Methodology:

Data were systematically compiled from legacy databases, published articles, institutional repositories, and cruise reports. Each record was assigned a unique Core ID (as provided by authors) and standardised metadata including geographic coordinates, water depth, collection date, collection method and citation. Manual data entry was largely required due to inconsistencies in carbon reporting and significant errors uncovered during compilation.

Carbon measurements were collated in 1cm intervals up to 1m below the seafloor. Lookup tables were used for sediment classification based on particle size thresholds following a current best practise adapted Folk classification system.

Data collection:

Due to the manual requirements of the compilation, data assembly was conducted through Microsoft Excel (version 2408). Geographic presentation of data was conducted using ArcGIS Pro 3.2.1.

Data quality:

In some instances data has been averaged to account for intervals presented in this study (e.g., 0 - 0.5cm and 0.5 - 1cm have been averaged to produce a 0 - 1cm value). This may cause under or overestimations of carbon values in some instances, but is a necessary step to simplify the usage and presentation of the data.

Where possible the database has been combed for null or obsolete values (e.g., negative or overly inflated carbon values, geolocations outside of scope), however due to the scale of data reported it is highly possible that errors will occur.

As this is a compilation of other works (secondary data), assurances for data accuracy cannot be given. Any discrepancies or concerns with data presented here should be directed to the authors or data holders.

When potential inaccuracies were uncovered during collation, authors were contacted for clarification. Vagueness (e.g., comments such as "most, but not all, of the cores in this study analyse the top 2cm of sediment") and seemingly incorrect values were continually present when identifying data for this dataset, therefore it is imperative to contact the cited authors if any additional clarification is required.

Unknown metadata values are either listed as NaN or specified as unknown. Where appropriate data values less than or equal to 0.0 have been listed as "NaN" as they are believed to be either incorrect (negative values) or redundant (0 values).