Introduction

Microplastics are defined as plastics that are smaller than 5 mm (0.20 in) and are a growing problem affecting coastal communities, marine ecosystems, marine life, and human health1,2,3,4,5. Microplastics have been found in multiple media such as in oceans, rivers, estuaries, lakes, the atmosphere, beaches, sea ice, and sediments6,7,8,9,10. These small plastics originate either as primary sources from terrestrial runoffs, littering, and industrial discharge of particulates in commercial products in which they occur or as secondary sources from the degradation of large plastics11,12,13,14 (macroplastics, i.e., >5 mm).

Microplastics affect both the environment and the organisms therein. Microplastics act as vectors for heavy metal contamination, and diseases, thus aggregating and increasing toxicity in the environment15,16,17. Aquatic biota such as plankton, fishes, crabs, clams, shrimps, and mussels ingest microplastics which clog their tissues and organs, thereby affecting their energy reserves, causing neurotoxicity, behavioral abnormalities, stunted growth, decrease reproductivity, and eventual death18,19. These ingested microplastics can also bioaccumulate in humans through the consumption of seafood, eventually leading to inflammation, cell damage, and oxidative stress in humans20,21. Recently, there have been reported findings of microplastics in human placenta with dire effects on fetal development22. The breakdown of microplastics can result in the leaching of toxins which seeps into sediments or kill organisms23,24.

In addition to the harm to aquatic organisms and the environment, microplastics pollution affects economies in many ways, including clean-up costs, decline in fisheries and coastal tourism25,26,27. Over time, lost fishing gear breaks down through abrasion and biofouling resulting in the release of microplastic fragments and fibers24. Fishes consuming these pieces of microplastics can expose themselves to toxic chemicals28,29. Seafood is the main source of animal protein for approximately 20% of the global population30 (1.4 billion people). Marine microplastics therefore endanger this source of protein by reducing the efficiency and productivity of aquaculture and commercial fisheries through fish mortality.

Borrelle et al.31 estimates that about 19 to 23 million metric tons, or 11%, of plastic waste (i.e., the main source of microplastics) generated globally in 2016, entered aquatic ecosystems, with this estimate expected to increase to 53 million metric tons per year by 2030. Beaumont et al.30 estimates a loss in marine ecosystem services between $3,300-$33,000 for each metric ton of plastic entering the ocean per year. At these rates, the economic cost of marine plastic pollution runs into several billions of dollars per year.

The increasing concern about microplastic pollution has led to a rapid research growth in this area in recent years, generating a large volume of data. To illustrate this trend, a Web of Science (WoS) database search using the keywords microplastic OR microplastics, along with the “All Fields” option was performed. Considering only English language “Articles” and “Review Articles” related to environmental microplastics, the search yielded 10,883 articles published between 1964 (first record of publication in WoS) and 2022 (Fig. 1). Among these articles, less than a hundred papers were published during the first four decades of the record keeping. Thereafter, the number of publications gradually increased until a rapid growth in the last five years. Indeed, the number of publications in 2022 (i.e., 3,405) was over three-fold that of 2019 (i.e., 1,042) (Fig. 1).

Fig. 1
figure 1

Number of microplastic publications between 1964–2022.

Full size image

Despite the growing awareness and increase in microplastic research, a lack of large-scale, long-term, comprehensive data hinder a complete understanding of the sources, distribution, and impacts of microplastics. Even when available, the management of marine debris data, from large size visual surveys along the coast and in the open ocean, to effects of microplastics on planktonic communities, the blue economy, among others, lags far behind the needs of the scientific, education, and decision-maker communities27,32. The European Union’s EMODnet (European Marine Observation and Data Network) marine litter database33 (https://emodnet.ec.europa.eu/en/chemistry) archives and offers downloadable microplastic data as part of its floating microlitter collection. This database is however limited to only data from European waters. Another product, LITTERBASE34 (https://litterbase.awi.de/), developed by the Alfred Wegener Institute Helmholtz Centre for Polar and Marine Research, Germany, offers a global map and data analysis of marine microplastics from peer-reviewed scientific publications and a limited number of reports. This product does not however include non-published data, archives original data nor offers users the ability to download the data. A proposed ocean surface microplastic database by the Ministry of Environment of Japan (MOEJ) is also yet to be launched. The lack of comprehensive data on the spatial and temporal variability of microplastics is also a challenge for numerical modeling of their occurrence as a way to effectively understand and forecast their origins, trajectory, and aggregation23,35. Subsequently, there is the need for a well curated, expansive, and FAIR36 (Findable, Accessible, Interoperable, and Reusable) database to facilitate the understanding and control of microplastic pollution.

The National Oceanic and Atmospheric Administration (NOAA) National Centers for Environmental Information (NCEI)’s microplastics data stewardship project was started in January 2020 to obtain, aggregate, and archive global microplastic data. The microplastics website and database were launched in July 2021. This database collates microplastic data from large ocean surveys, citizen-science led initiatives, and published literature sources, which provides students, scientists, environmentalists, policy makers, and others, a robust, and open access repository for archived information needed in marine microplastics debris monitoring. One priority in creating the NOAA NCEI microplastic database is data access. The increased awareness of microplastic impacts on the environment and human health has led to a surge in microplastic research. Therefore, open access to the large amount of data generated is crucial to enable a broad, comprehensive assessment of the environmental issue. A FAIR microplastic database will enhance a uniform global understanding of the environmental problem36,37,38,39. In turn, it will aid in formulating management policies around the generation, handling, and disposal of microplastics.

A recent study by Jenkins et al.39 reported that only 28.5% of microplastic publications since 2006 contained a data sharing statement. Of this number, 38.8% provided their study data in the paper’s supplementary material and 13.8% through a data repository. In summary, the need to improve open access to microplastic data is monumental. An overarching goal of the microplastics product is to establish NCEI as the primary location for open access, comprehensive, quality-controlled global microplastics data and information. This effort along with other NCEI archived data (e.g., Global Ocean Current Database, Blended Seawinds, World Ocean Database, etc.), will serve a diverse international customer base to attain a holistic understanding of the global microplastic problem. In this paper, we present the NOAA NCEI global marine microplastics database, its creation, quality control procedures, and future directions.

Results

Overview

The NOAA NCEI microplastics database contains only in-situ measured marine microplastic concentrations. Data from animal tissues, model output and laboratory experiments are not included. At present, the database contains data from only the surface ocean. Recognizing microplastics are not only in surface ocean waters, our future goal is to broaden the database to include data from different ocean depths, ocean sediments, and beaches. This expansion will enable a more comprehensive understanding of microplastics in the marine environment.

The database has two levels: archive and geodatabase. All microplastic data received are ingested into the NOAA NCEI archive after initial quality control and guaranteed to be available for at least 75 years. Next, the data are homogenized and added to the geodatabase which is displayed on the NCEI microplastics ArcGIS web portal. The archive provides more detailed information about individual datasets (Fig. 2), allowing in-depth exploration for interested categories of users such as scientists, graduate students, coastal managers, and policy makers. The ArcGIS geodatabase and web portal on the other hand is geared more towards a general audience. As such, not all metadata associated with an archived microplastic dataset is provided on the web portal.

Fig. 2
figure 2

An example of a screenshot from an archived dataset collected in the Southern Ocean from 2016-11-28 to 2017-07-27, showing detailed information on how the data was collected, quality-controlled and analyzed. (Credit9: https://www.ncei.noaa.gov/archive/accession/0253447).

Full size image

Archive display interface

A user-friendly interface displays the detailed metadata information about individual datasets in the archive. These information include a title for the data submission, investigators and their affiliations, package description, a map showing study area and sampling locations, data citation, temporal coverage, spatial coverage, platforms, keywords, identification information, funding information, and variable metadata section40,41. The variable metadata section contains details on how the data was collected, quality- controlled and analyzed (Fig. 2). The archive display interface also contains HTTPS and FTP links to download the data package.

To ensure uniformity and ease of use, the titles of archived datasets follow the following template: “[observed properties] collected from [research vessels or other platforms] in [sea names] from [start date] to [end date]. In the screenshot of an archive display interface shown in Fig. 2, the data package title is “Floating microplastics concentration collected from AKADEMIK TRYOSHNIKOV and S.A. AGULHAS II in the Southern Ocean from 2016-11-28 to 2017-07-27”9.

ArcGIS geodatabase and web portal

The web portal contains the homogenized microplastic data. This interface uses user-friendly features such as dropdown menus, display filters, selection and drawing tools, and maps, to enhance the user experience of searching for microplastic data. A detailed help document is provided on the web portal to help users to navigate the site and download data.

As of June 2023, the database contains about 14,000 microplastic records. Each data record represents the concentration of microplastics (counts of pieces/m³) in a given space and time. Other information provided include the sampling equipment, collecting organization, key words associated with the record (e.g., ship name), and reference to original sources including bibliographic digital object identifiers (DOI) (Table 1). The database is publicly accessible from https://experience.arcgis.com/experience/b296879cc1984fda833a8acc93e31476 and can be downloaded (CSV, JSON, and GeoJSON formats) in its entirety or subsampled using filters (e.g. date, oceans, and seas, microplastic concentration, or sampling methods). The database is currently updated quarterly.

Table 1 Description of fields in the database and map portal.
Full size table

The “NCEI Accession No. Link” directs the user to the original data package associated with the record in the NCEI archives. Here, the user can obtain in-depth information on how the record was obtained, quality controlled, and processed by the data collector.

With the “Concentration class range” and “Concentration class text”, we classify the microplastic concentrations (pieces/m³) in the database (Table 2). The classes are determined based on statistical characteristics and distributions of the database records such as minimum, mean, maximum, standard deviation, and interquartile range. The concentration class range and text of a record is therefore dynamic as more data is added and the statistical characteristics of the entire database change.

Table 2 Microplastic concentration class ranges and texts.
Full size table

Data sources

While it continues to grow, at the time of this manuscript writing, the NOAA NCEI microplastic database has collated information from 33 datasets, all from peer-reviewed published papers of 23 unique lead authors. 30 of the datasets were obtained by email solicitations while 3 were self-reported. 4 of the 33 datasets were collected by citizen science initiatives; The Ocean Race (formerly known as Volvo Ocean Race), Adventure Scientists, Surfing for Science, and Oceaneye Association. Most of the data records were collected from local and regional studies. Although the Ocean Race dataset provides a near-global snapshot of floating microplastic distribution, it does not cover all ocean sub-basins42.

Spatial and temporal coverage

The NOAA NCEI microplastic database is global, containing records from Arctic, Atlantic, Indian, Pacific, and Southern Oceans (Fig. 3). Most of the records are from the Atlantic Ocean (62%) with the least from the Southern Ocean (0.2%) (Table 3). At the time of this manuscript writing, the records were collected from 4/20/1972 to 10/5/2021, with the bulk (72%) collected in the post 2000 era (Fig. 4). Nearly all of the pre 2000 records were collected in the North Atlantic Ocean by the Sea Education Association (SEA), Massachusetts, USA1. The exceptions are ~45 records collected in the northeast Pacific Ocean in the 1970’s43.

Fig. 3
figure 3

A screenshot showing the NOAA NCEI microplastic database GIS web portal with microplastic concentrations.

Full size image
Table 3 Number of microplastic records in each ocean.
Full size table
Fig. 4
figure 4

Number of microplastic records in the NOAA NCEI database.

Full size image

Discussion

As described in the Methods section, several steps are taken to ensure that the microplastic concentrations ingested into the NCEI database are of the highest standards. The NOAA NCEI Send2NCEI44 (S2N) data submission platform includes fields that allow only certain values and formats. This minimizes data entry and spelling errors. In addition, data submitters are contacted on ambiguities in their data such as duplicates, and outliers. Furthermore, the dataset is checked by multiple curators and subject matter experts, prior to being served to the public.

The field of microplastics research is quite young. Although there has been immense expansion of research activities and volume of data generated in recent times, there are still no uniform standards for data collection, analyses, and reporting. The growing interest in this contaminant has led to the development of several microplastic study methods, each with its own strengths and weaknesses38. Due to the stark variations in microplastic origin, density, chemical properties, morphology, size and color, there is no single combination of methods for sampling, extracting, analyzing, and reporting38,39,45,46. Thus, the microplastic concentrations in the database may not always be comparable across studies. Users should consider using data records along with more detailed metadata in the archives (such as sampling protocols and instrumental analysis, e.g., shown in Fig. 2) for further investigation of data usability.

Importance of measuring and reporting standards

An example data compatibility issue observed while compiling the microplastics database is the inconsistency in data reporting standards such as the units of measurements. Units found in the literature include counts of pieces/m³, counts of pieces/km2, counts of pieces/km³, counts of pieces/g, g/km2, g/m3, among others. This lack of consistency creates problems for the research community and interest groups trying to compare records and to form composite datasets. Data harmonization will help merge multiple studies and synthesize information for a better understanding and regulation of the global microplastic problem. NCEI’s efforts to help address these shortcomings include providing a comprehensive microplastic database that gives an overview of the sampling efforts and helps identify the areas to standardize data collection and reporting to enable data harmonization. Standardization will help resolve the calibration needs for datasets with different methodologies, which will expand sharing, scalability, and utility of microplastic data. It will also enhance the fidelity and reproducibility of research results and success at obtaining grant funding for further studies. To achieve these, the standards ought to be consensus based, consistent, and based on best scientific practices.

The need and urgency to standardize and harmonize microplastic data collection, analysis, and reporting have led to a number of national and regional initiatives. Aside the NOAA NCEI’s effort, there are also the European Union’s EUROqCHARM (EUROpean quality Controlled Harmonization Assuring Reproducible Monitoring and assessment of plastic pollution; https://www.euroqcharm.eu/en) project, and the MOEJ guidelines for harmonizing ocean surface microplastic monitoring methods project32,47. On a global scale, the Global Partnership on Plastic Pollution and Marine Litter (GPML; https://www.gpmarinelitter.org/), a multi-stakeholder partnership under the United Nations Environmental Program, is leading efforts at bringing together all the aforementioned groups and others, unto a common platform for cooperation and coordination to share ideas, knowledge, experiences, and resources towards harmonizing microplastic data

Harmonization of current microplastic data products (i.e., EMODnet, LITTERBASE, and NCEI) starts by leveraging the common variables in the individual databases. These include sampling date, latitude, longitude, and sampling methods. Microplastic abundance is however not reported in common units among the different databases. Thus, data harmonization will involve performing unit conversions, among others, in order to have variables with a limited set of measurement units. Both the EMODnet and NCEI products provide users with access to the original and harmonized data while the LITTERBASE product does not archive the original data. In the case of the NCEI product, the archived data retains its original unit reported by the data owner while the harmonized data in the geodatabase (i.e., web portal) are converted to a common format (i.e., pieces/m3) where possible. For the LITTERBASE product, data is typically provided in units of items/km², items/km, items/m³ and other dimensions are converted to these units where possible for comparison. In a situation where microplastic measurements were provided in several dimensions (e.g., count and weight), LITTERBASE uses a preferred unit of items/km2. Also, for datasets that LITTERBASE considered to be spatially extensive, these were aggregated to means for subareas34. In summary, a unified guideline is needed in order to provide a FAIR and homogenized global microplastic data.

Citizen science vs professional scientific research studies

Most of the records we have were obtained from professional scientific research studies, which can be time consuming, expensive, challenging, geographically limited, and seasonally driven48,49. There is, however, a growing interest and potential from citizen science initiatives for microplastic data collection48,49,50,51,52. When properly trained and harnessed, the enthusiasm of these groups can generate substantial data which will contribute towards a more informed, comprehensive understanding of microplastic occurrence and distribution. Involving citizen scientists also creates awareness outside of the professional scientific research community, increases engagement on environmental issues and promotes a community-based approach to environmental pollution management48,49.

Citizen science initiatives often adopt innovative measures to involve individuals through social and sports activities to collect microplastic samples. For example, the Surfing for Science citizen science project attached affordable and easy to use manta trawls on paddle surf boards, kayaks, and rowing boats to acquire microplastic samples3. Similarly, The Ocean Race initiative used two yachts (Turn the Tide on Plastic and AkzoNobel) that were competing in a race around the world as ships of opportunity to collect 96 microplastic samples during their circumnavigation42. The Adventure Scientists initiative used trained citizen scientists for an opportunistic collection of 1,628 1-liter glass jar grab samples across several locations such as shorelines, estuaries and offshore49.

Methods

Data acquisition and submission

At a minimum, we require data with sampling dates (year, month, and day), sampling location geographic coordinates, mesh size, and microplastic concentrations. Data submission and inclusion in the NOAA NCEI microplastic database is freely opened to the public. It is not restricted to only US-based researchers, or projects funded by NOAA or other US funding agencies. Data generated from both grant funded, and non-grant funded projects are welcome. Likewise, data from professional or non-professional scientists (e.g., citizen scientists) are all welcome. Both published and unpublished microplastic datasets are accepted and included in the database. All of the above data sources and kinds are subjected to the same rigorous quality assessment and quality control standards.

We obtain microplastic data predominantly in two ways; self-reporting by data owners and email solicitation requests to data owners. Self-reporting is typically done through the NCEI S2N web portal. This is an archiving tool that allows the data owner to easily submit their data files, metadata, and related documentation to NCEI for long term preservation, stewardship, and access. S2N thus helps the data owner meet any funding requirements for data documentation, sharing, and archiving44. S2N includes controlled vocabularies that enables accurate data findability. It also allows the creation of a user profile which enhances data submitter’s ease of use by retaining records of previous submissions and allowing it to be duplicated to start a new submission.

Data acquisition through email requests begins by NCEI scientists identifying suitable microplastic datasets. The scientists perform literature searches from online reference and citation databases such as Web of Science, Scopus, and Google Scholar using the keywords microplastic, microplastics, plastic, and plastics in the title, abstract and keywords. Identified research papers are then reviewed to ensure they (1) contain microplastic data, (2) are collected from the ambient marine water environment, (3) do not include data from animal tissues, (4) are in-situ data and not model output or laboratory experiment, and (5) use appropriate sampling and analytical methodologies such as those outlined below. If a paper is suitable, the corresponding authors are contacted through emails to obtain their permissions for data to be included into the NOAA archive and geodatabase, and freely and openly redistributed without restriction. When permission is granted, the data are archived on behalf of the owner using S2N. If an identified, suitable research paper uses secondary data, we contact the original data owner for their permission and cite the original data owner.

In addition, we find unpublished data by making inquiries to specific Citizen science groups, initiatives, and researchers. This includes direct contact with presenters at webinars, workshops, and conferences. Those data sampling methods are reviewed against sampling protocols found in published literature. If the sampling methods and protocols are in line with those of peer review publications, they are ingested into both the archive and the geodatabase. If the methods used by a study are too different from what is widely adopted in the literature, the data is archived but not added to the geodatabase.

Data licensing

The NOAA NCEI microplastics database publishes only data that the owners have given explicit permissions to be made completely open and freely available to the public. All submitted data are under conditions of Creative Commons (CC) CC0 (i.e., open access) and CC-BY 4.0 (i.e., cite data source) licenses, or their equivalents, wherein the data is completely open, freely accessible to the public, and users are asked to cite the original data source. Any license assigned by the data source is identified in the metadata maintained and redistributed by NCEI. NCEI does not assign data licenses of any type to original data acquired by NCEI because only the data source can provide the license for the original data, not NCEI. NCEI may transform, reconfigure, or otherwise do quality checks/flags on original source data prior to including that source data into the microplastics database, thus adding value to the overall quality of output data from the microplastics database. NCEI applies a CC0 license to the NCEI microplastics database product, which provides specific attribution for each data package that was contributed to develop the NCEI microplastics product. Because NCEI does not include original data in the NCEI data product that applies a more restrictive license, there is little likelihood of a conflict between an originator’s license and the NCEI license.

There are instances where some scientific journals require researchers to submit their data to a repository prior to submitting their manuscripts. In this case, NCEI can archive the data and not make it discoverable to the public. After the publication of the said manuscript, the author informs NCEI, and the data then becomes discoverable and freely available to the public.

Each dataset archived at NCEI has an associated data citation. In both the archives and microplastic web portal, citation is given to the data owner. The data citation is consistent with the guidelines and recommendations of FORCE1153 and DataCite (https://datacite.org/), and contains information such as list of authors, title of the data package, publication year, data repository, NCEI accession number, and an optional DOI40,41. For a submitted data that already has a DOI, that DOI is maintained. While DOI is highly recommended for all submitted datasets, for those that do not have one, the data owner is given the option of whether a DOI should be minted for it or not.

Quality assessment and quality control

Evaluating sampling and analytical procedures

Both self-reported and solicited data are subjected to quality assessment and quality control to ensure correctness and completeness before archiving. At present, there are no globally-defined uniform standards for microplastic data. As such, we assess the study that collected the data by evaluating the sampling methods and strategy, sample size, sample handling, processing and storage, laboratory preparations, negative and positive controls, sample treatment, and particle and polymer identification38,46,54,55,56,57.

We check that the sampling methods and strategies are clearly defined and reproducible. Known microplastics sampling methods include selective sampling, volume-reduced sampling, and bulk sampling6,58. In selective sampling, microplastics are directly extracted from samples by visual identification. In volume-reduced sampling the samples are filtered or sieved at the sampling location and only the targeted components are transported to the laboratory. In bulk sampling, the entire volume of the sample is taken and is considered the best method when the abundance of microplastic is small6. Examples of instruments used for microplastics sampling include manta net, neuston net, plankton net, bongo net, multiple opening–closing net, continuous plankton recorder, aluminum bucket, stainless steel bucket, glass bottles and jars, and water pump/intake through vessel system2,4,38. We confirm that the mesh size used for sampling and/ filtering was less than 5 mm in order to capture microplastics. The most commonly used net mesh sizes are 333–335 µm59.

The water volume that was sampled should be reported to aid the computation of microplastic concentration. Sufficient water volume should be sampled as microplastics are heterogeneously distributed60. We assess that the sample volume size is representative of the sampling objectives, methods (instruments), strategy, and location. For example, grab sampling collects more microplastic particles than trawl nets. Also, smaller mesh sizes retain more microplastics than larger mesh sizes2,45,61. In one instance, Barrows et al.2 observed that grab sampling collected over three orders of magnitude more microplastics per volume of water and smaller sizes than neuston net sampling. Ideally, the study should collect replicate samples providing a measure of variability in sample collection and a statistically robust analysis of data62. The number of replicates and how they were nested within samples should also be reported.

We evaluate the procedures that were used to handle, store, and process the microplastic samples to ensure that contamination from the field and the laboratory (air, water, and materials) were eliminated. We ensure that the study used non-plastic instruments for data collection and for laboratory analysis6,37,46. Between the moment a sample is collected and examination in the laboratory, the sample should be stored on ice or frozen46,56. Samples can also be preserved in a glass container with ethanol, formalin, or formaldehyde56. Materials that were used such as equipment, tools, clothing, and work surfaces ought to be free of microplastics contamination. This includes wearing cotton or non-synthetic clothes, and thoroughly washing materials and cleaning work surfaces with ultrapure water (e.g., Milli-Q water) and filtered solvents6,63,64. The study must also report the use of field and laboratory blanks to account for procedural contamination46,65. The reported microplastic concentration should account for the controls by deducting the baseline by microplastic count, shape, color, and polymer type65.

We assess if the study adopted procedures that enhance particle identification and counting. Sample treatment includes organic digestion, density separation, sieving and filtering62,66,67. Sieving is usually enough for particles >300 µm as the sizes are large enough to allow for adequate sorting. Organic digestion may be needed to dissolve organic matter in some samples especially for the detection of small microplastics (typically <300 µm56). Organic digestion methods may include the wet peroxide oxidation (WPO) method which uses aqueous 0.05 M Fe (II) solution and 30% H2O2 solution to digest organic materials63. Other studies may involve the use of 10% KOH solution as well as enzymatic digestion methods68. Once organic materials are removed from the sample, the authors should mention what instruments were used for visual identification and quantification of microplastics. The instrument detection limits should also be reported.

We note if the study reports the shapes and polymer types of microplastics encountered. While not currently a focus in our database, it may be in the future as this field evolves. Microplastic shapes include fiber, fragment, film, foam, and pellet2,38,56. Microplastic polymer types include polypropylene (PP), low density polyethylene (LDPE), high density polyethylene (HDPE), polystyrene (PS), polyamide (PA; nylon), polyethylene terephthalate (PET), and polyvinyl chloride (PVC)46,66,69. Researchers should report confirmation of microplastics using chemical characterization methods such as Raman and Fourier-transform infrared (FTIR) spectroscopy6. Particle counts with confidence intervals, detection limits for the count and for minimum particle size, polymer types and percentages (of different polymer types, of synthetic vs natural material), and particle sizes should also be reported56. It is noted that not all samples collected in a study can be confirmed using these technologies due to logistical constraints, costs, etc. Nevertheless, a reasonable subsample should be confirmed for microplastic polymer type. Hermsen et al.56 recommends that for pre-sorted particles less than 100, all particles should be analyzed. For particles more than 100, at least 50% should be identified with a minimum of 100 particles.

Evaluating sampled data

After examining the sampling and analytical procedures, we evaluate the microplastic data. We check that the data contains the minimum requirements: sampling dates (year, month, and day), sampling location geographic coordinates, mesh size, and microplastic concentrations. Environmental (e.g., wind conditions) or logistical factors that may affect the interpretation of results should also be reported70,71. We check that the value of each record item matches the data type and confer with the data submitter on any ambiguity. We also verify that the data are plastics less than 5 mm, collected from the ocean surface and within valid geographical limits (i.e., latitude is between 90°S and 90°N and longitude is between 180°W and 180°E decimal degrees). Finally, we flag duplicate data for further consultation with the data submitter.

Sampled microplastic concentrations depend on factors such as study objectives, study area, sampling time, sampling instruments, sampling strategies, and analytical methods2,38,57,61. We ensure that the reported microplastic concentrations are within a reasonable range with respect to findings in published literature. Outlier data points (e.g., higher than usual ranges seen in published literature) are flagged for further consultation with the data submitter. We accept microplastic data that are reported in concentration units (i.e., counts of pieces per unit volume). Particle counts (as opposed to total mass/weight) are more convenient to link with toxicity studies since it makes it easier to calculate concentrations of specific microplastic types46,62. Concentration units other than counts of pieces/m³ (e.g., counts of pieces/km2, counts of pieces/km³) are converted to pieces/m³ (using information from the study such as dimensions of sample collection instrument) for data harmonization. Submitted microplastic data that are reported as weight are archived but not displayed on the geodatabase map portal due to harmonization challenges with other data.

Conversion of units from surface area (e.g., counts of pieces/km2) to volume (i.e., counts of pieces/m³) for data harmonization potentially creates biases and also limits comparison with some datasets. Microplastic measurements per unit area appears to be the commonly used unit for data collected with nets (i.e., areal sampling, e.g., Lavender Law et al.2; Reisser et al.11; Eriksen et al.13) while measurements per unit volume appears to be the commonly used unit for data collected by other means such as buckets, bottles, and pumps (i.e., point/station/grab sampling, e.g., Osorio et al.72; Setiti et al.73). Because our database contains data collected with all these different instruments and sampling methods, we convert to a common unit of measurements per unit volume for harmonization in the geodatabase (web portal), while maintaining the original unit in the archive. It should be mentioned that there are several datasets (e.g., Goldstein et al.43; Faure et al.50; de Haan et al.3; Suaria et al.9) where data was collected with nets and the submitted data from the owner are reported in both measurements per unit area and measurements per unit volume (i.e., the unit conversions in this instance were not done by NCEI).

Microplastic data unit conversion comes with challenges. For example, the water volume sampled by nets could be misrepresented as the position of a net’s frame varies over water surface, especially in the presence of waves, thus the net (or even a volumeter), may not be entirely submerged in the water. There are advantages and disadvantages of each of the different microplastic sampling methods (as we have previously mentioned) and the microplastic research community is still deliberating on a possible unified unit of measure and standards of reporting. One of our aims in creating this database is to aggregate the different data types and allied information, which will hopefully generate enough information to help the research and end-user communities reach a consensus on standards. We have a notice on our website and help pages alerting users to use the geodatabase alongside the archive which contains the data in its original units submitted by the data owner.