Earth is a slightly flattened sphere, as geodesy measures.

1. Pass Title matches the evidence

   Title 'Earth is a slightly flattened sphere, as geodesy measures.' matches the finding False and the three evidence items on gravity-field oblateness, meridian curvature and latitude-dependent gravity.

2. Pass No obvious contradictions

   Finding False, answer, summary and conclusion all state the flat-Earth claim is contradicted with High confidence; no conflicting numbers or facts found.

3. Pass Plausible justification

   Conclusion gives a clear reason for the rating: three independent geodetic methods converge on a rotating, oblate Earth, with limitations stated.

4. Pass Sources are relevant

   Sources (GFZ GRACE Follow-On, Sun/Ditmar/Riva GRACE oblateness study, BIPM metre history, IGN France geodesy history) are on topic for the evidence they support, judged by name, title and description only.

5. Pass Common questions answered

   Three common questions (satellite models, local flat horizon, early survey errors) each receive a responsive answer.

Coverage: Read the full article payload; sources judged by name/title/description/role only, per First Look's rules; no source snapshots read; no web access.

1. Concern (material) Claims are supported and labeled

   Cite sources that actually describe meridian-arc curvature results and latitude-dependent gravity; earth-3 and earth-4 cover neither, and no source supports evidence:earth-latitude-gravity.

   The article's confidence rests on 'three distinct methods' converging (summary, conclusion), but two of the three evidence lines are not supported by the sources they cite. - evidence:earth-meridian-triangulation cites earth-3 and earth-4. The earth-3 BIPM page read in full covers only the 1889 prototype, the 1960 krypton definition, the 1983 light-based definition and the 2018 rewording; it says nothing about a meridian survey, astronomical latitude, or curvature, contrary to its description ('Explains the metre's historical connection to measurement of a terrestrial meridian'). earth-4 (IGN) describes the Nouvelle Triangulation de la France (late 19th c. to 1991) and its transition to RGF93; it mentions the 'Meridienne de France' only as the first chain of the network and says nothing about latitude-distance relationships or Earth's figure. So the item's key assertion - 'The relationship between distance and change in latitude is a curvature measurement and helped establish Earth's non-flat figure' - has no cited support. - evidence:earth-latitude-gravity cites earth-1, earth-2, earth-4. None describes a latitude-dependent gravity pattern: earth-1 is a mission page, earth-2 analyses monthly C20 variations, earth-4 has no gravity content. The claim is correct in substance but rests on the author's assertion. - method:3 says the four sources were reviewed for 'the metre's geodetic history', which the earth-3 text does not contain. The underlying facts are well established, so the conclusion is not wrong, but a reader following the article's own verification checklist would find two of its three pillars unsupported, which changes how strongly the article's stated evidence base backs it. Confidence: High (all sources read in full).

   About evidence:earth-meridian-triangulation, evidence:earth-latitude-gravity, source:earth-3, source:earth-4, method:3 The finding

2. Concern (minor) Sourcing is strong and transparent

   Add a primary source stating Earth's static flattening and normal-gravity formula (e.g., GRS80/IERS reference ellipsoid) instead of relying on a study of J2 changes.

   Source types are good: an institutional mission page (earth-1), a peer-reviewed open-access paper with DOI (earth-2), and two national/intergovernmental institutional pages (earth-3, earth-4); none is partisan or anonymous. However, the only peer-reviewed source addresses month-to-month changes in J2 ('A new methodology is proposed to estimate changes in the Earth's dynamic oblateness'), not the static figure the article is about, and earth-1 is a mission description. Readily available primary references that directly state Earth's flattening and latitude-dependent normal gravity (geodetic reference ellipsoid definitions such as GRS80/WGS84, or IERS Conventions) are not used. The mismatch between earth-3/earth-4 and their claims is reported under claims-supported and not double-counted here. Confidence: Medium-High.

   About evidence:earth-satellite-gravity, source:earth-2, source:earth-1 The finding

3. Pass Central claims are accurate

   Summary: the article rates 'Earth is flat rather than a rotating, approximately spherical body' False, holding that satellite gravimetry, meridian triangulation and latitude-dependent gravity converge on a rotating, slightly oblate Earth. Claim ratings: 1) Finding False / answer 'a rotating, slightly flattened sphere' - True. Well-established, uncontested geodetic fact. 2) evidence:earth-satellite-gravity - Mostly True. earth-1 confirms GRACE-FO's objective is 'precise global and high-resolution models for both the static and the time variable components of the Earth's gravity field' via inter-satellite ranging. earth-2 estimates 'changes in the Earth's dynamic oblateness (dJ2 ...)' - it measures variations of J2, which presupposes a nonzero oblateness, but it does not itself report the static figure. The article's wording ('reports changes in Earth's dynamic oblateness') is accurate. 3) evidence:earth-meridian-triangulation - True as history (18th-century French arc measurements did establish curvature/oblateness), but Unverifiable from the cited sources (see claims-supported). 4) evidence:earth-latitude-gravity - True per standard geodesy (normal gravity rises from equator to poles), Unverifiable from the cited sources. 5) Objection responses - True and reasonable. No central claim is false or misleading; the finding points the same way as all evidence. Confidence: High.

4. Pass Context is complete

   The context and editorialNote explicitly limit scope to 'the geometric claim that Earth is flat' and to 'the stated geodetic evidence'. The article does not cover more intuitive lines of evidence (circumnavigation, lunar-eclipse shadows, time zones, horizon effects, GNSS) beyond the Apollo 17 image and an objection about the horizon, but given the declared scope this does not change a reader's impression. No contrary evidence of substance exists that is omitted. Confidence: High.

5. Pass Framing is fair and neutral

   The claim is stated plainly; objections ('Satellite results cannot count because they assume a globe in advance', 'A locally flat horizon...', 'Errors in early meridian surveys...') are presented in their strongest common form and answered without ridicule, conceding model dependence ('satellite evidence is not treated as self-sufficient'). The editorial note invites challenges. No loaded language or one-sided standards. Confidence: High.

6. Pass Uncertainty is stated

   Each evidence item has a specific limitation (model and processing dependence for GRACE; 'systematic errors' and single-arc limits for triangulation; local anomalies for gravimetry). The conclusion states 'individual measurements remain model- and instrument-dependent' and keyPoints note 'No single item is complete proof by itself.' High confidence is warranted for this uncontested fact. The independence claim between the satellite and latitude-gravity items is somewhat overstated given shared sources, but that does not change confidence. Confidence: High.

Coverage: Read the full article payload (v2) and all four cited sources via get_source_snapshot (all live copies, HTTP 200, none archived): earth-1 GFZ GRACE-FO mission page (digest sha256:9929c064...), earth-2 Sun, Ditmar & Riva 2015, J Geod, PMC full text (sha256:5b42f339...), earth-3 BIPM 'History of the SI - the metre' (sha256:590f08cf...), earth-4 IGN 'Histoire de la geodesie' (sha256:c6f3c6bf...). The resources list duplicates earth-1 and earth-2. No live web search; general geodesy knowledge used only to frame questions. The Apollo 17 image was not examined as evidence.

1. Concern (minor) Strongest objections answered

   Add an objection addressing the claim that all cited evidence depends on trusting institutions, and point to checks readers can make without trusting them.

   The objections cover model dependence ("Satellite results cannot count because they assume a globe in advance"), local flatness of the horizon, and errors in early meridian surveys. The answers rest on cross-checking between methods rather than bare assertion, which is adequate. The article does not address a common way the claim circulates: that all of the cited evidence comes from institutions and agencies (GFZ, GRACE, BIPM, IGN) the claimant already distrusts, so the evidence counts only if those bodies are trusted. The reader verification checklist (method step 5) points readers back to those same institutional sources and offers no observation a reader can check without trusting them. The article should answer whether its conclusion holds without institutional trust, for example by noting checks anyone can make independently. Its convergence argument partly blunts this objection, so a reader would likely come away with the same belief; this is a secondary gap rather than a missing strongest objection.

   About conclusion, method:5 The finding

2. Concern (minor) Evidence is independent

   Cite a gravimetry source separate from GRACE and the IGN history for earth-latitude-gravity, or describe the evidence as two independent methods rather than three.

   The item earth-latitude-gravity cites sources earth-1, earth-2 and earth-4, which are exactly the sources of earth-satellite-gravity (earth-1, earth-2) and part of earth-meridian-triangulation's (earth-4). Its institutions (GFZ Potsdam, GRACE research community, IGN France) are likewise the union of the other two items' institutions. Its summary says "The same broad figure is used in satellite gravity analysis and terrestrial geodetic reference work," so it repackages the other two items rather than offering a separate gravimetric dataset. Yet the description says the three methods "independently contradict" the flat-Earth model, and the summary counts "three distinct methods." Independent support still remains: satellite gravity (earth-1/earth-2, GFZ/GRACE) and meridian triangulation (earth-3/earth-4, BIPM/IGN) share no sources, institutions, or instruments. The overlap therefore weakens the "three independent methods" framing but not the conclusion.

   About evidence:earth-latitude-gravity, evidence:earth-satellite-gravity, evidence:earth-meridian-triangulation, source:earth-1, source:earth-2, source:earth-4, conclusion The finding

3. Pass Rates the real claim

   The claim as named is "Earth is flat rather than a rotating, approximately spherical body," which matches how the flat-Earth claim circulates: a reader who repeats it means the whole Earth is a plane, not merely that local terrain looks flat. The context limits scope to "the geometric claim that Earth is flat, not every claim about maps, satellite imagery, or institutions," and method step 1 defines it as "a comparison between a flat Earth and a rotating, curved global figure." The conclusion stays within that scope ("converge on a rotating, curved, slightly oblate Earth") and does not reach beyond the claim's time, place, or population. No strawman or inflated version is rated.

4. Pass Rating follows from the evidence

   Before reading the finding, taking the evidence as stated: satellite gravity analysis resolving Earth's "dynamic oblateness," meridian-arc triangulation showing that "the relationship between distance and change in latitude is a curvature measurement," and latitude-dependent gravity "associated with Earth's rotation and oblate figure." All of it points one way and none supports a flat plane. I would rate the claim False. The article's finding is "False," the answer ("show a rotating, slightly flattened sphere") is consistent with it, and the conclusion ("contradicted with high confidence") agrees. Nothing in the evidence argues for Mostly false/Misleading or a partially true reading.

5. Pass Confidence matches the evidence

   The article states "Confidence: High, based on converging geodetic methods; individual measurements remain model- and instrument-dependent." Each evidence item carries an explicit limitation: model dependence for GRACE, historic systematic errors and single-arc limits for triangulation, and local anomalies for gravimetry. Two genuinely different methods remain even after the overlap noted under evidence-independence: orbital gravity-field analysis (GFZ/GRACE, 2015-2026) and ground triangulation with astronomical latitude (BIPM/IGN history). That support, from different instruments and eras, is enough for High on this claim, and I would also give High. The description's "independently contradict" and the summary's "three distinct methods" slightly overstate how independent the third item is. That is handled under evidence-independence and does not change the finding's strength or the reader's confidence that Earth is not flat.

Coverage: Examined the claim, context, title, description, summary, key points, finding, answer, confidence, and conclusion. Also examined all three evidence items with their methods, institutions, years, limitations, and source IDs, the three objections and responses, the five method steps, and the four sources' names, roles, and titles. As the division of labor requires, I took the evidence exactly as stated and did not check whether the sources support it; I did not read source snapshots or use the web. Outside knowledge was used only to name the institutional-trust objection, which the article does not address. Images, disclosures, and changelog were read for context but are outside these checks.

1. Concern (minor) Verdict fields agree and stand alone

   Change "contradicted" to "false" in summary, conclusion and description to match the label, and word the claim as it circulates: "Earth is flat."

   Quoted alone: FactHub rates the claim "Earth is flat rather than a rotating, approximately spherical body." False (High confidence): Satellite and ground measurements show a rotating, slightly flattened sphere. The fields agree in substance and the answer stands alone. Two lapses under the rule: 1) Stale wording from the earlier label: the changelog records the rating was 'Contradicted' before standardization to False, yet the summary still says 'The flat-Earth claim is contradicted by three distinct methods', the conclusion opens 'The claim that Earth is flat is contradicted with high confidence', and the description says the evidence 'independently contradict a flat-Earth model'. The rule names exactly this ('a summary that still says "contradicted" after the label became False') as Minor. 2) The claim embeds its correction ('rather than a rotating, approximately spherical body'), so it is not worded the way it circulates (rule: 'the claim is the exact assertion, worded the way it circulates'). The meaning is still clear, so Minor.

   About conclusion The finding

2. Concern (minor) Each section does its job

   Delete the hand-written evidence-map sentence from context and the 'different methods, institutions, countries, and periods' keyPoint; trim summary to one sentence.

   Context ends: 'The evidence map represents three methods, institutions in Germany and France plus international metrology, and historical and modern periods.' The fourth keyPoint says 'The conclusion is strong because evidence from different methods, institutions, countries, and periods converges.' Both are hand-written independence/evidence-map notes that the page generates (rule: 'a hand-written independence note or evidence summary that repeats what the page generates' is Minor). The keyPoint is also labelled 'Important context:', i.e., context content in keyPoints. The summary is two sentences, while the rule asks for one sentence naming the strongest lines of evidence. The conclusion does synthesize (different dependencies, no single institution required, residual model dependence) and the editorialNote states scope, caveats and review status, so no Material lapse.

   About conclusion The finding

3. Concern (minor) Consistent house style

   Spell out GRACE (Gravity Recovery and Climate Experiment) and IGN at first use, explain 'dynamic oblateness', and remove the 'Fact check —' label from the start of context.

   Rule: 'an abbreviation is spelled out at first use' and 'Technical terms a general reader would not know are explained at first use.' In reader-visible text, 'GRACE Follow-On' appears in keyPoints and the satellite evidence summary without being expanded (only the source title spells it out), and 'IGN France' (source:earth-4 name) is never expanded. 'dynamic oblateness' (keyPoints, evidence summary) and 'gravimetry' are not explained, though 'slightly flattened' in the answer partly helps. Context also begins with 'Fact check —', a label that reads like template text rather than prose; it is not an instruction or placeholder, so Minor rather than Material. Units, dates (December 7, 1972) and names are otherwise consistent, American spelling throughout.

   About conclusion, evidence:earth-satellite-gravity, source:earth-4 The finding

4. Pass Title states the finding

   Title: 'Earth is a slightly flattened sphere, as geodesy measures.' It leads with what is true in positive form, names the subject (Earth) and resolves the flat-Earth claim without repeating the myth or using loaded words (rule: 'Lead with what is true... must resolve the checked claim'). seoTitle 'Is Earth Flat? Geodesy Shows a Curved, Oblate Earth' uses search wording but states the finding, which the rule allows.

5. Pass Limitations are specific

   Each limitation is item-specific: satellite gravity 'depends on satellite tracking, calibration, reference frames... GRACE also measures time-variable mass redistribution'; triangulation 'Historic instruments, baseline measurements, refraction... A single meridian arc cannot by itself determine the complete global figure'; gravimetry 'elevation, tides, geology... an isolated reading is not a shape test'. None is boilerplate.

6. Pass Tone is measured and respectful

   Language is measured ('contradicted with high confidence', 'No single method is complete proof by itself'). Objection responses are short and respectful, conceding valid points ('Orbit and gravity solutions do use models'; 'The early surveys did contain important systematic errors'). No ridicule, sarcasm or unjustified absolutes.

7. Pass Readers can verify it

   All four sources carry role labels ('Satellite geodesy mission', 'Peer-reviewed gravity analysis', 'Metrology history', 'Geodesy history'), full titles, and descriptions of why each was chosen; the peer-reviewed study has a DOI (10.1007/s00190-015-0852-y), and resources add the mission data page. Method steps cover claim definition (step 1), prioritizing independent methods (step 2), sources reviewed (step 3) and recorded limits (step 4). Whether the sources contain what is claimed is out of scope here.

Coverage: Reviewed the full article payload of version 2 only (title, seoTitle, description, claim, finding, answer, confidence, summary, keyPoints, context, conclusion, evidence and limitations, objections, method, editorialNote, sources, image caption/alt text). No sources or web consulted, per the auditor's scope. Metadata fields (dates, identifiers, URLs, changelog) excluded from house-style review.