A wide range of natural archives, including glaciers, cave formations, trees, lake and marine sediments, and corals, contain physical and chemical records of Earth’s hydroclimate history. Within these archives, scientists can measure the isotopic ratios of oxygen and hydrogen to deduce changes over time in the isotopic composition of environmental waters, such as precipitation, seawater, and groundwater.
Measuring changes in the relative abundances of heavy oxygen and hydrogen isotopes (δ18O and δD, respectively) recorded in these natural archives provides a way to reconstruct local, regional, and even global climate signals. Measuring these changes also provides invaluable comparison targets for global climate models. However, the data produced by individual studies are scattered over numerous locations and formats.
A new project is changing this by providing a comprehensive synthesis of δ18O and δD records in a format suitable for regional-scale climate reconstructions or for data-model comparisons. Called the Past Global Changes (PAGES) Iso2k Project (PAGES Iso2k), the project brings together records that will advance the community’s understanding of how hydroclimate varied over the past 2,000 years, also referred to as the Common Era. Scientists plan to use the Iso2k database to uncover how external climate forcings, anthropogenic forcings, and internal variability contributed to variations in the global water cycle.
The Iso2k Project
The Iso2k team began developing a comprehensive database of δ18O and δD records in 2015 as a formal project within the framework of the PAGES 2k initiative. This global database of paleo-δ18O and paleo-δD records is based on a wide variety of marine and terrestrial archives, with resolutions from subannual to centennial, that span at least part of the past 2,000 years.
The main goal of Iso2k is to identify regional- and global-scale features in Common Era hydroclimate and atmospheric circulation and then assess possible relationships with temperature reconstructions [PAGES 2k Consortium, 2013, 2017; Tierney et al., 2015; McGregor et al., 2015; Abram et al., 2016]. Other key scientific questions to be addressed include the following:
- How do water isotopic proxy records capture changes in tropical and extratropical modes of climate variability?
- Are modern temperature-hydrology relationships stable through space and time?
- How are changing oceanic conditions reflected in seawater δ18O and salinity variations?
To build the database, the team collated published records archived in the National Oceanic and Atmospheric Administration (NOAA) Paleoclimatology and German PANGAEA online data repositories, as well as additional data reported in tables of scientific papers (but not yet registered in global databases) and some data sets contributed directly by authors.
The team agreed upon intentionally broad data selection criteria to include as many records as possible. To be used in the compilation, records must be published, and data sets must be publicly available (i.e., in a public archive, in the publication itself, or promptly provided by the author when requested). Future versions of the Iso2k database will also screen and incorporate unpublished records contributed by the community.
Enabling Data Synthesis and Model Comparison
Once the Iso2k team developed the main list of records, they started to build metadata for each record. They designed a detailed template that created a common language among different proxy sensor archives to catalog the defining characteristics of each record. This template also incorporates all the information needed to facilitate process-based data syntheses and comparisons with climate models and reanalyses.
For example, instead of simply listing records by their inferred climatic variable (e.g., temperature), clear distinctions are made between the type of archive (e.g., speleothem), the observation (e.g., δ18O), the measured material (e.g., carbonate), the inferred material (e.g., groundwater), the authors’ interpretations of the dominant controls on the isotopic ratios of the inferred material (e.g., precipitation δ18O), and the authors’ interpretation of climate (e.g., summer monsoon precipitation amount).
The metadata richness is not meant to be comprehensive, but rather to facilitate wide use of the database for a range of users, from climate modelers with a limited knowledge of water isotopes to experts in isotopic proxy data. For example, data can be queried by “inferred material” to directly compare with isotope-enabled climate model simulations or to select appropriate proxy system models [Evans et al., 2013; Dee et al., 2015].
The rich metadata structure of the Iso2k database serves as a basis for future projects in the paleo community using Linked Paleo Data (LiPD) and Linked Earth. The database can also easily be used for future research by nonpaleo communities such as data assimilators, modelers, and policy makers.
Links to the Broader Community
The Iso2k effort is currently the only global hydroclimate database being constructed for the Common Era, and it has strong ties to other archive- and climate target–specific groups within PAGES. Examples of these groups include Speleothem Isotopes Synthesis and Analysis (SISAL) [Comas-Bru et al., 2017], Marine Annually Resolved Proxy Archives (MARPA) [Dassié et al., 2017], and a growing interest in a Hydro2k working group [Smerdon et al., 2017]. The Iso2k Project is also building strong ties outside of PAGES, for example, with the Paleoclimate Modelling Intercomparison Project (PMIP).
The Iso2k effort builds on the high-profile PAGES 2k temperature syntheses [PAGES 2k Consortium, 2013, 2017; Tierney et al., 2013; McGregor et al., 2015; Abram et al., 2016], and it represents the first steps toward a broader “hydroclimate 2k” synthesis. This big data approach will give us the tools necessary to investigate interannual- to centennial-scale variability, as well as possible secular trends in hydroclimate over the past 2,000 years.
The First Iso2k Science Team Meeting
The Iso2k team, whose more than 50 members come from more than 10 countries, collaborated remotely for 2 years. Then, on 14–15 May 2017, 28 members, including 10 early-career scientists, met in person for the first time at the Consejo Superior de Investigaciones Científicas (CSIC) offices in Zaragoza, Spain. The meeting directly followed the 5th PAGES Open Science Meeting.
At the team meeting, participants evaluated the climatic and isotopic patterns emerging from the beta version of the database, which was released internally in spring 2017. Participants then focused on troubleshooting database metadata and quality control issues, in anticipation of the upcoming wide release of the database.
Next Steps and Community Involvement
As of March 2018, the beta version of the database contained 629 individual isotopic records from more than 432 global locations (Figure 1). The database can be accessed using R, Python, or MATLAB. Version 1.0 of the database is close to completion, and the final quality control efforts are scheduled to be finished by summer 2018.
The team presented their first scientific results at the PMIP Phase 4 (PMIP4) meeting in Stockholm, Sweden, in September 2017 and presented further results at the American Geophysical Union’s 2017 Fall Meeting in New Orleans [e.g., Conroy et al., 2017; Konecky et al., 2017].
They will submit papers describing the database to Scientific Data and major findings to another journal in 2018, after which the database will be freely accessible for broad use.
The Iso2k Project welcomes new members from the community who are interested in using water isotopes to advance our understanding of Common Era climate. If you would like to know more about Iso2k and possibly get involved, visit our website.
We thank PAGES and the CSIC for Iso2k workshop support, the NOAA Paleoclimatology team for assistance in locating data, and National Science Foundation AGS-PRF grant 1433408 for support to B.K.