Predictions of Imminent Earth Impactors Discovered by LSST
Pith reviewed 2026-05-15 14:50 UTC · model grok-4.3
The pith
LSST will discover about one to two meter-sized imminent impactors per year.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
Using 343 CNEOS-recorded meter-size impactors, simulations with the Sorcha survey simulator under LSST's three-night and one-night strategies predict that LSST will discover ∼1-2 meter-size and larger imminent impactors per year. This represents ∼4% of all Earth impactors ≳1 m in diameter and would almost double the current discovery rate. The median time of discovery and median time of first observation for impactors discovered in our simulations are ∼1.57 and ∼3.06 days before impact, respectively.
What carries the argument
Sorcha survey simulator modeling LSST pre-impact observations of the 343 CNEOS impactors under three-night and one-night streak-matching strategies.
If this is right
- LSST will nearly double the current discovery rate of imminent impactors.
- Median discovery will occur 1.57 days before impact and first observation 3.06 days before.
- LSST detections will show a southern hemisphere bias, complementing northern surveys.
- The 11 previously discovered impactors are biased north; LSST will improve global coverage.
Where Pith is reading between the lines
- More frequent pre-impact characterization of individual NEOs from space to meteorite will become routine.
- Better sampling of the small NEO population could tighten models of impact frequency at meter scales.
- One-night streak strategies may be adopted by other surveys to catch fast objects.
Load-bearing premise
The 343 CNEOS-recorded impactors are statistically representative of the meter-scale population and the Sorcha simulator correctly captures LSST detection efficiency for fast-moving streaked objects.
What would settle it
The actual count of imminent impactors LSST discovers in its first year of operation compared against the predicted rate of 1-2 per year.
read the original abstract
Imminent impactors are natural bodies discovered in space before impacting the Earth. They provide a rare opportunity to characterize individual near-Earth objects (NEOs) in great detail as asteroids in space, meteors in Earth's atmosphere and meteorites on the ground. The Vera C. Rubin Observatory's upcoming Legacy Survey of Space and Time (LSST) is expected to transform our understanding of the NEO population. In this work, we evaluate LSST's expected discovery performance for imminent impactors using $343$ meter-size objects previously recorded in NASA's CNEOS database as fireballs impacting Earth's atmosphere. We simulate pre-impact observations of these CNEOS impactors with the Sorcha survey simulator under LSST's default three-night discovery strategy and a one-night strategy for fast-moving objects that relies on matching aligned streaks in two exposures on the same night. We estimate that LSST will discover $\sim1-2$ meter-size and larger imminent impactors per year, representing $\sim4\%$ of all Earth impactors $\gtrsim1$ m in diameter and almost doubling the current discovery rate of imminent impactors. The median time of discovery and median time of first observation for impactors discovered in our simulations are $\sim1.57$ and $\sim3.06$ days before impact, respectively. The spatial distribution of the 11 previously discovered imminent impactors is biased towards the Northern Hemisphere, where the observatories that discovered them are located. We find a similar trend towards Southern Hemisphere impacts in our simulated LSST detections of the CNEOS impactors, suggesting Rubin will provide a powerful counterpart to existing asteroid surveys primarily located in the Northern Hemisphere.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper simulates pre-impact observations of 343 historical CNEOS fireballs using the Sorcha survey simulator to predict LSST's performance for imminent impactors. It estimates that LSST will discover ∼1-2 meter-size and larger imminent impactors per year (∼4% of all Earth impactors ≳1 m), nearly doubling the current rate, with median discovery and first-observation times of ∼1.57 and ∼3.06 days before impact. The work also examines hemispheric biases, noting LSST's potential to complement northern-hemisphere surveys.
Significance. If the simulation assumptions hold, the quantitative predictions would be useful for LSST operations planning and for assessing the survey's contribution to the imminent impactor discovery rate. The forward-modeling approach that replays real historical events through an independent survey simulator is a methodological strength, as it grounds the rates in observed data rather than purely synthetic distributions.
major comments (3)
- [Abstract] Abstract: the ∼1-2 per year discovery rate and 4% fraction are stated without error bars, confidence intervals, or sensitivity tests on the streak detection efficiency (a free parameter in the simulation). This is load-bearing because the rate is obtained by scaling the 343-event sample through the simulator.
- [Simulation setup and results] Simulation setup and results: no validation or cross-check is presented for the one-night streak-matching strategy against real LSST-like data or against the three-night baseline for fast-moving streaked sources. The abstract notes this strategy is key to the higher discovery numbers, yet its fidelity is untested.
- [Discussion] Representativeness discussion: the claim that the CNEOS sample yields an unbiased extrapolation to the full meter-scale population is not supported by any test for selection biases in entry angle, time of day, or brightness that could differ between CNEOS detections and LSST-detectable objects.
minor comments (2)
- [Abstract] The abstract introduces the Sorcha simulator without a citation; a specific reference should be added.
- Figure captions (where present) would benefit from explicit statements of the assumed streak detection efficiency and the exact LSST cadence parameters used.
Simulated Author's Rebuttal
We thank the referee for their constructive comments on our manuscript. We address each major comment below and have made revisions where appropriate to improve the clarity and robustness of our results.
read point-by-point responses
-
Referee: [Abstract] Abstract: the ∼1-2 per year discovery rate and 4% fraction are stated without error bars, confidence intervals, or sensitivity tests on the streak detection efficiency (a free parameter in the simulation). This is load-bearing because the rate is obtained by scaling the 343-event sample through the simulator.
Authors: We agree that uncertainty estimates and sensitivity analyses would enhance the manuscript. The discovery rate is calculated by determining the detection fraction from the 343 simulated CNEOS events and scaling to the known annual impact rate for meter-sized objects. We will add Poisson-based error bars to the rate and 4% fraction. Additionally, we will include sensitivity tests varying the streak detection efficiency parameter by ±10% and ±20% from the nominal value, reporting the resulting range in discovery rates. These will be incorporated into the abstract and a new subsection in the results. revision: yes
-
Referee: [Simulation setup and results] Simulation setup and results: no validation or cross-check is presented for the one-night streak-matching strategy against real LSST-like data or against the three-night baseline for fast-moving streaked sources. The abstract notes this strategy is key to the higher discovery numbers, yet its fidelity is untested.
Authors: Direct validation against LSST data is not feasible as the survey has not yet begun operations. The one-night streak-matching approach is adapted from methods used in other wide-field surveys for detecting fast-moving objects. We will expand the methods section to describe the algorithm in more detail and add a comparison of the one-night versus three-night strategies, quantifying the increase in discoveries. We will also reference literature on streak detection performance in surveys like Pan-STARRS to support the strategy's applicability. revision: partial
-
Referee: [Discussion] Representativeness discussion: the claim that the CNEOS sample yields an unbiased extrapolation to the full meter-scale population is not supported by any test for selection biases in entry angle, time of day, or brightness that could differ between CNEOS detections and LSST-detectable objects.
Authors: We will add a new subsection in the discussion analyzing potential selection biases in the CNEOS fireball sample. Specifically, we will compare the distributions of entry angles, impact times (local solar time), and estimated brightnesses in the CNEOS events to those expected for objects detectable by LSST based on the simulation. This will allow us to assess and discuss any biases that might affect the extrapolation to the full population. revision: yes
Circularity Check
No circularity: forward simulation of external CNEOS catalog through independent Sorcha model
full rationale
The paper's rate estimate (∼1-2 discoveries per year, ∼4% fraction) is obtained by injecting the 343 CNEOS fireballs into the Sorcha survey simulator and counting detections under two LSST strategies. This is a standard forward-modeling pipeline that takes an external, independently recorded catalog as input and applies an external detection-efficiency model; no parameter is fitted to the output rate, no equation reduces the prediction to a fitted input by construction, and no self-citation chain is required to justify the core numerical result. The derivation therefore remains self-contained against external benchmarks.
Axiom & Free-Parameter Ledger
free parameters (1)
- streak detection efficiency
axioms (1)
- domain assumption The 343 CNEOS impactors form a representative sample of the meter-scale Earth-impactor population
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.