Geng Integrity Report: Wang Mingxing (2026) 'Vegetation Coverage Change and Driving Mechanism in Lianyungang, Jiangsu Province'

Academic fraud report · Geng Detector

Summary

Verdict: CLEAR (no academic fraud detected). The Geng review examined the paper published in Beijing Surveying and Mapping (DOI: 10.19580/j.cnki.1007-3000.2026040117) and found no indicators of data fabrication, image manipulation, or logical inconsistency. Key confirmations: (1) Tabular data are internally consistent — trend areas in Table 3 sum to exactly 7626.00 km², matching the study area stated in Section 1.1, with percentages summing to 100.00%; (2) the land-use transfer matrix in Table 5 cross-validates with only ~0.01 km² rounding error, and reported contribution rates (e.g., 59.33% = 360.30/607.24) are arithmetically correct; (3) the authors honestly report a near-zero R² (0.0016) for the long-term NDVI trend, declining to inflate statistical significance; (4) submission (2026-04-27) and online publication (2026-06-18) dates align plausibly with citations to early-2026 literature. Limits of the review: no image-level analysis was possible (Figures 1–6 not provided as pixel data), so duplication, rotation, or splicing cannot be ruled out. Confidence in the textual/numerical findings is high.

Verdict

CLEAR — No evidence of academic fraud. The paper passes consistency, arithmetic, and timeline checks.

Key findings

Arithmetic self-consistency (Table 3): Trend areas sum to 7626.00 km² (1741.97 + 2161.72 + 1344.24 + 1277.43 + 1100.64), matching exactly the study-area size of 7626 km² stated in Section 1.1; percentages sum to 100.00% (22.84% + 28.35% + 17.63% + 16.75% + 14.43%).
Transfer-matrix integrity (Table 5): Row/column cross-totals are correct within ~0.01 km² rounding tolerance; the contribution rate 360.30/607.24 = 59.33% (耕地 cropland contribution) is numerically accurate.
Honest reporting of weak trends: Authors report R² = 0.0016 (0.16×10⁻²) with an annual change of 0.7×10⁻⁴/a and explicitly state the long-term trend is not significant — a counter-fabrication pattern rather than inflated statistics.
Timeline plausibility: Submission 2026-04-27; online first 2026-06-18; references [2], [22], [24] cite January–March 2026 works, consistent with pre-submission literature access. Datasets used (MOD13Q1; Peng et al. 2019 meteorological data) predate the 2000–2024 study window.
Image-based checks: Not performed — only text and tabular excerpts were available, so duplication, splicing, or PS artefacts in Figures 1–6 cannot be assessed.

Evidence highlights

DOI: 10.19580/j.cnki.1007-3000.2026040117
Table 3 area sum: 7626.00 km² ≡ study-area size
Table 5: row/column cross-validation error ≤ 0.01 km²
Reported R² = 0.0016 with explicit "not significant" conclusion
Cropland contribution: 360.30 / 607.24 = 59.33% (verified)
Reference window: 2026-01 to 2026-03 citations; submission 2026-04-27

Notes

The review is limited to textual and tabular evidence. A 100% integrity assessment would require the original TIFFs/PDFs of Figures 1–6 for pixel-level duplication and splicing analysis.
The reviewer recommends retaining this baseline and, if warranted, performing image forensics on Figures 2–6 before final publication-stage certification.
As with all automated checks, false negatives cannot be excluded.