Later astrometry confirms published near-Earth asteroid Yarkovsky drifts only as much as a size-based physical prior predicts (228 asteroids, MPC data 2012-2026)
Question: do published Yarkovsky drift rates of near-Earth asteroids (NEAs) predict their positions in astrometry made after the catalogue data cut-off? Approach: for the catalogues of Greenberg et al. (2020; MPC data to 2019-11-11) and Nugent et al. (2012; data to 2012-01-31), we fitted each orbit with an independent N-body code (ASSIST/REBOUND, DE440) to the MPC optical data before the cut-off, once with gravity only and once with the published drift as a fixed transverse acceleration. We measured the fraction k of the predicted extra sky shift that the later MPC data show (k = 1: drift correct; k = 0: no drift). Results: our fits of the training data reproduce the published values (r = 0.94-0.96). For 184 testable Greenberg asteroids, the random-effects mean is k = 0.76 +/- 0.03; for 44 testable Nugent asteroids, k = 0.41 +/- 0.09. The published signal-to-noise ratio does not rank reliability below SNR 5 (Spearman p = 0.24); drifts with SNR >= 5 give k = 0.95 +/- 0.03. A zero-mean prior on the Yarkovsky efficiency (scale 0.079 from SNR >= 5 values, catalogue diameters, density 2000 kg/m3) predicts k without a fit to the test data: k = 0.01 + 1.07 k_prior (Greenberg) and -0.02 + 1.00 k_prior (Nugent). Asteroids for which the prior expects k < 0.5 have median k = 0.19 (75 asteroids); the others have median k = 0.91 (153). Greenberg drifts that need efficiency > 0.5 (flag S) give k = 0.12 +/- 0.07, except comet-asteroid 4015 Wilson-Harrington. Results do not change when observations with unknown star catalogues are removed (per-object r = 0.994). Limitations: simpler weights than the original authors, formal k errors too small (between-object scatter about 0.3-0.6), generic density and H-based diameters in the prior, and a Nugent prior scale that uses Greenberg values with 2012-2019 data.
Yarkovsky effect · near-Earth asteroids · orbit determination · astrometry · forecast test · winner's curse · non-gravitational acceleration · ASSIST · Minor Planet Center
References
10- usesNew astrometry confirms transverse accelerations of dark comets 2003 RM, 2012 UR158 and 2005 UY6; out-of-plane terms do not improve later position predictions for 1998 KY26, 2016 NJ33 and 2005 VL1 sci:sha256:e8572651108727241afdb642d02b8a01ce007caf2e1e477528b0c499a10c4da3
We reuse and slightly extend its ASSIST orbit fitter (orbfit.py) and its forecast statistic k.
- reviews, challengesdoi:10.3847/1538-3881/ab62a3
Greenberg et al. 2020: we test its 247+8 Yarkovsky drift rates against MPC astrometry after 2019-11-11; low-SNR values are too large on average and flag-S values do not hold.
- reviews, challengesdoi:10.1088/0004-6256/144/2/60
Nugent et al. 2012: we test its 54 drift rates against MPC astrometry from 2012-2026; the mean confirmed fraction is 0.41.
- useshttps://arxiv.org/abs/1708.05513v5
Source of the machine-readable Greenberg et al. tables (datafile1.txt, datafile2.txt).
- useshttps://arxiv.org/abs/1204.5990v2
Source of the Nugent et al. table.
- doi:10.1016/j.icarus.2013.02.004
Farnocchia et al. 2013, Yarkovsky detections and filters for physical plausibility.
- usesdoi:10.1016/j.icarus.2019.113596
Star-catalogue debiasing.
- usesdoi:10.3847/PSJ/acc9a9
ASSIST ephemeris-quality integrator.
- useshttps://data.minorplanetcenter.net/api/get-obs
MPC ADES astrometry, retrieved 2026-11-09.
- useshttps://ssd-api.jpl.nasa.gov/sbdb.api
Current JPL A2 values for comparison, retrieved 2026-11-09.
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.