Devesh Nandal

The Institute for Theory and Computation at Harvard-Smithsonian identifies Nandal as a Swiss National Science Foundation Postdoctoral Fellow working on next-generation stellar-evolution models for massive and supermassive stars. His research includes rotation, angular-momentum transport, stellar mass loss, dark-matter effects on stellar structure, Population III stars and predicted chemical signatures in high-redshift galaxies. [S1]

The Galileo Project gives the same conventional scientific background and states that he defines and conducts research associated with the project. [S2] His recent mainstream publications demonstrate a substantial astrophysical record independent of UAP work. In February 2026, Nandal and Avi Loeb published an Astrophysical Journal paper modelling whether supermassive stars could reproduce the unusual spectra of JWST "little red dots". [S7] In July 2026, Nandal and collaborators published an Astrophysical Journal Letters study modelling pulsational mass loss from supermassive stars and its possible connection to the compact shells inferred around little-red-dot sources. [S8][S9]

These papers are relevant to his UAP profile because they establish the type of work he normally performs: numerical modelling in which observables are compared against conditional physical models. That methodological background is transferable to UAP analysis, but it does not make every reconstruction unique or correct.

Much of astrophysics is an inverse problem. Researchers observe light, spectra or motion and attempt to infer an underlying physical system. Multiple models can sometimes reproduce the same observation.

UAP video analysis often has exactly this structure. A camera records angular movement in a two-dimensional image. The analyst may want physical speed, object size, altitude or trajectory. Those quantities cannot always be recovered uniquely because range, platform motion and viewing geometry are unknown.

Nandal's PR043 analysis is therefore closely related to his conventional scientific skill set. The central question is not whether mathematics can be applied to the clip. It can. The question is whether the released observations constrain the mathematical model tightly enough to make one physical interpretation substantially more likely than alternatives.

Nandal appeared in the initial public UAPSAC roster in June 2026 with "numerical analysis and astrophysics" assigned as his area. [S3][S4] DefenseScoop separately reported that an ODNI official confirmed the wider UAP Governance Board and its relationship with external advisory groups including UAPSAC. [S4] This gives Nandal's council role a stronger institutional basis than a private research-group appointment alone. However, the public record does not show that council members possess unrestricted access to all classified UAP information.

Early council statements instead emphasised obtaining additional unclassified data and developing future scientific collection. Nandal's public UAP work to date should therefore be judged mainly from the datasets and analyses that are actually available.

PR043 was released in May 2026 as part of the PURSUE process. The official record describes an unresolved UAP report from Djibouti/Africa in 2025 and provides a short looped infrared video. Public material does not include a witness narrative, target range, full sensor metadata or a paired incident report. [S5][S6]

The visible target is small and only briefly distinguishable. This creates a difficult measurement problem. Image displacement can be measured directly in pixels. Physical speed cannot be obtained without some model of geometry and distance.

In August 2026, Nandal and Loeb matched visible terrain in the video to an area in southern Djibouti and used ground features as control points for a photogrammetric reconstruction. [S5] They estimated a conditional sensor-platform altitude and velocity, tracked the target across frames and modelled the physical path implied by different assumed target altitudes. Their analysis concluded that a distant, fast target consistent with a tactical missile was physically possible and described a likely fast-missile interpretation.

The missile comparison was not a serial-number identification. The public clip did not contain a direct range measurement and the target itself was unresolved or only marginally resolved. The authors therefore produced a family of conditional solutions rather than one direct measured target velocity. [S5] This distinction is central.

The phrase "likely missile" can sound stronger than the public dataset warrants. Nandal and Loeb's own modelling shows that inferred physical speed changes substantially with assumed target altitude and range. [S5] A different geometry can move the solution toward a smaller, slower and more conventional nearby object.

Independent analyst Mick West subsequently published a different camera/terrain fit that produced a much slower solution, illustrating that the reconstruction is sensitive to assumptions not fixed by the public clip. UAP Globe's detailed case review records both positions and concludes that neither has been settled by the currently released evidence. [S6]

This does not make Nandal's analysis invalid. It means the result should be described as a model-supported interpretation rather than a unique measurement.

Nandal's PR043 work is significant even if the missile interpretation remains uncertain. UAP research often receives attention when an object appears impossible. A strong scientific programme must also attempt to show when an unresolved government case is compatible with ordinary technology. The Nandal–Loeb analysis asks whether the observed motion can fit the physical envelope of a tactical missile.

That is a falsifiable conventional hypothesis. If additional sensor telemetry, range or platform data become available, the model can be tested more decisively. This is preferable to leaving the case in a vague category of "unexplained" without attempting quantitative reconstruction.

The PR043 analysis also generated a secondary methodological dispute concerning the terrain geolocation used in the reconstruction. UAP Globe records that an identical coordinate had appeared in an earlier Metabunk analysis and that Nandal later stated that he had independently derived the location using Google Earth and Google Maps and had not seen the earlier forum work. [S6] Nandal also stated that he would acknowledge the earlier geolocation if the work proceeded to journal publication.

The coordinate overlap alone is insufficient evidence of plagiarism or misconduct. The relevant research lesson is simpler: when public analysis communities are independently working on the same released material, citation and provenance should be documented carefully so that readers can reconstruct which steps were independently derived and which were inherited.

As of 4 September 2026, the PR043 work appears publicly as a Galileo Project/Nandal–Loeb analysis report and Loeb explanatory article rather than as a completed peer-reviewed journal paper. [S5] This should remain explicit. Peer review would not automatically settle the target identity, but it could clarify:

• uncertainty propagation;
• sensitivity to terrain-control selection;
• range degeneracy;
• sensor-model assumptions;
• competing camera solutions. The assessment concerns the cited preprint rather than a final journal version.

Nandal's established astrophysics career is particularly useful when evaluating his UAP role because it gives a clear comparison standard. His stellar-evolution papers present model assumptions, equations, simulated outputs and comparisons to observable astronomical data. The same standard should apply to UAP reconstruction. A compelling visual match is not enough.

The analysis must expose the parameters that drive the result and show how much the conclusion changes when those parameters vary. The PR043 work moves in that direction, but the lack of range and complete telemetry remains decisive.