# Citation Audit — DOWNMASS-01

**Candidate:** DOWNMASS-01
**Artifact audited:** `dissertation.md` reference list (entries 1-89)
**Method:** Crossref `api.crossref.org/works/<doi>` resolution for every DOI; DataCite `api.datacite.org/dois/<doi>` + `doi.org` HEAD cross-check for every Crossref-404; ARGOS `near_miss_resolver.py` run on every Crossref-404; live GET for the one non-DOI repository URL.
**Date:** 2026-06-18
**Disposition:** READ-ONLY. No change was made to the dissertation. This file is a repair worklist.


## Header Summary

| Metric | Count |
|---|---|
| Total references | 89 |
| Of which carry a DOI | 88 |
| Of which carry a non-DOI URL (ref 25) | 1 |
| **Resolved (Crossref 200)** | **80** |
| **Unverifiable** (Crossref 404 but confirmed REAL at DataCite / doi.org) | **8** |
| **Flagged-dead** (confirmed 404/410 with no real registration anywhere) | **0** |
| Non-DOI URL reachable (ref 25) | 1 (reachable, HTTP 200) |

### Counts by classification (real defects only)

| Classification | Count |
|---|---|
| fabricated_neighbor / not_found (fabricated) | **0** |
| mistyped_recoverable (real defect) | **0** |
| misattributed (real defect) | **0** |
| **Total real defects** | **0** |

**Bottom line: no fabricated, dead, mistyped, or misattributed citation was found.** Every one of the 88 DOIs is a real, correctly-cited identifier. 80 are registered in Crossref; the other 8 are registered in DataCite (arXiv preprints, two university repositories, and the DLR repository) and are therefore invisible to a Crossref-only probe but fully real. Per the audit guardrail, a Crossref 404 that is confirmed registered at another registry is **unverifiable-by-Crossref, never dead.** The single non-DOI URL (ref 25, Texas Tech institutional repository) is live.


## Guardrail Note — why the 8 Crossref-404 entries are NOT defects

The ARGOS `near_miss_resolver.py` resolves DOIs against **Crossref only.** arXiv DOIs (`10.48550/arxiv.*`) and institutional-repository DOIs are registered at **DataCite,** not Crossref, so the resolver returns HTTP 404 for them and then falls through to a Crossref title search, which it (mis)labels `mistyped_recoverable` or `misattributed`. That label is a false positive of the Crossref-only tool, not a real defect. I verified each of the 8 directly against the DataCite registry: in every case the DataCite record exists and its title matches the cited title. The resolver's suggested "topic neighbors" are off-topic or are merely the later journal-published version of the same paper, never a correction of a wrong identifier. The cited DOIs are correct as written.

This is exactly the case the standing guardrail anticipates: do not flag a real, registry-resolvable DOI as fabricated because one registry's API does not hold it.


## Table of every flagged reference (all UNVERIFIABLE-by-Crossref, all confirmed REAL)

| Ref | Author-year | Cited DOI | Crossref | Classification | Suggested fix |
|---|---|---|---|---|---|
| 23 | Rao et al. 2023, "OPUS: An integrated assessment model for satellites and orbital debris" | `10.48550/arxiv.2309.10252` | 404 | unverifiable (real @ DataCite) | **No fix needed.** DataCite confirms title "OPUS: An Integrated Assessment Model for Satellites and Orbital Debris", publisher arXiv; `doi.org` -> arxiv.org/abs/2309.10252. DOI is correct. |
| 47 | Callaway, Goodman-Bacon & Sant'Anna 2021, "Difference-in-Differences with a Continuous Treatment" | `10.48550/arxiv.2107.02637` | 404 | unverifiable (real @ DataCite) | **No fix needed.** DataCite title matches exactly (arXiv). Optional: a journal/working-paper version now exists in Crossref (`10.3386/w32117`, NBER 2024); the arXiv DOI as cited is valid. |
| 48 | Butts 2021, "Difference-in-Differences Estimation with Spatial Spillovers" | `10.48550/arxiv.2105.03737` | 404 | unverifiable (real @ DataCite) | **No fix needed.** DataCite title matches exactly (arXiv). Resolver's Crossref "neighbors" are unrelated China land-use papers; ignore. DOI is correct. |
| 49 | Roth 2024, "Interpreting Event-Studies from Recent Difference-in-Differences Methods" | `10.48550/arxiv.2401.12309` | 404 | unverifiable (real @ DataCite) | **No fix needed.** DataCite title matches exactly (arXiv). Optional: journal version exists (`10.1007/s42973-026-00235-x`, Japanese Economic Review 2026); arXiv DOI as cited is valid. |
| 62 | Hobe 2026, "The Cologne Manual on Space Traffic Management (CM-STM)" | `10.18716/cm25` | 404 | unverifiable (real @ DataCite) | **No fix needed.** DataCite confirms the full CM-STM title, publisher "Institute of Air Law, Space Law and Cyber Law, University of Cologne; DLR"; `doi.org` -> CM-STM.pdf. DOI is correct (matches Appendix C note that it was DataCite-confirmed). |
| 78 | Tüllmann et al. 2018, "On the Implementation of a European Space Traffic Management System I. A White Paper" | `10.7892/boris.111870` | 404 | unverifiable (real @ DataCite) | **No fix needed.** DataCite confirms the exact title, publisher "University of Bern" (BORIS); `doi.org` -> boris-portal.unibe.ch record. Resolver's Crossref "neighbors" (e.g. a colorectal-surgery manual) are noise; ignore. DOI is correct. |
| 80 | Kaltenhäuser et al. 2026, "Uncontrolled Re-entry Risk for Aviation and the benefits of real time information services" | `10.60575/e98z-c874` | 404 | unverifiable (real @ DataCite) | **No fix needed.** DataCite confirms the exact title, publisher DLR. `doi.org` HEAD returns 401 (DLR landing-page auth wall), which is a publisher access control, not a dead DOI — DataCite registration is the authority. DOI is correct (matches Appendix C note that it was DataCite-confirmed). |
| 84 | Ashruf et al. 2024, "Deciphering Solar Cycle Influence on Long-Term Orbital Deterioration..." | `10.48550/arxiv.2405.08837` | 404 | unverifiable (real @ DataCite) | **No fix needed.** DataCite title matches exactly (arXiv); `doi.org` -> arxiv.org/abs/2405.08837. Optional: a journal version now exists (`10.3389/fspas.2026.1797886`, Frontiers 2026); arXiv DOI as cited is valid. |


## Resolver evidence (verbatim, for the record)

The ARGOS resolver was run on all 8 Crossref-404 DOIs. Its Crossref-only verdicts and the independent DataCite ground-truth are reconciled below. The resolver verdict is recorded for transparency; the DataCite check is the binding determination.

- **Ref 23** — resolver: `misattributed` (note: "real papers on this topic exist under different DOIs (e.g. '1 Orbital Debris', 10.1007/978-1-4614-1830-6_3)"). DataCite: REAL, title "OPUS: An Integrated Assessment Model for Satellites and Orbital Debris", arXiv. **Resolver false positive; DOI correct.**
- **Ref 47** — resolver: `mistyped_recoverable` (similarity 1.00 to NBER `10.3386/w32117`). DataCite: REAL, exact-title arXiv preprint. **DOI correct; NBER is a later version, not a correction.**
- **Ref 48** — resolver: `misattributed` (neighbors are China green-productivity papers). DataCite: REAL, exact-title arXiv preprint. **Resolver false positive; DOI correct.**
- **Ref 49** — resolver: `mistyped_recoverable` (similarity 1.00 to Japanese Economic Review `10.1007/s42973-026-00235-x`). DataCite: REAL, exact-title arXiv preprint. **DOI correct; the JER entry is a later version.**
- **Ref 62** — resolver: `mistyped_recoverable` (similarity 0.69 to IAF paper `10.52202/083086-0055`). DataCite: REAL, exact CM-STM title, Cologne/DLR. **DOI correct; the IAF item is a different paper about the manual, not the manual.**
- **Ref 78** — resolver: `misattributed` (neighbors include a "?" EU-grant record and a colorectal manual). DataCite: REAL, exact title, University of Bern. **Resolver false positive; DOI correct.**
- **Ref 80** — resolver: `misattributed` (neighbors include an EGU abstract and a recidivism thesis). DataCite: REAL, exact title, DLR. **Resolver false positive; DOI correct.**
- **Ref 84** — resolver: `mistyped_recoverable` (similarity 0.95 to Frontiers `10.3389/fspas.2026.1797886`). DataCite: REAL, exact-title arXiv preprint. **DOI correct; the Frontiers entry is the journal version.**


## Advisory (optional, not a defect)

Three preprints now have peer-reviewed journal versions in Crossref that the candidate could optionally cite alongside or in place of the arXiv DOI, if the committee prefers the version of record. None of these are corrections; the arXiv DOIs are valid as cited.

- Ref 47 -> NBER WP `10.3386/w32117` (2024)
- Ref 49 -> Japanese Economic Review `10.1007/s42973-026-00235-x` (2026)
- Ref 84 -> Frontiers in Astronomy and Space Sciences `10.3389/fspas.2026.1797886` (2026)

This is consistent with Appendix C of the dissertation, which already documents that refs 62 and 80 were DataCite-confirmed and that IEEE/arXiv/Springer DOIs returning non-Crossref status codes were registered. The audit corroborates that disclosure: the reference list is clean.