Abstract:
This database is a curated compilation of stable oxygen isotope (delta18O) and barium measurements from seawater, freshwater, rivers, fjords, and precipitation, covering the pan-Arctic region (here defined as north of 60 degrees North). Data were assembled from 99 published and archived source datasets spanning the Arctic Ocean, Beaufort Sea, Bering Sea, Chukchi Sea, Davis Strait, Barents Sea, Laptev Sea, Kara Sea, East Siberian Sea, Greenland Sea, Svalbard, Greenland fjords, and major Arctic rivers (Yenisey, Lena, Ob, Mackenzie, Yukon, Kolyma). Measurements were collected between December 1967 and August 2025, totalling 45,737 data points, of which 6504 are freshwater samples and 5,286 are land-based.
Each record includes the delta18O value, the measurement standard used (predominantly the Vienna Standard Mean Ocean Water (VSMOW)), and associated ancillary data including geographic coordinates, sampling depth, temperature, and salinity where available. 5531 out of these delta18O records contain total barium measurements. All source datasets have been harmonised to a common schema and subjected to quality assurance and quality control (QA/QC) processes, with World Ocean Circulation Experiment (WOCE) flags applied where present in the original data. The database was constructed as part of the Artificial Intelligence for Stable Isotope Tracers (AISIT) programme, funded by the Natural Environment Research Council (NERC) and the Engineering and Physical Sciences Research Council (EPSRC), to provide an Artificial Intelligence (AI) ready, Findable Accessible Interoperable and Reusable (FAIR) compliant resource for the study of Arctic freshwater sources, hydrological processes, and climate change. The database is designed to support reproducible scientific analysis, isotope-salinity relationships, and machine learning applications.
The AISIT project was funded by NERC and EPSRC with the grant no. NEB2678.
Keywords:
AISIT, Arctic Ocean, BIOPOLE, barium, delta18O, freshwater tracers, pan-Arctic, seawater, stable oxygen isotope
Thorpe-Morgan, C., Rowlands, E., Sanders, R.N.C., ten Hoopen, P., & Tarling, G.A. (2026). AISIT Database of Pan-Arctic stable oxygen isotope (delta18O) and barium measurements from seawater, freshwater, rivers, fjords, and precipitation north of 60 degrees North, 1967-2025 (Version 1.0) [Data set]. NERC EDS UK Polar Data Centre. https://doi.org/10.5285/0dfb14d6-73a9-4273-99ed-b231ca581b1d
| Access Constraints: | Under embargo until 01/05/2026. |
|---|---|
| Use Constraints: | The AISIT Database is supplied under Open Government Licence v.3 http://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/. Individual source datasets retain their original licences (CC-BY-4.0, CC0, OGL, and others as documented in the database). Users should consult the Licence and Data_Use_Statement fields for each contributing dataset, which are indicated in the file AISIT_Database_source_datasets_info.csv. |
| Creation Date: | 2026-04-07 |
|---|---|
| Dataset Progress: | Complete |
| Dataset Language: | English |
| ISO Topic Categories: |
|
| Parameters: |
|
| Personnel: | |
| Name | UK Polar Data Centre |
| Role(s) | Metadata Author |
| Organisation | British Antarctic Survey |
| Name | Charles Thorpe-Morgan |
| Role(s) | Investigator, Technical Contact |
| Organisation | British Antarctic Survey |
| Name | Emily Rowlands |
| Role(s) | Investigator |
| Organisation | British Antarctic Survey |
| Name | Petra ten Hoopen |
| Role(s) | Investigator |
| Organisation | British Antarctic Survey |
| Name | Rachael N C Sanders |
| Role(s) | Investigator |
| Organisation | British Antarctic Survey |
| Name | Geraint A Tarling |
| Role(s) | Investigator |
| Organisation | British Antarctic Survey |
| Name | Katharine R Hendry |
| Role(s) | Investigator |
| Organisation | British Antarctic Survey |
| Name | Michael P Meredith |
| Role(s) | Investigator |
| Organisation | British Antarctic Survey |
| Parent Dataset: | N/A |
| Reference: | This database is part of the AISIT programme (https://aisit.ac.uk/) and is associated with the BIOPOLE National Capability programme (https://biopole.ac.uk/). Source datasets are listed in full, with DOIs and citations provided for each, in the file AISIT_Database_source_datasets_info.csv. Complementary datasets are the Global Network of Isotopes in Precipitation (GNIP) Database https://www.iaea.org/services/networks/gnip, Global Network of Isotopes in Rivers (GNIR) Database (https://www.iaea.org/services/networks/gnir and PAGES CoralHydro2k Seawater d18O Database (https://doi.org/10.25921/ap7d-2k16). Klein, E. S., Cherry, J. E., Young, J., Noone, D., Leffler, A. J., & Welker, J. M. (2015). Arctic cyclone water vapor isotopes support past sea ice retreat recorded in Greenland ice. Scientific Reports 2015 5:1, 5(1), 10295-. https://doi.org/10.1038/srep10295 Ahktar et al. (2024). Croissant: A Metadata Format for ML-Ready Datasets. https://arxiv.org/abs/2403.19546. |
|
|---|---|---|
| Quality: | All data were subjected to QA/QC processes. WOCE quality control flags are included where provided by source datasets; where non-standard flags were used by the source, these have been translated to WOCE equivalents (recorded in Delta_O18_WOCE_Flag_Translated) and the originals preserved (Delta_O18_WOCE_Flag_Original). Missing values are represented by empty fields. Datetime entries without a day have been assigned the first day of the month; entries without a time have been assigned 00:00:00. The delta18O standard is recorded per sample; entries where the standard could not be determined are flagged "VSMOW (assumed)" and should be treated with caution. Land-based sample positions were flagged using an approximate point-in-polygon method and should be verified manually where precision is critical. Some source datasets were obtained via private communication and are unpublished; these are identified in the Data_Provenance and Database_Citation fields. There is no fixed spatial or temporal resolution, data are discrete point observations from individual water samples. Spatial and temporal coverage is uneven and dependent on source dataset sampling strategies. |
|
| Lineage/Methodology: | Source datasets were identified through a systematic survey of international resources. Open data repositories searched include the NOAA National Centers for Environmental Information, Arctic Data Center, British Oceanographic Data Centre, Goddard Institute for Space Studies (GISS) Global Seawater Oxygen-18 Database, University of Utah Waterisotopes Database waterisotopes.org, GEOTRACES Intermediate Data Product 2025 (IDP2025), PANGAEA, Laboratoire d'Oceanographie et du Climat: Experimentations et Approches Numeriques (LOCEAN), the NSF Arctic Data Center, Arctic GRO, PAGES CoralHydro2k Seawater delta18O database and Global Ocean Data Analysis Project Version 2 (GLODAP). Due to their QA adjustments to the data over time, we opted not to use data directly from GLODAP in this study. Instead, we utilised the database to identify additional cruises available to AISIT, and then sought the original data from the source. Data was also obtained through direct private communication with data holders. The Global Network of Isotopes in Precipitation (GNIP) and Global Network of Isotopes in Rivers (GNIR) represent large scale freshwater databases with global coverage. Due to the data use agreement of these repositories, we were unfortunately unable to include these data into the AISIT database. However, we have highlighted these databases to potential users as a large repository of, mostly freshwater, O18 data and encouraged those using the AISIT database for scientific studies to supplement the data therein with that of the GNIP and GNIR surveys. Each source dataset was assessed for data availability, licensing, and methodological compatibility. Data were extracted and harmonised into a common schema, including standardisation of date-time formats (ISO 8601), coordinate systems (decimal degrees), depth fields (metres below surface, derived via the Thermodynamic Equation Of Seawater - 2010 Gibbs Seawater height from pressure (TEOS-10 gsw_z_from_p), where pressure was available), and salinity (PSS-78). Delta18O values are reported relative to the standard specified in the O18_Standard column (predominantly VSMOW; entries where the standard could not be confirmed are labelled "VSMOW (assumed)"). Quality control flags follow the WOCE convention. Boolean flags identify freshwater samples (Freshwater), BIOPOLE-associated samples (BIOPOLE), and land-based sample locations (Land_Based, determined via a point-in-polygon test against a 10 m global coastline using Shapely/GEOS Python package). The database was built using Python, follows the Croissant metadata standard for AI-readiness, adheres to the Climate and Forecast (CF) conventions and Unified Code for Units of Measure (UCUM) standard and is compliant with ISO 19115 Geographic Information Metadata. |
|
| Temporal Coverage: | |
|---|---|
| Start Date | 1967-12-01 |
| End Date | 2025-08-15 |
| Spatial Coverage: | |
| Latitude | |
| Southernmost | 60 |
| Northernmost | 90 |
| Longitude | |
| Westernmost | -180 |
| Easternmost | 180 |
| Altitude | |
| Min Altitude | N/A |
| Max Altitude | N/A |
| Depth | |
| Min Depth | 0 m |
| Max Depth | 4451 m |
| Location: | |
| Location | Arctic |
| Detailed Location | Canadian Arctic Archipelago |
| Location | Canada |
| Detailed Location | N/A |
| Location | Norway |
| Detailed Location | N/A |
| Location | Arctic Ocean |
| Detailed Location | N/A |
| Location | Barents Sea |
| Detailed Location | N/A |
| Location | Beaufort Sea |
| Detailed Location | N/A |
| Location | Bering Sea |
| Detailed Location | N/A |
| Location | Chukchi Sea |
| Detailed Location | N/A |
| Location | Arctic |
| Detailed Location | Davis Strait |
| Location | Greenland |
| Detailed Location | N/A |
| Location | United States Of America |
| Detailed Location | Alaska |
| Location | Arctic |
| Detailed Location | Fram Strait |
| Location | Arctic |
| Detailed Location | Laptev Sea |
| Location | Arctic |
| Detailed Location | Kara Sea |
| Location | Arctic |
| Detailed Location | East Siberian Sea |
| Location | Arctic |
| Detailed Location | Greenland Sea |
| Location | Arctic |
| Detailed Location | Labrador Sea |
| Location | Arctic |
| Detailed Location | Baffin Bay |
| Location | Arctic |
| Detailed Location | Siberia |
| Location | Arctic |
| Detailed Location | Yenisey, Lena, Ob, Mackenzie, Yukon, Kolyma |
| Location | Arctic |
| Detailed Location | Svalbard |
| Location | Iceland |
| Detailed Location | N/A |
| Data Collection: | Instrumentation varies by source dataset and includes CTD rosette systems, Niskin and other water sampling bottles, underway towfish systems, and Remotely Operated Vehicles (ROVs). Stable oxygen isotope analyses were carried out by laboratories associated with each contributing dataset; analytical methods are described in the respective source dataset publications. Specific instrument models are documented at the source dataset level. Database construction used Python 3.12. Software packages used: pandas, netCDF4, openpyxl, gsw/TEOS-10, Shapely/GEOS, numpy, pyarrow and matplotlib. |
|---|
| Data Storage: | All files are provided in open formats. The database is provided as a single XCSV file and a CF-1.9/ACDD-1.3 compliant NetCDF4 file. Additionally the data are provided in PARQUET format. Two CSV metadata documents record the AISIT database schema and contributing datasets. Metadata are also available as a machine-readable Croissant file (.json). |
|---|