Enabling FAIR data in Earth and environmental science with community-centric (meta)data reporting formats

The study addresses the challenge of making Earth and environmental science data more Findable, Accessible, Interoperable, and Reusable (FAIR). While repositories and search engines have improved preservation and access, interoperability and reuse remain difficult due to data heterogeneity and limited community resources for data management. The authors propose community-centric reporting formats as practical, domain-informed tools to harmonize (meta)data and reduce barriers to synthesis and reuse. They present 11 reporting formats spanning cross-domain elements (dataset, file-level, CSV, samples, terrestrial model archiving, location) and domain-specific data (amplicon tables, leaf-level gas exchange, soil respiration, water/soil chemistry, hydrologic monitoring). The goal is to standardize key variables and metadata to facilitate consistent data organization, archiving, and reuse in interdisciplinary Environmental System Science research, and to offer a process model for communities to develop and adopt such formats.

The paper situates its contribution within existing efforts on data and metadata standards, noting the role of formal standards (e.g., ISO 8601 for timestamps, OGC Sensor Observation Service) and community reporting formats in specific domains (e.g., FLUXNET, marine observations, solid Earth geoscience). The authors reviewed 112 existing standards and resources across agencies and organizations (e.g., EML, JSON-LD, CF conventions, FGDC standards, WQX, EarthChem, Genomic Standards Consortium MIxS/MIMARKS, CUAHSI ODM, NEON, AmeriFlux conventions). They highlight gaps in coverage and suitability for ESS needs, complexity of some standards, and the need for pragmatic, scientist-friendly formats that can map to existing standards while supporting community workflows.

The authors implemented a community-centric development process to design reporting formats for common Earth science (meta)data. Key steps included: 1) comprehensive review of existing standards and resources relevant to each data type; 2) constructing crosswalks to map variables, terms, and metadata elements across standards, identifying gaps and harmonization opportunities; 3) iterative development of templates and documentation with feedback from prospective users via pilots, workshops, and open review; 4) defining minimum required and optional (meta)data fields to balance low barrier adoption with machine-actionability; 5) harmonizing conventions across formats (e.g., date as YYYY-MM-DD; latitude/longitude in decimal degrees; CSV structural guidance); and 6) publishing and versioning the formats across ESS-DIVE (dataset publications for citation and preservation), GitHub (ongoing edits, issues, version control), and GitBook (rendered documentation). Methods per format: - Dataset metadata: aligned with EML and JSON-LD to ensure findability/citation and interoperability. - File-level metadata: synthesized practices from six organizations to describe file contents within datasets. - CSV guidelines: domain-agnostic rules to improve tabular data machine-readability (e.g., no mixed types, explicit missing value conventions). - Sample IDs and metadata: aligned with IGSN to support globally unique sample identifiers and consistent sample tracking, incorporating biological and environmental sample interoperability considerations. - Terrestrial model data archiving: community guidelines to determine what model components and artifacts to preserve, with focus on usability and longevity. - Location metadata: generalized guidance from CF conventions, FGDC, and OGC to describe research locations across projects. - Amplicon abundance tables: requirements for sequencing and bioinformatic processing metadata mapped to Genomic Standards Consortium specifications; CSV structural adherence. - Leaf-level gas exchange: variable naming based on instrument outputs and plant trait databases; modular templates for raw/processed data and experimental metadata. - Soil respiration: integrated nine existing guidelines addressing various gas exchange data types and timestamp conventions. - Sample-based water and soil chemistry: harmonized chemical concentration reporting drawing from WQX and EarthChem while optimizing for scientific lab outputs. - Hydrologic monitoring (water level/sonde): harmonized key variables (e.g., level, temperature, pH) and site/time metadata across commonly used community standards. Community engagement involved 247 individuals from 128 institutions providing input; consensus building emphasized usability for field personnel, lab analysts, data managers, and modelers.

- Reviewed 112 pre-existing standards/resources; none fully met ESS community needs, necessitating development of all 11 formats. - Produced 11 reporting formats: 6 cross-domain (dataset metadata, file-level metadata, CSV, sample metadata/IDs, terrestrial model data archiving, location metadata) and 5 domain-specific (amplicon abundance tables, leaf-level gas exchange, soil respiration, sample-based water and soil chemistry, hydrologic monitoring). - Harmonized key conventions across formats: dates in YYYY-MM-DD; spatial variables as latitude/longitude in decimal degrees with bounds; CSV structural norms; optional linkage to IGSN for samples. - Ensured machine-actionability via minimal required fields and data dictionaries, while retaining low adoption barriers. - Multi-platform dissemination: each format is archived as an ESS-DIVE dataset (citable), maintained on GitHub (versioning and feedback), and rendered on GitBook (accessible documentation). - Community participation: 247 individuals from 128 institutions contributed to iterative development and testing. - Demonstrated immediate uptake: multiple datasets in ESS-DIVE adopted one or more formats (e.g., FTICR/NPOC/TN datasets, soil enzyme kinetics, leaf gas exchange, wildfire stream studies), enabling automated metadata quality checks, DOI assignment, and discoverability via DataONE, OSTI, DataCite, and Google Dataset Search. - Alignment with FAIR principles: enhanced findability (identifiers, searchable metadata), accessibility (repository and APIs), interoperability (crosswalks to standards), and reusability (consistent templates and required metadata).

The reporting formats directly address the interoperability and reuse challenges in Earth and environmental sciences by providing pragmatic, community-developed templates that integrate with scientific workflows. Benefits include better organization of field/lab data, reduced time for cross-study synthesis, and enabling value-added services (e.g., QA/QC, processing pipelines, automated metadata validation). Case examples (AmeriFlux FP-in, Watershed Function SFA, ESS-DIVE automated checks and DOIs) illustrate how adoption yields improved curation and discoverability. The authors balanced FAIR goals with usability, adopting globally unique identifiers for samples (IGSN) and mappable metadata while simplifying terminology and reducing template complexity for time-limited researchers. Crosswalks facilitate compatibility with existing standards and future tool development. Incentives for adoption include community ambassadorship by domain scientists, webinars, tutorials, and early demonstration datasets. The work charts a path toward further automation (format validators, unit conversion tools, parsing and integration services) to make datasets fully searchable and more machine-actionable, expanding the impact across repositories and platforms.

This work delivers 11 community-centric (meta)data reporting formats spanning cross-domain and domain-specific Earth science data, alongside a replicable process for community development, documentation, and dissemination. The formats enhance FAIRness by standardizing essential variables and metadata, improving findability, interoperability, and reuse in interdisciplinary research. Future directions include automated format validation, conversion tools from instrument outputs, repository-side data integration and fusion services, and parsers enabling advanced queries within files to achieve fuller machine-actionability and searchability across datasets and systems.

- Full FAIR machine-readability is not yet achieved for all formats; many rely on human-readable data dictionaries rather than formal ontologies. - Some existing standards are complex and require significant learning; the authors intentionally simplified terminology, which may reduce direct one-to-one mapping to formal schemas. - Current ESS-DIVE checks include manual validation; automated format checkers are still under development. - Domain coverage is limited to 11 data types prioritized by the ESS community; additional formats will be needed to cover other data types. - Adoption depends on community incentives and resources; limited time and funding for data management may constrain uptake.