NEO: Near-Earth Objects Overview

• NEO refers to near-Earth objects, defined by a perihelion distance of ≤1.3 au, and includes subpopulations like Atiras, Atens, Apollos, and Amors.
• Detection techniques leverage both visible-light and thermal-IR observations to determine size, albedo, and orbits, addressing challenges such as rotational variability.
• Survey advancements with missions like NEO Surveyor and LSST aim to close completeness gaps for objects >140 m, crucial for improving planetary defense models.

NEO most commonly denotes a near-Earth object: an asteroid or comet whose perihelion distance satisfies $q \le 1.3$ au. In current observational practice the term is dominated by near-Earth asteroids (NEAs), the non-cometary component of the population, because they account for most known objects and most of the present planetary-defense burden. More than 32,000 NEAs were already cataloged by early June 2023, yet the catalog is strongly size-dependent: it is near complete at kilometer scale but only $\sim 38\%$ complete for objects larger than 140 m, the regime associated with severe regional damage and the focus of the U.S. Congressional survey mandate (Grav et al., 2023).

1. Definition, orbital classes, and hazard thresholds

Near-Earth objects are defined dynamically, not compositionally. The standard criterion is perihelion distance $q \le 1.3$ au. Within this set, NEAs are commonly partitioned into four dynamical sub-populations according to semimajor axis $a$, perihelion $q$, and aphelion $Q$ (Grav et al., 2023).

Class	Orbital criterion	Dynamical description
Atiras	$Q \le 0.983$ au	Orbits entirely inside Earth’s orbit
Atens	$a < 1$ au, $Q > 0.983$ au	Mostly inside Earth’s orbit but crossing it
Apollos	$a > 1$ au, $\sim 38\%$0 au	Earth-crossers, mostly outside Earth’s orbit
Amors	$\sim 38\%$1 au	Just outside Earth’s orbit, do not currently cross

The hazard logic is primarily size driven. Impact energy scales roughly as $\sim 38\%$2, so the transition from tens of meters to hundreds of meters corresponds to a sharp increase in destructive potential. Historically, survey policy first targeted $\sim 38\%$3 km objects, but subsequent analyses placed most residual impact risk in the 140 m–1 km interval. The 2005 George E. Brown, Jr. Near-Earth Object Survey Act therefore mandated detection and tracking of $\sim 38\%$4 of NEOs larger than 140 m (Grav et al., 2023).

A second key geometric quantity is the Minimum Orbit Intersection Distance (MOID). In classical planetary-defense usage, potentially hazardous asteroids are often defined by $\sim 38\%$5 au together with $\sim 38\%$6, roughly corresponding to $\sim 38\%$7 m for typical albedo assumptions. In NEO Surveyor performance modeling, any object with $\sim 38\%$8 au is treated as a PHA regardless of $\sim 38\%$9, specifically to include dark 140 m objects that can be as faint as $q \le 1.3$0 mag (Ieva et al., 2018, Grav et al., 2023).

Because visible surveys measure absolute magnitude $q \le 1.3$1 rather than diameter directly, the diameter–albedo relation is fundamental: $q \le 1.3$2 The wide albedo range observed for NEAs, approximately $q \le 1.3$3–60%, makes a fixed mapping such as $q \le 1.3$4 m too simplistic. Using the measured albedo distribution, the practical proxy adopted for $q \le 1.3$5 m in current survey-completeness work is $q \le 1.3$6 mag (Grav et al., 2023).

2. Catalog growth, present discovery state, and completeness

The known NEA catalog has expanded rapidly. Annual discoveries rose from 27 in 1992 to 485 in 2002, 991 in 2012, and 3,189 in 2022. Yet the size distribution of newly found objects has shifted downward: about $q \le 1.3$7 of 2012 discoveries had $q \le 1.3$8, whereas by 2022 only about $q \le 1.3$9 did so, meaning that recent surveys are increasingly finding smaller objects rather than closing the 140 m completeness gap (Grav et al., 2023).

Catalog quality is also heterogeneous. Of the $a$0 NEAs in the MPC catalog as of early June 2023, roughly 3,000 were numbered, about 2,000 had arcs longer than 10 years, only about one third had multi-opposition orbits, and about one half had been observed for less than 7 days and were effectively lost pending rediscovery. This makes the distinction between “known,” “securely recoverable,” and “physically characterized” operationally important (Grav et al., 2023).

Current completeness is best summarized in both diameter space and $a$1-space. The Known Object Model (KOM), calibrated to the historical discovery record and applied to ten realizations of the NEO Surveyor reference population, yields $a$2 completeness for $a$3 km and

$a$4

This is consistent with Harris-style estimates once the conversion between diameter and $a$5 is treated with a realistic albedo distribution rather than a fixed $a$6 cutoff (Grav et al., 2023).

The policy consequence is straightforward. Using the cited estimate of $a$7 NEAs with $a$8 mag and the MPC’s $a$9 known objects in that regime, reaching $q$0 completeness would require about 35,550 objects, or roughly 22,000 more. At the then-current discovery rate of about 731 NEAs per year with $q$1, the timescale would exceed 30 years, and the rate would likely slow as completeness increases (Grav et al., 2023).

3. Detection physics and physical characterization

Visible-light surveys remain the dominant discovery channel, but they are intrinsically albedo biased because reflected-light brightness is set by $q$2. Infrared surveys address this by observing thermal emission, which depends primarily on size, distance, and temperature. Warm Spitzer’s NEOSurvey was designed around this principle: it used 4.5 $q$3m photometry to derive diameters and albedos for a flux-limited sample of known NEOs, with the explicit goal of creating a large and uniform catalog of physical properties. In that analysis the dominant uncertainty source was the NEATM beaming parameter $q$4; diameter uncertainties were typically about $q$5 and albedo uncertainties about $q$6, with more than $q$7 of the diameter error budget and $q$8–$q$9 of the albedo error budget driven by $Q$0 (Trilling et al., 2016).

Taxonomic characterization remains more observationally demanding than orbit determination. A visible-photometry survey of 67 previously unclassified NEOs found 41 S-complex, 11 C-complex, 7 X-complex, 4 V-type, 3 D-type, and 1 A-type objects, and showed that compositionally informative follow-up remains sparse compared with the rate of discovery. The same study emphasized that less than $Q$1 of NEOs had been characterized at that time and that smaller objects, especially $Q$2 m, were the highest-priority regime for mitigation-relevant physical data (Ieva et al., 2018).

Rotation complicates color-based taxonomy whenever filters are not acquired simultaneously. The MANOS survey, using non-simultaneous $Q$3 color sequences for 189 NEOs, showed that ignoring lightcurve variability can bias both individual classifications and ensemble results. In the 49-object subset with dedicated lightcurve modeling, $Q$4 changed taxonomic type when rotational variability was not properly handled, and the uncorrected sample showed a systematic over-abundance of C-complex classifications. This established rotation-corrected spectro-photometry as a methodological requirement rather than a refinement (Moskovitz et al., 16 Mar 2026).

At still smaller sizes, direct pre-impact discovery becomes inefficient. The NELIOTA lunar monitoring system addresses the decimeter-to-meter regime indirectly by treating the Moon as an impact detector. In its first year it reported 31 validated lunar impact flashes, reached an effective limiting magnitude of $Q$5 mag for low lunar phase at $Q$6, and inferred a detection rate of $Q$7. This extends NEO population constraints into a size regime largely inaccessible to conventional telescopic surveys (Xilouris et al., 2018).

4. Discovery pipelines, confirmation, and survey-system architecture

Short-arc classification is central to operational NEO discovery. The digest2 code takes tracklets—typically 2–4 detections over minutes to hours—and computes a pseudo-probability $Q$8 that the object belongs to the NEO class by sampling bound orbits consistent with the observations and weighting them with a synthetic Solar System model. The MPC operational threshold is $Q$9: tracklets above this value are eligible for posting to the Near-Earth Object Confirmation Page (NEOCP). In the January 2017 MPC sample discussed in the digest2 paper, $Q \le 0.983$0 of NEO tracklets and $Q \le 0.983$1 of non-NEO tracklets had $Q \le 0.983$2 at the time of detection; across a ten-year synthetic survey, $Q \le 0.983$3 of NEOs reached $Q \le 0.983$4 at least once, indicating that survey cadence and sky coverage, rather than the classifier itself, dominate residual incompleteness (Keys et al., 2019).

The Rubin Observatory LSST will alter this pipeline quantitatively. Simulations using the baseline cadence and current NEOCP criteria show that Rubin would contribute about 129 new NEOCP candidates per night in its first year, an increase of roughly $Q \le 0.983$5 over present-day traffic, but only about $Q \le 0.983$6 of those candidates would be true NEOs. The dominant contaminants are faint, previously undiscovered main-belt asteroids. A self-recovery prediction algorithm can reduce the list requiring external follow-up to about 64 candidates per night, but purity remains only about $Q \le 0.983$7, so follow-up selection becomes an algorithmic triage problem rather than a simple confirmation workflow (Wagg et al., 2024).

Within the broader observing ecosystem, emerging facilities are sharply differentiated by function. LSST is the major future optical discovery engine; NEOCam/NEO Surveyor-class thermal-IR missions are the most direct route to size-limited completeness; JWST and WFIRST/Roman are characterization platforms rather than survey systems; SPHEREx contributes spectral context at scale; and radar remains indispensable for precision orbits, shapes, and in favorable cases densities. Comparative assessments of these systems consistently identify combined visible and thermal-IR architectures as the most effective route to the 140 m completeness goal (Milam et al., 2019, Myhrvold, 2015).

A long-standing instrumental argument complements this systems view. A dedicated low-resolution visible spectrograph with $Q \le 0.983$8 on a 1–2 m telescope was proposed specifically because many NEOs are too faint and too briefly observable for efficient characterization with conventional higher-resolution instruments. The core idea was that simultaneous low-resolution spectra avoid the rotational phase mismatches inherent in sequential broadband colors while maintaining throughput high enough to characterize faint targets on small telescopes (Trueblood et al., 2010).

5. NEO Surveyor, the Known Object Model, and projected completeness

NEO Surveyor is NASA’s dedicated space-based infrared survey mission for planetary defense. It is designed around a 50 cm telescope at Sun–Earth L1 with two infrared channels near 4.6 and 8 $Q \le 0.983$9m, a field of regard spanning solar elongations $a < 1$0–$a < 1$1 and ecliptic latitudes within $a < 1$2, and a cadence in which each side of the field of regard is covered in about 6–7 days and revisited roughly two weeks later. The mission’s primary objective is to find the majority of objects large enough to cause severe regional impact damage, specifically $a < 1$3 m, within its five-year baseline survey; the mission paper projects $a < 1$4–300,000 new NEO discoveries down to sizes as small as $a < 1$5 m, together with thousands of comets (Mainzer et al., 2023).

The Known Object Model was constructed to place that mission in the context of the existing catalog. It uses a synthetic NEA reference population, a time-dependent model of historical survey performance from 1970 onward, and an epoch-based selection function defined by limiting magnitude, field of regard, and discovery probability. This partitions the reference population into objects statistically corresponding to those already discovered and those still undiscovered, allowing the overlap between current surveys and NEO Surveyor to be computed explicitly rather than by subtraction (Grav et al., 2023).

The resulting performance numbers are the central quantitative benchmark for present NEO policy. For $a < 1$6 m, NEO Surveyor alone reaches about $a < 1$7 completeness in five years. The current catalog contributes about $a < 1$8 completeness, but because of overlap the combined completeness is not their sum; it rises to about $a < 1$9. Approximately $Q > 0.983$0 of the currently cataloged $Q > 0.983$1 m objects are expected to be re-detected and tracked by NEO Surveyor, while the remaining $Q > 0.983$2 of currently known objects are not seen by the spacecraft but still contribute $Q > 0.983$3–$Q > 0.983$4 of the final completeness because they are already cataloged. For PHAs with $Q > 0.983$5 m, the expected five-year completeness is about $Q > 0.983$6 (Grav et al., 2023).

The orbital-class dependence of this performance is astrophysically revealing. Over five years, NEO Surveyor detects about $Q > 0.983$7 of Atiras, $Q > 0.983$8 of Atens, $Q > 0.983$9 of Apollos, and $a > 1$0 of Amors for $a > 1$1 m. The relatively low Amor fraction is geometric: many Amors remain at distances or elongations less favorable for the spacecraft’s field of regard, whereas interior-Earth and Earth-crossing populations benefit strongly from the near-Sun observing geometry (Grav et al., 2023).

This performance still falls short of the statutory $a > 1$2 goal on the nominal five-year timeline. The same analysis therefore identifies extended operations, follow-on missions, or complementary survey assets as necessary to close the remaining gap, especially for high-eccentricity, high-inclination, and some outer-crossing Amor populations that remain difficult after the baseline mission (Grav et al., 2023).

6. Composition, size dependence, and remaining planetary-defense questions

Compositional statistics now indicate that the NEO population is not self-similar across size. The NEOShield-2 photometric survey already suggested a deficit of small C-complex NEOs relative to S-complex objects: within its 67-object sample, C-complex bodies were almost absent in the $a > 1$3 m bin, whereas S-complex objects were present across small and medium size bins. The same study highlighted two carbonaceous PHAs with very low MOID and high inclination as especially challenging mitigation targets, because low albedo complicates size inference from $a > 1$4 and high inclination raises mission $a > 1$5 (Ieva et al., 2018).

MANOS extended this issue into the meter-to-decameter regime. Combining 189 MANOS color targets with MITHNEOS, NEOSHIELD2, and MANOS spectroscopy produced a consistent visible-taxonomy sample of 831 NEOs spanning $a > 1$6–29. In that combined dataset, the ordinary-chondrite-like S+Q fraction falls from about $a > 1$7 at $a > 1$8 to about $a > 1$9 at $\sim 38\%$00, while the X-complex fraction rises from roughly $\sim 38\%$01–$\sim 38\%$02 to $\sim 38\%$03. The paper’s synthesis sharpens this further: the abundance of S-complex or ordinary-chondrite-like NEOs decreases by a factor of two from $\sim 38\%$04 of the population at kilometer scales to about one third at sizes $\sim 38\%$05 m. The authors evaluate and reject simple explanations based on source region alone, thermal modification, discovery bias, tidal resurfacing, regolith grain size, and impact shock darkening as primary drivers, and instead favor a genuine compositional gradient tied to delivery from young asteroid families and size-dependent transport into resonance pathways (Moskovitz et al., 16 Mar 2026).

This has direct implications for meteorite interpretation and hazard modeling. If the small-NYO population is already more X- and primitive-rich than the meteorite collection, then atmospheric filtering is only part of the reason ordinary chondrites dominate recovered meteorites. A plausible implication is that probabilistic impact-risk models should treat composition as a function of size rather than assuming that decameter-scale impactors are scaled-down versions of kilometer-scale S-type bodies. That matters because density, porosity, fragmentation behavior, and momentum-transfer efficiency in deflection scenarios all depend on composition and internal structure (Moskovitz et al., 16 Mar 2026).

The remaining survey gaps are correspondingly structured. After the currently modeled five-year NEO Surveyor mission, the hardest residual objects are high-eccentricity NEAs, high-inclination populations, certain Amors, and the vast population below 140 m. The detection problem at that point is no longer merely one of collecting more objects, but of covering the remaining regions of orbital phase space while simultaneously scaling taxonomic, thermal, and dynamical characterization fast enough that the known catalog becomes a physically characterized hazard inventory rather than only an astrometric one (Grav et al., 2023).