BUSCADOR RECOLECTA

This dataset reports for the classical rosette bottle data for the following Essential Ocean Variables: salinity, dissolved oxygen, total alkalinity, pH, nitrates and silicic acid. It also reports on the measurements of the CTD-O2 probe at the bottle levels: pressure, temperature, salinity and dissolved oxygen, The GEOVIDE cruise was carried out coast to coast between Portugal and Newfoundland via the south tip of Greenland, following the OVIDE line in the eastern part and crossing the Labrador Sea in the western part. The classical hydrographic rosette was cast 163 times at 78 different geographical positions called stations. While the CTD-O2 probe acquired continuous profiles of the “physical” variables (pressure, temperature, salinity and dissolved oxygen), 22 Niskin bottles were closed at different levels during the upcast to provide samples for biogeochemical analysis. After calibration, we find precisions for pressure, temperature, salinity and dissolved oxygen that fit the GO-SHIP international quality requirements. In parallel, but not simultaneously, a trace-metal rosette (TMR) was cast 53 times, also acquiring profiles of physical variables, and equipped with 24 Go-Flo bottles adapted for the sampling of trace metals. Depending on the number of operations, stations were identified as “Short” (one single CTD cast), “Large” (3 CTD casts), “XLarge” (up to 6) and “Super” (up to 11). All along the track of the ship, current magnitude and direction was measured by Ship Acoustic Doppler Current Profilers, down to 1000m depth, No

This dataset is composed of 2 files: a database (in csv format) with 695 records of biogeochemical variables (temperature, salinity, oxygen, nutrients, pH and total alkalinity) analyzed in water samples collected at the GIFT time series and a Readme (txt) file that includes a short description of the variables provided., If the dataset is used, please consider citing Flecha et al., (2019) (doi: 10.1038/s41598-019-52084-x)., This data set includes recently published data used to assess the temporal evolution of pH in Atlantic and Mediterranean water masses exchanging at the Strait of Gibraltar (Long:-5.345, Lat: 36.137, Datum:WSG84) during the decade 2005-2015 and to calculate the magnitude of natural and anthropogenic components on total pH changes (Flecha et al., 2019). The database provides measurements of carbon system parameters in water samples collected at 3 stations that form the marine time series GIFT during 26 oceanographic campaigns conducted over the decade 2005–2015. Geographic coordinates of sampling stations are provided. Some physical data (i.e. pressure, temperature and salinity) are also included. During the cruises, a temperature and salinity profile in each station was obtained with a Seabird 911 Plus CTD probe connected to a rosette sampler. Conductivity measurements were converted into practical values of the salinity scale with the UNESCO equation (1986). Seawater was subsequently collected for biogeochemical analysis using Niskin bottles immersed in the oceanographic rosette at variable depths (from 5 to 8 levels) depending on the instant position of the interface between the Atlantic and Mediterranean flows that was identified by CTD profiles. The biogeochemical variables shown in the database are pH in total scale at 25 °C (pHT25), total alkalinity (AT), Dissolved Oxygen (DO) and inorganic nutrients (nitrate, NO3− and Silicate, SiO44−). pHT25 data were obtained by the spectrophotometric method with m-cresol purple as indicator (Clayton & Byrne 1993) with an addition of 0.0047 (DelValls & Dickson, 1998). Samples were taken directly from the oceanographic bottles in 10 cm path-length optical glass cells and measurements were carried out with a Shimadzu UV-2401PC spectrophotometer containing a 25 °C-thermostated cells holder. Samples for AT analysis were collected in 500-ml borosilicate bottles, and poisoned with 100 μl of HgCl2-saturated aqueous solution and stored until measurement in the laboratory. AT was measured by potential titration according to Mintrop et al. (2000) with a Titroprocessor (model Metrohm 794). DO concentration was obtained through automated potentiometric modification of the original Winkler method using the Titroprocessor. Upon collection, flasks were sealed, stored in darkness and measured within 24 h. Water samples (5 mL, two replicates) for inorganic nutrients determination were taken, filtered immediately (Whatman GF/F, 0.7 μm) and stored frozen for later analyses in the shore-based laboratory. Nutrients concentrations were measured with a continuous flow auto-analyzer using standard colorimetric techniques (Hansen & Koroleff 1999). More details on procedures and data structure are given in a single README file (txt). The data are provided as [space] delimitated plain text files., Plan Estatal de I+D+i, European Commission, CSIC. CARBOOCEAN (FP6-511176), SESAME (FP6-036949), CARBOCHANGE (FP7-264879), PERSEUS (FP7-287600), COMFORT (H2020-820989), CTM2006-28141-E/MAR, CTM2016-75487-R., 1 data csv‘GIFT_carbonparameteres_2005_2015.csv’ file and 1 readme.txt file., Peer reviewed

The item is made of 6 files: 1) README.txt; 2) TCO2_NNGv2LDEO_climatology.nc contains the climatology of TCO2 centered in 1995 and computed with NNGv2LDEO in netcdf4 format; 3) pCO2_NNGv2LDEO_climatology.nc contains the climatology of pCO2 centered in 1995 and computed with NNGv2LDEO and NNGv2 of Broullón et al. (2019) in netcdf4 format ; 4) NNGv2LDEO.mat is the neural network object used to create the climatology of TCO2; 5) TCO2NNWOA13.mp4 is a video of the surface climatology, 3 vertical sections in the Pacific Ocean, Atlantic Ocean and Indian Ocean and, the variation in depth of one month (April); 6) Example.rar contains an example matrix of inputs and targets to the neural network, the NNGv2LDEO.mat and a MATLAB script to compute TCO2 with NNGv2LDEO, This research was supported by Ministerio de Educación, Cultura y Deporte (FPU grant FPU15/06026), Ministerio de Economía y Competitividad through the ARIOS (CTM2016-76146-C3-1-R) project co-funded by the Fondo Europeo de Desarrollo Regional 2014-2020 (FEDER) and EU Horizon 2020 through the AtlantOS project (grant agreement 633211), Peer reviewed

This dataset is composed of 2 files. Main one is the data set itself (arios_database_hy1.csv) in WHP-Exchange bottle format, with 1421 samples of temperature, salinity, oxygen, nutrients, pH, alkalinity, and chlorophyll. The other file (readme_ARIOSDATABASE.txt) includes a short description of the the database and calculated variables, The data are provided under an Attribution-ShareAlike 4.0 International license. However, if you use the data, so as to support the authors, please consider citing the above mentioned article where data collection and analytical techniques are given in detail. Here we only give a brief details and a guide to the contents of the data files, The ARIOS project (Acidification in the Rias and the Iberian Continental Shelf) included among its tasks the compilation and analysis of the historical record of the measurements of the carbon system and associated parameters carried out by the Instituto de Investigaciones Mariñas (IIM-CSIC) in Vigo (Spain) that gathers 3,343 oceanographic stations and 17,653 discrete samples between 1976 and 2018, This unique collection that gathers biogeochemical data from 24 oceanographic projects and cruises between 40º N and 45º N, 11ºW and the Galician coast in the last 40 years is a departure point to evaluate the acidification in the Galician coastal upwelling system. All ARIOS measurements passed a first quality control of hydrographic data through the specific software of the GO-SHIP program (Global Program of Hydrographic Investigations in Oceanographic Ships) that compiles various quality control procedures (Velo et al., 2019). The database also displays a quality indicator for each measurement. The different methodologies used in the ARIOS database changed over time following the scientific and technical advances in chemical oceanography from 1976 to 2018. The most important changes are as follows: Temperature: - From 1976 to 1984, the temperature measured with a Wallace and Tiernan bathylograph - From 1984 to 1990, inverted thermometers were used, correcting the temperature with protected and unprotected thermometers according to Anderson (1974). - After 1990, different CTD models were used to obtain the thermohaline profile. Salinity: - Salinity measurements were made with an AUTOSAL 8400A inductive salinometer calibrated with normal IAPSO water and using the equations provided by UNESCO (1981). The estimated analytical error was 0.003. Oxygen: - In 1981, the oxygen concentration was measured by first time using the Winkler's titration method following Culberson et al. (1991) - During the 80s and early 90s, the titration was performed with Metrohm instruments (E-425 or E-473) with an analytical error of 1 μmol kg-1. - After 1997, the oxygen concentration was estimated using a Titrino 720 analyzer (Metrohm) with an accuracy of 0.5 μmol kg-1. Nutrients: - Nitrate: Measured with an ALPKEM autoanalyzer as described in Strickland and Parsons (1968) using Cd-Cu reduction (Mouriño and Fraga 1985). Accuracy of 0.1 μmol / kg. - Phosphate: Measured with an ALPKEM autoanalyzer following Hansen and Grasshof (1983) and achieving an accuracy of 0.01 μmol / kg, - Silicate: Measured with an ALPKEM autoanalyzer following Hansen and Grasshof (1983) and achieving an accuracy of 0.05 μmol / kg. Total alkalinity: - In 1981, alkalinity was measured for the first time by potentiometric titration with 0.1N HCl at final pH 4.44, following Pérez and Fraga (1987a) with an analytical error of 2 μmol / kg and an accuracy of 0.1%. - Beginning in 2001, certified reference material (CRM) samples provided by Dr. A. Dickson from the University of California were used in the alkalinity analysis that improved the accuracy to ± 1.4 μmol/kg and accuracy to <0.1%. pH: - From 1976 to 1986, the pH was measured with a Metrohm E-510 pH instrument with a glass and reference reference Ag/ClAg calibrated with 7.413 NBS buffer. All pH values ​​were converted to values ​​at 15 ºC using two different temperature corrections: 1) between 1976 and 1984, the Buch and Nynas tables published by Barnes (1959) and 2) between 1984 and 1986 following Pérez and Fraga (1987 ). - From 1976 to 2001, pH was measured with a Metrohm E-654 pH meter with Orion 81-04 Ross combined glass electrode. The pH was converted to SWS scale using the hydrogen activity coefficient provided by Mehrbach et al. (1973) at 25ºC with the parameterization given by Pérez and Fraga (1987b). The error of the potentiometric method was 0.010. - After 2001, the pH measurements were spectrophotometric following Clayton and Byrne (1993) and subsequently adding 0.0047 to the pH value to do so (DelValls and Dickson, 1998). The precision of the spectrophotometric measurements was 0.003 pH units. Chla: - Chla was measured using a Turner 10000R fluorometer. Between 1976 and 1984 on Schleicher and Scholl 602eh filters of 6 cm in diameter and later with Whatman GF / C filters of 2.5 cm. The precision was 0.05 g / L, CSIC y Plan Nacional de I+D del Gobierno de España, 1 data csv ‘ARIOSDATABASE.csv’ file and 1 readme.txt file, Peer reviewed

The item is made of 6 files: 1) README.txt; 2) INTECMAR_NN-database.csv: Dataset containing all the input variables used compute the time series of AT and pH as well as these two computed variables; 3) Training_database.xlsx: Dataset containing the data to train and test the neural networks; 4) pH_NN.mat is the neural network object used to compute the pH time series; 5) AT_NN.mat is the neural network object used to compute the total alkalinity time series; 6) Source_code.rar contains the MATLAB files to configure, train and validate the neural networks created in this study, This research was supported by Ministerio de Educación, Cultura y Deporte (FPU grant FPU15/06026) and Ministerio de Economía y Competitividad through the ARIOS (CTM2016-76146-C3-1-R) project co-funded by the Fondo Europeo de Desarrollo Regional 2014-2020 (FEDER), No