1 of 28

12-14 November 2024, Costa Rica and Online

RDA 23rd Plenary Meeting (RDA P23) | Sustainable Science

Connecting and Governing Ocean Data

for Sustainability

Impacts and Challenges of Curating Oceanographic Data

RDA 23rd Plenary Meeting

www.rd-alliance.org

2 of 28

Why is this important?

The Oceans…

Supply our oxygen
Play a vital role in the water cycle
Drive our climate system
Hold biodiversity
Provide nutrition
Offer recreation opportunities

cloud over ocean image here

fish image

reef image

diver/surfer mage

NOAA.gov

WHOI Image Archive

RDA 23rd Plenary Meeting

www.rd-alliance.org

3 of 28

Why is this important?

Unprecedented change makes it imperative to document and understand past and current states, and fluxes to better predict future rates of change and their impacts.

fish kill image

dead reef image

tornado image

plastics image

wikipedia

Michael Hooper, USGS

Nat’l Geographic

Kai Kopecky et al

NOAA.gov

RDA 23rd Plenary Meeting

www.rd-alliance.org

4 of 28

Why is this important?

The Science We Need for the Ocean We Want

Conserve and sustainably use the oceans, seas and marine resources for sustainable development.

To achieve these goals, researchers will need to share and access high quality oceanographic data for solution science and continued discovery.

RDA 23rd Plenary Meeting

www.rd-alliance.org

5 of 28

Oceanographic Research Data Challenges

One water sample

100K mass spectra

≥1TB

Imaging Flow Cytobot database

> 1billion images

>7TB

Ocean Metaproteomics

Phytoplankton Ecology

RDA 23rd Plenary Meeting

www.rd-alliance.org

Emerging research practices and instruments are adding to an already highly heterogeneous data landscape, posing challenges to share, discover and use data.

Examples:

Metaproteomics - study of proteins in the ocean, can inform scientists of what organisms are present and their activity wrt biogeochemical cycling w/in the oceans. Novel and dense information type need to be saved and shared to facilitate discovery and stewardship.

Each sample can contain 10,000-100,000 mass spectra, producing TB sized files

Imaging Flow Cytobot instrument can take continuous high-resolution (hourly) measurements of phytoplankton cells (such as diatom and dinoflagellate cells 10 to 150 microns in size) to help determine community structure and its seasonal variation. The database >10yrs old contains > 1 Billion images and is growing from 7TB

6 of 28

Oceanographic Research Data Challenges

Oceanographic Data are:

Big (high volume, velocity, variety)
Unshared
Of unknown quality
Siloed
Un-FAIR

But, there’s more…

Brett, A., et Al. (2020). Ocean data need a sea change to help navigate the warming world. In Nature (Vol. 582, Issue 7811, pp. 181–183) doi: 10.1038/d41586-020-01668-z

RDA 23rd Plenary Meeting

www.rd-alliance.org

7 of 28

Oceanographic Research Data Challenges

Hypothesis-driven research questions are often complex
Data are highly heterogeneous in file type, scale, and format
A single collaborative research project may involve several investigators each responsible for each data type

RDA 23rd Plenary Meeting

www.rd-alliance.org

8 of 28

Oceanographic Research Workflow

Oceanographic Research Data Challenges

Physical and Chemical

Water Properties

Organism Molecular Data

Ship-based Data (direction, speed)

Physical Samples (E.g., Sediment Cores)

RDA 23rd Plenary Meeting

www.rd-alliance.org

9 of 28

Oceanographic Research Workflow

Oceanographic Research Data Challenges

Oceanographic

Repository

(bco-dmo.org)

Biomolecular

Repository (ncbi.nim.nih.gov)

Ship-based Underway Data

Repository

(rvdata.us)

Physical Sample

Registry

(geosamples.org)

RDA 23rd Plenary Meeting

www.rd-alliance.org

10 of 28

Oceanographic Research Workflow

Oceanographic Research Data Challenges

Oceanographic

Repository

(bco-dmo.org)

Biomolecular

Repository (ncbi.nim.nih.gov)

Ship-based Underway Data

Repository

(rvdata.us)

Physical Sample

Registry

(geosamples.org)

RDA 23rd Plenary Meeting

www.rd-alliance.org

11 of 28

Role of Disciplinary Repositories

Respond to research community needs
Leverage community practices and standards
Address “I” & “R” to shorten time to science
Apply CARE Principles

demo.bco-dmo.org

https://www.gida-global.org/care

RDA 23rd Plenary Meeting

www.rd-alliance.org

12 of 28

Role of Disciplinary Repositories

Deep Disciplinary Curation:

Data Professionals with domain knowledge AND information science skills
Collaboration with the data originators
Time!

BCO-DMO Data Curators (total 5.5 FTE)

RDA 23rd Plenary Meeting

www.rd-alliance.org

13 of 28

Role of Disciplinary Repositories

Deep Disciplinary Curation:

Applying community accepted quality control practices
Harmonizing data values and formats to community norms
Augmenting files as necessary to improve reuse.
Apply authoritative vocabulary mappings
Document and share provenance
Apply licensing
Archive

Incoming data submission in Excel, multiple combined tables, undescribed color formatting, local PI terms, embedded undescribed plots…

RDA 23rd Plenary Meeting

www.rd-alliance.org

14 of 28

Role of Disciplinary Repositories

Deep Disciplinary Curation:

Applying community accepted quality control practices
Harmonizing data values and formats to community norms
Augmenting files as necessary to improve reuse.
Apply authoritative vocabulary mappings
Document and share provenance
Apply licensing
Archive

Photosynthetically Active Radiation (PAR) from

three different BCO-DMO datasets:

harmonized to standard .csv tabular format
augmented to include observational metadata (i.e., x, y, z, t) in community formats

RDA 23rd Plenary Meeting

www.rd-alliance.org

15 of 28

Role of Disciplinary Repositories

Science-on-schema

Google Dataset Search

Ocean Data & Information System

General Discovery

Community Discovery

RDA 23rd Plenary Meeting

www.rd-alliance.org

16 of 28

Role of Disciplinary Repositories

Improve Discovery and Interoperability

Notebooks (direct access, ingest)
ERDDAP - (download in variety of formats)
APIs (Rest, SPARQL endpoint)

RDA 23rd Plenary Meeting

www.rd-alliance.org

Once those data are processed, we make them accessible…

We meet researcher needs for access and interoperability - wherever they are technically - by providing access to data in a variety of ways and formats and facilitating ingest into community relevant tools of choice.

For example:

BCO-DMO has adopted ERDDAP to provide a simple, consistent way to access and download datasets in several common file formats (e.g, CSV, NetCDF, or JSON), with the capability of subsetting and creating graphs and maps. ERDDAP's web services can be used in several different applications and utilities such as Python, R, Javascript, and MATLAB.

We will soon be providing sample notebooks for use by end-users to allow them to access analyze and visualize our datasets.

API’s both restful and sparql allow more advanced users direct query and access capabilities

17 of 28

Role of Disciplinary Repositories

The Power of the PID:

Augmenting metadata, leveraging PIDs to pull in complementary data and information, improving fitness for purpose determination and understanding full context of data within the project.

DOI
ISBN
ORCiD
FundRef
Accessions

Datasets

Software

Publications

RDA 23rd Plenary Meeting

www.rd-alliance.org

18 of 28

Role of Disciplinary Repositories

In situ Environmental Datasets

Genetic Data

RDA 23rd Plenary Meeting

www.rd-alliance.org

19 of 28

Role of Disciplinary Repositories

In situ Environmental Datasets

Genetic Data

RDA 23rd Plenary Meeting

www.rd-alliance.org

20 of 28

Role of Disciplinary Repositories

In situ Environmental Datasets

Genetic Data

RDA 23rd Plenary Meeting

www.rd-alliance.org

21 of 28

Role of Disciplinary Repositories

In situ Environmental Datasets

Genetic Data

RDA 23rd Plenary Meeting

www.rd-alliance.org

22 of 28

Role of Disciplinary Repositories

In situ Environmental Datasets

Genetic Data

RDA 23rd Plenary Meeting

www.rd-alliance.org

23 of 28

Role of Disciplinary Repositories

In situ Environmental Datasets

Genetic Data

RDA 23rd Plenary Meeting

www.rd-alliance.org

24 of 28

Emerging Community Needs

O’Brien, M, et al., 2024 Earth Science Data Repositories: Implementing the CARE Principles. Data Science Journal, 23: 37, pp. 1–29. DOI: https:// doi.org/10.5334/dsj-2024-037

CARE:

Requires new infrastructure
Requires FAIR first
Will be challenging for rapid uptake

Roadmap for Repositories

RDA 23rd Plenary Meeting

www.rd-alliance.org

FAIR curation improves quality and efficiency of data use. But, when considering governance we need to include CARE: repositories can incorporate traditional knowledge into their metadata…

Aligns data use with indigenous rights:

Where Indigenous Peoples’ data comprise (1) information about the environment, lands, skies, resources, and non-humans with which they have relations; (2) information such as administrative, census, health, social, commercial, and corporate and, (3) and cultural oral histories, ancestral and clan knowledge, cultural sites, and stories, belongings.

NEED the Persistent and innovativative COMMUNITY to provide the services that meet needs

If solution science is important to address today’s global sustainability challenges, as an international organization RDA holds value as a place to collaborate inform and adopt output within our own communities.

25 of 28

Emerging Community Needs

Implementing CARE within repositories…

GEOME is a success story within oceanography, applying notices and subsequent labels, through collaboration with the community and Local Contexts.

https://localcontexts.org/about/

https://support.datacite.org/docs/local-contexts-notices-and-labels

Learn More:

RDA 23rd Plenary Meeting

www.rd-alliance.org

26 of 28

Emerging Community Needs

Implementing CARE within repositories…

GEOME is a first success story within oceanography, applying notices and subsequent labels, through collaboration with the community and Local Context.

RDA 23rd Plenary Meeting

www.rd-alliance.org

27 of 28

Linking it all together: Shortening time to science

Acknowledging indigenous rights

BCO-DMO

Repositories leverage:

Best practices for publishing data on the web
Community accepted formats and standards
Persistent Identifiers (PIDs)

Partner to access and share related content
Provide discovery and access to content related to the datasets

RDA 23rd Plenary Meeting

www.rd-alliance.org

28 of 28

THANK� YOU

RDA 23rd Plenary Meeting

www.rd-alliance.org