Simulated LSST Observations of Real Metre-scale Impactors

Using real metre-sized asteroid Earth impactors from the last decade, we ask the question: ``If the Vera C. Rubin Observatory’s Legacy Survey of Space and Time (LSST) had been operating over the last 10 years, how many imminent impactors would it have observed and discovered pre-impact?’’ We feed 216 large impacts detected by global coverage orbital sensors through the LSST Solar System Survey Simulator Sorcha, and find that it would have made 99 observations of 28 unique objects, and discovered one object four days pre-impact. Recently proposed pipeline modifications would increase this discovery number to seven. Scaling our results to take into account the biases of our reference population, we estimate that LSST will discover 12 +/- 3 imminent impactors over its nominal 10 year survey, with an average warning time of 3.5 days. While this is at the low end of previous estimates of 1 - 10 discoveries per year, the significant increase in warning time compared to the current average (9 hours across 11 impactors) will bring significant opportunities for follow-up telescopic observations, deployment of specialised equipment for fireball observations, and planetary defence operations. We also show that the LSST will provide substantial precovery data for impactors discovered by other surveys, instantly lengthening observation arcs and thereby reducing the orbital and impact location uncertainties. In some cases, these observations may also enable the linkage of telescopic observations with observed fireballs post-impact, providing valuable pre-impact astrometric and photometric data. This has significant implications for both asteroid research and planetary defence.

💡 Research Summary

This paper investigates how many imminent metre‑scale Earth impactors the Vera C. Rubin Observatory’s Legacy Survey of Space and Time (LSST) would have detected and discovered if it had been operating over the past decade. The authors assembled a reference population of real impactors recorded by U.S. Government (USG) sensors between 1 January 2015 and 31 December 2024. From an initial 372 USG fireball events, they removed those with incomplete orbital solutions or physically impossible orbits, leaving 216 objects with reliable pre‑impact orbital elements. For each object the authors estimated mass from the reported kinetic energy, assumed a bulk density of 1500 kg m⁻³, derived a diameter, and then computed an absolute magnitude using an S‑type albedo of 0.18 and the standard HG phase function (G = 0.15). Colours for the LSST filter set were taken from published S‑type colour tables.

The LSST Solar System Survey Simulator “Sorcha” was then used to generate a 10‑year ephemeris for each test object with the REBOUND N‑body integrator (IAS15) and the AS‑SIST planetary ephemerides. The simulated observing schedule employed the v5.0.1 baseline (30 s exposures in g, r, i, z, y and 38 s in u) and the exact LSSTCam focal plane geometry, including chip gaps. An observation was recorded whenever the object fell within the camera footprint and was brighter than the 5‑σ limiting magnitude for that exposure. The resulting database contains the time, filter, limiting depth, and measured trailed‑source magnitude for each detection.

The simulated detections were fed to the LSST Solar System Processing (SSP) pipeline. The current SSP links pairs of detections obtained on the same night into “tracklets”. A discovery is declared when at least three tracklets, obtained on three separate nights within a 14‑day window, can be linked into a heliocentric orbit. The authors set the linking efficiency to 0.95, reflecting the minimum efficiency required for the real survey. They also explored a modified pipeline that would immediately report tracklets containing three or more observations (rather than waiting for a second night’s pair), a change that is being discussed within the LSST community.

Results: Of the 216 injected impactors, 28 (≈13 %) were observed at least once, yielding a total of 99 detections. The mean trailed‑source magnitude was 22.1 mag; the faintest detection was r = 24.6 mag and the brightest y = 18.0 mag. Most objects (23) received 1–4 observations; five objects received five or more, with the most observed object accumulating 14 detections over a 20‑day interval. Under the current SSP rules only one object (USG 2015‑05‑10T07‑45) satisfied the discovery criteria, being observed 14 times from 21 days to 1 day before impact and linked into three tracklets within a 14‑day window, resulting in a pre‑impact discovery four days before impact.

When the modified SSP that accepts three‑plus‑observation tracklets is applied, seven objects would be flagged as discoveries, indicating a substantial increase in LSST’s pre‑impact detection capability. The authors note that the USG dataset is limited to events with impact energies > 0.05 kT TNT (≈1 m diameter) and that early‑year velocity vectors can have errors up to 10 km s⁻¹ and 90°, but post‑2018 data show markedly improved accuracy, reducing systematic bias in the simulation.

To extrapolate to the future LSST survey, the authors correct for the selection bias inherent in the USG sample (which under‑represents smaller, fainter impactors) and estimate that LSST will discover 12 ± 3 imminent impactors over its nominal 10‑year mission, corresponding to roughly 1–2 per year. The average warning time is projected to be 3.5 days, a dramatic improvement over the current 9‑hour average for the 11 impactors discovered to date. This longer lead time would enable coordinated follow‑up with ground‑based telescopes, high‑speed fireball cameras, radar facilities, and even rapid deployment of specialized instrumentation, greatly enhancing scientific return (e.g., pre‑impact photometry, spectroscopy, and size‑albedo constraints) and planetary‑defence response (e.g., impact‑location refinement, evacuation planning).

The paper also highlights LSST’s role in “precovery”: for impactors first discovered by other surveys (e.g., ATLAS, CSS, Pan‑STARRS), LSST’s archival images will often contain earlier detections, instantly extending the observational arc and reducing orbital uncertainties. In some cases this may allow a direct link between telescopic astrometry and post‑impact fireball observations, providing a rare full‑life‑cycle dataset from space‑based detection to atmospheric entry.

In summary, the study demonstrates that LSST, even with its baseline processing, will modestly increase the number of metre‑scale impactors discovered pre‑impact, but with a warning time an order of magnitude longer than today. Planned pipeline enhancements and systematic exploitation of LSST’s deep, high‑cadence imaging could raise the discovery count to several per year and substantially improve planetary‑defence readiness while delivering valuable scientific data on the physical properties of small near‑Earth objects.