https://njuastrometry.github.io/en/author/mu-zi-li/Mu-Zi LiHugo Blox Builder (https://hugoblox.com)en-USWed, 01 Jul 2026 00:00:00 +0000https://njuastrometry.github.io/media/icon_hu1909758375025618910.pnghttps://njuastrometry.github.io/en/author/mu-zi-li/https://njuastrometry.github.io/en/publication/2026-universe-12-197/Wed, 01 Jul 2026 00:00:00 +0000https://njuastrometry.github.io/en/publication/2026-universe-12-197/<p><strong>Motivation and approach</strong></p> <p>Stellar-aberration-based navigation (StarNAV) estimates spacecraft velocity from velocity-induced changes in the apparent angular separations between stars. Achieving metre-per-second velocity sensitivity requires milliarcsecond-level angular measurements, so uncertainty in the transformation from focal-plane coordinates to celestial directions must be assessed alongside sensor noise and astrometric catalog errors.</p> <p>We select 477,502 Gaia DR3 stars with $G&lt;10$ mag and propagate their positions and covariances to J2026.0. Using HEALPix with $N_{\mathrm{side}}=64$, we divide the sky into 49,152 fields of approximately $0.84,\mathrm{deg}^{2}$ each, representing the adopted approximately one-degree field scale. The median field contains about eight reference stars. We compare nine plate models to quantify the effects of model complexity, reference-star density, and spatial distribution on plate-solution uncertainty.</p> <p><strong>Main results</strong></p> <ul> <li>Under the plate-constant-variance metric, the four-parameter linear model is the most stable option for the adopted sparse fields. The median propagated positional uncertainty is <strong>0.95 mas</strong>, and the 95th percentile is <strong>1.7 mas</strong>.</li> <li>More complex models are more sensitive to limited reference-star counts and uneven spatial distributions. Additional distortion terms can increase variance or cause poorly conditioned solutions.</li> <li>Using the first-order scaling $\delta v \sim c,\delta\theta$, the reported angular-uncertainty range of <strong>0.6–1.7 mas</strong> corresponds to an approximate velocity-error scale of <strong>0.9–2.5 m/s</strong>, with a median of about <strong>1.4 m/s</strong>.</li> <li>Low plate-constant variance does not guarantee adequate correction of real optical distortion. A linear model can leave nonlinear distortion biases that are not included in the variance calculation.</li> </ul> <p>The velocity conversion is an order-of-magnitude estimate of the plate-solution contribution, not a prediction or lower bound for the complete navigation solution. Future StarNAV systems should jointly account for plate-solution uncertainty, residual distortion bias, measurement geometry, sensor noise, and catalog errors.</p>