Integrity Review Report: Optimization and accuracy validation of near-space 1.27 μm O2(a1Δg) dayglow forward model based on limb observation mode (DOI: 10.1080/10095020.2026.2633870)

Academic fraud report · Geng Detector

Summary

Verdict: No evidence of academic fraud was identified. This review examined a 2026 Geo-spatial Information Science paper by Guoyin Zeng et al. on a near-space O2(a1Δg) dayglow forward model. Three focus areas were assessed: (1) authenticity of the sensitivity analysis data in Table 1, where three wavelengths at 25 km showed nearly identical relative radiance changes (-3.79%, -3.79%, -3.80%) under a +10% O3 cross-section perturbation, along with strong, physically expected spectral response patterns (e.g., -5.29% to -5.32% impact on the Q branch from FWHM broadening); (2) timeline and external data source coherence, confirming that 2010 SCIAMACHY/Envisat and SABER/TIMED datasets are appropriate for validation given Envisat's 2012 decommissioning; and (3) figure reuse concerns regarding Figures 6, 7, and 9, which are consistent with legitimate data overlay operations rather than image duplication. Confidence is moderate-to-high for textual/logical checks; image-level pixel comparison was not performed due to lack of source figures. No action recommended.

Verdict

No fraud detected (清白). The paper and its evidence appear authentic. Three specific concerns raised in the review were assessed and each resolved in favor of the authors.

Key findings

Table 1 sensitivity values (e.g., -3.79%, -3.79%, -3.80% at 25 km under +10% σ_O3 perturbation) display near-linear response with sub-percent truncation differences, characteristic of a real numerical model rather than fabricated data.
The FWHM (+0.18 nm) perturbation produces a much larger negative response on the Q branch (1269.2 nm: -5.29% to -5.32%) than on the P/R branches (-0.80% to -1.33%), consistent with established spectroscopic physics and argues against manual fabrication.
Use of 2010 SCIAMACHY (Envisat) and SABER (TIMED) observations is appropriate; Envisat ceased operations in 2012, so historical archival data is the correct validation choice.
Citation of He et al. (2025) and Li et al. (2025) together with the submission date (July 2025) and acceptance date (February 2026) forms a coherent timeline.
Figures 6, 7, and 9 represent a legitimate superimposition workflow (Airglow + Scatter → Entrance pupil), not image reuse.
First author Guoyin Zeng's career stage (B.S. 2022, current Ph.D.) is consistent with publishing such a model-optimization study.

Evidence highlights

DOI: 10.1080/10095020.2026.2633870
Sensitivity numerics preserved from Table 1, Page 16: 1264.6 nm = -3.79%, 1269.2 nm = -3.79%, 1274.5 nm = -3.80% (25 km, +10% σ_O3 & Φ perturbation); FWHM +0.18 nm perturbation at Q branch: -5.29% to -5.32%.
Satellite data: SCIAMACHY (Envisat) 2010, SABER (TIMED) 2010 (Sections 4.2 / 4.2.3).
Citation window: submission July 2025; acceptance February 2026; references include 2025 publications.
Figure relationship: Figure 6 (airglow) + Figure 7 (scatter) → Figure 9 (entrance pupil), stated in the text as a superimposition of simulated airglow and scattered absorption radiance spectra.

Notes

Pixel-level image forensics were not performed; figures were evaluated only through the textual description. This is a known limitation of text-only review.
The figures are physics-simulation spectral curves generated from code, which structurally limits the "Photoshop-style" fabrication patterns often seen in biological Western blots and flow cytometry plots.
A minor suggestion (non-fraud-related) is for domain experts to verify the integration limits in the radiative-transfer equation in the appendix code; this is academic discussion, not misconduct.
This report is AI-assisted and intended for academic discussion; final determinations of misconduct require investigation by the responsible institution.