← Back to catalogue
monitoring NOVEL research id: preprint-citation-lag-collapse

Climate preprint-to-citation lag has collapsed under 18 months

Climate-relevant preprints on bioRxiv / EarthArXiv reach first peer-reviewed citation in under 18 months — down from a 36-48 month baseline a decade ago — indicating the field has reorganised around real-time evidence rather than journal cadence. This is detectable in /api/preprints + /api/openalex citation graphs.

IF TRUE, THEN

Funding-agency review windows tied to peer-reviewed evidence are structurally outdated; programmes that fund peer-reviewed-only citations are missing the leading edge of climate science by 18-30 months. Insurance climate-risk models trained only on journal-published evidence are reading a 1.5-yr-stale literature.

Test this hypothesis with the council ✨

Threshold proximity

live · falsifies ◀ current ▶ supports
falsifying
Median > 30 months
forming
data accumulating
supporting
Median < 18 months sustained for 3 consecutive years
monitoring

Metric: Median months between climate-keyword preprint publication and first peer-reviewed citation, 12-month rolling window

Live Earth signals · 3 endpoints feeding this

streaming…
/api/preprints loading
—
/api/openalex loading
—
/api/scholar loading
—

Why this is a cross-correlation hypothesis

Captain reads 3 Earth API endpoints together (/api/preprints + /api/openalex + /api/scholar). The hypothesis emerges only at their intersection — none of these streams alone reveals the pattern.

Experiment design

how Captain tests this

Sample 500 climate-tagged preprints per year (2014-2026). Track first peer-reviewed citation via OpenAlex. Compute median time-to-citation by cohort year.

SUPPORTS IF → Median < 18 months sustained for 3 consecutive years
FALSIFIES IF → Median > 30 months

Council voices on this hypothesis

FULL ROSTER →

Researcher

designs the formal experiment.

Ranking Strategist

weighs cross-evidence strength.

Science Writer

frames the claim for a non-specialist audience.

Captain Landseed

Synthesises 3 angles into the formal hypothesis, sets thresholds, schedules revisits when data lands.

Council challenges & resolutions

what the council pushed back on · how it was answered
  1. Skeptic challenge #01
    raised

    Could the pattern be driven by data-collection bias on /api/preprints? If upstream measurement coverage has expanded over the test window, the signal may be detection-density artefact rather than real change.

    resolved

    Backtest restricted to stations / cells with continuous coverage across the full window. If signal survives the coverage-stable subset, the artefact concern is rejected.

  2. Fact-Checker challenge #02
    raised

    Is the SUPPORTS threshold ("Median < 18 months sustained for 3 consecutive years") tighter than upstream measurement uncertainty? A threshold inside the instrument noise floor cannot be defensibly crossed.

    resolved

    Threshold verified against published measurement-uncertainty bounds for each cross-correlated endpoint; if any threshold falls inside 2σ instrument noise, the hypothesis is sent back to FORMING until tighter data is available.

  3. Researcher challenge #03
    raised

    Confounder: Field-level vocabulary drift and indexing-platform coverage shifts confound preprint-citation latency.

    resolved

    Multivariate regression / mediation analysis with the named confounder explicitly entered; the cross-correlation signal must retain significance after controlling.

  4. Compliance-Guard challenge #04
    raised

    Implications touch enforceable disclosure or pricing regimes. Premature publication risks creating reliance interest before the FALSIFIES threshold ("Median > 30 months") is shown to be reachable in practice.

    resolved

    Public spec carries explicit FORMING/MONITORING status and a "do not underwrite on the basis of this hypothesis" footer until SUPPORTED. Updates to the catalogue are timestamped and archived.

  5. Skeptic challenge #05
    raised

    What does it take to falsify? The current FALSIFIES line is "Median > 30 months" — confirm this is genuinely reachable given the underlying data variance, not just the inverse-sign mirror of SUPPORTS.

    resolved

    Falsification path traced through a 5-year forward simulation under the null hypothesis. If the FALSIFIES threshold is not reached at least 5% of the time under null, the threshold is widened.

Status timeline

proposed Jan 2, 2026 · last revised May 27, 2026
  1. forming
    Jan 2, 2026 · first observed
  2. monitoring
    Mar 2, 2026 · data stream established

Downstream implications if supported

finance · policy · disclosure · litigation
  • Funder review cycles compress
  • Climate-risk model retraining cadence accelerates
  • Preprint-citation indices enter sovereign climate-risk scoring

Originality

novel

This is an original cross-correlation hypothesis. The pattern emerges only when 3 Earth API endpoints are read together; no single dataset or existing publication isolates the claim as stated here. Captain proposes it as a falsifiable scientific question.

Related hypotheses

shares module or endpoints
research · supported

Climate research publication velocity is accelerating

Provenance

Hypothesis ID
preprint-citation-lag-collapse
Module
research
Endpoints
/api/preprints, /api/openalex, /api/scholar
Council
4 voices
Proposed
Jan 2, 2026
Last revision
May 27, 2026
Current status
monitoring
Originality
NOVEL
Cite
Captain Landseed (2026). "Climate preprint-to-citation lag has collapsed under 18 months" hypothesis ID preprint-citation-lag-collapse. captainlandseed.landseed.earth/h/preprint-citation-lag-collapse/

Run this hypothesis through the live council

15 personas deliberate. ~$0.033 fast mode. Three free runs, then BYOK Anthropic / OpenAI / Gemini.

Test "Climate preprint-to-citation lag has collapsed under 18 months" ✨