🔭 astrophysics

Simulated LSST Observations of Real Metre-scale Imminent Impactors

This study simulates the Vera C. Rubin Observatory's LSST survey over the past decade using real meteor data to predict that it would have discovered approximately eight imminent Earth impactors about four days before impact, significantly improving early warning capabilities for planetary defense and scientific follow-up compared to current systems.

Original authors: Michael A. Frazer, Hadrien A. R. Devillepoix, Sophie E. Deam

Published 2026-05-11

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Michael A. Frazer, Hadrien A. R. Devillepoix, Sophie E. Deam

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the Earth is a busy city, and every year, about 35 to 40 small rocks (roughly the size of a car or a small house) crash into our atmosphere. Most of them burn up like shooting stars, but some are big enough to create a loud "boom" (an airburst) or even land as meteorites.

For a long time, we've been like a city without a weather forecast for these rocks. We usually only find out about them 2 to 20 hours before they hit, which is like being told a storm is coming just as the first raindrop hits your nose.

This paper asks a "What if?" question: What if we had a super-powered, all-seeing camera (the Vera C. Rubin Observatory, or LSST) watching the sky over the last 10 years? Would it have spotted these incoming rocks earlier?

Here is the breakdown of their findings, using simple analogies:

1. The Experiment: The "Time-Travel" Simulation

The researchers didn't wait for the new telescope to start working. Instead, they used a sophisticated computer program called Sorcha (think of it as a "video game engine" for space) to simulate how the LSST would have seen the sky between 2015 and 2024.

They fed the program data on 216 real rocks that actually hit Earth during that decade. They asked the computer: "If the LSST had been looking at the sky back then, how many of these rocks would it have seen?"

2. The Results: The "Spotting" Game

The results were a mix of "not bad" and "could be better," depending on how you play the rules:

The "Eyes" (Observations): The LSST would have seen 30 of those 216 rocks. It would have taken a picture of them 130 times in total.
- Analogy: Imagine a security camera that takes a photo of a car passing by. It caught 30 cars, but missed 186 of them because they were too small, too dark, or moving too fast for the camera's settings.
The "Brain" (Discovery): Seeing a rock isn't the same as knowing it's coming. The computer needs to link several photos together to say, "Hey, that's a rock, and it's heading for us!"
- Standard Rules: Using the current, strict rules for linking photos, the LSST would have only discovered 2 of the rocks before they hit.
- New Rules: The researchers tested a "faster" way of linking photos (better for fast-moving objects). With this tweak, the LSST would have discovered 8 rocks.

3. The Big Win: Time is Money

The most exciting part isn't just how many they found, but when they found them.

Current Reality: We usually get a warning of about 9 hours (sometimes as little as 2 hours).
LSST Prediction: The LSST would have given a warning about 4 to 5 days in advance.
- Analogy: Currently, we are like a driver who sees a pothole only when their tire hits it. The LSST is like a driver who sees the pothole from a mile away, giving them plenty of time to slow down, steer around it, or even send a crew to fix it.

4. Why Did It Miss Some?

The paper explains that the LSST isn't a dedicated "rock hunter" like a specialized security guard. It's a massive survey camera that scans the whole sky slowly and methodically.

The Cadence Problem: The LSST looks at a spot in the sky twice a night, then moves on. To "discover" a rock, it usually needs to see it on three different nights. By the time it gets that third look, the rock might have already hit Earth.
The Speed Problem: Fast-moving rocks are harder to link together in the computer's database.

5. The "Ghost" Observations (Precovery)

Even for the rocks the LSST saw but didn't officially "discover" (because it didn't link the photos in time), the data is still gold.

Analogy: Imagine the LSST took a blurry photo of a rock 3 days before it hit, but didn't realize what it was. Later, another telescope finds the rock and says, "Hey, that's a rock!" The LSST can then look back at its old photos and say, "Oh, I saw that too!"
This allows scientists to combine the data, creating a much longer "movie" of the rock's path, which helps predict exactly where it will land with much higher accuracy.

Summary

If the Vera C. Rubin Observatory had been running over the last decade, it likely would have discovered about 8 of the metre-sized rocks that hit Earth. While that sounds like a small number, the real victory is time. Instead of a panic-inducing 9-hour warning, we would have had several days to prepare, study the rock, and plan our response.

The paper concludes that while the LSST won't replace the specialized "rock hunters" that find things at the last minute, it will act as a powerful early-warning system, giving us a "head start" on the next incoming space rock.

Technical Summary: Simulated LSST Observations of Real Metre-scale Imminent Impactors

Problem Statement
As of mid-2026, only 11 metre-scale objects have been discovered prior to impacting Earth, typically with warning times of 2–20 hours. While the Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST) is expected to revolutionize Solar System discovery, its specific capability to detect and discover imminent metre-scale impactors—objects that will strike Earth within days or weeks—remains uncertain. Previous estimates vary, and it is unclear whether the LSST's survey strategy (optimized for deep, wide-field imaging rather than rapid NEO tracking) will allow for the discovery of these fast-moving, faint objects before impact. This study seeks to quantify how many such objects the LSST would have observed and discovered had it been operational over the last decade.

Methodology
The authors utilized the Sorcha (Solar System Survey Simulator) to model LSST observations against a population of real, historical impactors.

Input Data: The study used 216 metre-scale impactor events recorded by U.S. Government (USG) space-based sensors between January 1, 2015, and December 31, 2024. These events were selected from a total of 372 USG records; 216 had sufficient velocity data to derive pre-impact orbits.
Physical Modeling: Impactor masses and diameters were estimated using kinetic energy relations ( $E = \frac{1}{2}mv^2$ $E = \frac{1}{2} m v^{2}$ ) and a bulk density of 1500 kg m $^{-3}$ $^{- 3}$ . Two population scenarios were simulated:
1. "Best Case": All objects assumed to be bright S-type asteroids (albedo $p_v = 0.24$ ).
2. "Realistic Case": An even split between S-type and dark C-type asteroids (albedo $p_v = 0.04$ ).
Simulation Parameters: The simulation ran the LSST pointing baseline (modified to cover 2015–2024) against the derived orbits. Objects were recorded as "observed" if they fell within the LSSTCam footprint and were brighter than the 5- $\sigma$ limiting magnitude.
Linking Algorithms: The study evaluated two discovery criteria:
1. Default SSP: Requires three tracklets (pairs of observations) linked over three nights within a 14-day window.
2. Modified Algorithm: Allows for the immediate linking of single tracklets containing three or more observations (potentially occurring in a single night or over a shorter baseline), a proposed change to the Solar System Processing (SSP) pipeline.

Key Results

Observation Rates: In the "best case" (all S-type) simulation, the LSST would have observed 30 of the 216 objects (13.9%), generating 130 total observations. Most objects received 2–4 observations.
Discovery Rates:
- Using the default 3-tracklet algorithm, only 2 objects (0.9%) would have been discovered pre-impact.
- Using the modified 3-observation linking method, the number of discoveries increased to 8 objects (3.7%).
Warning Times: The median warning time for the 8 discoveries in the best-case scenario was 5.0 days (excluding one object discovered 3.5 years in advance). This is substantially longer than the current average of ~9 hours and the maximum of 21 hours (2014 AA).
Physical and Orbital Biases: Observed objects were predominantly larger ( $H_r > 31$ , $D > 2$ m), slower (impact velocities $< 25$ km s $^{-1}$ ), and on evolved, Earth-like orbits (low inclination and eccentricity). The LSST's Southern Hemisphere location resulted in a bias toward Southern Hemisphere impacts.
Realistic Predictions: Scaling the results to account for the realistic S/C-type mix and the 58% availability of velocity data in the USG dataset, the authors predict the LSST will observe 34 ± 2 unique metre-scale impactors over its 10-year survey. Under the modified linking algorithm, this yields 8 ± 2 pre-impact discoveries, with a median warning time of 4.3 ± 3.2 days.

Significance and Claims
The paper argues that while the LSST may not replace dedicated NEO surveys (like ATLAS or CSS) for the volume of imminent impactor discoveries, it will provide a distinct advantage in lead time.

Extended Warning: Discovering impactors days in advance, rather than hours, offers significant opportunities for telescopic follow-up, targeted fireball observations, and improved planetary defence planning.
Precovery Potential: Even for objects not discovered by the LSST, the survey will likely capture "precovery" observations (faint, unlinked detections). If these objects are later identified by other surveys or via fireball networks, these historical LSST data points can be linked, drastically extending the observation arc and reducing orbital and impact location uncertainties.
Scientific Value: A sample of pre-impact observations linked to recovered meteorites or fireballs will allow for the comparison of impact-derived masses with pre-impact photometry, potentially refining estimates of asteroid densities and albedos.

The authors conclude that the LSST will likely discover approximately one imminent impactor per year (assuming the modified linking algorithm is implemented), providing a critical new data source for understanding the meteoroid-fireball-meteorite connection and enhancing planetary defence capabilities.