epsl-reporting-and-reproducibility

Reporting & Reproducibility (epsl-reporting-and-reproducibility)

The letters format pushes almost everything load-bearing but non-narrative — full analytical tables, method details, standards data, model configurations — into the supplement and the repositories. That makes the supplement and deposit part of the reviewed science, not an afterthought. EPSL sits inside Elsevier's research-data framework (data statements, repository linking), and its communities run mature FAIR infrastructure; reviewers know the canonical archive for each data type and notice absences. Re-check the current mandate wording on the official Guide for Authors.

When to trigger

• Assembling supplementary text, tables, and figures
• Writing the data-availability statement and choosing repositories
• Depositing geochemical tables, waveforms, paleomagnetic data, model code, or mission-derived products
• Making sure a reader could re-derive every figure and headline number

What to build

1. A methods supplement that stands alone. Instrumentation, chemistry/procedures, standards with measured-vs-accepted values, blank records, correction sequence, and reduction software with versions — enough that a peer lab could replicate the campaign.
2. Complete data tables. Every analysis (not just plotted ones), every correction applied, with units, σ-levels, and metadata columns (coordinates, sample IDs, stratigraphic height); machine- readable formats, not PDF-only.
3. A data-availability statement with resolvable identifiers. DOIs/accessions, not "available on request" — Elsevier's data policy and the community both treat that phrase as a red flag.
4. Code and models. Reduction scripts, inversion configurations, and model input decks deposited (Zenodo or equivalent) with versions pinned; state the random seed for stochastic steps.
5. Restricted material handled explicitly. Mission data under embargo or proprietary industry data: say what is withheld, why, and the release path.

Deposition routing: data type → community repository

Data type	Expected archive	Note
Geochemistry / isotope tables	EarthChem Library (DOI per dataset)	the NSF-backed geochemistry standard
Marine, paleoclimate, observational Earth data	PANGAEA	editor-curated, DOI-issuing, FAIR-aligned
Seismic waveforms / station products	IRIS / EarthScope services	cite network DOIs
Paleomagnetic / rock-magnetic data	MagIC	the community's designated archive
Planetary mission data / derived products	NASA PDS (or ESA PSA)	cite the archive bundle, instrument, version
Geochronology reduction files	EarthChem / Zenodo	include tracer + constants metadata
Code, model configs, notebooks	Zenodo / GitHub+Zenodo DOI	pin versions; archive the release, not the repo URL

Worked micro-example (illustrative — the deposit for the ash-bed geochronology Letter)

A reviewer-ready package for the extinction-tempo study (illustrative):

• Supplement: full CA-ID-TIMS procedure, tracer and decay constants with citations, standards table (measured vs accepted, per session), blank records, and the age-model priors.
• EarthChem: the complete grain-by-grain U-Pb table — including the two excluded inherited grains — with coordinates and bed heights, under its own DOI.
• Zenodo: the Bayesian eruption-tempo model config and scripts, seeded, versions pinned; rerunning it regenerates Figure 3 and the headline 66.021 Ma age exactly.
• Statement: "All isotopic data are available at EarthChem (DOI: illustrative); model code at Zenodo (DOI: illustrative)." One sentence, two resolvable identifiers.

The self-test before submission: hand the supplement + deposits to a colleague and ask them to re-derive the abstract's numbers. Anything they cannot re-derive is under-reported.

Referee-pushback patterns and the venue-specific fix

• "Data availability inadequate." → Replace "on request" with the discipline archive + DOI; EPSL's communities have a canonical home for nearly every data type.
• "Methods insufficient to assess the ages." → Move the full traceability chain (standards, blanks, constants, software versions) into the supplement.
• "Figures cannot be reproduced from the tables." → Deposit the reduction scripts and confirm the pipeline regenerates every exhibit.

Anti-patterns

• "Data available from the corresponding author upon reasonable request"
• Plotted analyses only, with rejected/excluded data invisible everywhere
• PDF-only data tables that cannot be re-used
• Waveforms, paleomag, or mission products parked in a generic repository when a designated community archive exists
• A supplement assembled the night before submission, drifting from the manuscript's numbers

Output format

【Supplement】methods + standards + blanks + full tables? [Y/N]
【Deposit】data type → repository + DOI/accession (list)
【Statement】written with resolvable identifiers? [Y/N]
【Code/models】archived, versioned, seeded? [Y/N or N/A]
【Re-derivable】abstract numbers regenerate from deposit? [Y/N]
【Next】epsl-writing-style

Supplementary resources

• ../../resources/external_tools.md — repositories and reproducibility tooling by data type
• ../../resources/official-source-map.md — Elsevier research-data policy links for EPSL