Zooplankton, phytoplankton and pervasive anthropogenic particles along a zonal North Atlantic transect: a multi-vessel citizen-science image survey

This dataset contains the observational data, analysis code and image archive underlying the study "Zooplankton, phytoplankton and pervasive anthropogenic particles along a zonal North Atlantic transect: a multi-vessel citizen-science image survey" (Dryden, Duncan and the GOES Foundation citizen-science consortium; submitted to Ocean Science, Copernicus; manuscript egusphere-2026-3979). Between 8 January and 6 February 2022, thirteen sailing vessels crossing the trade-wind North Atlantic, from the Canary Islands and Cape Verde to the Caribbean and Suriname, collected surface water samples on a fixed twice-daily schedule using a standardised 0.5-litre cartridge filter (20 micrometre mesh). One vessel, Emma, additionally sailed a higher-latitude return leg from 19 May to 31 July 2022, contributing 271 further photographs. Each filter was photographed under a common microscope design (at least 5 megapixels, with a polarising filter) at a 5 mm field of view, and an automated image pipeline located the filter apertures, sized every particle larger than 20 micrometres, and classified it as zooplankton, phytoplankton, plastic particle, plastic fibre, black (combustion) carbon, mineral, plant debris or unidentified. In total the crossing yielded 1,466 accepted photographs across 282 stations and 18,142 classified objects; a two-scorer eye check against the machine classification is included for transparency. Two files make up this record. zenodo_upload_v146_ALL.zip holds the material added for the current manuscript revision: the station-level position table for all 1,466 crossing photographs and all 282 stations (built from the participating crews' field logs), the seven figures reported in the manuscript, a PDF snapshot of the volunteer field guide describing the sampler build and protocol, the crews' field logs themselves, and the original data workbook GOES_ALL_DATA.xlsx. That workbook is retained for provenance rather than as a live source: it reflects an earlier stage of the project (eight vessels, 352 stations, 51,547 particles before later curation and the addition of five further vessels) and its particle counts do not match the totals reported in the current manuscript. 22167365.zip holds everything from the record's earlier versions: the automated image-classification pipeline and its complete per-image output, the frame-curation list with a stated reason for every excluded file, the two-scorer eye-check data, and specimen images of the animals reported in the manuscript (two copepods and three shrimps). The per-image output in this archive is the current pipeline result and reproduces the 18,142-object total reported in the manuscript. Notes on interpretation: identifications derive from citizen-science microscopy. The plastic-versus-natural-particle partition is provisional pending FTIR/Raman confirmation, and "black carbon" is an optical description (dark, colourless particles) that may also include dark plastic or mineral grains. The crews' hand-logged counts are qualitative context and are not used in the quantitative analysis, which rests on the automated image pipeline and an independent surface net tow. Positions are from image GPS where available and from the field logs otherwise.