Coastal waters are the most biogeochemically dynamic areas of the ocean, exhibiting the highest standing stocks, process rates, and transport fluxes of carbon (C), nitrogen (N), and phosphorus (P) per unit area (Bauer et al., 2013; Mackenzie et al., 2011). In these areas, organic matter plays a critical role in numerous biogeochemical processes, serving as both a C, N, and P reservoir and substrate (Carreira et al., 2021).

Organic material found in the marine environment is commonly distinguished by its size; material retained on a filter with a pore size typically between 0.2 and 0.7 µm is classified as particulate organic matter (POM), whereas organic matter that passes through the filter is referred to as dissolved organic matter (DOM). This partitioning is operational but has implications for biogeochemical cycling: POM can be suspended in the water column or sink to the sediments controlled by its size, shape, and density (Laurenceau-Cornec et al., 2015), whereas DOM is a solute that mostly remains in the water column. In most coastal waters, DOM concentrations are higher than POM, with POM having a larger proportion of known biochemical classes (e.g. carbohydrates, proteins) than the dissolved fraction, suggesting that, generally, DOM is more reworked and recalcitrant (Boudreau and Ruddick, 1991; Lønborg et al., 2018; Benner and Amon, 2015).

The DOM pool consists mainly of C (DOC), N (DON), and P (DOP), but it also includes other elements such as oxygen, sulfur, and trace elements (Lønborg et al., 2020). In coastal waters, DOM originates from multiple sources. Internal, or autochthonous, sources include planktonic organisms (Lønborg et al., 2009; Carlson and Hansell, 2015), benthic microalgae, macrophytes, and sediment porewater (Burdige and Komada, 2014; Wada et al., 2008). On the other hand, DOM from external, or allochthonous, sources, has mainly terrestrial origins, including wetlands, river and surface runoff, groundwater discharges, and atmospheric deposition (Iavorivska et al., 2016; Raymond and Spencer, 2015; Taniguchi et al., 2019; Santos et al., 2021). The main sinks for DOM from the water column in coastal waters are the following: (1) bubble coagulation and abiotic flocculation (Kerner et al., 2003) or sorption to particles (Chin et al., 1998), (2) sunlight-mediated photodegradation (Mopper et al., 2015), and (3) microbial degradation by mainly heterotrophic prokaryotes (Lønborg and Álvarez-Salgado, 2012).

Given the importance of DOM as a source of nutrients and for coastal biogeochemical cycling in general, numerous studies have measured the C, N, and P contents of the DOM pool over the last few decades (e.g. García-Martín et al., 2021; Cauwet, 2002; Osterholz et al., 2021). Most data, however, are often unavailable or stored in an inaccessible manner, making it difficult to, for example, analyse global spatial and temporal patterns effectively. Global open ocean DOM data compilation for DOC, total dissolved nitrogen (TDN) (Hansell et al., 2021), and DOP (Liang et al., 2022; Karl and Björkman, 2015) already exists and contains a few coastal samples (<200 m) (Hansell et al., 2021), but there are no compilations specifically focused on coastal waters. Hence, there is a clear need for a comprehensive global and integrated database of DOC, DON, and DOP measurements for coastal waters. To address this need, we have prepared the first edition of a coastal DOM database (named CoastDOM v1), by compiling both previously reported as well as unpublished data. These data have been obtained from authors of the original studies or extracted directly from the original studies. In order to allow the DOM measurements to be interpreted across larger scales, and to better understand their relationship with local environmental conditions, we have included concurrently collected ancillary data (such as physical and/or chemical seawater properties) whenever available. The objective of this database is multifaceted. Firstly, we aimed to compile all available coastal DOM data into a single repository. Secondly, our intention was to make these data easily accessible to the research community. And thirdly, we sought to achieve long-term consistency of the measurements to enable data intercomparison and establish a robust baseline for assessing, for example, the impacts of climate change and land use changes.

The data compiled in CoastDOM v1 were collected, analysed and processed by different laboratories; however, all data included have undergone quality control measures, either by using reference samples or internal quality assurance procedures. While many of the included DOC and TDN data have been systematically compared against consensus reference material (CRM), mainly provided by the University of Miami's CRM programme (Hansell, 2005), there is a limitation in CoastDOM v1 regarding the inter-calibration across different measurement systems used for both DOP and DON determination. While the CRM could be used for DOC, DON, and DOP measurements, this has not yet been attempted for DOP, and measurement uncertainties increase in the sequence DOC > DON > DOP. Although some of the reported measurements have quantified the DOP recovery based on commercially available DOP compounds such as adenosine triphosphate (ATP), it is not known if these were conducted systematically in all cases. Therefore, we strongly recommend undertaking further inter-calibration across laboratories for future measurements of TDP, as has been done for DOC and TDN measurements (e.g. Sharp et al., 2002). Since additional quality control is not possible in retrospect, we assessed the quality of CoastDOM v1 based on its internal consistency.

In CoastDOM v1, we defined “coastal water” as encompassing estuaries (salinity >0.1) to the continental shelf break (water depth <200 m). However, some locations, such as deep fjords which are close to the coast, cannot be classed as coastal due to bathymetry (deeper than >200 m). Therefore, we evaluated the inclusion of some datasets on a case-by-case basis. For inclusion in the database, each DOM measurement needed, at a minimum, to contain the following information (if reported in the original publication or otherwise available):

country where samples were collected,

To ensure the inclusion of only high-quality data, we only accepted entries with specific World Ocean Circulation Experiment (WOCE) quality codes: “2: Acceptable measurement” and “6: Mean of replicate measurements”. In our quality control assessments, we carefully avoided overly strict criteria, known as “data grooming”, which could potentially overlook genuine patterns and changes in the dataset that may be significant over longer temporal and/or wider spatial scales. Coastal waters are known to exhibit a wide range of environmental concentrations, influenced by factors such as seasonality and local anthropogenic activities. Consequently, these data points may encompass a wide concentration range. However, obtaining consistent long-term datasets is important to enable data intercomparison and establish a robust baseline. Such long-term consistency can be achieved by using the CRM standards provided by the Hansell laboratory for DOC and TDN. Another helpful approach is comparing the DOM concentrations obtained by different laboratories in the same study area and time of year.

The dataset is available at the PANGEA database (10.1594/PANGAEA.964012; Lønborg et al., 2023). The file can be downloaded as a *.csv merged file and is available in full open access.

In CoastDOM v1, we have compiled available coastal DOM data in a single repository, making it openly and freely available to the research community. This compilation has established a consistent global dataset, serving as a valuable information source to investigate a variety of environmental questions and to explore spatial and temporal trends. We suggest a set of recommendations for the future expansion of this global dataset. First, our analysis highlights a spatial bias, with a concentration of sampling efforts and/or data availability predominantly concentrated in the Northern Hemisphere. The data gap in coastal DOM measurements in the Southern Hemisphere needs to be addressed to provide a more representative global understanding of the role of DOM in coastal water biogeochemistry. Additionally, increased sampling efforts especially around Africa and South America as well as island nations are warranted due to the vulnerability of many coastal areas to climate change and intensifying human activities, which will undoubtedly impact DOM biogeochemistry. Furthermore, there are comparatively few data from coastal waters affected by river discharge into the tropics, e.g. the Amazon or Indian and Indonesian rivers that together dominate freshwater inputs to the coastal ocean. Second, there is a need for more comprehensive temporal and spatial datasets to capture the variability of DOM concentrations in highly dynamic and productive coastal systems. Focused efforts should be made to resolve these temporal and spatial changes. Third, only a fraction of data entries report paired DOC, DON, and DOP measurements; we encourage that these be measured and reported together in order to better determine changes in stoichiometry and composition. Fourth, collecting and reporting ancillary data, such as temperature, salinity, nutrient measurements, and particulate components, are important to provide context and to better understand the underlying processes driving the observed DOM concentrations. Fifth, studies need to collect a minimum of metadata and report them in a standardized manner. Lastly, we recommend regular inter-calibration exercises to establish standardized and interoperable methods and data, particularly for DON and DOP measurements. This will ensure the comparability and reliability of data across different studies and enhance our understanding of DON and DOP dynamics in coastal waters.

In light of ongoing global environmental changes, the mobilization and open sharing of existing data for important biogeochemical variables, such as the DOM pool, are crucial for establishing baselines and determining global trends and changes in coastal waters. The aim is to publish an updated version of the database periodically to determine global trends of DOM levels in coastal waters, and we therefore encourage researchers to submit new data to the corresponding author. The CoastDOM v1 dataset was developed according to the FAIR principles regarding findability, accessibility, interoperability, and reusability of data. Thus, CoastDOM v1 will serve as a reliable open-source information resource, enabling in-depth analyses and providing quality-controlled input data for large-scale ecosystem models.