The LiDAR used is the Titan DW 600 from Teledyne Optech of the “Platform topo-bathymétrique aéroportée Nantes-Rennes”.  Both 532-nm and 1064-nm LiDAR swath were acquired 6th and 7th of October 2017, at 400 m above ground with a FOV of 28° giving strip image width of 199.5 m with a nominal footprint of 0.49 m along-track and 0.29 m cross-track. A mirror compensation and a sidelap of 30% was used to prevent empty space between strips. The effective incident angle range was from 4 to 20°. The aircraft speed being 201.7 km/h the down track footprint rise up to 0.98 m on sideway what justify a final spatial resolution of 1 m. The laser pulse frequency was set to 50 kHz (100 kHz locally) giving an overall mean point density of 3.52 points/m². The full-waveform (FWF) was recorded only in 532-nm with a maximum length of 60 m, at 1 GHz frequency, resulting in a vertical resolution of 0.15 m. The Optech LiDAR Mapping Suite (LMS) combines a Global Positioning System (GPS) and Inertial Measurement Unit (IMU) to provide a georeferenced point cloud and associated FWF with strip optimization. To compute the complete point cloud, LMS uses an analogous approach to photogrammetric block adjustment. On overlap areas between each flight line, planes (typically roofs in land) are extracted and used in a least squares adjustment model provided by LMS. The trajectory accuracy provided by GEOFIT Company was better than 0.15 m in planimetry and 0.08 m on elevation. The set of data is stored in a zip file composed of the following image files associated with ENVI headers which are text files describing the image format. - LaBaule_2017106-7-orthophoto: contain the mosaic of RGB image acquired with the camera associated to the LiDAR. - LaBaule_2017106-7_536nm_raytracing-incident_angles_2m: contains the effective incident angle of each pixel retained in the mosaic with a pixel FOV of 0.3° and with a tilling of image strips in the top of each other from the North to the South. - LaBaule_2017106-7_1064nm_raytracing_discrete_echo_ranges_2m: contains the elevation of the water surface of each pixel. - LaBaule_2017106-7_536nm_raytracing_FWF_2m: contains the row FWF projected on the image plane at 0 m NGF with an offset of 30 m for easy spectral processing of the FWF (true range are stored in band names). - LaBaule_2017106-7_536nm_raytracing_FWF_2m_intensity_stats: contains the basic statistics of the FWF intensities. - LaBaule_2017106-7_536nm_raytracing_FWF_2m_range_stats: contains the min, max and thickness of the FWF ranges. - LaBaule_2017106-7_536nm_raytracing_FWF_2m_row_results: contains the row results of water surface and method 1 (dddNCFWF echoes) and method 2 (NCFWFT-1 bottoms) bathymetries as defined in https://doi.org/10.3390/rs11020117. - LaBaule_2017106-7_536nm_raytracing_FWF_4m_final_bathymetry: contains the final (cleaned NCFWFT-1 bottoms) bathymetry. The same data were acquired the 11th of August 2018 during a low tide with the same settings without RGB camera but with a new FWF recorder giving a lot less noisy signal allowing a constant spatial resolution of 1 m. The set of data is stored in a zip file composed of the following image files associated with ENVI headers which are text files describing the image format. - LaBaule_20180811_536nm_raytracing-incident_angles_1m: contains now 3 channels for the effective incident angle, the strip number and the GPS time. - LaBaule_20180811_1064nm_raytracing_discrete_echo_ranges_1m: contains the elevation of the water surface and all echoes above it in each pixel. - LaBaule_20180811_536nm_raytracing_FWF_1m: contains the row FWF projected on the image plane at 0 m NGF with an offset of 30 m for easy spectral processing of the FWF (true range are stored in band names). - LaBaule_20180811_536nm_raytracing_FWF_1m_intensity_stats: contains the basic statistics of the FWF intensities. - LaBaule_20180811_536nm_raytracing_FWF_1m_range_stats: contains the min, max and thickness of the FWF ranges. - LaBaule_20180811_536nm_raytracing_FWF_1m_row_results: contains the row results of water surface and method 1 (dddNCFWF echoes) and method 2 (NCFWFT-1 bottoms) bathymetries as defined in https://doi.org/10.3390/rs11020117. - LaBaule_20180811_536nm_raytracing_FWF_1m_final_bathymetry: contains the best of dddNCFWF echoes and NCFWFT-1 bottoms simply presented with the application of a 5x5 median filter. -  

The present dataset is based on a nine site study of fine seabed topography in intertidal zones. Four coral sites (Maupiti A, B and C and Niau islands) and five rocky sites (Ars en Ré, Socoa, Parlementia A and B and Banneg island) have been explored. The data has been gathered using on-foot GNSS RTK for all sites (Trimble R8/R8S and Leica sytems) except Banneg island, where aerial Lidar data from Litto3D program has been used. The horizontal resolution varies between 3.8 and 12cm allowing to describe a wide range of spatial scales (generally over 3 spectral decades). The data has been processed to explore the statistical and spectral metrics which can be used to characterize the architectural complexity of seabeds.

These data were collected as part of the Boundary Layer Turbulence and Abyssal Recipes (BLT Recipes) Experiment which focused on near-bottom processes within a continental slope canyon on the eastern side of the Rockall Trough, off Ireland. This specific dataset contains a bathymetry product based on data collected with the shipboard multibeam on RRS Discovery during cruises in 2021 and 2022.

In the framework of the ERC FOCUS project, the Geosciences Ocean Laboratory (LGO) organized, from July 22 to 26, 2019, an experiment of seafloor geodesy in the Bay of Brest, entitled Geodesea-2019.  The aim was to test acoustic beacons acquired for the FOCUS project and to test experimental protocols for seafloor positioning. The experiment was carried out from the Albert Lucas research station-vessel. The experiment was conducted in collaboration with the Laboratoire Environnement et Sociétés (LIENSs) from la Rochelle and the iXBlue company, in Brest. Relative acoustic ranging was carried out during four days, each beacon ranging the other ones at a regular time interval, while acquiring auxiliary data (temperature, pressure, sound-speed) to be able to convert travel-times into distances. Absolute positioning of the beacons on the seafloor was also carried out using a small Unmanned Surface Vehicle (USV) on which a compact GNSS/Acoustic system was mounted, combining an Ultra Short Baseline (USBL), an inertial system (INS) and a GNSS receiver.

A 30-year (1971-2000) temperature and salinity climatology is presented for surface and near-bed regions of the NW European shelf seas, with a resolution of 1/6 longitude by 1/10 latitude. The data have been extracted from the International Council for the Exploration of the Sea (ICES) data centre and supplemented by additional records from the World Ocean Data Centre (WODC). From the original data, which are irregularly distributed in space and time, the mean monthly temperature and salinity are calculated, as well as the climatic mean annual cycle. The climatology presented here is an improvement upon all existing climatologies presented in the literature for the NW European shelf; covering a wider area on a finer scale and including the surface and near-bed distribution of both temperature and salinity. Comparison of our data with existing climatologies shows good agreement, with differences occurring where our climatology is an improvement. This climatology, which will prove to be valuable to many users in the marine community will be regularly updated and made available to all users via the ICES data centre.

As part of the EPIGRAM project, two moorings, Aspex 9 and 10, equipped with temperature, conductivity and pressure sensors (Seabird microcat 37SM), were deployed offshore Aquitaine for a period of 2 years (2009-2011). The probes were calibrated in metrology and provide salinity values derived from calculations based on Fofonoff and Millard (1983) [Fofonoff, P. and Millard, R.C. Jr UNESCO 1983. Algorithms for computation of fundamental properties of seawater. UNESCO Tech. Pap. in Mar. Sci., No. 44, 53 pp. Eqn.(31) p.39. http://unesdoc.UNESCO.org/images/0005/000598/059832eb.pdf]. Several oceanographic missions ASPEX1, 2, 3 and 4, were dedicated to deploying and recovering these instruments. The frame of the mooring (Figure B) is laid on the seafloor with sensors at 0.5 m above it. Aspex 9 and 10 moorings are located at the continental shelf (at 148 m water depth) and continental slope (at 472 m water depth), respectively (Fig. A). Aspex 9 mooring recorded data from 18.07.2009 at 23:40:00 to 01.07.2010 at 09:18:00 (with a measurement every 120 seconds) and from 04.09.2010 at 23:30:00 to 10.08.2011 at 06:57:00 (with a measurement every 180 seconds). Aspex 10 mooring recorded data from the 18.07.2009 at 21:40:00 to 22.07.2010 at 23:18:01 and from 05.09.2010 at 01:20:00 to 10.08.2011 at 04:58:00 (with a measurement every 120 seconds). This ASPEX dataset can be downloaded from here: a zip file containing one .txt file with indications of recording time, location and water depth of the two moorings and four data files (one .txt file per mooring and per recorded period) with for each measurement the year (YYYY), month (MM), day (DD), hour (hh), minute (mm) second (ss) and pressure (P), temperature (T), conductivity (C) and salinity (S) values. As part of the PAMELA project, 18 XBT (eXpendable BathyThermograph) probes (Sippican) were launched in 2013 (July to September) offshore Aquitaine Basin. Seawater temperature profiles were thus acquired through the water column from different locations at the shelf area (at 140 m water depth), the continental slope and the basin (at 1213 m water depth) (Fig. A). Some of these XBT profiles are accessible via the Coriolis platform (http://www.coriolis.eu.org/). The complete GAZCOGNE1-2 dataset can be downloaded from here: a zip file containing one .txt file with indications of time, location and water depth for the XBTs and 18 data files (one .txt file per XBT profile). All times are in UTC.

Maupiti ("the Stuck Twins'') is a diamond-shaped island located in the western part of the Society archipelago in French Polynesia. The present study focuses on the data recovered over a single cross-barrier transect located in the south-west barrier during the MAUPITI HOE field campaign, from 5 to 18 July 2018. The studied area is representative of the reef structure observed along the 4km-long southwestern barrier reef, showing an alongshore-uniform structure exposed to swell approaching with weak incident angles, a healthy reef colony. In the cross-barrier direction, the reef displays a clear partitioning of bottom roughness that ranges from low-crested compact structures at the reef crest to higher and sparser coral bommies on the backreef. The experimental setup was specifically designed to analyse and differentiate the dynamics over three roughness-contrasting sections found over the barrier reef.

Pockmarks are defined as depressions on the seabed and are usually formed by fluid expulsions. Recently discovered, pockmarks along the Aquitaine slope within the French EEZ, were manually mapped although two semi-automated methods were tested without convincing results. In order to potentially highlight different groups and possibly discriminate the nature of the fluids involved in their formation and evolution, a morphological study was conducted, mainly based on multibeam data and in particular bathymetry from the marine expedition GAZCOGNE1, 2013. Bathymetry and seafloor backscatter data, covering more than 3200 km², were acquired with the Kongsberg EM302 ship-borne multibeam echosounder of the R/V Le Suroît at a speed of ~8 knots, operated at a frequency of 30 kHz and calibrated with ©Sippican shots. Precision of seafloor backscatter amplitude is +/- 1 dB. Multibeam data, processed using Caraibes (©IFREMER), were gridded at 15x15 m and down to 10x10 m cells, for bathymetry and seafloor backscatter, respectively. The present table includes 11 morphological attributes extracted from a Geographical Information System project (Mercator 44°N conserved latitude in WGS84 Datum) and additional parameters related to seafloor backscatter amplitudes. Pockmark occurrence with regards to the different morphological domains is derived from a morphological analysis manually performed and based on GAZCOGNE1 and BOBGEO2 bathymetric datasets. The pockmark area and its perimeter were calculated with the “Calculate Geometry” tool of Arcmap 10.2 (©ESRI) (https://desktop.arcgis.com/en/arcmap/10.3/manage-data/tables/calculating-area-length-and-other-geometric-properties.htm). A first method to calculate pockmark internal depth developed by Gafeira et al. was tested (Gafeira J, Long D, Diaz-Doce D (2012) Semi-automated characterisation of seabed pockmarks in the central North Sea. Near Surface Geophysics 10 (4):303-315, doi:10.3997/1873-0604.2012018). This method is based on the “Fill” function from the Hydrology toolset in Spatial Analyst Toolbox Arcmap 10.2 (©ESRI), (https://pro.arcgis.com/en/pro-app/tool-reference/spatial-analyst/fill.htm) which fills the closed depressions. The difference between filled bathymetry and initial bathymetry produces a raster grid only highlighting filled depressions. Thus, only the maximum filling values which correspond to the internal depths at the apex of the pockmark were extracted. For the second method, the internal pockmark depth was calculated with the difference between minimum and maximum bathymetry within the pockmark. Latitude and longitude of the pockmark centroid, minor and major axis lengths and major axis direction of the pockmarks were calculated inside each depression with the “Zonal Geometry as Table” tool from Spatial Analyst Toolbox in ArcGIS 10.2 (©ESRI) (https://pro.arcgis.com/en/pro-app/tool-reference/spatial-analyst/zonal-statistics.htm). Pockmark elongation was calculated as the ratio between the major and minor axis length. Cell count is the number of cells used inside each pockmark to calculate statistics (https://pro.arcgis.com/en/pro-app/tool-reference/spatial-analyst/zonal-geometry.htm). Cell count and minimum, maximum and mean bathymetry, slope and seafloor backscatter values were calculated within each pockmark with “Zonal Statistics as Table” tool from Spatial Analyst Toolbox in ArcGIS 10.2 (©ESRI). Slope was calculated from bathymetry with “Slope” function from Spatial Analyst Toolbox in ArcGIS 10.2 (©ESRI) and preserves its 15 m grid size (https://pro.arcgis.com/en/pro-app/tool-reference/spatial-analyst/slope.htm). Seafloor backscatter amplitudes (minimum, maximum and mean values) of the surrounding sediments were calculated within a 100 m buffer around the pockmark rim.

This benchmark dataset contains the physical data used as predictors to reconstruct global chlorophyll-a concentrations (Chl, a proxy of phytoplankton biomass) in Roussillon et al., as well as the reference satellite Chl target fields. The nine physical predictors' data (Short-Wave radiations, Sea Surface Temperature, Sea Level Anomaly, Zonal and meridional surface currents, Zonal and meridional surface wind stress, Bathymetry, Binary continental mask) were extracted from publicly available datasets over [1998-2015] and resampled to the same spatio-temporel resolution as Chl, i.e. monthly on a 1°x1° grid between 50°N and 50°S. Missing values were gap-filled using the heat diffusion equation. Each variable was normalized by substracting its mean from the original values and dividing by its standard deviation over [1998-2015]. This dataset was used to train and validate the Multi-Mode Convolutional Neural network (CNNMM8) introduced in Roussillon et al. ; reconstructed monthly Chl fields over the [2012-2015] test period are also provided here. We hope this benchmark dataset can help to promote the improvements of methods as well as the emergence of new ideas, as building datasets is sometimes more time-consuming than the implementation of machine learning tools themselves. This would also facilitate the quantitative comparison of models performances' on the exact same datasets.

The mooring was deployed on 15 September 2017 from Norwegian Research Vessel Lance at 80.6°N and 7.26°E (depth of 730 m) in the Yermak Pass over the Yermak Plateau north of Svalbard. It comprised 3 instruments: an upward-looking RDI 75kHz, a Long Ranger Acoustic Doppler Current Profiler (ADCP) at 340 m with 16 m vertical resolution (25 bins of 16 m each) and a 2-hour sampling time; a Seabird SBE37 measuring temperature, salinity and pressure at 348 m with 10-minute sampling time; and an Aquadopp current meter at 645 m with a 2-hour sampling time. The mooring was retrieved on the 19 July 2020 by Norwegian Icebreaker K.V. Svalbard. The present dataset features: The ADCP 50-hour high pass filtered velocities and  the Aquadopp 50-hour high pass filtered velocities.