EMODnet Chemistry aims to provide access to marine chemistry datasets and derived data products concerning eutrophication, acidity and contaminants. The importance of the selected substances and other parameters relates to the Marine Strategy Framework Directive (MSFD). This aggregated dataset contains all unrestricted EMODnet Chemistry data on eutrophication and acidity, and covers the Northeast Atlantic Ocean (40W). Data were aggregated and quality controlled by 'IFREMER / IDM / SISMER - Scientific Information Systems for the SEA' in France. ITS-90 water temperature and water body salinity variables have also been included ('as are') to complete the eutrophication and acidity data. If you use these variables for calculations, please refer to SeaDataNet for the quality flags: https://www.seadatanet.org/Products/Aggregated-datasets. Regional datasets concerning eutrophication and acidity are automatically harvested, and the resulting collections are aggregated and quality controlled using ODV Software and following a common methodology for all sea regions ( https://doi.org/10.13120/8xm0-5m67). Parameter names are based on P35 vocabulary, which relates to EMODnet Chemistry aggregated parameter names and is available at: https://vocab.nerc.ac.uk/search_nvs/P35/. When not present in original data, water body nitrate plus nitrite was calculated by summing all nitrate and nitrite parameters. The same procedure was applied for water body dissolved inorganic nitrogen (DIN), which was calculated by summing all nitrate, nitrite, and ammonium parameters. Concentrations per unit mass were converted to a unit volume using a constant density of 1.025 kg/L. The aggregated dataset can also be downloaded as an ODV collection and spreadsheet, which is composed of a metadata header followed by tab separated values. This spreadsheet can be imported to ODV Software for visualisation (more information can be found at: https://www.seadatanet.org/Software/ODV).

Seasonal climatology of Water body phosphate for Loire river for the period 1950-2021 and for the following seasons: - winter: January-March, - spring: April-June, - summer: July-September, - autumn: October-December. Observation data span from 1950 to 2021. Depth levels (m): [0.0, 2.0, 4.0, 6.0, 8.0, 10.0, 15.0, 20.0, 25.0, 30.0, 35.0, 40.0, 45.0, 50.0, 60.0, 70.0, 80.0, 90.0, 100.0, 110.0, 120.0, 130.0]. Data sources: observational data from SeaDataNet/EMODNet Chemistry Data Network. Description of DIVAnd analysis: the computation was done with DIVAnd (Data-Interpolating Variational Analysis in n dimensions), version 2.7.4, using GEBCO 15 sec topography for the spatial connectivity of water masses. The horizontal resolution of the produced DIVAnd maps is 0.01 degrees. Horizontal correlation length is defined seasonally (in meters): 230000 (winter), 264000 (spring), 140000 (summer), 135000 (autumn). Vertical correlation length was optimized and vertically filtered and a seasonally-averaged profile was used (DIVAnd.fitvertlen). Signal-to-noise ratio was fixed to 1 for vertical profiles and 0.1 for time series to account for the redundancy in the time series observations. A logarithmic transformation (DIVAnd.Anam.loglin) was applied to the data prior to the analysis to avoid unrealistic negative values. Background field: the vertically-filtered data mean profile is substracted from the data. Detrending of data: no, advection constraint applied: no. Units: umol/l.

This dataset comprises the global frequency, classification and distribution of marine heat waves (MHWs) from 1996-2020, in Chauhan et al. 2023 (https://doi.org/10.3389/fmars.2023.1177571). The classification was done based on their attributes and using different baselines. Daily SST values were extracted from the NOAA-OISST v2 high-resolution (0.25°) dataset from 1982-2020. MHWs were detected using the method presented by Hobday et al. 2016 (https://doi.org/10.1016/j.pocean.2015.12.014), and by using the 95th percentile of the accumulated temperature distribution to flag the extreme events. A shifting baseline of 8-year rolling period was selected between the years 1982-1996, since this period shows relatively stable maximum values of temperature across different ocean regions. The shifting baseline aims to account for the decadal changes of westerly winds, temperatures and ocean gyres circulations. The classification was done using the KMeans clustering algorithm to identify the relevant features of MHWs and classify them into separate groups based on feature similarities. This algorithm takes MHW features, namely duration, maximum intensity, rate onset and rate decline, as input vectors and applies clustering in the 4-dimensional feature space where each data point represents an MHW event. Note that all the MHWs features are standardized because unequal variances can put more weight on variables with smaller variances. This record comprehends the geospatial datasets of: Average number of MHW days per year (i.e., the sum of all MHW days divided by the total number of years, 1996-2020). Average cumulative intensity per year (i.e., the sum of cumulative intensity divided by the total number of years, 1996-2020). Total number of MHW events across the different periods averaged on the total number of years (1989-2020). The period 1982-1988 was only used as an initial baseline without calculating MHWs. Spatial distribution of three MHW categories: moderate MHWs, abrupt and Intense MHWs and extreme MHWs; displaying the total number of MHW days normalized by the number of years considered (i.e., 1989-2020). Distribution of Extreme MHWs across the different periods (A) 1989-1996, (B) 1997-2004, (C) 2005-2012, (D) 2013-2020. The relative frequency (γ) is a ratio of extreme MHWs in a specific period and all extreme MHWs in the same cluster for all periods.

Modelled density of the seapen Kophobelemnon stelliferum in the North East Atlantic. The Random Forest density model trained on data collected by an ROV was constrained by an ensemble of Maxent and Random Forest presence-absence model trained on a larger dataset also collected by an ROV. This species provides structural complexity in an environment where it is lacking and, thus, promotes higher biodiversity where they settle. They are vulnerable to mechanical disturbance of the sediment by fishing gear and a better understanding of their distribution will lead to better management of their population. This work was performed at the University of Plymouth in 2021.

This visualization product displays the density of seafloor litter per trawl. EMODnet Chemistry included the collection of marine litter in its 3rd phase. Since the beginning of 2018, data of seafloor litter collected by international fish-trawl surveys have been gathered and processed in the EMODnet Chemistry Marine Litter Database (MLDB). The harmonization of all the data has been the most challenging task considering the heterogeneity of the data sources, sampling protocols (OSPAR and MEDITS protocols) and reference lists used on a European scale. Moreover, within the same protocol, different gear types are deployed during fishing bottom trawl surveys. In cases where the wingspread and/or number of items were unknown, data could not be used because these fields are needed to calculate the density. Data collected before 2011 are affected by this filter. When the distance reported in the data was null, it was calculated from: - the ground speed and the haul duration using this formula: Distance (km) = Haul duration (h) * Ground speed (km/h); - the trawl coordinates if the ground speed and the haul duration were not filled in. The swept area is calculated from the wingspread (which depends on the fishing gear type) and the distance trawled: Swept area (km²) = Distance (km) * Wingspread (km) Densities have been calculated on each trawl and year using the following computation: Density (number of items per km²) = ∑Number of items / Swept area (km²) Percentiles 50, 75, 95 & 99 have been calculated taking into account data for all years. More information on data processing and calculation are detailed in the document attached. Warning: the absence of data on the map doesn't necessarily mean that they don't exist, but that no information has been entered in the Marine Litter Database for this area.

'''DEFINITION''' The OMI_EXTREME_SL_MEDSEA_slev_mean_and_anomaly_obs indicator is based on the computation of the 99th and the 1st percentiles from in situ data (observations). It is computed for the variable sea level measured by tide gauges along the coast. The use of percentiles instead of annual maximum and minimum values, makes this extremes study less affected by individual data measurement errors. The annual percentiles referred to annual mean sea level are temporally averaged and their spatial evolution is displayed in the dataset omi_extreme_sl_medsea_slev_mean_and_anomaly_obs, jointly with the anomaly in the target year. This study of extreme variability was first applied to sea level variable (Pérez Gómez et al 2016) and then extended to other essential variables, sea surface temperature and significant wave height (Pérez Gómez et al 2018). '''CONTEXT''' Sea level (SLEV) is one of the Essential Ocean Variables most affected by climate change. Global mean sea level rise has accelerated since the 1990’s (Abram et al., 2019, Legeais et al., 2020), due to the increase of ocean temperature and mass volume caused by land ice melting (WCRP, 2018). Basin scale oceanographic and meteorological features lead to regional variations of this trend that combined with changes in the frequency and intensity of storms could also rise extreme sea levels up to one meter by the end of the century (Vousdoukas et al., 2020, Tebaldi et al., 2021). This will significantly increase coastal vulnerability to storms, with important consequences on the extent of flooding events, coastal erosion and damage to infrastructures caused by waves (Boumis et al., 2023). The increase in extreme sea levels over recent decades is, therefore, primarily due to the rise in mean sea level. Note, however, that the methodology used to compute this OMI removes the annual 50th percentile, thereby discarding the mean sea level trend to isolate changes in storminess. The Mediterranean Sea shows statistically significant positive sea level trends over the whole basin. However, at sub-basin scale sea level trends show spatial variability arising from local circulation (Calafat et al., 2022; Meli et al., 2023). '''COPERNICUS MARINE SERVICE KEY FINDINGS''' The completeness index criteria is fulfilled by 41 stations in 2023, 3 more than in 2022, including the first station in the African coast, in the Alboran Sea (Melilla). The mean 99th percentiles reflect the spatial variability of the tide, a microtidal regime, along the Spanish, French and Italian coasts, ranging from around 0.20 m above mean sea level in Sicily and the Balearic Islands (e.g.: 0.22 m in Porto Empedocle; 0.23 m in Ibiza) to around 0.60 m above mean sea level in the Northern Adriatic Sea (e.g.: 0.63 m in Trieste, 0.61 m in Venice). The annual 99th percentiles standard deviation ranges between 2 cm in the Alboran Sea and Sicily to 8 cm in Marseille. The 2023 99th percentile anomalies present positive values in the central and northern part of the Mediterranean Sea, with the exception of Ibiza, in the Balearic Islands, and zero or slightly negative anomalies in the Spanish coast and South of Italy. However, these anomalies are only significant, when compared with the standard deviation of the annual percentiles in the record, at a few stations: Marseille (+12 cm), Ibiza (+8 cm), Trieste (+8 cm) and Venice (+7 cm). '''DOI (product):''' https://doi.org/10.48670/moi-00265

Species distribution models (GAM, Maxent and Random Forest ensemble) predicting the distribution of Solitary Scleractinian fields assemblage in the Celtic Sea. This community is considered ecologically coherent according to the cluster analysis conducted by Parry et al. (2015) on image sample. Modelling its distribution complements existing work on their definition and offers a representation of the extent of the areas of the north-east Atlantic where they can occur based on the best available knowledge. This work was performed at the University of Plymouth in 2021.

This visualization product displays the spatial distribution of seafloor litter density per trawl. EMODnet Chemistry included the collection of marine litter in its 3rd phase. Since the beginning of 2018, data of seafloor litter collected by international fish-trawl surveys have been gathered and processed in the EMODnet Chemistry Marine Litter Database (MLDB). The harmonization of all the data has been the most challenging task considering the heterogeneity of the data sources, sampling protocols (OSPAR and MEDITS protocols) and reference lists used on a European scale. Moreover, within the same protocol, different gear types are deployed during fishing bottom trawl surveys. In cases where the wingspread and/or number of items were unknown, data could not be used because these fields are needed to calculate the density. Data collected before 2011 are affected by this filter. When the distance reported in the data was null, it was calculated from: - the ground speed and the haul duration using this formula: Distance (km) = Haul duration (h) * Ground speed (km/h); - the trawl coordinates if the ground speed and the haul duration were not filled in. The swept area is calculated from the wingspread (which depends on the fishing gear type) and the distance trawled: Swept area (km²) = Distance (km) * Wingspread (km) Densities have been calculated on each trawl and year using the following computation: Density (number of items per km²) = ∑Number of items / Swept area (km²) Then a grid with 30km x 30km cells is used to calculate the weighted mean of densities in each cell from the formula : Weighted mean (number of items per km²) = ∑ (Distance (km) * Density (number of items per km²)) / ∑ Distance (km) Percentiles 50, 75, 95 & 99 have been calculated taking into account data for all years. More information on data processing and calculation are detailed in the document attached. Warning: the absence of data on the map doesn't necessarily mean that they don't exist, but that no information has been entered in the Marine Litter Database for this area. This work is based on the work presented in the following scientific article: O. Gerigny, M. Brun, M.C. Fabri, C. Tomasino, M. Le Moigne, A. Jadaud, F. Galgani, Seafloor litter from the continental shelf and canyons in French Mediterranean Water: Distribution, typologies and trends, Marine Pollution Bulletin, Volume 146, 2019, Pages 653-666, ISSN 0025-326X, https://doi.org/10.1016/j.marpolbul.2019.07.030.

This visualization product displays the total abundance of marine macro-litter (> 2.5cm) per beach per year from Marine Strategy Framework Directive (MSFD) monitoring surveys. EMODnet Chemistry included the collection of marine litter in its 3rd phase. Since the beginning of 2018, data of beach litter have been gathered and processed in the EMODnet Chemistry Marine Litter Database (MLDB). The harmonization of all the data has been the most challenging task considering the heterogeneity of the data sources, sampling protocols and reference lists used on a European scale. Preliminary processing were necessary to harmonize all the data: - Exclusion of OSPAR 1000 protocol: in order to follow the approach of OSPAR that it is not including these data anymore in the monitoring; - Selection of MSFD surveys only (exclusion of other monitoring, cleaning and research operations); - Exclusion of beaches without coordinates; - Some categories & some litter types like organic litter, small fragments (paraffin and wax; items > 2.5cm) and pollutants have been removed. The list of selected items is attached to this metadata. This list was created using EU Marine Beach Litter Baselines and EU Threshold Value for Macro Litter on Coastlines from JRC (these two documents are attached to this metadata); - Normalization of survey lengths to 100m & 1 survey / year: in some cases, the survey length was not exactly 100m, so in order to be able to compare the abundance of litter from different beaches a normalization is applied using this formula: Number of items (normalized by 100 m) = Number of litter per items x (100 / survey length) Then, this normalized number of items is summed to obtain the total normalized number of litter for each survey. Finally, the median abundance for each beach and year is calculated from these normalized abundances per survey. Sometimes the survey length was null or equal to 0. Assuming that the MSFD protocol has been applied, the length has been set at 100m in these cases. Percentiles 50, 75, 95 & 99 have been calculated taking into account MSFD data for all years. More information is available in the attached documents. Warning: the absence of data on the map doesn't necessarily mean that it doesn't exist, but that no information has been entered in the Marine Litter Database for this area.

Seasonal climatology of Water body dissolved oxygen concentration for Loire river for the period 1950-2021 and for the following seasons: - winter: January-March, - spring: April-June, - summer: July-September, - autumn: October-December. Observation data span from 1950 to 2021. Depth levels (m): [0.0, 2.0, 4.0, 6.0, 8.0, 10.0, 15.0, 20.0, 25.0, 30.0, 35.0, 40.0, 45.0, 50.0, 60.0, 70.0, 80.0, 90.0, 100.0, 110.0, 120.0, 130.0]. Data sources: observational data from SeaDataNet/EMODNet Chemistry Data Network. Description of DIVAnd analysis: the computation was done with DIVAnd (Data-Interpolating Variational Analysis in n dimensions), version 2.7.4, using GEBCO 15 sec topography for the spatial connectivity of water masses. The horizontal resolution of the produced DIVAnd maps is 0.01 degrees. Horizontal correlation length is defined seasonally (in meters): 150000 (winter), 100000 (spring), 70000 (summer), 115000 (autumn). Vertical correlation length was optimized and vertically filtered and a seasonally-averaged profile was used (DIVAnd.fitvertlen). Signal-to-noise ratio was fixed to 1 for vertical profiles and 0.1 for time series to account for the redundancy in the time series observations. A logarithmic transformation (DIVAnd.Anam.loglin) was applied to the data prior to the analysis to avoid unrealistic negative values. Background field: the vertically-filtered data mean profile is substracted from the data. Detrending of data: no, advection constraint applied: no. Units: umol/l.