Simulated LSST Observations of Real Metre-scale Imminent Impactors

Hadrien A. R. Devillepoix; Michael A. Frazer; Sophie E. Deam

arxiv: 2602.06314 · v2 · pith:QUKFUSWEnew · submitted 2026-02-06 · 🌌 astro-ph.EP · astro-ph.IM

Simulated LSST Observations of Real Metre-scale Imminent Impactors

Michael A. Frazer , Hadrien A. R. Devillepoix , Sophie E. Deam This is my paper

Pith reviewed 2026-05-16 07:01 UTC · model grok-4.3

classification 🌌 astro-ph.EP astro-ph.IM

keywords LSSTimminent impactorsfireballsplanetary defensenear-Earth objectsprecoverylinking algorithmsSorcha simulator

0 comments

The pith

LSST simulations with real fireballs predict eight imminent impactors discovered four days before impact over a 10-year survey.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The authors take 216 real metre-scale fireballs recorded over the past decade and feed them through the LSST Solar System Survey Simulator to ask how many would have been spotted before hitting Earth. The survey would have observed 30 of the objects, most with only two to four visits. Standard linking catches just two of them pre-impact, but a modified algorithm tuned for fast movers raises the count to eight. This yields a forecast of 8 ± 2 discoveries across the full survey, with an average four-day warning instead of the current few hours. The earlier notice would allow telescopic follow-up, better planetary-defense coordination, and precovery that lengthens arcs for objects found by other surveys.

Core claim

Using a population of 216 real fireballs observed by orbital sensors over the last decade and the LSST Solar System Survey Simulator Sorcha, the study finds that LSST would have observed 30 of these objects pre-impact. The default linking algorithm would have discovered only two pre-impact, but a modified linking algorithm suited to fast-moving objects raises the number to eight. This leads to the prediction that LSST will discover 8 ± 2 imminent impactors over its nominal 10-year survey, with an average discovery time of four days before impact.

What carries the argument

The LSST Solar System Survey Simulator Sorcha run on 216 real fireballs from the past decade, together with a modified linking algorithm optimized for fast-moving objects.

If this is right

Significant opportunities for telescopic follow-up, targeted fireball observations, planetary defence planning, and public engagement.
Potential for precovery of impactors observed by LSST but discovered by other surveys, instantly lengthening observation arcs and reducing orbital and impact-location uncertainties.
In some cases, linkage of LSST telescopic observations with post-impact fireballs to supply pre-impact astrometric and photometric data.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Longer observation arcs from precovery could shrink impact-location uncertainties enough to improve evacuation planning for the few events that do strike.
If the modified linking algorithm works on real LSST data, similar fast-mover pipelines could be adopted by other wide-field surveys.
The lower discovery rate relative to some prior estimates may indicate that the metre-scale near-Earth population is sparser than current models assume.

Load-bearing premise

The sample of 216 real fireballs from the last decade is representative of the metre-scale impactor population LSST will encounter.

What would settle it

The actual number of metre-scale imminent impactors discovered pre-impact by LSST during its first decade of operation, if it falls well outside the range 6–10.

read the original abstract

As of mid-2026, 11 objects have been discovered prior to impacting the Earth, with warning times between 2 - 20 hours. Using real metre-sized 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 decade, how many imminent impactors would it have observed and discovered pre-impact, and how early would these discoveries have been?'' We use the LSST Solar System Survey Simulator Sorcha and a population of real fireballs observed by orbital sensors over the last decade to investigate which events would have been observed pre-impact. We find that the LSST would have observed 30 (13.9%) of the 216 simulated objects, with most objects receiving 2 - 4 observations. Using the default linking algorithm, only two (0.9%) of these objects would have been `discovered' pre-impact. Using a modified linking algorithm better suited to fast moving objects, this increases to eight (3.7%). Based on this, we predict that the LSST will discover 8 +/- 2 imminent impactors over its nominal 10 year survey, at the low end of previous estimations. However, we predict these objects to be discovered 4 days pre-impact, substantially earlier than the current average. This will bring significant opportunities for telescopic follow-up, targeted fireball observations, planetary defence planning, and public engagement. There is also significant potential for precovery for impactors observed by the LSST but 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.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Real fireball data fed into Sorcha yields a concrete LSST forecast of 8 discoveries at 4-day lead times, but the jump depends on an untested linking tweak.

read the letter

The main thing to know is that this paper runs 216 real metre-scale fireballs through the Sorcha LSST simulator and comes out with 30 observed events and, after a modified linking step, 8 discovered pre-impact over a 10-year survey, at an average 4 days before impact. That is lower than some earlier synthetic estimates but with earlier warning than the current 2-20 hour record for small impactors. Using actual observed events instead of purely modeled populations is the step that produces these specific numbers. The public simulator and forward simulation keep the work traceable and avoid obvious circular fitting. The discussion of precovery opportunities and longer arcs for follow-up is straightforward and practical. The soft spot sits in the linking algorithm. The default version recovers only 2 objects while the fast-mover modification reaches 8, yet the text gives no injection-recovery numbers, false-positive rates, or completeness curves for high-velocity metre-scale cases. Without those checks the 3.7 percent fraction and the +/-2 error bar rest on an assumption that may not hold. The 216-event sample is also taken as representative without much discussion of selection effects from the original sensors. This work is aimed at planetary-defense and LSST operations people who need ballpark yields for small impactors. A reader planning follow-up strategies or survey cadence would get usable numbers to think with. The core approach is honest and grounded in real data, so it deserves a serious referee even if the linking validation needs tightening in revision.

Referee Report

1 major / 2 minor

Summary. The paper uses the Sorcha LSST Solar System Survey Simulator to forward-model observations of 216 real metre-scale fireballs recorded by orbital sensors over the last decade. It reports that LSST would detect 30 (13.9%) of these objects pre-impact, with only 2 (0.9%) linked into discoveries under the default algorithm but 8 (3.7%) under a modified linking algorithm tuned for fast movers. Scaling this fraction yields a prediction of 8 ± 2 imminent impactor discoveries over the nominal 10-year LSST survey, occurring on average 4 days before impact—earlier than the current 2–20 hour warning times—and discusses follow-up and precovery opportunities.

Significance. If the results hold after validation, the work supplies a concrete, data-driven forecast grounded in real events rather than purely synthetic populations, placing the expected LSST yield at the low end of prior estimates while highlighting substantially earlier discovery times. The use of the public Sorcha simulator and an independently observed sample of fireballs adds reproducibility and avoids circular fitting to the target discovery numbers.

major comments (1)

[§4] §4 (Results, modified linking algorithm): The headline 8/216 (3.7%) discovery fraction—and therefore the entire 8 ± 2 ten-year forecast—arises only after replacing the default linking algorithm with a modified version for fast movers. The manuscript states that the default recovers 2 objects while the modification recovers 8, yet reports no injection-recovery tests, completeness curves versus velocity or arc length, or false-positive rates for this modification on synthetic metre-scale fast-moving trajectories. Because the central claim scales directly from this unvalidated step, quantitative performance metrics on fast-mover cases are required.

minor comments (2)

[Abstract] Abstract: The ±2 uncertainty on the predicted yield should be explicitly derived (e.g., Poisson statistics on the 8/216 fraction, bootstrap resampling of the 216 events, or propagation of simulator uncertainties) rather than stated without justification.
[§2] §2 (Sample selection): The 216-object sample is described as “real fireballs observed by orbital sensors,” but selection criteria, completeness limits, and any velocity or size biases relative to the metre-scale impactor population are not quantified; a short discussion of representativeness would strengthen the extrapolation.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their constructive review and positive evaluation of the manuscript's significance. We address the single major comment below and will revise the manuscript accordingly to strengthen the validation of the modified linking algorithm.

read point-by-point responses

Referee: [§4] §4 (Results, modified linking algorithm): The headline 8/216 (3.7%) discovery fraction—and therefore the entire 8 ± 2 ten-year forecast—arises only after replacing the default linking algorithm with a modified version for fast movers. The manuscript states that the default recovers 2 objects while the modification recovers 8, yet reports no injection-recovery tests, completeness curves versus velocity or arc length, or false-positive rates for this modification on synthetic metre-scale fast-moving trajectories. Because the central claim scales directly from this unvalidated step, quantitative performance metrics on fast-mover cases are required.

Authors: We agree that the modified linking algorithm requires quantitative validation to support the discovery fraction and the resulting 8 ± 2 forecast. In the revised manuscript we will add a dedicated subsection to §4 that presents injection-recovery tests performed on a synthetic population of fast-moving metre-scale trajectories. These tests will report completeness as a function of on-sky velocity and arc length, together with false-positive rates under the modified linking criteria. The new results will be used to justify the choice of the modified algorithm and to propagate any associated uncertainties into the final 10-year prediction. revision: yes

Circularity Check

0 steps flagged

No significant circularity; prediction is direct count-based extrapolation from independent real-event simulation

full rationale

The derivation counts simulated discoveries (8 out of 216 real fireballs) under default and modified linking, then scales the 3.7% fraction to a 10-year LSST forecast of 8 +/- 2. This is a forward simulation on externally observed historical data using Sorcha, with no parameters fitted to LSST discovery numbers, no equations that reduce the output to the input count by construction, and no load-bearing self-citations or ansatzes. The modified algorithm choice is an explicit modeling decision whose validation is a separate assumption, not a self-referential reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The simulation rests on the representativeness of the real fireball sample and the fidelity of the Sorcha tool plus linking algorithms; no free parameters or invented entities are explicitly introduced in the abstract.

axioms (1)

domain assumption The 216 real fireball events accurately represent the metre-scale imminent impactor population that LSST would encounter.
Used as the input population for all simulated observations and discovery statistics.

pith-pipeline@v0.9.0 · 5664 in / 1228 out tokens · 31381 ms · 2026-05-16T07:01:20.965817+00:00 · methodology

Simulated LSST Observations of Real Metre-scale Imminent Impactors

Core claim

What carries the argument

If this is right

Where Pith is reading between the lines

Load-bearing premise

What would settle it

discussion (0)