ±0.2 cm s−1for the WH master processed individually and -3.1 ±0.2 / -1.2 ±0.2 cm s−1for the WH master/slave combination (Fig. 2.3c). Considering that the mean difference is smaller for the WH master, we have chosen this instrument for the correction of the geostrophic velocities, processed individually without the WH slave data. We are assuming here that WH master data are better quality because the measurements are closer to the bottom-track data and SADCP data than the other instruments. This is mostly seen in the ocean interior, while the WH master/slave combination provides better results on the western boundary. Unfortunately, a study using the WH in master/slave mode is not possible due to the fact that the uplooking WH (slave) underwent some data reception errors during the cruise and was therefore not used after Station 81 (24.5◦N, 44.9◦W) near the eastern 2.2. MATERIAL AND METHODS 19 end of the survey. Only the velocity component that is perpendicular to the section is considered further. Thus, a velocity rotation was applied in the non-zonal parts of the section, near the eastern and western boundaries. 2.2.3 Tidal barotropic component correction The barotropic tidal component was subtracted from the LADCP, bottom-track and SADCP velocity measurements. It was calculated using the OSU (Oregon State University) TOPEX/POSEIDON global tidal model (TPXO) [Egbert et al., 1994; Egbert and Erofeeva, 2002]. This global model of ocean tides best fits, in a leastsquares sense, the Laplace Tidal Equations and along track averaged data from TOPEX/POSEIDON and Jason (on TOPEX/POSEIDON tracks since 2002) obtained with OTIS (OSU Tidal Inversion Software). The time considered for the tidal prediction is the bottom track time, which is half way through the time spent at the station. Once calculated, the tidal barotropic component (Fig. 2.4) was subtracted from the velocity measurements taken by the LADCP, including bottom-track measurements and SADCP profiles. 2.2.4 Reference velocity field The initial geostrophic velocity field was calculated with a zero-velocity reference layer at 1000 m at Stations 1-44 (79◦W to 69.5◦W) and 3000 m at Stations 45-125 (69.1◦W to 13.4◦W) following the study carried out by Bryden et al. [2005]. When the deepest common depth was less than the reference level (e.g. near the eastern boundary), the bottom was considered as the zero velocity layer. As already mentioned, the reference velocity was obtained using WH master bottom-track data. Each station pair has a geostrophic velocity profile and two bottom track data profiles located on the two stations surrounding it. Therefore, the correction was applied calculating 20 CHAPTER 2. METHODOLOGY −80 −70 −60 −50 −40 −30 −20 −10 −20 0 20 40 Longitude (°E) Velocity (cm/s) Figure 2.4: Tidal barotropic component calculated from the OSU TPXO tide prediction model. This velocity is subtracted from the velocity measurements of the LADCP and SADCP profiles. differences between geostrophic and bottom-track profiles at the corresponding depth range. The differences obtained were averaged and a final mean of the contributions of the stations on either side was used (Fig. 2.5). Therefore, a comparison can be made between each station pair’s geostrophic profile and the LADCP data, by considering this reference velocity derived from the mean of the LADCP measurements from the stations on either side. In station pairs with LADCP data only available for one station, this is taken as the whole correction contribution. 2.3 Results The initial geostrophic velocity profile (referenced to 1000dbar) and those using the LADCP information were compared as shown in Figure 2.5, which corresponds to the station pair 22-23 (located at 26.5◦N, at 76.1◦W and 75.9◦W respectively). Individual plots, as shown in Figure 2.5, were drawn to compare the initial velocity profile at each station pair with the ADCP-referenced, as well as the bottom track data taken into account for each correction, corresponding to the data from the previous and following station. SADCP was also included in these plots in order to check the resemblances between the available data and the corrections made. A slight disagreement between 2.3. RESULTS 21 −800 −600 −400 −200 0 −8 −6 −4 −2 0 2 4 6 8 10 12 −4700 −4600 −4500 Velocity (cm/s) −4000 −3500 −3000 −2500 −2000 −1500 −1000 Depth (m) Initial geostrophic velocity Corrected velocity WH master bottom−track st. 022 WH master bottom−track st. 023 SADCP velocity − mean st. 022−023 Figure 2.5: Comparison between the initial geostrophic profile and the ADCPreferenced for station pair 22-23 (located at 26.5◦N, at 76.1◦W and 75.9◦W respectively). The dashed line represents the initial calculation while the solid one is corrected using ADCP data. Stars and circles correspond to bottom track records of Stations 22 and 23 respectively. Triangles show the SADCP velocity calculated as the mean of the measurements taken during each cast. Note that the vertical scale is different for each depth range. them can be seen due to the fact that measurements at the bottom were used as the reference in the correction. As this was applied to the whole water column, there are reasonable differences considering there is high ageostrophic behaviour at the surface. SADCP data can be included in the study in two ways: firstly, by taking the value for each station pair as the averaged measurements during each cast, and secondly, by taking the values obtained while the ship sailed between stations and averaging them. In both cases, ageostrophic features are included in the data. Thus, the first option was chosen so as not to include the possible ageostrophic features located between stations registered while sailing, but those registered on the station’s position. 22 CHAPTER 2. METHODOLOGY −80 −70 −60 −50 −40 −30 −20 −10 −20 0 20 40 Velocity offset (cm/s) a) −80 −70 −60 −50 −40 −30 −20 −10 −20 0 20 40 Longitude (°E) Velocity (cm/s) b) Initial ADCP−referenced Figure 2.6: a) velocity obtained from the depth-averaged offset of each geostrophic profile, considered as the reference velocity providing the new corrected calculations; b) final velocity estimated at the sea bottom before (gray line) and after (black line) the referencing ADCP data contribution was taken into account. The reference velocity was plotted for each station pair throughout the section to examine the corrections made (Fig. 2.6a). As already mentioned, the initial reference velocity was considered to be zero in the geostrophic calculations [Bryden et al., 2005]. In Figure 2.6a, the reference velocity shown reflects the behaviour of some oceanographic features. On the western boundary, high positive (northward) velocities across the section represent the contribution of the Florida Current at approximately 80◦W. Moving eastwards, negative (southward) velocities represent a recirculation structure and the Deep Western Boundary Current (DWBC) located near 77◦W. Moving further eastwards, some minor structures are present with offsets of less than 5 cm s−1. The positive and negative velocities switching between adjacent groups of stations represent the mesoscale contribution. The LADCP provided 2.3. RESULTS 23 reference level velocities. To visualize them, final velocity estimations at the sea bottom at each station pair were plotted for the initial case considering zero reference velocity and for the LADCP-referenced velocity field (Fig. 2.6b). The contribution of the LADCP data can be clearly seen (black line). Figure 2.7 shows vertical sections highlighting the structure of the velocity field across the section before and after the correction. As expected, no qualitative changes are apparent near the surface, where alternating northward and southward flows are a feature of each velocity field. The differences between the two velocity fields are most apparent near the original zero reference velocity level. Almost-barotropic currents are clearly seen along the section including the northward/southward adjacent currents that represent the eddy oceanic field. The same final velocity field would be obtained when LADCP-referenced velocities are obtained, regardless of the initial reference layer proposed for the geostrophic calculations. The reference velocity provided by the bottom track data would change in order to adjust the different initial velocity profiles to the LADCP measurements. Longitude (°E) Depth (m) b) −70 −60 −50 −40 −30 −20 −6000 −4000 −2000 2 14 24 35 45 55 65 75 85 95 105 115 124 Depth (m) a) −70 −60 −50 −40 −30 −20 −6000 −4000 −2000 Positive (Northward) Negative (Southward) Zero 2 14 24 35 45 55 65 75 85 95 105 115 124 Figure 2.7: Velocity fields (cm s−1) contoured a) before and b) after the correction that applied ADCP referencing. 24 CHAPTER 2. METHODOLOGY 2.4 Discussion Calculating a reference velocity using bottom track LADCP data from cruise D279 along 24.5◦N has been described. The results indicate that this methodology is useful for obtaining absolute geostrophic estimations from direct velocity observations during the hydrographic measurements. The choice of different initial zero velocity reference levels in the geostrophic calculations modifies the reference velocity obtained in order to apply the correction shown in this work. However, the LADCP-referenced velocity field is the same regardless of this choice, because the LADCP data to which the initial field is referenced is the same. The most significant advance made in relation to previous works concerning the D279 cruise data is the statistical analysis carried out in order to determine which instrument performed best. The final results were not masked by constraints in the LADCP data processing or the conditioning of the performance of the different devices. These data will be used further by the Ocean Observation and Climate team at the National Oceanography Centre, Southampton (NOCS) to study the Atlantic Meridional Overturning Circulation (AMOC) quantification and dynamics. In conclusion, the ADCP is a valuable tool for measuring ocean deep velocity profiles. In this study we demonstrate its usefulness for correcting standard hydrographic measurements and calculations. LADCP data processing can provide independent constraints in the study of ocean dynamics through inverse methods. There are still some deficiencies in the resulting data but continuous improvements are being made in the performance of the instruments, as well as their processing and further applications. Chapter 3 THE CANARY CURRENT Mean mass transport and variability of the Canary Current Abstract The Canary Current constitutes the major Eastern Boundary Current of the North Atlantic subtropical gyre, playing an essential role in the large-scale oceanic circulation. Winter mean mass transport across a section north of the Canary Islands is quantified using four hydrographic cruises carried out during the winter season of 1997, 2006, 2008 and 2009. Variability over these years is also evaluated. Transports are estimated geostrophically, and both lowered and shipboard Acoustic Doppler Current Profiler (ADCP) data are included in referencing these calculations to non-zero reference level velocities. Results show a characteristic southward thermocline flow of -3.5 Sv (1 Sv ∼109kg s−1) in winter, that corresponds to the Canary Current. The inclusion of ADCP velocities suggests a stretching of the circulation, spanning across the entire Canary archipelago. The averaged stream function is also shown at intermediate (-1.1 Sv) and deep (-3.3 Sv) layers, computed by dividing the water column into layers as a function of neutral density (γn). Variability between the sampled years is studied at the different depth ranges, revealing a permanent southward flow in the surface and deep waters, and different water mass circulations at intermediate 25 26 CHAPTER 3. THE CANARY CURRENT layers. This pattern is not observed in the 2008 data. A comparison between the methodology applied and the use of inverse models is also carried out. The inverse model leads to near zero mass transports below the surface layer, whereas the southward flow for thermocline waters obtained by the inverse model is corroborated by ADCP-referenced transports to the south. The channel between the archipelago and the African coast is also studied in detail through comparisons between stream functions obtained from CTD and current meter data at surface and intermediate depths. The ADCP-referenced mean mass transport in the surface layer and that using current meter data are -1.4 Sv and -1.3 ±1.29 Sv, respectively. This study leads to mass transport values in accordance with our knowledge of the Canary Current and in good agreement with current meter records at its eastern margin. Additionally, we offer a new picture of the horizontal extension of the thermocline flow in winter. Keywords Canary Current, North Atlantic subtropical gyre, ADCP, Mass transport, Inverse model, Eastern Boundary Current I. Comas-Rodr´ıguez, A. Hern´andez-Guerra, P. V´elez-Belch´ı, E. Fraile-Nuez, A. Mart´ınezMarrero, and F. L´opez-Laatzen (2011). In prep. 3.2. DATA 33 0 250 500 750 1000 0100200300400500600 18 17 13 12 11 9 123456789101112131415161718192021 123456789101112131415161718192021 0 kmkm Depth (m) 1000 2000 3000 4000 5000 14151617181920 14151617181920 8 7 5 4 3 2 20 | 1 3 6 9 12 15 17 18 Potential Temperature (oC) Longitude (oW) Figure 3.3: Mean potential temperature (◦C) vertical section. Note that the vertical scale is different for each depth range. 34 CHAPTER 3. THE CANARY CURRENT African coast. Fig. 3.4 shows the averaged salinity along the section. Values in the range from 34.90 to almost 37 are revealed. Relatively high values are found between stations 8 (∼14.7◦W) and 15 (∼17.3◦W), corresponding to the Mediterranean Water contribution. This warm and salty core extends horizontally about 200 km and vertically from around 900 to 1400 m. Also at intermediate layers, relatively low values are found in the Lanzarote Passage (stations 1-5, ∼13-13.7◦W). Those values around 35.4 can be identified as AAIW, as previously supported by the Θ-S relationships plotted in Fig. 3.2. Apart from this usual presence of AAIW in the Passage [Knoll et al., 2002; Hern´andez-Guerra et al., 2003; Mach´ın et al., 2006], the contribution of this water mass appears as a low salinity patch near the western end of Fig. 3.4, between stations 19-21 (∼19-20◦W). In Fig. 3.5 the computed neutral density is shown. The isoneutral surfaces allow us to divide the ocean into multiple layers in terms of the limits between the presence of different water masses. Following the work of Ganachaud [2003], 14 layers are considered for this region in the North Atlantic Ocean, as reflected in Fig. 3.5 and shown in Table 3.2. Layers 1-4 cover surface and thermocline waters, while layers 5-7 are representative of intermediate waters, and layers 8-14 stand for deep waters. Due to the peculiar cooling that occurred during two (2006 and 2009) of the four cruises here studied, the average γn-section does not show the upper layer (26.44 kg m−3), even though it is considered in velocity and mass transport calculations for the 1997 and 2008 cruises. The isoneutral distribution gives us a first glimpse at the transport behaviour across the section. The absence of steep slopes in Fig. 3.5 suggests a lack of high transport values along the open-ocean part of the section. The upper panel suggests near-zero transport estimates, except for stations 1-11 (∼13-15.8◦W), where sloping isoneutrals are observed. Also, in the deepest layers a slight incline can be seen, revealing low transports at these depths too. 3.3 Methodology Geostrophic velocities are obtained using the thermal wind equation. To integrate this equation, a reference level (where the velocity should be known) is required. 3.3. METHODOLOGY 35 0 250 500 750 1000 0100200300400500600 36.5 36 35.7 35.6 35.5 35.4 35.4 123456789101112131415161718192021 123456789101112131415161718192021 0 kmkm Depth (m) 1000 2000 3000 4000 5000 14151617181920 14151617181920 35.3 35.2 35 34.95 37 | 34.9 34.95 35.2 35.4 35.5 35.6 35.7 36.5 Salinity Longitude (oW) Figure 3.4: Mean salinity vertical section. Note that the vertical scale is different for each depth range. 36 CHAPTER 3. THE CANARY CURRENT 0 250 500 750 1000 0100200300400500600 26.85 27.162 27.38 27.62 123456789101112131415161718192021 123456789101112131415161718192021 0 kmkm Depth (m) 1000 2000 3000 4000 5000 14151617181920 14151617181920 27.82 27.922 27.975 28.008 28.044 28.072 28.11 28.1295 | 26.44 26.85 27.38 27.922 γn (kg m−3) Longitude (oW) Figure 3.5: Mean neutral density (kg m−3) vertical section. Note that the vertical scale is different for each depth range. 3.3. METHODOLOGY 37 The level of no-motion (zero-velocity) has been chosen following Mach´ın et al. [2006], where a sensitivity analysis was carried out by studying the reference level velocity uncertainty obtained considering different choices and their results through inverse modelling. The chosen values locate the open-ocean stations reference level at γn=27.922 kg m−3(roughly 1600 m) and the shallow-water stations reference level at γn=27.38 kg m−3(roughly 700 m). This last reference level separates layers 4 and 5, and thus represents the limit between central and intermediate waters. This level of no-motion has also been used by Fraile-Nuez et al. [2010] at the Lanzarote Passage. γn=27.922 kg m−3is the interface between layers 7 and 8, separating intermediate and deep waters (Table 3.2). During the calculations, when the deepest common depth is less than the corresponding reference level, the bottom is considered as the zero-velocity layer. Table 3.2: Neutral surface layers (kg m−3). Layer Lower interface Water masses 1γn= 26.44 2γn= 26.85 NACW 3γn= 27.162 4γn= 27.38 5γn= 27.62 6γn= 27.82 AAIW/MW 7γn= 27.922 8γn= 27.975 9γn= 28.008 10 γn= 28.044 NADW 11 γn= 28.072 12 γn= 28.0986 13 γn= 28.11 14 γn= 28.1295 In order to solve the problem attached to the unknown value of a reference layer 38 CHAPTER 3. THE CANARY CURRENT in the water column, ADCP (LADCP and SADCP) data are used in order to obtain a reference velocity. ADCP-referenced geostrophic calculations and their associated mass transports are also computed separately for the built average transect and for the four separate yearly cruises. 3.3.1 Average transect ADCP-referencing Fig. 3.6 shows the initial geostrophic velocity section along the whole transect, obtained using the neutral density computed through the averaged values of potential temperature and salinity mentioned before. Alternating northward and southward flows are appreciated throughout the velocity field. Vertical zero-velocity isolines are distributed separating these northward/southward adjacent currents, which represent the high eddy mesoscale field [Mason et al., 2011]. Horizontal null values in velocity are also seen around 700 m depth for the Lanzarote Passage and at 1600 m for the rest of the section. This corresponds to the no-motion reference velocity location for the shallow-water and open-ocean stations, respectively. Through the whole section, a slight southward flow is revealed, corresponding to the Canary Current. Meanwhile, the northward flow near the bottom in stations 1-5 would represent the AAIW circulation. In Fig. 3.7 the entire ADCP-referencing process is summarized following ComasRodr´ıguez et al. [2010]. First, processed LADCP data are averaged for the 1997, 2006 and 2009 LADCP measurements, together with the bottom-track data available for the two last cruises. SADCP data relative to 2008 are not included in the averaging due to their high scatter near the surface. In order to get a mean ensemble velocity section, LADCP profiles provide a better quantification of the water column velocities. Only the velocity component that is perpendicular to the section is further considered. Thus, a velocity projection is applied in the non-zonal parts of the transect (stations 1-10, up to ∼15.5◦W). Fig. 3.7a highlights the change in velocity performed after adjusting the geostrophic profile to the LADCP profile for station pair 2-3 (∼13.2-13.4◦W), located in the Lanzarote Passage. Initial geostrophic and 3.3. METHODOLOGY 39 0 1000 2000 3000 4000 5000 5 2.5 2 1.5 1.5 1 0.2 0.1 0.1 −9 −3 −3 −2 −1.5 −0.5−0.5 −0.4 −0.4 −0.2 −0.1 Initial geostrophic velocity (cm/s) Depth (m) 123456789101112131415161718192021 123456789101112131415161718192021 14151617181920 14151617181920 Longitude (oW) Positive Negative Zero − Figure 3.6: Mean geostrophic velocity (cm s−1) vertical section. Isolines are drawn as follows: every 0.1 cm s−1from 0 to 0.5 and every 0.5 cm s−1from 0.5 to 20 (positive range of velocities); every 1 cm s−1from -20 to -6, every 0.5 cm s−1from -6 to -0.5 and every 0.1 cm s−1from -0.5 to 0 (negative range of velocities). Positive/negative values stand for northward/southward velocities. 40 CHAPTER 3. THE CANARY CURRENT −20 0 20 −1000 −500 0 Velocity (cm/s) Depth (m) a−20 0 20 −3000 −2000 −1000 0 Depth (m) b Initial geostrophic velocity Averaged LADCP velocity a)st.2−3 b)st.8−9 ADCP referenced velocity LADCP bottom track a)st.2 b)st.8 LADCP bottom track a)st.3 b)st.9 20 18 16 14 −10 −5 0 5 10 Mean tidal barotropic velocity (cm/s) c1 3 5 7 9 111315171921 20 18 16 14 −10 −5 0 5 10 Longitude (°W) Mean reference velocity (cm/s) d1 3 5 7 9 111315171921 Figure 3.7: ADCP-referencing process summary for the averaged data. a) Comparison between the initial and the ADCP-referenced geostrophic profiles for station pair 23 (located in the Lanzarote Passage). b) Comparison between the initial and the ADCP-referenced geostrophic profiles for station pair 8-9. c) Mean tidal barotropic component calculated from the OSU TPXO prediction model. d) Mean velocity at the reference level for the average cruise. The stars mark the station pairs corrected with bottom-track data. The dots represent station pairs where the LADCP profile is used. ADCP-referenced velocities are shown. Full-depth LADCP measurements, as well as bottom-track velocities, are plotted for comparison. LADCP profiles have been averaged for each station pair. Fig. 3.7b shows the same comparison for paired stations 8-9 (∼14.7-15.1◦W). The barotropic tidal component (Fig. 3.7c) is subtracted from 3.3. METHODOLOGY 41 each LADCP (including its bottom-track) velocity profile. This component is calculated using the OSU (Oregon State University) TOPEX/POSEIDON global tidal model (TPXO) [Egbert et al., 1994; Egbert and Erofeeva, 2002]. This global model of ocean tides best fits, in a least-squares sense, the Laplace Tidal Equations and along track averaged data from TOPEX/POSEIDON and Jason (on TOPEX/POSEIDON tracks since 2002) obtained with OTIS (OSU Tidal Inversion Software). The time considered for the tidal prediction at each yearly cruise is the bottom track time, which is the midpoint of time spent at the station. Whenever the available data quality permits, referencing is done using bottomtrack data, assuming these to be the best estimate for velocity corrections [ComasRodr´ıguez et al., 2010]. In cases where bottom-track data seem to register some feature in disagreement with the full-depth profile (e.g., baroclinic component of high frequency inertial velocities), the full water column depth LADCP measurements are used. Both the 200 m near the surface and the seafloor are discarded due to their dependence on the atmosphere directly above and the bottom-track data, respectively. In both cases the LADCP data are averaged for each station pair and differences are obtained by comparing geostrophic and absolute velocities within the coincident depth range. A final mean of these differences is obtained which is considered to be the station pair reference velocity. Fig. 3.7d shows the reference velocity calculated, which replaces the initial zero-velocity. Finally, an ADCP-referenced velocity section is shown in Fig. 3.8. As expected, no qualitative changes are apparent near the surface. Very similar values are apparent in the adjacent northward/southward fluxes already seen in Fig. 3.6. The differences are most visible near the original zero-velocity reference level. The eddy oceanic field is clearly seen along the section, flowing in opposite directions and representing the eddy oceanic field. Slight changes can be seen in the Lanzarote Passage, where the presence of a northward current gains importance and reaches to significantly shallower waters at the western margin. This leads to a better appreciation of the AAIW flow when using ADCP data to correct geostrophic calculations. 42 CHAPTER 3. THE CANARY CURRENT 0 1000 2000 3000 4000 5000 3 2 2 1.5 1.5 1 1 0.3 −11 −5 −4 −2.5 −2 −2 −1.5 −0.3 ADCP − referenced velocity (cm/s) Depth (m) 123456789101112131415161718192021 123456789101112131415161718192021 14151617181920 14151617181920 Longitude (oW) Positive Negative Zero − Figure 3.8: Mean ADCP-referenced geostrophic velocity (cm s−1) vertical section. Isolines are drawn as follows: every 0.1 cm s−1from 0 to 0.5 and every 0.5 cm s−1 from 0.5 to 20 (positive range of velocities); every 1 cm s−1from -20 to -6, every 0.5 cm s−1from -6 to -0.5 and every 0.1 cm s−1from -0.5 to 0 (negative range of velocities). Positive/negative values stand for northward/southward velocities. 3.4. MEAN MASS TRANSPORT AND VARIABILITY 49 −10 −5 0 5 Surface layer (γn<27.38) 1 3 5 7 9 111315171921 −10 −5 0 5 Mean Mass Transport (Sv) Intermediate layer (27.38<γn<27.922) Initial geostrophic ADCP−referenced 20 19 18 17 16 15 14 13 −10 −5 0 5 Deep layer (γn>27.922) Longitude (°W) a b c Figure 3.14: Mean accumulated mass transport (Sv). 50 CHAPTER 3. THE CANARY CURRENT explained by the not-well sampled eddy observed at the end of the section. Concerning the Lanzarote Passage, the AAIW undercurrent can be seen flowing northwards with an initial value of 0.4 Sv, and 0.7 Sv after including the reference velocity. This flow pattern confirms the presence of AAIW proposed earlier (Figs. 3.2 and 3.4). After applying the new reference velocities, a slightly southward mass transport of -1.1 Sv is also seen west of Lanzarote, up to station 17 (∼18◦W), whereas the accumulated geostrophic transport is nearly zero. The stream function at the deepest layer (Fig. 3.14c) carrying NADW is relatively weak up to station 11 (∼15.8◦W). The deep pattern is quite irregular, with an increasing southward flow while moving westwards from station 7 (∼14.4◦W) on. Null mass transports are seen in stations 1-5 because, due to their shallowness, the casts never get to those layers considered deep. As already mentioned, the presence of a not-well resolved eddy masks the transport calculations to the west of station 17 (∼18◦W). The lack of more stations after 20◦W hinders a good definition of the complete eddy structure and its associated flow. If we assume that the transport west of 18◦W is due to an eddy signal, the geostrophic mass transport in the deep layers is 0.7 Sv, whereas the ADCP velocity contribution raises the southward flow up to -3.3 Sv. 3.4.2 Variability between the 1997, 2006, 2008 and 2009 cruises In order to evaluate variability between the cruises, each year’s section is considered individually. Surface, intermediate and deep layers are shown separately for comparison. Mass transports corresponding to geostrophic and ADCP-referenced sections are distinguished. A summary of these results is presented in Table 3.3. The errors estimated in the calculation of reference level velocities (seen in Fig. 3.12) are taken into account at every depth range. Errorbars are shown for the accumulated mass transport in 2006 and 2009. Fig. 3.15 shows the accumulated mass transport computed for the thermocline layer for each year. The main pattern observed in waters down to γn<27.38 kg m−3is a southward flow (except in 2008), as well as important 3.4. MEAN MASS TRANSPORT AND VARIABILITY 51 mesoscale activity revealed by the saw-like structures exhibited in each year’s panel. In the 1997 stream function, the southward flow begins in the Lanzarote Passage (stations 1-5), with higher values in the ADCP-referenced data. However, west of Lanzarote the initial values show a net zero mass transport distribution, whereas the corrected estimates suggest a net southward flow up to the end of the section. If we suppose that transport at the last station pair is due to a not-well resolved eddy and therefore is excluded in the calculations, the mass transport in 1997 reaches initial values of -0.4 Sv and a flow of -4.2 Sv is obtained including the reference velocities acquired from ADCP data. Within the Lanzarote Passage, the mass transport is -1.9 Sv and -3.7 Sv, before and after the correction, respectively. In 2006, transport estimates near the African coast change considerably after introducing ADCP reference velocities. A southward geostrophic flow in the Lanzarote Passage is developed, followed by a slightly southward net mass transport west of station 6 (∼14.1◦W). The corrected estimates present a slightly northward flow east of Lanzarote, which turns into a southward flow following the descending slope pattern when moving westwards towards the end of the section. Thus, the stream function in 2006 shows -4.1 Sv and -2.4 ±1.4 Sv, before and after application of the correction, of which -2.5 Sv and 0.7 ±0.2 Sv, respectively, flow between Lanzarote and the African coast. As shown in Fig. 3.15, the 2008 data register a different pattern, with a southward flow in the Lanzarote Passage followed by zero mass transport for the geostrophic calculations. Regarding the ADCP-referenced calculations, the increasing slope to the west of station 6 (∼14.1◦W) suggests a northward mass transport that counteracts with the Passage flow, ending in zero net accumulated mass transport at station 16 (∼17.6◦W). Mass transport values for the whole section in 2008 are detailed as follows: -1.0 Sv / -0.05 Sv corresponding to the non-referenced / referenced values through ADCP measurements. As seen in Fig. 3.15, the initial stream function in 1997 is very similar to that in 2008. However, the stream functions after the ADCP correction are different. This difference in the main pattern across the section can be attributed to the use of SADCP data in the geostrophic velocity correction for 2008, instead of LADCP data (not available for this cruise). Thus, only the first 800 m of the water column 52 CHAPTER 3. THE CANARY CURRENT contribute to the calculation of reference level velocities, not considering the intermediate and deep water behaviour. Therefore, the reference level velocities inferred from SADCP data alone are probably underestimated. In 2008, the mass transport across the Lanzarote Passage is estimated as -0.8 Sv / -1.2 Sv before and after the ADCP correction. Finally, the 2009 data present a strong southward flow through the thermocline layer in the Lanzarote Passage in both stream functions. Both nonreferenced and ADCP-referenced stream functions show zero net mass transport west of Lanzarote up to station 16 (∼17.6◦W). To the west, geostrophic calculations are flat, whereas the corrected estimates suggest a slight increase in the southward flow. Geostrophic mass transport values obtained are -5.0 Sv for the whole section, of which -3.1 Sv passes east of Lanzarote. On the other hand, the ADCP-referenced solution reveals an integrated mass transport of -6.7 ±1.6 Sv (-4.0 ±0.3 Sv across the Passage). It is appreciated how the ADCP-referenced values follow the already registered structure obtained of the initial geostrophic calculations. On all four cruises, Ekman transport has been added to the shallowest layer. The period considered is the whole month in which each cruise took place. In Fig. 3.16, mass transport at intermediate layers is displayed. The 1997 and 2008 ADCP-referenced stream function end approximately at the longitude of 17.6◦W, whereas 2006 and 2009 data reach near-zero values around 20◦W. However, the most important role at intermediate layers seems to be played by the flows that take place within the Lanzarote Passage. Concerning the stream function behaviour, it is observed that 1997 and 2009 data at intermediate layers both follow a pattern of southward flow after the ADCP correction. On the other hand, 2006 and 2008 mass transports behave in the opposite way, giving importance to a northward flow at intermediate depths. In the following, accumulated mass transport values are shown (geostrophic estimates / ADCP-refererenced calculations) for each individual cruise. In the first case, a near-zero net mass transport is revealed along the whole section for 1997. This pattern remains clear for the geostrophic transport, whereas the presence of mesoscale activity at intermediate depths forces a southward net flow to the 3.4. MEAN MASS TRANSPORT AND VARIABILITY 53 −10 −5 0 5 1997 1 3 5 7 9 111315171921 2006 1 3 5 7 9 111315171921 20 18 16 14 −10 −5 0 5 Longitude (°W) Surface layer (γn<27.38) Mass Transport (Sv) 2008 Initial geostrophic ADCP−referenced 20 18 16 14 2009 Figure 3.15: Accumulated mass transport (Sv) in the surface layer (γn<27.38 kg m−3) for the four winter cruises. west of station 8 (∼14.7◦W). In that year, 0.8 Sv / -2.3 Sv flow across the section, out of which 0.4 Sv / -0.2 Sv correspond to the transect east of Lanzarote. The non-well sampled eddy near the end of the section probably contributes to the high net accumulated mass transport observed in the figure. Following the similarities mentioned above, the 2009 plot presents 1.1 Sv / -1.0 ±2.2 Sv (stations 1-21, ∼13-20◦W) and 0.8 Sv / -0.2 ±0.2 Sv (stations 1-5, ∼13-13.7◦W). The non-referenced northward flow is corrected to turn slightly southwards after station 9 (∼15.1◦W) after the ADCPreferencing. The transport west of the Passage is marked by the eddy field and the net contribution is nearly zero. For the second case (2006 and 2008 data), the 2006 54 CHAPTER 3. THE CANARY CURRENT dataset reveals a flat net accumulated mass transport for the geostrophic calculations. However, the ADCP-referenced stream function shows a northward transport from the African coast up to station 9 (∼15.1◦W). Moving westwards a counteracting southward flow is observed up to station 16 (∼17.6◦W). Once null transport values are reached, zero net contribution is observed at stations 16-21 (∼17.6-20◦W). The values obtained are 0.7 Sv / 0.6 ±1.8 Sv for the whole section, while the mass transport between stations 1 and 5 (∼13-13.7◦W) increases from 0.2 Sv to 2.0 ±0.1 Sv when applying the reference velocity calculation. Finally, in the 2008 transports shown in Fig. 3.16, a zero initial net accumulated mass transport can be seen throughout the section. The rise from the referencing procedure is included (-0.2 Sv becomes 1.4 Sv for the whole transect, while -0.06 Sv becomes -0.3 Sv between the archipelago and the African coast). It is shown that mass transport across intermediate layers is not significant when the error is known (2006 and 2009). Variability in the deep layers among 1997, 2006, 2008 and 2009 is presented in Fig. 3.17. As observed in Fig. 3.14 for the average built section, the deep pattern is quite noisy. No deep waters are found in the Lanzarote Passage due to its shallow bathymetry. Therefore, no mass transport can be observed near the continental slope, but a southward current to the west of station 6 (∼14.1◦W) (except in 2008, that shows a northward mass transport). This difference may be caused, as stated before, by the use of SADCP data alone in the referencing process, which seems not to be as good as LADCP for referencing purposes. The main transport patterns can be compared up to station 16 (∼17.6◦W), because of the lack of further stations in 1997 and 2008. Due to the choice of the no-motion reference level, no high geostrophic mass transport values are shown in the deep layers. After applying the ADCP-referencing, both 1997 and 2006 corrected results show an increase from the initial near-zero accumulated mass transport to a marked southward transport. The 1997 stream function is modified by the presence of an eddy from station 9 (∼15.1◦W) to the end of the section, while the 2006 mass transport also reflects the mesoscale contribution along the whole section, highlighted 3.4. MEAN MASS TRANSPORT AND VARIABILITY 55 −10 −5 0 5 1997 1 3 5 7 9 111315171921 2006 1 3 5 7 9 111315171921 20 18 16 14 −10 −5 0 5 Longitude (°W) Intermediate layer (27.38<γn<27.922) Mass Transport (Sv) 2008 Initial geostrophic ADCP−referenced 20 18 16 14 2009 Figure 3.16: Accumulated mass transport (Sv) in the intermediate layer (27.38< γn<27.922 kg m−3) for the four winter cruises. by the presence of an eddy beginning at station 18 (∼18.5◦W) but not well sampled at its western margin. Mass transport estimates are shown before / after the ADCPreferencing. In 1997, 0.6 Sv / -4.7 Sv are registered, while -1.5 Sv / -5.3 ±3.2 Sv are obtained from the 2006 cruise. These differences are probably due to the not-well sampled eddy present at the end of the survey. Stream function values corresponding to 2008 behave differently, showing a northward trend for both non-referenced and ADCP-referenced solutions, which are 1.4 Sv / 3.5 Sv, respectively. On the contrary, in 2009 there is a northward current shown up to station 16 (∼17.6◦W), but after 18◦W the flow turns southwards ending in an accumulated mass transport of -3.3 Sv / -5.0 ±3.9 Sv across the whole section. The presence of a non-completely sampled 56 CHAPTER 3. THE CANARY CURRENT −10 −5 0 5 1997 1 3 5 7 9 111315171921 2006 1 3 5 7 9 111315171921 20 18 16 14 −10 −5 0 5 Longitude (°W) Deep layer (γn>27.922) Mass Transport (Sv) 2008 Initial geostrophic ADCP−referenced 20 18 16 14 2009 Figure 3.17: Accumulated mass transport (Sv) in the deep layer (γn>27.922 kg m−3) for the four winter cruises. eddy is revealed in both the geostrophic and corrected transports, contributing to the high values observed. As already mentioned, the high mass transport values obtained from the ADCP-referencing procedure are also a consequence of the large mass involved in the deep layer range (from 1600 m to the sea bottom). A small error in the LADCP velocity estimates results in a high error in the deep layer mass transport. This causes the reference velocities calculated to affect the stream function, magnifying the structures present in the deep layers. 3.4. MEAN MASS TRANSPORT AND VARIABILITY 57 Table 3.3: Summary of accumulated mass transport results (Sv). Year Longitude* Thermocline Intermediate Deep 1997 Geostrophic -0.4 0.8 0.6 17.6◦W LADCP-ref. -4.2 -2.3 -4.7 Inverse model -3.7 ±0.8 -0.4 ±1.0 0.4 ±1.8 LP Geostrophic -1.9 0.4 LADCP-ref. -3.7 -0.2 Inverse model -1.0 ±0.7 0.8 ±1.1 2006 Geostrophic -4.1 0.7 -1.5 20◦W LADCP-ref. -2.4 ±1.4 0.6 ±1.8 -5.3 ±3.2 LP Geostrophic -2.5 0.2 LADCP-ref. 0.7 ±0.2 2.0 ±0.1 2008 Geostrophic -1.0 -0.2 1.4 17.6◦W SADCP-ref. -0.05 1.4 3.5 LP Geostrophic -0.8 -0.1 SADCP-ref. -1.2 -0.3 2009 Geostrophic -5.0 1.1 -3.3 20◦W LADCP-ref. -6.7 ±1.6 -1.0 ±2.2 -5.0 ±3.9 LP Geostrophic -3.1 0.8 LADCP-ref. -4.0 ±0.3 -0.2 ±0.2 Averaged Geostrophic -2.9 0.4 0.7 18◦W LADCP-ref. -3.5 -1.1 -3.3 LP Geostrophic -2.1 0.4 LADCP-ref. -1.4 0.7 *Longitude up to which results are shown. See text for explanation. LP stands for Lanzarote Passage (up to 13.7◦W). 58 CHAPTER 3. THE CANARY CURRENT 3.5 LADCP vs. inverse model solution In the framework of the CANIGO project, an inverse box model was carried out in order to determine reference-level velocities in the region [Mach´ın et al., 2006]. The inverse model assumes that the ocean is in hydrostatic and geostrophic balance and permits the exchange of properties between adjacent layers. The goal of the inverse box model is to find an optimal but minimum solution for reference-level velocities that conserves mass transport (and other properties) within a closed volume of water [Wunsch, 1996]. Our transect corresponds to the southern section of the closed box formed by stations occupied during CANIGO. In this section, a comparison between the mass transport computed using LADCP data and using the reference velocities from Mach´ın et al. [2006] is carried out. The same comparison is not done for the RAPROCAN cruises because a closed-box is necessary to carry out a box inverse model. Fig. 3.18 shows the stream function corresponding to the inverse box model solution by Mach´ın et al. [2006] (dashed lines) and the ADCP-referenced procedure (solid lines). Surface (1-4, γn<27.38 kg m−3), intermediate (5-7, 27.38< γn<27.922 kg m−3) and deep layers (8-14, γn>27.922 kg m−3) are considered separately. The January 1997 cruise reached the longitude of 17.6◦W (station 16 in this study, see Fig. 3.1). The inverse model calculations provided accumulated mass transport values of -3.7 ±0.8 Sv for thermocline waters, -0.4 ±1.0 Sv for intermediate waters and 0.4 ±1.8 Sv for deep waters. Thus, only the southward transport in the surface layers is significantly different than zero. On the other hand, the ADCP-referenced accumulated mass transport provides a different solution. The stream function in this case reveals values of -4.2 Sv, -2.3 Sv and -4.7 Sv at surface, intermediate and deep layers, respectively. Both methods here compared provide reference level velocities which are incorporated into the geostrophic calculations leading to new estimates in the water column. It has been therefore shown that the results from both methodologies in shallower waters represent the Canary Current dynamics quite well. It is worth remembering that the inverse model solution resolves the minimum possible reference 4.1. INTRODUCTION 65 4.1 Introduction The Azores Current (AzC) is the northernmost current of the North Atlantic subtropical gyre (NASG). It originates as a branch of the Gulf Stream, heading southeastwards and crossing the Mid-Atlantic Ridge south of the Azores [K¨ase and Siedler, 1982; Gould, 1985]. The AzC flows eastwards as a zonal jet, associated with the Azores Front (AzF), and divides into three main branches that turn southwards [Stramma and Siedler, 1988b; Klein and Siedler, 1989]. The easternmost branch feeds the Canary Current (CC) that flows across the Canary Islands [Stramma, 1984; Stramma and M¨uller, 1989; New et al., 2001; Mach´ın et al., 2006]. The three branches feed the North Equatorial Current, that flows to the west closing the NASG [Hern´andezGuerra et al., 2005]. The eastward mass transport of the AzC has been reported to be about 10-12 Sv (1 Sv ≈109kg s−1), mainly through the upper 1000 m of the ocean [Gould, 1985; Sy, 1988; Stramma and M¨uller, 1989]. Its formation and variability have been studied both through historical datasets and circulation models [Paillet and Mercier, 1997; Pingree et al., 1999; Alves et al., 2002; P´erez et al., 2003]. There are two hypotheses about the driving mechanism of the AzC. The first one is that the Azores Current is driven by the wind stress curl [K¨ase and Krauss, 1996]. The second AzC formation mechanism is believed to imply a water mass transformation associated with the Mediterranean outflow in the Gulf of C´adiz [Jia, 2000; ¨ Ozg¨okmen et al., 2001], relying on the dynamical concept of β-plumes. A combination of both has also been considered by Lamas et al. [2010], providing higher transport estimates (16.5 Sv for the first 1500 m from Argo data). The AzC variability seems to depend mainly on the magnitude of the exchange through the Strait of Gibraltar. Some studies have also demonstrated the presence of a recirculation as a westward counterflow adjacent to the AzC, called the Azores Countercurrent (AzCC) as defined by Onken [1993], who attributes the existence of the AzCC to a feature in the meridional gradient of the wind stress curl. Cromwell et al. [1996] used altimetry 66 CHAPTER 4. THE AZORES CURRENT SYSTEM and hydrography data to indicate that such westward flow north of the AzC appears to be a persistent feature of the circulation in this region. In line with the AzC formation, two mechanisms have been suggested to drive the AzCC. Alves and Colin de Verdi`ere [1999] supported an eddy-driven mechanism, which is expected to form two westward countercurrents, north and south of the AzC due to geostrophic turbulence rectification. In contrast, the second hypothesis suggests the formation of a single westward countercurrent to the north of the AzC, due to the aforementioned topographic β-plume [Kida et al., 2008; Volkov and Fu, 2010]. According to these studies, the resulting principal current flows westwards, transporting 2-7 Sv. Also Fiekas et al. [1992] observed that the initial eastward-flowing 12 Sv registered in the upper 700 m were reduced to 7 Sv when considering the adjacent counterflow. However, most of the literature focuses on modeling or limited sets of data. Several pending questions about these currents, like the observational verification of their vertical extension and the horizontal structure, demand further in-situ measuring efforts. To fulfill this objective, an extensive hydrographic dataset that can be used to verify the conclusions drawn from modeling has been missing. This paper provides a high spatial resolution synoptic survey expressly designed to measure the both the AzC and AzCC and resolve the mesoscale. The main aim of the present study is to delimit (both horizontally and vertically) and quantify the Azores Current System (ACS) through hydrographic data. We present a quantitative study that constitutes a reference in the hydrographic description of the water column in the region while providing a new insight into the meridional horizontal structure of the system. The paper is structured as follows. After presenting the collected hydrographic data and describing the methodology in section 4.2, the water mass distribution and circulation in the Azores Current System is thoroughly studied in section 4.3. Our final discussion is given in section 4.4. 4.2. DATA AND METHODS 67 4.2 Data and methods The ORCA cruise was carried out between 15 October and 11 November 2009 onboard the BIO Hesp´erides. The survey comprised three large-scale sections, two zonal sections at nominal latitudes of 29◦and 37◦N, joined by a meridional section at the longitude of 24.5◦W. In order to locate the AzC and the AzCC, we will focus on this meridional section between latitudes 37◦and 33◦N (stations 39-55, see Fig. 4.1a), as will be later justified. These 17 SeaBird 911+ CTD stations were carried out with a spatial separation of 15 nm (∼28 km) between 26 October and 1 November 2009. At each cast, dual sensors of temperature and conductivity acquired data from the surface down to 15-20 m above the bottom. The temperature and pressure sensors were calibrated before the cruise at the SeaBird facilities, whereas the conductivity sensor was calibrated on board with bottle sample salinities (hereinafter salinity is expressed in the Practical Salinity Scale). Additionally, neutral density (γn) is computed following Jackett and McDougall [1997]. The Lowered Acoustic Doppler Current Profiler (LADCP) system was mounted on the rosette and deployed at each CTD cast. The LADCP consists of two 300 kHz Teledyne/RDI Workhorses (WH) run in master/slave mode. The data are processed using the Visbeck software developed at Columbia University [Fischer and Visbeck, 1993]. Continuous current measurements were also made in the upper 700 m using a 75 kHz Ocean Surveyor Shipboard ADCP (SADCP). These data are properly calibrated and GPS-referenced, as well as processed with the CODAS (Common Ocean Data Access System) processing toolbox. The Visbeck software adjusts the shallowest LADCP data to the SADCP data and the barotropic tidal component is removed from the ADCP data. Continuous temperature records were also acquired using an SBE-21 thermosalinometer and averaged onto a 1/10◦grid. The near-real time merged altimeter product (which uses Jason-1, Envisat, GFO, ERS-1, ERS-2 and Topex/Poseidon data) was produced by Ssalto/Duacs and distributed by Aviso with support from Cnes. Absolute geostrophic velocities, computed from absolute dynamic topography, have been acquired on a 1/4◦grid along the 24.5◦W meridional 68 CHAPTER 4. THE AZORES CURRENT SYSTEM section (averaging the data provided by Aviso at 24.75◦and 24.25◦W from 26 October to 1 November). Ocean surface wind stress data used to obtain Ekman transport come from the QuickScat scatterometer. 25.4 26.5 27.3 27.6 27.9 28.1 Salinity θ (°C) NACW MW SAIW NADW LSW b 34.5 35 35.5 36 36.5 37 5 10 15 20 25 26oW 25oW 24oW 32oN 33oN 34oN 35oN 36oN 37oN 38oN 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 Azores −2000 −4000 −3000 −5000 −5000 a Figure 4.1: a) Location of the CTD stations occupied at the Azores Current System during the ORCA cruise in the fall of 2010. For reference, 200-, 1000-, 2000-, 3000-, 4000and 5000-m isobaths are shown [Smith and Sandwell, 1997]. b) Θ-S diagram. NACW, SAIW, MW (dark gray, stations 41-43), NADW and LSW (light gray, stations 39-43) are shown. Mass transport is calculated by dividing the water column into 14 neutral density layers, following the criteria proposed by Ganachaud [2003] for the Atlantic Ocean, but with a slight modification for central waters. Thus, each thermocline layer covers 4.3. RESULTS 69 roughly a comparable depth range to other layers. Geostrophic relative velocities are obtained using the thermal wind equation. To integrate this equation, the reference level of no-motion has been located at γn=28.072 kg m−3(roughly 3000 m). An estimation of the absolute velocity is obtained using LADCP data [McDonagh et al., 2008] following the procedure described by Comas-Rodr´ıguez et al. [2010]. 4.3 Results 4.3.1 Water mass distribution Potential temperature (Θ) and salinity (S) data obtained during the cruise allow water mass identification as shown in Fig. 4.1b. In this Θ-S diagram, the five main water masses present at the section are marked. A slight scattering is seen above the seasonal thermocline, due to heating, precipitation and evaporation that take place at shallow depths. Fig. 4.2 shows a) Θ, b) salinity and c) γnvertical sections. Central waters extend to a density level of γn<27.38 kg m−3(roughly 700 m), and define the thermocline layer, occupied by North Atlantic Central Water (NACW). The AzF is found near station 52 (∼33.75◦N, marked by a black dot in Figs. 4.2a and 4.2b), corresponding to the 16.2◦C isotherm and 36.2 isohaline at 150 dbar as suggested by P´erez et al. [2003]. In these vertical sections, temperature and salinity gradients can be seen where the AzF is located. Therefore, the AzF is associated with the maximum baroclinic velocities registered in the Azores Current. In contrast, other frontal zones such as the Cape Verde Frontal Zone (CVFZ), although known for their characteristic fluctuations of isolines [Mart´ınez-Marrero et al., 2008], are not associated with maximum velocity values. Below the central waters, at intermediate layers (27.38< γn<27.922 kg m−3, roughly 700-1600 m), relatively warmer and saltier (>35.7) Mediterranean Water (MW) is found between stations 41 and 43 (dark gray in Fig. 4.1b). In Fig. 4.2b, this MW core (36.5◦-36◦N) is located at approximately 1000 m, whereas values above 35.6 spread horizontally up to station 49 (34.5◦N). Between stations 50 and 53 (34.25◦-33.5◦N) relatively cooler and fresher (>35.5) mixed SubArctic Intermediate Water (SAIW) is also found, approximately in the 700-900 m 70 CHAPTER 4. THE AZORES CURRENT SYSTEM 0 150 300 19 15 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 5539 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 23 __ 1000 2000 3000 4000 5000 ||||| 11 9 6 4 3 2.5 1 3 6 9 12 15 17 19 21 a) Potential Temperature (oC) 0 150 300 ||||||||||||||||| 36.4 36 37 __ 1000 2000 3000 4000 5000 ||||| 35.6 35.6 35.4 3535 34.98 34.94 34.92 34.9 34.8 35 35.2 35.4 35.5 35.6 35.7 36.5 b) Salinity 0 1000 2000 3000 4000 5000 ||||| ||||||||||||||||| 26.85 27.162 27.38 27.82 27.922 27.975 28.044 28.072 28.0986 28.11 28.11 __ 25.05 26.85 27.38 27.92 c) γn (kg m−3) 37 36 35 34 3337 36 35 34 33 Latitude (oN) Depth (m) Figure 4.2: Vertical sections for a) Θ (◦C), b) salinity and c) γn(kg m−3). Note the different vertical scales in plots a) and b). The black dots mark the location of the AzF as suggested by P´erez et al. [2003]. 4.3. RESULTS 71 range. The upper limit of the SAIW domain is defined by Arhan [1990] at the 27.3 isopycnal, but ‘pure’ SAIW temperature and salinity values are much lower than those shown here. The deep layers, from approximately 1600 m to the ocean bottom (γn>27.922 kg m−3), are composed of North Atlantic Deep Water (NADW) and Labrador Sea Water (LSW). LSW is found from stations 39 to 43 (light gray in Fig. 4.1b, 37◦-36◦N in Fig. 4.2). The presence of LSW is evidenced by the rise of isohalines (Fig. 4.2b) and bending of isoneutrals (Fig. 4.2c) below 2000 m, as well as Θ values above 2◦C and salinity around 34.9 [van Aken, 2000]. 4.3.2 The Azores Current and Countercurrent System In order to horizontally delimit the AzC and AzCC domains, velocity from the SADCP is shown in Fig. 4.3. Velocity averaged over nearly the 700-m depth range reached by the instrument (Fig. 4.3a) and averaged velocity up to 50 m (Fig. 4.3b), are plotted separately for comparison. Fig. 4.3 clearly reveals a cyclonic recirculation at the southern part of the section, at latitudes south of 33◦N (station 55). This study only covers stations 39 (located at the northernmost point of the section) to 55 in order to focus on the AzC and AzCC . In the following, the transect is divided into four different sectors. From station 39 (37◦N) and up to station 42 (36.25◦N), a westward velocity is present which, as we will show, belongs to an anticyclonic recirculation located north of the section. Between stations 42 and 46 (36.25◦-35.25◦N) we find the westward-flowing AzCC at subsurface levels, unnoticeable at the surface, as shown by the lack of flow in Fig. 4.3b. To the south, enclosed by stations 46 and 49 (35.25◦-34.5◦N), there is a cyclonic eddy forced by the strong horizontal shear in this transition area. Finally, the AzC can be identified between stations 49 to 53 (34.5◦-33.5◦N), flowing eastwards in the surface layer and below (Figs. 4.3a and 4.3b). Fig. 4.4 shows the reference velocity obtained for the section. Each value, corresponding to a station pair, represents the velocity applied to the geostrophic profile to match the LADCP data. As described in Comas-Rodr´ıguez et al. [2010], either bottom-track or LADCP full-depth profiles are used when referencing the geostrophic 72 CHAPTER 4. THE AZORES CURRENT SYSTEM 26oW 25oW 24oW 30oN 32oN 34oN 36oN 38oN 39 49 55 10 cm s−1 Latitude Mean velocity (<700m) a 26oW 25oW 24oW 30oN 32oN 34oN 36oN 38oN 39 49 55 10 cm s−1 Longitude Surface mean velocity (<50m) b Figure 4.3: a) Full-depth averaged and b) mean surface velocities from the SADCP. For reference, locations of the CTD stations and main isobaths are shown. 4.3. RESULTS 73 calculations. Reference velocities included in the corrected transport estimates are quite small (see Fig. 4.4). Fig. 4.5 shows the accumulated mass transport along the section, considering surface (γn<27.38 kg m−3), intermediate (27.38< γn<27.922 kg m−3) and deep (γn>27.922 kg m−3) layers separately. Ekman transport has been added to the shallowest layer. Both geostrophic (dashed lines) and LADCPreferenced (solid lines) stream functions are shown separately in these calculations. A large water volume is comprised between γn=27.922 kg m−3and the bottom (from approximately 1600 m to almost 5000 m depth). Thus, the transport in the deep water layers is slightly different, magnifying the reference velocity contribution. In the following, we will discuss the flow resulting from the LADCP-referenced mass transport estimates. 33 34 35 36 37 −3 −2 −1 0 1 2 3 Latitude(°N) Reference Velocity (cm/s) LADCP profile corrected LADCP bottom−track corrected 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 Figure 4.4: Reference level velocities (cm s−1)obtained from LADCP bottom-track (stars) and full-depth (dots) profiles. 74 CHAPTER 4. THE AZORES CURRENT SYSTEM Fig. 4.5 presents a quantification of the flow within the AzC and AzCC. The negative (westward) flow from stations 39 to 42 (37◦-36.25◦N) is caused by an anticyclonic recirculation, as will be discussed later. Concerning the AzCC domain, between stations 42 and 46 (36.25◦-35.25◦N), a prominent westward flow is noticeable at surface and intermediate layers. Mass transport at both the surface (-3.4 Sv) and intermediate (-3.1 Sv) layers contributes to the AzCC, with a westward flow of 6.5 Sv. The cyclonic eddy located south of station 46 (35.25◦N) is evident until station 48 (34.25◦N), thereafter the mass transport is nearly zero (from surface to γn=27.922 kg m−3) until station 49 (34.5◦N). Between stations 49 and 53 (34.5◦-33.5◦N), the AzC transports a total of 14.9 Sv (10.7 Sv carried at thermocline layers and 4.2 Sv at intermediate depths) eastwards. With the abovementioned results, an eastward net mass transport of 8.4 Sv flows across the meridional section at 24.5◦W. Fig. 4.5 confirms that the widths of the AzC and AzCC are about 110 km each. The maximum eastward flow of the AzC is associated with the AzF, located near station 52 (∼33.75◦N) as seen in Fig. 4.2. Fig. 4.6 reveals the vertical structure of the AzC and AzCC from SADCP data (Figs. 4.6a and 4.6b) and the calculated mass transports (Figs. 4.6c and 4.6d). Both the geostrophic (dashed line) and LADCP-referenced (solid line) estimates are shown (see caption in Fig. 4.6). A very small vertical shear in the westward velocity corresponding to the AzCC (Fig. 4.6a) is seen between 150 m and 650 m depth, with a slight maximum registered just below the surface. Integrated mass transport in this same area (Fig. 4.6c) show that the highest transport is located roughly between 500 and 700 m (27.162< γn<27.38 kg m−3). Fig. 4.6b shows a baroclinic eastward flow corresponding to the AzC, as also seen in Fig. 4.6d. Figs. 4.6c and 4.6d show significant mass transport up to approximately 2000 m, including the thermocline as well as intermediate layers (up to γn=27.975 kg m−3). The eastward flow seen at the deepest layers in Fig. 4.6c is probably associated with the presence of LSW on the northern margins of the AzCC domain, also observed in Fig. 4.1b, which flows across the section in the deepest layers. 82 CHAPTER 5. DISCUSSION the addition of a baroclinic velocity component (dependent on depth) and a reference level velocity. The assumption of a no-motion reference level leads to the initial calculation of mass transports. In this thesis, ADCP (Acoustic Doppler Current Profiler) data are used to estimate the velocity at the reference layer and, consequently, to calculate the absolute mass transports. This is a significant advance on this topic in comparison with previous studies, due to the statistical analysis carried out in order to determine which instrument provides a better direct velocity measurement. The dataset acquired along 24.5◦N permitted this comparison between Shipboard ADCP (SADCP) (down to ∼700 m), Lowered ADCP (LADCP) bottom track (from ∼200 m above the bottom), and full-depth profiles. The LADCP data processing provides a range of options, allowing the resulting velocity profiles to be constrained by CTD, GPS, bottom-track or even SADCP measurements. In this thesis, different choices based on the proposed methodology are taken, considering the quality of each dataset. Confidence in the estimated reference level velocities depends on the quality of the employed data. Critically reviewing the acquired ADCP data shown in this thesis, some deficiencies related to instrument performance should be mentioned. To overcome these impediments, the use of these data has been carefully studied throughout. However, it is a fact that continuous improvements are being made in the performance of the instruments. Their data processing and further application are being improved thanks to the current studies carried out regarding this topic. The suggested methodology is applied to four deep hydrographic sections from the African coast to 20◦W to compute absolute mass transport across the Canary Islands. The section was repeatedly occupied during the winter seasons of 1997, 2006, 2008 and 2009. The presence of mesoscale features is evident in the flow estimates for each yearly realization. An averaged section of these four surveys is built, decreasing the eddy field. The use of ADCP-referenced calculations gives a new view of the Canary Current. The Canary Current transports similar values before (-2.9 Sv) and after (-3.5 Sv) applying the proposed methodology. The transport estimates are coherent with previous studies in the region. However, a horizontal stretching 5.1. GENERAL DISCUSSION 83 of the Canary Current is seen when the ADCP velocities are used. Therefore, an important contribution of this study is this redefinition of the Canary Current horizontal extension, that was initially believed to be restricted to the easternmost part of the Canary Islands (between 13◦and 14.7◦W), but finally revealed to flow across the entire archipelago (up to 18◦W). Considering the results obtained for each individual cruise, variability between 1997, 2006, 2008 and 2009 can be appreciated in the central waters. Once again, the sampled eddies make the inference of a net Canary Current transport from each yearly cruise difficult. By discarding the unresolved mesoscale features at the end of each section and paying attention to the trends in the thermocline flow, some patterns can be seen. According to the results shown, ADCP-referenced transports are -4.2 Sv in 1997, -2.4 Sv±1.4 in 2006 and -6.7±1.6 Sv in 2009. However, the flow in 2008 is nearly zero. These estimates give a first approach to the evaluation of the mass transport variability in the region. The study of the Azores Current System sheds some light upon the matter concerning the origins of the Canary Current. In the study shown here, it is demonstrated that the Azores Current System displaces eastwards a total mass transport of 8.4 Sv (across thermocline and intermediate layers). Another important advance is made by locating and quantifying the Azores Countercurrent, essential in the understanding the contribution of this current system to the regional circulation at the eastern boundary of the NASG. We have considered the idea that an Azores Current branch turns southwards and feeds the Canary Current. Taking into account only the thermocline transport, 10.7 Sv are carried eastwards, while -3.4 Sv flow to the west, driven by the Azores Countercurrent. Hence, the surface net mass transport is 7.3 Sv to the east across the nominal longitude of 24.5◦W. According to previous studies estimating the characteristic mass transport across the Canary Islands, this net value by itself could feed the southward current flow of the Canary Current. Further studies using a box 84 CHAPTER 5. DISCUSSION 28oW 24oW 20oW 16oW 12oW 8oW 28oN 30oN 32oN 34oN 36oN 38oN Figure 5.1: Geographical location of the area of study including the sections occupied during the ORCA (red dots) and RAPROCAN (black dots) cruises. layout accomplished during the ORCA cruise (Fig. 5.1) would confirm that there is nearly zero net flow entering the northern section, demonstrating the hypothesis that the Portugal Current is not a source for the Canary Current. In this thesis a lack of westward flow south of the Azores Current System through the western section is also shown. The main part of the flow coming from the Azores Current crosses the Canary archipelago across the southern section of the ORCA box grid, feeding the Canary Current. 5.2. CONCLUSIONS 85 5.2 Conclusions The conclusions that arise from this thesis are: 1. The Canary Current is a fundamental surface current in the North Atlantic subtropical gyre circulation. Our previous knowledge about its southward flow and mass transport is confirmed by the values given in the studies carried out within this thesis. The need for a continuously repeated deep hydrographic section is important in order to measure its mean flow and its variability. 2. The calculation of reference velocities replacing the no-motion level assumption provides a new approach to the estimation of flows in the region. The most accurate data are bottom-track records, but their use is conditioned by coherency with the behaviour over the water column. The ADCP-referencing method performs better when applying LADCP full-depth profiles that have been processed by including the bottom-track and SADCP measurements as constraints. 3. The periodical repetition of a hydrographic section north of the Canary Islands (29◦N) provides the possibility of discerning a mean circulation pattern. The averaging of the data acquired in winter of 1997, 2006, 2008 and 2009 probably helps to erase the mesoscale interaction over a section which is contaminated by eddies during each yearly realization. The ADCP-referencing procedure does not significantly vary the thermocline mass transports estimated in previous studies in the region. However, a new result concerns the Canary Current horizontal extension. 4. No significant changes are proposed here in the quantification of the mass transport across the Canary archipelago during the winter season among the four annual studies (1997, 2006, 2008 and 2009). Although there are some variations in the estimates, the presence of an important eddy field in the region and the need for the use of different data sources in the ADCP-referencing makes the reaching of definite conclusions in terms of the variability of the Canary Current difficult. 86 CHAPTER 5. DISCUSSION 5. It is demonstrated that the Azores Current System plays a fundamental role in the Canary Basin circulation dynamics. The Azores Countercurrent is confirmed north of the Azores Current and quantified, reducing the net eastward transport observed across the section along 24.5◦W. Preliminary conclusions can be drawn to suggest that the main part of this flow may turn southwards and feed the surface flow across the Canary archipelago. 5.3. FURTHER RESEARCH 87 5.3 Further research Results presented in this thesis constitute the basis for future studies of the regional circulation in the Canary Basin. Undoubtedly, an improvement in velocity-measuring instrument (ADCP) performance is needed to reduce the error attached to the use of their data as a reference for geostrophic calculations. The addition of error bars to the final mass transports in future studies might lead to a better assessment of the range of uncertainty introduced by the calculated reference velocities. This thesis demonstrates the possibility of studying the net flow pattern, avoiding the introduction of unresolved mesoscale features. Nonetheless, the evaluation of the mesoscale contribution needs revision, so that no eddies would affect the estimated flow. Future research will be focused on applying the presented methodology to alternative datasets. This includes the study of measurements carried out along the RAPROCAN section during the summer season. This will lead to the possibility of inferring seasonal changes as well as examining the variability affecting the Canary Current between the surveyed years. To confirm the Azores Current as a main source for the Canary Current, a study is being carried out considering the full ORCA station grid. Transport will be assessed across the box layout, based on an inverse model, that support the preliminary conclusions obtained in this thesis regarding the origins of the Canary Current. Appendix A Resumen en espa˜nol /Spanish summary A.1 INTRODUCCI´ ON GENERAL A.1.1 Localizaci´on oceanogr´afica El giro subtropical de Atl´antico Norte (NASG, North Atlantic subtropical gyre, en ingl´es) se extiende desde 15◦hasta 45◦N, encontr´andose el continente americano en su margen occidental. Por el este, est´a limitado por Europa, el estrecho de Gibraltar (que conecta el Atl´antico con el mar Mediterr´aneo) y ´ Africa. Su importancia en la circulaci´on del Atl´antico Norte ha despertado nuestro inter´es durante d´ecadas, dando lugar a estudios en ambas fronteras del giro. En el margen oriental del NASG se encuentra la Cuenca de Canarias, la cual se extiende entre 20◦y 40◦N, con un ´area total de 7 millones de km2, localizada desde 10◦hasta 40◦W. Debido a su grandes dimensiones, pueden encontrarse diferentes masas de agua en la Cuenca de Canarias. Una masa de agua est´a formada por un volumen de agua que puede ser identificado por su lugar de formaci´on y se diferencia por sus valores caracter´ısticos de temperatura y salinidad. Generalmente, las masas de agua se ven 89 90 APPENDIX A. RESUMEN EN ESPA ˜ NOL /SPANISH SUMMARY afectadas por la interacci´on atm´osfera-oc´eano en el lugar de su hundimiento, se establecen en un rango de profundidades dependiendo de su densidad y fluyen a lo largo de las isopicnas. En la Cuenca de Canarias podemos encontrar las siguientes masas de agua: El Agua Central del Atl´antico Norte (NACW, North Atlantic Central Water, en ingl´es) es una masa de agua central localizada entre la superficie y los 600-800 metros de profundidad. Puede ser sencillamente identificada en un diagrama Θ-S por una l´ınea bien definida entre los 26.5 y 27.3 kg m−3(σθ). Por debajo de las aguas centrales se encuentran las aguas intermedias. En algunos casos, ´estas se forman a altas latitudes y, por lo tanto, se caracterizan por su alta densidad. En el rango de los 700-900 metros, podemos encontrar Agua Ant´artica Intermedia (AAIW, Antarctic Intermediate Water, en ingl´es), la cual es reconocida por ser relativamente m´as fr´ıa y menos salina que las aguas que la rodean. Por el contrario, el Agua Mediterr´anea (MW, Mediterranean Water, en ingl´es) es f´acilmente identificada por sus altos valores de temperatura y salinidad. Aproximadamente entre los 800-1400 metros, se localizan n´ucleos aislados de MW, conocidos como Meddies (remolinos de agua mediterr´anea, o Mediterranean eddies, en ingl´es). Por debajo de los 1500 metros, el Agua Profunda del Atl´antico Norte (NADW, North Atlantic Deep Water, en ingl´es) ocupa el mayor volumen de agua en el Oc´eano Atl´antico, extendi´endose por encima de los 4500 metros de profundidad. La NADW incluye adem´as el Agua del Mar del Labrador (LSW, Labrador Sea Water, en ingl´es), la cual se forma durante el invierno por convecci´on profunda en el Mar del Labrador. En la capa m´as profunda de esta cuenca, tambi´en se encuentra Agua Ant´artica de Fondo (AABW, Antarctic Bottom Water, en ingl´es) diluida. Localizadas en el margen oriental de la cuenca canaria, aproximadamente a 100 kil´ometros de la costa de ´ Africa, se encuentran las Islas Canarias. Son un archipi´elago espa˜nol situado entre 27◦y 29◦N, desde 18◦hasta 13◦W. En el norte de la Cuenca de Canarias, se encuentra el archipi´elago de Azores, a unos 1500 kil´ometros al oeste de la costa portuguesa. Con una orientaci´on de oeste-noroeste a este-sureste, se extiende A.1. INTRODUCCI ´ ON GENERAL 91 28oW 24oW 20oW 16oW 12oW 8oW 28oN 30oN 32oN 34oN 36oN 38oN Azores Canary Islands Madeira Figura A.1: Localizaci´on geogr´afica del ´area de estudio. Las principales is´obaras se muestran como referencia seg´un la base de datos de Smith-Sandwell [Smith and Sandwell, 1997]. en el rango de latitudes de 36.5◦a 40◦N, entre los 31.5◦y 24.5◦W. El ´area de estudio cubierta por esta tesis se ubica en la Cuenca de Canarias, desde la corriente superficial que fluye a trav´es de las Islas Canarias hasta el sistema de corrientes que se localiza al sur del archipi´elago de Azores (Fig. A.1). 92 APPENDIX A. RESUMEN EN ESPA ˜ NOL /SPANISH SUMMARY A.1.2 Las Corrientes de Canarias y Azores como parte del giro subtropical del Atl´antico Norte Figura A.2: Las corrientes superficiales del Oc´eano Atl´antico seg´un Tomczak and Godfrey [1994]. Los vientos predominantes al norte de los tr´opicos son los alisios (vientos del este) y los vientos del oeste a latitudes medias. ´ Estos crean convergencia de Ekman en el Atl´antico Norte, localizando un centro de altas presiones (anticicl´onico) a latitudes medias, el cual provoca el giro subtropical. La circulaci´on alrededor de un anticicl´on es horaria en el Hemisferio Norte debido al efecto de Coriolis. En el margen occidental A.2. METODOLOG´ IA 99 A.2 METODOLOG´ IA Las observaciones de la velocidad mediante ADCP proporcionan una herramienta oceanogr´afica para el estudio de muchos aspectos de la din´amica de los oc´eanos. Los ADCP instalado en el casco del barco (SADCP, Shipboard ADCP, en ingl´es) dan una detallada comprensi´on de la din´amica de la parte superior del oc´eano hasta una profundidad m´axima de aproximadamente 1000 metros. Los ADCP instalados en la roseta hidrogr´afica (LADCP, Lowered ADCP, en ingl´es) proporcionan perfiles de velocidad en todo el rango de profundidades de una estaci´on hidrogr´afica est´andar. Las medidas directas de velocidad ya se han utilizado para estudiar la circulaci´on oce´anica. Por ejemplo, datos de LADCP en la Corriente de Agulhas revelaron una estructura vertical muy diferente a la dada por la idea tradicional de nivel de nomovimiento en una capa profunda. El volumen total de transporte de la Corriente de Agulhas fue recalculada, revelando la presencia de una corriente hacia el noreste [Beal and Bryden, 1997]. Los datos de LADCP tambi´en han sido utilizado para corregir los c´alculos geostr´oficos iniciales en diferentes regiones oceanogr´aficas [Joyce et al., 2001; McDonagh et al., 2008]. En estos dos casos, las observaciones de velocidad se utilizaron para estimar la velocidad en el nivel de referencia como condici´on inicial para un modelo inverso. Para ilustrar el uso de la metodolog´ıa propuesta en esta tesis, se emplear´an los datos de la campa˜na D279, llevada a cabo en la primavera de 2004 a bordo del RRS Discovery. El principal objetivo de este muestreo fue estimar la circulaci´on a trav´es de la secci´on zonal a 24.5◦N. Esta secci´on ya ha sido estudiada previamente en 1957 [Fuglister, 1960], 1981 [Roemmich and Wunsch, 1985], 1992 [Parrilla et al., 1994] y 1998 [Baringer and Molinari, 1999]. La campa˜na D279 incluy´o, por primera vez para esta secci´on, perfiles de LADCP que se pueden utilizar para estimar velocidades de referencia. La circulaci´on a trav´es del paralelo de 24.5◦N, as´ı como su variabilidad, ya han 100 APPENDIX A. RESUMEN EN ESPA ˜ NOL /SPANISH SUMMARY sido estudiadas con anterioridad empleando datos de la campa˜na D279. Estos c´alculos se encontraban limitados por la incertidumbre asociada a la obtenci´on de velocidades de referencia, consideradas nulas en torno a los 1000 decibares entre 80◦W y 70◦W y a los 3000 decibares para el resto de la secci´on. Asimismo se a˜nadi´o una velocidad de referencia uniforme con el fin de forzar el transporte geostr´ofico ya conocido para la Corriente del Golfo [Bryden et al., 2005]. Mediante la metodolog´ıa propuesta en esta tesis [Comas-Rodr´ıguez et al., 2010], se aplicar´a un desplazamiento a los perfiles geostr´oficos bas´andonos en las medidas promediadas en profundidad de LADCP, corrigiendo las velocidades iniciales y dando lugar a estimaciones absolutas. A.2. METODOLOG´ IA 101 A.2.1 Adquisici´on de datos El transecto transatl´antico realizado por la campa˜na D279 a lo largo de 24.5◦N se muestra en la Fig. A.4. ´ Esta tuvo lugar en primavera de 2004 (del 4 de abril al 10 de mayo) y se tomaron medidas hidrogr´aficas, de velocidad y bioqu´ımicas de toda la columna de agua, desde las aguas someras de la costa oriental de los Estados Unidos de Am´erica hasta las del oeste de ´ Africa [Cunningham, 2005]. Se realizaron 125 estaciones hidrogr´aficas hasta el fondo, con un CTD (Seabird 911+) con sensores duplicados de temperatura y salinidad. Las conductividades medidas por el CTD fueron calibradas compar´andolas a los valores obtenidos a partir de muestras de agua tomadas durante el tramo de subida de la roseta hidrogr´afica. Para tener en cuenta la deriva de los sensores, se corrigi´o su pendiente. 90oW 72oW 54oW 36oW 18oW 0o 20oN 24oN 28oN 32oN 36oN 40oN Atlantic Ocean Longitude Latitude Figura A.4: Posici´on de las estaciones durante la secci´on hidrogr´afica transatl´antica llevada a cabo durante la campa˜na D279. Se realizaron 125 estaciones a lo largo de la latitud de 24.5◦N. Tres cabezales de LADCP fueron instalados en la roseta utilizada para las estaciones. Adem´as, medidas continuas de velocidad fueron recogidas en los primeros 1000 metros mediante un SADCP instalado en el casco del buque oceanogr´afico. Los LADCP consist´ıan en un Broadband (BB) individual a 150 kHz con el cabezal apuntando al fondo marino (con sus propias bater´ıas) y dos Workhorse (WH) a 300 kHz, empleados en modo maestro/esclavo, con uno apuntando a la superficie del oc´eano (esclavo; slave, en ingl´es) y otro hacia el fondo (maestro; master, en ingl´es). Estos dos 102 APPENDIX A. RESUMEN EN ESPA ˜ NOL /SPANISH SUMMARY compart´ıan un paquete de bater´ıas. El SADCP instalado fue un Ocean Surveyor a 75 kHz con un ´angulo de haz de 30◦. Fue configurado para tomar medidas a intervalos de 120 segundos, con 60 celdas de 16 metros de grosor. Los datos se almacenaron empleando el software de adquisici´on de RDI (Teledyne RD Instruments), siendo luego promediados a 2 y 10 minutos. A.2. METODOLOG´ IA 103 A.2.2 Procesado de datos de LADCP Los datos de LADCP se procesaron empleando el software de Visbeck, desarrollado en la Universidad de Columbia [Fischer and Visbeck, 1993]. El procedimiento fue aplicado a diferentes combinaciones de datos con el fin de determinar cu´al de los cabezales tom´o mejores medidas durante la campa˜na. Para ello, se procesaron los datos siguiendo tres distintas combinaciones de los datos: datos del BB, del WH “master” y del acoplamiento “master/slave” del WH. Adem´as, se referenciaron los datos mediante medidas de GPS para obtener velocidades absolutas. Las opciones del software para restringir el procesado de LADCP a las medidas cerca del fondo (bottom-track, en ingl´es) o al SADCP [Visbeck, 2002] no fueron utilizados en este caso. En su lugar, mantuvimos ambos tipos de datos independientes del resultado del procesamiento para poder compararlos con los perfiles de LADCP. La Fig. A.5 muestra los datos procesados para cada combinaci´on de instrumentos en la estaci´on 23 (26.5◦N, 75.9◦W). Adem´as se muestran los datos de “bottom-track” en el rango cercano al fondo y de SADCP en los primeros 1000 metros de la columna de agua. Pueden apreciarse notables diferencias entre las medidas del “bottom-track” y el perfil de velocidades. De acuerdo al software de procesado, el error promedio estimado es de 1.8 cm s−1para los datos de “bottom-track” en la estaci´on 23. De manera preliminar, se asume que es preferible emplear los datos de “bottom-track” para el c´alculo de velocidades de referencia; no obstante, ser´a necesario un estudio estad´ıstico que cuantifique estas diferencias y determine qu´e instrumento obtuvo mejores medidas durante la campa˜na. Para ello, unas diferencias muy peque˜nas (casi nulas) entre los “bottom-track” del BB y del WH “master” implicar´an una mejor calidad en los datos (Fig. A.6a). Cada perfil de LADCP fue comparado con su “bottom-track”. Adem´as, el tramo de descenso de la roseta, as´ı como el de ascenso y el promedio de ambos, fueron considerados por separado. Asimismo, los datos de los cabezales BB, WH “master” y el par WH “master/slave” se estudiaron de manera independiente. Las medidas en el rango del fondo se compararon con el “bottom-track” (Fig. A.6b). Se ha asumido 104 APPENDIX A. RESUMEN EN ESPA ˜ NOL /SPANISH SUMMARY −800 −600 −400 −200 0a) −4000 −3500 −3000 −2500 −2000 −1500 −1000 Depth (m) −30 −20 −10 0 10 20 −4700 −4600 −4500 Eastwards Velocity (cm/s) −800 −600 −400 −200 0 b) −4000 −3500 −3000 −2500 −2000 −1500 −1000 SADCP BB mean (upcast & downcast) BB bottom−track WH master mean (upcast & downcast) WH master bottom−track WH master/slave mean (upcast & downcast) WH master/slave bottom−track −30 −20 −10 0 10 20 −4700 −4600 −4500 Northwards Velocity (cm/s) Figura A.5: Datos de LADCP procesados mediante el m´etodo Visbeck en la estaci´on 23 (26.5◦N, 75.9◦W): a) componente zonal de la velocidad, b) componente meridional de la velocidad. BB y WH “master” (maestro) representan los datos procesados de manera individual, mientras que WH “master/slave” (maestro/esclavo) corresponde con los datos combinados de los cabezales orientados en sentidos inversos del descenso de la roseta. Debe apreciarse que la escala vertical de los ejes es diferente para los distintos rangos de profundidades. Los perfiles medios (mean, en ingl´es) son el promedio entre el recorrido de descenso y de ascenso de la roseta. que las medidas del perfil de LADCP que se aproximan a las del “bottom-track” tienen mejor calidad (la diferencia entre estas medidas es, por tanto, m´as cercana a cero). Adem´as se compararon los datos de SADCP con los perfiles de LADCP en el rango cerca de la superficie (Fig. A.6c). Para todas estas diferencias se obtuvo una media y su desviaci´on est´andar. Consideramos que el instrumento cuyas medidas son mejores son aquellas cuya diferencia media y desviaci´on est´an m´as pr´oximas a cero. A la hora de obtener las mencionadas diferencias entre las velocidades medidas, la frontera oeste y el oc´eano interior han sido considerados separadamente debido a A.2. METODOLOG´ IA 105 −20 0 20 a) −10 0 10 −20 0 20 Velocity (cm/s) b) −10 0 10 −81 −79 −20 0 20 c) −70 −60 −50 −40 −30 −20 −10 −10 0 10 Longitude (°E) BB WH master WH master/slave Figura A.6: An´alisis estad´ıstico para los diferentes instrumentos empleados: a) media y desviaci´on est´andar de la diferencia entre las medidas de “bottom-track” del BB y del WH “master”; b) media y desviaci´on est´andar de las diferencias entre cada perfil de LADCP y su “bottom-track”; c) media y desviaci´on est´andar de cada perfil de LADCP y las medidas de SADCP en la zona superficial de la columna de agua. las diferentes din´amicas que implican. A continuaci´on, se muestran dos valores por instrumento, correspondiendo el primero a la frontera oeste de la secci´on (cerca del continente americano) y el segundo al oc´eano interior. Las diferencias obtenidas entre las medidas de “bottom-track” del BB y del WH “master” son 8.8 ±0.1 / 1.0 ± 0.2 cm s−1(Fig. A.6a). Por otro lado, la diferencia promedia entre el perfil de cada instrumento y su “bottom-track” ser´ıa -12.5 ±0.1 / -0.2 ±0.2 cm s−1para el BB, -3.1 ±0.1 / -0.1 ±0.1 cm s−1para el WH “master” de manera individual, y -2.6 ±0.0 / -0.7 ±0.1 cm s−1para el acoplamiento WH “master/slave” (Fig. A.6b). Finalmente, 106 APPENDIX A. RESUMEN EN ESPA ˜ NOL /SPANISH SUMMARY las diferencias medias obtenidas entre cada perfil de LADCP y el SADCP son -3.7 ± 0.1 / 1.3 ±0.3 cm s−1para el BB, 5.9 ±0.2 / 0.4 ±0.2 cm s−1para el WH “master” de manera individual, y -3.1 ±0.2 / -1.2 ±0.2 cm s−1para el acoplamiento WH “master/slave” (Fig. A.6c). Teniendo en cuenta que la diferencia media es menor para el WH “master”, escogemos este instrumento para obtener velocidades en el nivel de referencia con que corregir los c´alculos geostr´oficos iniciales, sin tener en consideraci´on al WH “slave”. Asumimos que estos datos son los de mejor calidad porque sus medidas se aproximan mejor a los datos de “bottom-track” y SADCP que los dem´as instrumentos. Esta apreciaci´on es notable en el oc´eano interior; sin embargo, puede observarse como la combinaci´on WH “master/slave” reproduce mejor el comportamiento en la frontera oeste del transecto. Desafortunadamente, para este conjunto de datos en particular, debemos descartar la posibilidad de emplear los datos combinados de WH “master/slave” ya que el cabezal esclavo sufri´o errores en la recepci´on de datos durante la campa˜na y dej´o de ser utilizado despu´es de la estaci´on 81 (24.5◦N, 44.9◦W). Para poder realizar una comparaci´on entre las velocidades medidas in-situ y las geostr´oficas, s´olo podremos emplear la componente de la velocidad que sea perpendicular al transecto. Por ello, se aplica una rotaci´on a las medidas en los tramos de la secci´on oblicuos, cerca de las fronteras este y oeste del muestreo. A.2. METODOLOG´ IA 107 A.2.3 Correcci´on de la componente barotr´opica de la marea La componente barotr´opica de la marea ha sido eliminada de las medidas de LADCP, su “bottom-track” y SADCP. Para calcularla se emple´o un modelo de marea global (TPXO, TOPEX/POSEIDON global tidal model, en ingl´es) de la Universidad del Estado de Oreg´on (OSU, Oregon State University, en ingl´es) seg´un Egbert et al. [1994]; Egbert and Erofeeva [2002]. Este modelo realiza una aproximaci´on por m´ınimos cuadrados de las ecuaciones de marea de Laplace, empleando datos de los sat´elites TOPEX/POSEIDON y Jason obtenidos mediante el software OTIS (OSU Tidal Inversion Software). El instante de tiempo considerado para la predicci´on de la marea es la hora de llegada al fondo del perfil de la roseta, coincidiendo con la mitad del tiempo que el buque se detuvo en cada estaci´on. Una vez calculada, la componente barotr´opica de la marea (Fig. A.7) se elimina de las medidas de velocidad in-situ tomadas por los diferentes corrent´ımetros ac´usticos Doppler. −80 −70 −60 −50 −40 −30 −20 −10 −20 0 20 40 Longitude (°E) Velocity (cm/s) Figura A.7: Componente barotr´opica de la marea calculada del modelo de predicci´on OSU TPXO. Esta velocidad ser´a eliminada de las medidas de velocidad obtenidas por el LADCP y el SADCP. 108 APPENDIX A. RESUMEN EN ESPA ˜ NOL /SPANISH SUMMARY A.2.4 Velocidades en el nivel de referencia −800 −600 −400 −200 0 −8 −6 −4 −2 0 2 4 6 8 10 12 −4700 −4600 −4500 Velocity (cm/s) −4000 −3500 −3000 −2500 −2000 −1500 −1000 Depth (m) Initial geostrophic velocity Corrected velocity WH master bottom−track st. 022 WH master bottom−track st. 023 SADCP velocity − mean st. 022−023 Figura A.8: Comparaci´on entre el perfil geostr´ofico inicial y la velocidad absoluta tras incluir la velocidad de referencia calculada, para el par de estaciones 22-23 (situado en 26.5◦N, a 76.1◦W y 75.9◦W respectivamente). La l´ınea discontinua representa los c´alculos iniciales mientras que la continua ha sido corregida gracias a los datos de ADCP. Los asteriscos y los c´ırculos corresponden a las medidas de “bottom-track” de las estaciones 22 y 23, respectivamente. Los tri´angulos muestran la velocidad del SADCP como la media de las medidas tomadas durante el tiempo que el barco permaneci´o en cada estaci´on. Debe apreciarse que la escala vertical de los ejes es diferente para los distintos rangos de profundidades. El campo de velocidades geostr´ofico inicial fue calculado suponiendo un nivel de no-movimiento a 1000 metros para las estaciones 1-44 (79◦W a 69.5◦W) y a 3000 metros para las estaciones 45-125 (69.1◦W a 13.4◦W) siguiendo el trabajo de Bryden et al. [2005]. Cuando la mayor profundidad del par de estaciones es menor que el nivel de referencia (por ejemplo, cerca de la frontera este de la secci´on), el fondo marino es considerado como la capa de no-movimiento. Como se hab´ıa concluido, el c´alculo de las velocidades de referencia se realiza empleando los datos de “bottom-track” del A.3. RESULTADOS 115 Figura A.12: Diagrama Θ-S promedio. Las estaciones situadas en el Pasaje de Lanzarote (estaciones 1-5) se muestran en gris. (aproximadamente 700 metros), donde encontramos la NACW, claramente identificada por los valores propuestos por Harvey [1982]. Por debajo de las aguas centrales, encontramos las capas intermedias (27.38< γn<27.922 kg m−3, aproximadamente entre los 700-1600 metros), compuestas por AAIW, relativamente menos salina (<35.4) y MW, relativamente m´as c´alida y salina (>35.5). El AAIW se encuentra mayormente en la regi´on del Pasaje de Lanzarote (la rama se˜nalada en la Fig. A.12 se corresponde con la estaci´on 2), mientras que el MW se halla en estaciones en oc´eano abierto. Por ´ultimo, en capas profundas, desde aproximadamente 1600 metros hasta el fondo oce´anico (γn>27.922 kg m−3), encontramos NADW. A partir de las medidas de Θ y S, la densidad neutral (γn) fue calculada seg´un Jackett and McDougall [1997]. De este modo, podemos dividir la columna de agua en diferentes capas. Siguiendo el criterio propuesto por Ganachaud [2003] para el 116 APPENDIX A. RESUMEN EN ESPA ˜ NOL /SPANISH SUMMARY Oc´eano Atl´antico, hemos considerado 14 capas de densidad neutral (Tabla A.2). Las capas 1-4 abarcan aguas superficiales y la termoclina, de la 5 a la 7 representan a las aguas intermedias y el rango 8-14 son aguas profundas. Siguiendo el trabajo de Mach´ın et al. [2006], el nivel de referencia se estableci´o a γn=27.922 kg m−3(aproximadamente 1600 metros) para las estaciones profundas (6-21) y a γn=27.38 kg m−3 (en torno a 700 metros) para las m´as someras, las del Pasaje de Lanzarote. Cuadro A.2: Capas seg´un densidad neutral (kg m−3). Capa L´ımite inferior Masa de agua 1γn= 26.44 2γn= 26.85 NACW 3γn= 27.162 4γn= 27.38 5γn= 27.62 6γn= 27.82 AAIW/MW 7γn= 27.922 8γn= 27.975 9γn= 28.008 10 γn= 28.044 NADW 11 γn= 28.072 12 γn= 28.0986 13 γn= 28.11 14 γn= 28.1295 A.3. RESULTADOS 117 Secci´on promedio De las estimaciones de densidad neutral calculadas para la secci´on promedia, se obtienen velocidades geostr´oficas. Asimismo, se calculan perfiles de LADCP promedios para cada estaci´on teniendo en cuenta los datos de 1997, 2006 y 2009; as´ı como sus medidas de “bottom-track”. A continuaci´on, se aplica el procedimiento propuesto en esta tesis. Siempre que sea posible, se emplean los datos de “bottom-track” para el c´alculo de las velocidades en el nivel de referencia. Cuando se elige utilizar el perfil completo de LADCP, se descartan los 200 primeros y ´ultimos metros por su dependencia con la atm´osfera y los datos de “bottom-track”, respectivamente. La Fig. A.13 muestra el transporte de masa acumulado para la secci´on promedio. Se consideran por separado los tres grupos de capas, atendiendo a la separaci´on por densidades especificada anteriormente. Adem´as se a˜nade el transporte de Ekman, calculado mediante datos de QuickScat, a la primera capa. Se emplea un valor promedio de los transportes de Ekman individuales, obtenidos para los datos de viento del mes en que tuvo lugar cada campa˜na. En la Fig. A.13, las l´ıneas discontinuas muestran el transporte geostr´ofico inicial y las l´ıneas continuas, el transporte absoluto, incluyendo las velocidades de referencia calculadas. Adem´as, en el Cuadro A.3 se facilita un resumen general de los valores de transporte obtenidos para este estudio, teniendo en consideraci´on la extensi´on hasta donde podemos discernir la corriente, descartando las estructuras mesosescalares no resueltas al final del transecto. El transporte de muestra en Sverdrups (Sv, donde 1 Sv ∼109kg s−1). El patr´on general del transporte inicial en la secci´on promedio para capas superficiales (Fig. A.13a) es un flujo hacia el sur hasta la estaci´on 8 (∼14.7◦W), seguido por una estructura pr´acticamente plana hasta el final del transecto. En cambio, las estimaciones de transporte absoluto presentan una mayor pendiente a lo largo de la secci´on, observ´andose el flujo negativo de la NACW desde el oeste de Lanzarote (estaci´on 6) hasta aproximadamente 18◦W (estaci´on 17). Al oeste de esta longitud, se observa una estructura mesoescalar que no contribuye al incremento del transporte neto de la corriente. La forma de dientes de sierra caracter´ıstica de las funciones de 118 APPENDIX A. RESUMEN EN ESPA ˜ NOL /SPANISH SUMMARY −10 −5 0 5 Surface layer (γn<27.38) 1 3 5 7 9 111315171921 −10 −5 0 5 Mean Mass Transport (Sv) Intermediate layer (27.38<γn<27.922) Initial geostrophic ADCP−referenced 20 19 18 17 16 15 14 13 −10 −5 0 5 Deep layer (γn>27.922) Longitude (°W) a b c Figura A.13: Transporte de masa acumulado (Sv) para la secci´ on promedio. A.3. RESULTADOS 119 corriente en esta secci´on se corresponden con la se˜nal de remolinos presentes en el ´area de estudio. Por otro lado, las capas intermedias mostradas en la Fig. A.13b presentan un transporte inicial muy d´ebil a trav´es de la secci´on. Cerca del margen oriental del transecto se aprecia una notable diferencia en las estimaciones una vez incluimos las velocidades en el nivel de referencia, lo que parece deberse a la presencia de un remolino que no ha sido bien resuelto durante el muestreo. El hecho de que en el Pasaje de Lanzarote se obtenga un transporte hacia el norte se explica mediante la presencia de AAIW, que fluye en dicha direcci´on. En lo que respecta a las aguas profundas (Fig. A.13c), el transporte de NADW es relativamente d´ebil hasta la estaci´on 11 (∼15.8◦W). Adem´as, el patr´on de circulaci´on es bastante irregular, desvelando un aumento del flujo hacia el sur al oeste de la estaci´on 7 (∼14.4◦W). Podemos asumir que el transporte al oeste de los 18◦W se debe a la presencia de un remolino, que no fue muestreado en su totalidad al terminar la secci´on a 20◦W. Secciones anuales Los datos hidrogr´aficos de las cuatro campa˜nas (1997, 2006, 2008 y 2009) han sido adem´as estudiados por separado con el fin de discernir una posible variabilidad entre el comportamiento de la corriente en la estaci´on de invierno de estos cuatro a˜nos. De nuevo, la metodolog´ıa sugerida es aplicada a cada conjunto de datos, esta vez de manera individual e independiente. Mientras que en 2006 y 2009 se emplean los registros de “bottom-track” siempre que sea posible, para 1997 s´olo disponemos del perfil de LADCP de la columna de agua para la estimaci´on de velocidades en el nivel de referencia. Adem´as, en 200, las ´unicas medidas de velocidad directa disponibles proceden del SADCP, y ser´an las empleadas en la estimaci´on de velocidades en el nivel de referencia, descartando los 200 metros de datos m´as cercanos a la superficie oce´anica. 120 APPENDIX A. RESUMEN EN ESPA ˜ NOL /SPANISH SUMMARY La Fig. A.14 muestra la componente barotr´opica de la marea, que es calculada mediante el modelo y restada de todas las medidas de velocidad de su correspondiente secci´on anual. Adem´as, en la Fig. A.15 podemos ver la velocidad calculada en el nivel de referencia para cada uno de los casos. Los c´ırculos marcan los pares de estaciones para los que se estim´o dicha velocidad mediante el perfil completo de LADCP (o SADCP para 2008), mientras que los asteriscos marcan aquellas comparaciones realizadas entre los datos de “bottom-track” y el rango de velocidades del perfil geostr´ofico coincidente a esas profundidades. Para los datos de LADCP de 2006 y 2009, disponemos de una estimaci´on del error en el c´omputo de velocidades, el cual empleamos para la elaboraci´on de las mostradas barras de error, como la varianza de las diferencias medias consideradas velocidades de referencia para cada par de estaciones. 1 3 5 7 9 111315171921 20 18 16 14 −10 −5 0 5 10 Tidal barotropic velocity (cm/s) 20 18 16 14 Longitude (°W) −10 −5 0 5 10 1 3 5 7 9 111315171921 1997 2006 2008 2009 Figura A.14: Componente barotr´opica de la marea calculada a partir del modelo de predicci´on OSU TPXO para cada campa˜na. A.3. RESULTADOS 121 2006 1 3 5 7 9 111315171921 20 18 16 14 −10 −5 0 5 10 Longitude (°W) Reference velocity (cm/s) 2008 20 18 16 14 2009 −10 −5 0 5 10 1997 1 3 5 7 9 111315171921 Figura A.15: Velocidad en el nivel de referencia estimado para cada una de las cuatro campa˜nas. Los asteriscos representan las velocidades obtenidas mediante datos de “bottom-track”. Los c´ırculos marcan aquellas velocidades calculadas a partir de perfiles de LADCP o SADCP (para 2008). Se muestran las barras de error calculadas para estas estimaciones. A continuaci´on, se eval´uan los transporte de masa acumulados para cada a˜no y cada rango de profundidades. Para las campa˜nas de 2006 y 2009, puede derivarse una estimaci´on del error en el transporte a partir del error mostrado en la obtenci´on de las velocidades de referencia (Fig. A.15). La Fig. A.16 muestra los resultados para las capas superficiales, observ´andose un patr´on general hasta γn<27.38 kg m−3de un flujo hacia el sur (excepto en 2008), as´ı como la importante actividad mesoescalar presente en la secci´on (por las estructuras de dientes de sierra). En la funci´on de corriente de 1997, el flujo hacia el sur comienza en el Pasaje de 122 APPENDIX A. RESUMEN EN ESPA ˜ NOL /SPANISH SUMMARY −10 −5 0 5 1997 1 3 5 7 9 111315171921 2006 1 3 5 7 9 111315171921 20 18 16 14 −10 −5 0 5 Longitude (°W) Surface layer (γn<27.38) Mass Transport (Sv) 2008 Initial geostrophic ADCP−referenced 20 18 16 14 2009 Figura A.16: Transporte de masa acumulado (Sv) en capas superficiales (γn<27.38 kg m−3) para las cuatro campa˜nas. Lanzarote (estaciones 1-5), con valores m´as altos en las estimas derivadas del LADCP. Sin embargo, al oeste de Lanzarote, los valores iniciales muestran un transporte neto de masa nulo, mientras que las estimaciones corregidas muestran un flujo hacia el sur hasta el final de la secci´on. En 2006, las estimaciones de transporte cerca de la costa africana cambian considerablemente despu´es de la introducci´on de las velocidades en el nivel de referencia. Un flujo geostr´ofico hacia el sur en el Pasaje de Lanzarote se desarrolla, seguido de un transporte de masa neto hacia el sur al oeste de la estaci´on 6 (∼14.1◦W). Las estimaciones corregidas presentan un peque˜no flujo hacia el norte al este de Lanzarote, que se convierte en un flujo hacia el sur hacia el final de la secci´on. Como se muestra en la Fig. A.16, los datos de 2008 registran un patr´on diferente, A.3. RESULTADOS 123 con un flujo hacia el sur en el Pasaje de Lanzarote seguido por transporte de masa cero para los c´alculos geostr´ofico. En cuanto a los c´alculos absolutos, la pendiente aumenta al oeste de la estaci´on 6 (∼14.1◦W), sugiriendo un transporte hacia el norte que contrarresta el flujo en el Pasaje, concluyendo en un transporte acumulado de masa nulo en la estaci´on 16 (∼17.6◦W). Como se observa en la Fig. A.16, la funci´on de corriente inicial en 1997 es muy similar a la de 2008. Sin embargo, los valores de transporte despu´es de la introducci´on de medidas de ADCP son diferentes. Esta diferencia se puede atribuir al uso de los datos de SADCP en la correcci´on de la velocidad geostr´ofica para el a˜no 2008, en lugar de datos de LADCP (no disponibles para esta campa˜na). Por lo tanto, s´olo los primeros 800 metros de la columna de agua contribuyen al c´alculo de las velocidades de nivel de referencia, sin considerar el comportamiento de aguas intermedias y profundas. Por lo tanto, las velocidades de nivel de referencia inferidas a partir de ´unicamente datos de SADCP probablemente est´an siendo subestimadas. Por ´ultimo, los datos de 2009 presentan un fuerte flujo hacia el sur a trav´es de capas superficiales en el Pasaje de Lanzarote en ambos transportes, que adem´as presentan transporte de masa cero al oeste de Lanzarote, hasta la estaci´on 16 (∼17.6◦W). Al oeste, los c´alculos geostr´oficos muestran una estructura plana, mientras que las estimaciones corregidas sugieren un ligero aumento en el flujo hacia el sur. Se aprecia c´omo los valores corregidos siguen los patrones ya registrados mediante los c´alculos geostr´oficos iniciales. En la Fig. A.17 se muestra el transporte de masa en las capas intermedias. El 1997 y 2008 se dispone de datos hasta aproximadamente 17.6◦W, mientras que 2006 y 2009 se aproximan a valores nulos en torno a 20◦W. No obstante, el papel m´as importante en las capas intermedias parece ser jugado por los flujos que tienen lugar en el Pasaje de Lanzarote. En cuanto al comportamiento de la funci´on corriente, se observa que en 1997 y 2009 los datos en capas intermedias siguen un patr´on de flujo hacia el sur despu´es de la estimaci´on de velocidades absolutas mediante LADCP. Por otra parte, el transporte de masa en 2006 y 2008 se comporta de la manera opuesta, dando importancia a un flujo hacia el norte a profundidades intermedias. 124 APPENDIX A. RESUMEN EN ESPA ˜ NOL /SPANISH SUMMARY −10 −5 0 5 1997 1 3 5 7 9 111315171921 2006 1 3 5 7 9 111315171921 20 18 16 14 −10 −5 0 5 Longitude (°W) Intermediate layer (27.38<γn<27.922) Mass Transport (Sv) 2008 Initial geostrophic ADCP−referenced 20 18 16 14 2009 Figura A.17: Transporte de masa acumulado (Sv) en capas intermedias (27.38< γn<27.922 kg m−3) para las cuatro campa˜nas. En el primer caso, un transporte de masa neto casi nulo se pone de manifiesto a lo largo de toda la secci´on para el a˜no 1997. Este patr´on queda claro para el transporte geostr´ofico, mientras que la presencia de estructuras mesoescalares en profundidades intermedias fuerza un flujo hacia el sur al oeste de la estaci´on 8 (∼14.7◦W). A ra´ız de las similitudes mencionadas anteriormente, la figura para 2009 presenta que el flujo hacia el norte es corregido, virando a un leve transporte hacia el sur despu´es de la estaci´on 9 (∼15.1◦W). El transporte al oeste del pasaje est´a marcado por el campo de remolinos y la contribuci´on neta es casi cero. Para el segundo caso (datos de 2006 y 2008), el conjunto de datos de 2006 muestra un transporte de masa inicial pr´acticamente plano. Sin embargo, una vez referenciados, los c´alculos muestran un transporte