Physics / Cosmology / Simulation

The Invisible Interstellar Traffic Through Our Solar System

Report summary

A detailed, evidence-based report on trajectories, abundances, gravitational focusing, detection biases, and what three discoveries really tell us

Status
Research archive item
Category
Physics / Cosmology / Simulation
Length
12,810 words
Reading time
59 minutes
Report type
evaluation

Key topics

  • Physics / Cosmology / Simulation
  • Physics
  • Cosmology
  • Simulation
  • .NET
  • Angular
  • Python
  • Runtime
  • Research Archive

Research provenance

Archive status
Research archive item
Content identity
sha256:3bd18a7bdcf27794b18f7046636f70229fe9bb685b2f44db08ba800c7246168e

For citation, use the report title and canonical URL. Archival presence does not establish authorship or promote report statements into portfolio evidence.

This page renders the archived Markdown as safe, formatted HTML. It is background research and does not become a portfolio claim without evidence review.

Full report

On this page

A detailed, evidence-based report on trajectories, abundances, gravitational focusing, detection biases, and what three discoveries really tell us

Research cutoff: November 7, 2026 Edition: Expanded and corrected revision 2, November 7, 2026 Status: Scientific synthesis with reproducible illustrative calculations; not a census of observed objects Sources: 34 numbered bibliography entries (32 unique works/resources; two entries cross-reference a publication cited elsewhere) Central question: If interstellar objects travel through three-dimensional space in many directions, why do illustrations show them converging on the Sun, and how many actually pass through our Solar System unseen?

The key physical correction: Interstellar objects are not generally pointed toward the Sun. Far from the Sun, they occupy a vast range of positions and velocities. A typical object travels along an approximately straight line, possibly far from the planetary system. The Sun deflects the relatively small subset that comes sufficiently close. Solar gravity increases the cross-sectional area for encounters with a specified inner region; it does not make the entire interstellar population converge on the Solar System.


Revision 2: what changed and why

This revision retains the strongest parts of the earlier report but audits its assumptions and broadens its analysis. The most important correction is a units-and-population issue: the familiar \(n\sim0.1\,\mathrm{AU}^{-3}\) figure describes an ʻOumuamua-like brightness/size scale, not a rigorously observed \(D\ge100\)-meter diameter distribution. ʻOumuamua's effective radius was itself inferred to be of order 80–100 m, conditional on albedo, shape and lightcurve modeling [R01] [R02] [R04]. Accordingly, all diameter-extrapolation examples now state an explicit hypothetical 200-m reference threshold; they cannot be interpreted as measured size-bin rates.

Other improvements include: a flux-angle derivation showing why most tracks are oblique instead of Sun-pointed; a calculation of the fraction of flybys that bend by 10°, 30°, or 90°; the probability and detection-efficiency implications of just one ʻOumuamua-like discovery; real-world differences between Earth encounters and Sun-centered sphere crossings; a more careful treatment of the Sun's kinematic headwind, moving stellar populations and heliospheric dust; newly checked 2025–2026 meteor and capture studies; and a source-traceability audit. The number tables are independently recalculated from the stated equations, and their interpretation is tightened.

Scope warning: This is a research synthesis, not a newly calibrated estimate of the real population. It combines primary data and prior published inferences with explicitly labeled calculations. Independent literature results sometimes conflict because they choose different source populations, sizes, albedos, velocities and survey detection criteria; those disagreements are reported, not averaged away.

Abstract

The discovery of 1I/ʻOumuamua (2017), 2I/Borisov (2019), and 3I/ATLAS (2025) establishes a population of macroscopic objects arriving from outside the Solar System. Because telescopes discover only a small and strongly selected subset, three detections are compatible with many thousands of undiscovered objects passing inside the spherical volume bounded by Neptune's orbital radius. This is an inference, not a direct count.

An often-used benchmark density for roughly ʻOumuamua-scale objects (not a well-measured cumulative abundance above a precise diameter threshold) is 0.1 objects per cubic astronomical unit (AU⁻³), with an early study estimating 0.2 AU⁻³. Under an explicitly simplified steady-state model using 0.1 AU⁻³, a single inbound speed of 30 km/s, an isotropic distribution of far-field directions, and solar gravitational focusing, we obtain approximately 1,907 inward crossings per year inside 30 AU, 74 inside 5.2 AU, and 5.9 inside 1 AU. The corresponding mean numbers present at once are 11,851, 73.5, and 0.823. These numbers scale linearly with the assumed population density. They should not be interpreted as measured annual arrivals or as a validated confidence interval. [R01] [R02]

A particularly revealing consequence is that only ~0.31% of those modeled objects entering the sphere of radius 30 AU subsequently enter the sphere of radius 1 AU. Therefore, a physically correct drawing cannot show every interstellar object in the neighborhood curving sharply toward the Sun. Nearly all of the modeled outer-region encounters remain outer-region flybys.

The real incoming velocity distribution is not perfectly isotropic. Galactic stellar motions, the Sun's motion relative to nearby debris, and survey selection functions create structure. In particular, the **solar apex is an expected incoming radiant direction**, not the antapex; the velocity vectors of objects approaching from that direction point approximately opposite the Sun's velocity. The distinction between a sky direction from which something comes and the direction its velocity vector points is essential. [R09] [R10] [R11]


Executive summary: what the evidence does and does not establish

StatementAssessmentReason
Natural bodies from other planetary systems pass through the Solar System.EstablishedThree confirmed macroscopic visitors; also direct measurements of interstellar dust.
They can arrive on very differently oriented 3D trajectories.EstablishedOrbital inclinations of the confirmed visitors differ substantially.
All interstellar objects approach the Sun's center.FalseMost paths have nonzero impact parameters; many miss the planetary region entirely.
The Sun bends some flyby paths.EstablishedClassical hyperbolic two-body gravitational scattering.
There may be thousands of ʻOumuamua-like interstellar bodies inside 30 AU at once.Plausible model inferenceFollows from a sparse-data survey-inferred density; a single sharp diameter cutoff is not established.
About 1,900 such objects enter inside 30 AU each year.Conditional calculationTrue for the particular benchmark density/speed model; not an observation.
Thousands of bodies must pass near Earth every year.Not supportedCrossing a sphere around the Sun is not a close encounter with Earth.
Small interstellar rocks and grains are even more abundant.Qualitatively plausible, quantitatively unresolvedSize extrapolations are strongly model-dependent; dust and meteor surveys are not fitted by one universal power law.
The observed orbital inclinations prove objects prefer the ecliptic.Not supportedOnly 3I/ATLAS is nearly coplanar; three detections cannot determine the population.
The Sun's Galactic motion affects incoming directions.Supported by kinematic modelsThe incoming radiant distribution is structured, often enhanced toward the solar apex.
Rubin Observatory should greatly improve the census.Strong expectation, uncertain yieldSurvey simulations give substantially different discovery forecasts depending on assumptions.

Primary foundation: Do, Tucker & Tonry (2018); Jewitt & Seligman (2023); Dorsey et al. (2025); Marčeta & Seligman (2023). [R01] [R02] [R10] [R12]


Part I — What the space traffic really looks like

1. An interstellar object does not start by aiming for the Sun

An object ejected by a distant planetary system generally follows an orbit through the Galaxy. Near the Sun, well outside the region where the Sun's gravity significantly alters its trajectory, its short segment of motion can be approximated as

\[ \mathbf r(t)\simeq\mathbf r_0+\mathbf v_\infty t. \]

Both initial position \(\mathbf r_0\) and incoming velocity \(\mathbf v_\infty\) matter. Knowing only a direction on the sky is not enough to tell whether an object will pass close to the Sun.

Imagine a very large, nearly empty, three-dimensional field containing the Sun. A huge variety of straight tracks might traverse that field:

  • Some pass thousands of AU above the ecliptic.
  • Some pass to the Sun's left or right without entering the giant-planet region.
  • Some cut through the sphere of radius 30 AU but stay outside Jupiter's orbit.
  • A smaller fraction come close enough for solar gravity to produce conspicuous bends.
  • An extremely small subset might collide with the Sun, a planet, or a moon.

A 2D sketch of one small cross-section might look like the following. Tracks are schematic, not to scale; the true population fills 3D space and includes a broad distribution of velocity directions.

                path that misses 30 AU entirely
   ----------------------------------------------------->

   ----------->        [ 30 AU spherical region ]
                      .-------------------------.
   ----------------> /                           \ -------->
                    |        .-------.            |
   <--------------- | ---    | Sun * | ---        | <-------
                    |        '-------'            |
   ----------------> \       ) curved (          / -------->
                      '-------------------------'
                 \  a relatively close flyby  /
                  \___________________________/

    Other tracks move into/out of this page, not just left-to-right.

The figure illustrates impact parameters, not a collection of arrows pointing at the Sun. It is conceptually closer to cars crossing a large region at various lanes and headings than to raindrops somehow attracted to a central drain.

Is the distribution truly random?

Not necessarily. If we imagine a mathematically ideal isotropic background, incoming velocity directions are uniform over the celestial sphere. In reality, populations orbiting within the Milky Way are correlated with their parent stellar populations. In the Sun's frame, the Sun's motion creates a preferred incoming radiant; local stellar streams and age-dependent velocity dispersions produce additional structure. None of this requires objects to target the Sun. [R09] [R10] [R11]

2. Six different geometrical quantities often get confused

QuantityMeaningWhat it does not mean
Incoming radiantSky direction the object comes from, as seen before solar deflectionNot automatically its orbital-plane tilt
Incoming velocity vector \(\mathbf v_\infty\)Vector of motion far from Sun; points opposite the incoming radiantDoes not automatically point at the Sun
Impact parameter \(b\)Offset of the unperturbed incoming straight-line trajectory from the SunNot the actual closest approach after focusing
Perihelion distance \(q\)Minimum Sun-object separation along its actual orbitNot the distance to Earth
Orbital inclination \(i\)Tilt of the full orbital plane relative to the ecliptic, with motion sense encodedNot the object's incoming direction's latitude
Orbital angular momentum \(\mathbf h=\mathbf r\times\mathbf v\)Vector normal to the orbital planeNot a direct measurement of which part of the sky it entered from

An orbital plane is defined by the Sun and the path's angular momentum. An arrival direction alone does not uniquely determine the plane because the sideways offset vector also matters. For the same incoming velocity vector, different impact-parameter directions can produce different orbital planes.

Counterexample: A body arriving along a direction lying exactly in the ecliptic may nevertheless follow a trajectory whose closest approach is far above the ecliptic, producing a steeply inclined orbital plane. Conversely, an orbit with \(i\approx180^\circ\) is nearly coplanar but retrograde: it travels against the planets' usual orbital direction.

2A. A quantitative answer to “wouldn't most paths be almost straight?”

Yes. A natural population is best described in phase space: six variables per object (three positions and three velocity components). Even in the artificial isotropic model, uniform sky radiants do not mean the objects are on radial rays aimed at the Sun. A line's direction and its perpendicular offset are independent ingredients.

The missing entry-angle distribution

At the surface of a large spherical target of radius \(R\), assume trajectories are locally straight and the background is isotropic. Let \(\alpha\) be the angle between the incoming velocity and the inward surface normal. An object crossing exactly radially would have \(\alpha=0^\circ\); one barely grazing the surface has \(\alpha\) near \(90^\circ\). Geometrical solid angle supplies a factor \(\sin\alpha\); the projected flux through the surface supplies another \(\cos\alpha\). Normalizing gives

\\[ p(\alpha\mid\text{inward crossing})=2\sin\alpha\cos\alpha=\sin(2\alpha),\qquad 0\leq\alpha\leq90^\circ. \\]

Therefore, in the unfocused approximation, only \(\sin^2(15^\circ)\approx6.7\%\) of inward crossings are within 15° of directly radial, while roughly 93% are more oblique than that. This is a statistical statement about surface crossings, not an exact description of the velocity distribution inside 1 AU after strong gravitational focusing. It is particularly useful at outer radii (30 AU) where focusing changes the cross section by only about 6.6% in the benchmark.

Even when viewing a narrow sample with almost the same velocity direction, many parallel tracks will miss the Sun by entirely different distances. A directional headwind changes the relative numbers of orientations, not the need for a wide distribution of positions and impact parameters.

Perihelion statistics within a chosen spherical encounter region

Define the gravitational length \(\ell\equiv GM_\odot/v_\infty^2\). From the impact-parameter relation in Section 3,

\\[ b^2=q^2+2\ell q,\quad d\sigma=2\pi b\,db=2\pi(q+\ell)\,dq. \\]

Conditional on the object reaching a perihelion no larger than \(R\), this gives

\\[ \boxed{p(q\mid q\le R)=\frac{2(q+\ell)}{R^2+2\ell R},\qquad 0\le q\le R.} \\]

At 30 km/s, \(\ell\approx0.9857\) AU. In other words, even among the selected objects entering inside 30 AU, a large number have perihelia in the outer part of that region. The cumulative fraction penetrating inside any smaller radius \(x\le R\) is precisely

\\[ P(q\le x\mid q\le R)=\frac{x^2+2\ell x}{R^2+2\ell R}. \\]

That expression is the basis for the 0.310% inner-1-AU fraction, and makes clear why the original infographic's inward-pointing arrows were physically inaccurate.

How many 30-AU visitors are strongly deflected?

For a Sun-only encounter, the total turn angle is \(\delta=2\arctan(\ell/b)\). Among objects that enter the 30-AU sphere, \(b\) is distributed in area out to \(b_{\max}=\sqrt{30^2+2\ell(30)}\approx30.97\) AU. Therefore:

Total deflection thresholdFraction of 30-AU entrants whose trajectories turn at least that much
At least 10°13.2%
At least 30°1.41%
At least 60°0.304%
At least 90°0.101%

Illustrative model; turn angle refers to the angle between far-inbound and far-outbound velocity asymptotes. Thresholds follow from \(b_{\delta}=\ell/\tan(\delta/2)\) and the impact-parameter area distribution. These fractions are not a measurement of the Galaxy's ISO population.

So “the Sun slightly adjusts some trajectories” is a much better visual default than “everything converges on the Sun.” Close passes can bend dramatically—but close passes are an exceptional subset of all objects traversing a large Solar System volume.


Part II — Gravity bends selected paths; it does not collect everything

3. The simplest gravitational scattering calculation

For an object far from the Sun, let:

  • \(v_\infty\): incoming speed before solar gravitational acceleration;
  • \(b\): impact parameter (the offset of the original straight-line track);
  • \(q\): actual perihelion distance;
  • \(\mu=GM_\odot\): solar gravitational parameter;
  • \(\delta\): total deflection between inbound and outbound velocity asymptotes.

In an ideal Sun-plus-test-particle system, conservation of energy gives

\[ \frac12v_q^2-\frac{\mu}{q}=\frac12v_\infty^2, \qquad v_q^2=v_\infty^2+\frac{2\mu}{q}. \]

Conservation of angular momentum gives \(bv_\infty=qv_q\). Combining them,

\[ \boxed{b^2=q^2\left(1+\frac{2\mu}{qv_\infty^2}\right)}. \]

Thus an incoming trajectory with \(b>q\) can nevertheless reach perihelion \(q\), because gravity deflects it inward. The hyperbolic-scattering angle is

\[ \boxed{\delta=2\arctan\left(\frac{\mu}{bv_\infty^2}\right)}. \]

These expressions follow from Newtonian two-body mechanics. They assume the object is gravitationally small compared with the Sun and ignore planetary gravity and outgassing.

Worked example: an object arriving at 30 km/s

Actual perihelion \(q\)Original straight-line offset \(b\)Total idealized deflection \(\delta\)Interpretation
0.1 AU0.455 AU130.4°Very close solar flyby; sharp turn
1.0 AU1.724 AU59.5°Noticeable bending
5.2 AU6.107 AU18.3°Moderate bending
30 AU30.970 AU3.65°Almost straight on Solar System scale

Important: This table selects trajectories by perihelion; it does not say most objects have one of these perihelia. The table describes what gravity would do if the object reached that particular distance.

Very distant flybys with still larger impact parameters are deflected even less. High incoming speeds also produce smaller deviations for comparable impact parameters.

4. Gravitational focusing is an enlargement of the target cross section

For an isotropic flux of interstellar test bodies, the number crossing inward through a sphere of radius \(R\) per unit time is

\[ \dot N(<R)=n_\infty v_\infty\,\sigma_{\rm eff}(R), \]

with effective cross section

\[ \boxed{\sigma_{\rm eff}(R)=\pi R^2\left(1+\frac{2GM_\odot}{Rv_\infty^2}\right)}. \]

Interpretation: The Sun acts like a gravitational lens for trajectories, increasing the number whose perihelia fall within a specified radius. It is misleading to draw this as the Sun accelerating all surrounding objects directly inward.

At \(v_\infty=30\) km/s, the focusing factor is approximately 20.7 at 0.1 AU, 2.97 at 1 AU, 1.38 at 5.2 AU, and 1.066 at 30 AU. The inner-region cross section is enhanced more strongly than the outer-region cross section, but the inner region is still vastly smaller.

The missing concept in the incorrect infographic: broad impact parameters

For uniformly distributed incoming tracks, the number of paths in an impact-parameter ring \([b,b+db]\) is proportional to its area:

\[ d\sigma=2\pi b\,db. \]

For the set of trajectories entering a given maximum impact parameter \(b_{\max}\), the conditional distribution is

\[ p(b\mid b\le b_{\max})=\frac{2b}{b_{\max}^2}. \]

That makes large offsets common within any chosen encounter cross section. There is no reason to make all the tracks point centrally. The only selection for entering an inner region is whether \(b\) is below its gravitationally focused threshold.

For any set of objects entering the 30-AU sphere in this benchmark model, the fractions that subsequently reach smaller spheres are:

Smaller sphere reachedFraction of the 30-AU entrants
Inside Jupiter's orbital radius (5.2 AU)3.89%
Inside Earth's orbital radius (1 AU)0.310%
Inside 0.1 AU0.0216%

Hence ~99.69% of incoming 30-AU entrants do not reach inside 1 AU. This is the central quantitative demonstration that the Solar System is not a funnel forcing every object into the middle.

When does solar focusing matter over Galactic distances?

The Sun influences motion at all distances in the idealized inverse-square problem, but the acceleration becomes extremely weak at large distances. There is no hard boundary where the gravitational force turns on. At approximately 30 AU, the speed increase due to the Sun is modest for a 30-km/s object, and the change in direction of a trajectory with perihelion there is only a few degrees. Beyond the planetary region, Galactic tides and stellar encounters become increasingly important over long times.


Part III — Estimating the number of unseen macroscopic objects

5. What observations actually constrain the abundance?

Two influential density estimates are often quoted:

  1. Do, Tucker & Tonry (2018) modeled the Pan-STARRS sensitivity and inferred about 0.2 AU⁻³ for ʻOumuamua-like objects. It is not a direct volume census; it depends on the object's brightness, approach geometry, and survey sensitivity. [R01]
  2. Jewitt & Seligman (2023) reviewed an order-of-magnitude value of approximately 0.1 AU⁻³ for ʻOumuamua-scale objects. [R02]

The two values are reasonably described as a benchmark scale, not a tight measurement. The uncertainties are much broader than the apparent precision of the decimals. With one detected inactive-looking object underlying much of the early estimate, Poisson counting uncertainty alone is substantial; systematic uncertainty from albedo, physical size, lightcurve, and detection pipeline is also large.

An earlier survey-based upper limit from Engelhardt et al. (2017) applied to a differently modeled, larger/brighter comet-like population and is therefore not interchangeable with a small, inactive ʻOumuamua-like abundance estimate. [R03]

Why the three discovered objects are not three identical statistical trials

  • 1I/ʻOumuamua: unresolved and inactive-looking, despite measurable nongravitational acceleration. Its characteristic radius is often estimated near 0.08–0.10 km; it is incorrect to silently treat this as a strictly measured 100-m diameter cutoff. [R04] [R05]
  • 2I/Borisov: an active comet, its coma boosting visibility. [R06]
  • 3I/ATLAS: an active comet, independently much larger than ʻOumuamua according to a 2026 Hubble extraction under a specified albedo assumption. [R07] [R08]

Treating the detections as identically selected, equally bright asteroids with the same precisely known diameter would yield misleading population statistics. A 2025 Hubble study highlighted how strongly the coma complicates turning 3I's detection into a nucleus-size population estimate. [R07]

6. Explicit assumptions for a reproducible toy model

We use one clean baseline to make the physical meaning of the estimates transparent.

ParameterBaselineMeaning and caveat
Far-field number density \(n_\infty\)0.1 AU⁻³Illustrative ʻOumuamua-like normalization from the literature, not an observed cumulative \(D\ge 100\) m distribution
Incoming speed \(v_\infty\)30 km/s = 6.3285 AU/yrOne speed rather than a realistic Galactic distribution
Directional distributionIsotropicUseful baseline; reality is structured
Gravitational dynamicsSolar two-body gravityIgnores perturbations from planets and other stars during the passage
PopulationUnbound macroscopic ISOs resembling the discovery-inferred brightness/size scale of ʻOumuamuaNot dust, nor all sizes added together
ProcessSteady stateAssumes density and speed distribution do not vary significantly over crossing times

The Sun's gravitational parameter is \(GM_\odot\simeq39.4769\) AU³/yr², and 1 AU is approximately 149.6 million km.

6.1 Annual inward entry rate

\[ \boxed{\dot N(R)=n_\infty v_\infty\pi R^2\left(1+\frac{2GM_\odot}{Rv_\infty^2}\right)}. \]

Here \(\dot N\) counts inbound passages into a sphere. It does not count an outbound passage again. It also does not mean the object intersects the ecliptic or comes near a planet.

6.2 Number present inside the sphere at one instant

For an isotropic, mono-speed far-field distribution, conservation of phase-space density gives the spatial number-density profile

\[ \boxed{n(r)=n_\infty\sqrt{1+\frac{2GM_\odot}{rv_\infty^2}}}. \]

The total expected instantaneous number inside radius \(R\) is

\[ \boxed{N(<R)=4\pi\int_0^R n(r)\,r^2\,dr}. \]

This is a separate calculation from the annual crossing rate. Counting objects present now and counting new objects arriving each year are not interchangeable; a body can spend days to years inside a particular radius before leaving.

6.3 Baseline results

Sun-centered spherical boundaryInbound crossings/yearMean number present at any instantMean residence time inside sphere
0.1 AU0.4120.002272.0 days
1 AU — radius of Earth's orbit5.910.82350.9 days
5.2 AU — radius of Jupiter's orbit74.173.50.992 year
30 AU — radius of Neptune's orbit1,90711,8516.21 years

Caution about interpretation: Those are model expectations. There are not literally 0.823 objects at an instant; 0.823 is the ensemble average. The actual count fluctuates between integers, and real-world parameters introduce much larger uncertainty than sampling fluctuations alone.

In particular: A sphere with a radius of 30 AU has a vast 3D volume. The planetary system is a thin set of orbital tracks inside that volume. An object may pass 25 AU above the Sun, never come near a planet, and still be counted as inside the 30-AU sphere.

6.4 Sensitivity to population density

Every count and annual rate above is directly proportional to the assumed \(n_\infty\). For example:

Assumed densityInbound/year inside 1 AUInbound/year inside 30 AUPresent within 30 AU
0.02 AU⁻³1.183812,370
0.10 AU⁻³5.911,90711,851
0.20 AU⁻³11.823,81423,702

This is a sensitivity table, not a 95% confidence interval. The real size-dependent population could lie outside these examples.

6.5 Sensitivity to incoming speed

Keeping \(n_\infty=0.1\) AU⁻³ but changing the single representative speed:

Incoming speedInside 0.1 AU /yrInside 1 AU /yrInside 5.2 AU /yrInside 30 AU /yr
20 km/s0.607.2066.41,369
30 km/s0.415.9174.11,907
50 km/s0.275.67101.83,053
60 km/s0.245.94117.73,637

This nonmonotonic response in the inner Solar System is physically informative: at low speeds, gravitational focusing becomes more effective; at high speeds, more bodies cross a large target area per unit time. Both effects contribute.

For a real velocity distribution \(f(\mathbf v)\), we would calculate

\[ \dot N(R)=n_\infty\int d^3v\,f(\mathbf v)\,|\mathbf v|\,\pi R^2\left(1+\frac{2GM_\odot}{R|\mathbf v|^2}\right), \]

where \(f\) is normalized over velocities in the Sun's frame. This is one reason a single “objects per year” figure should be treated as a model output rather than as a precisely known astronomical fact.

6.6 A useful independent consistency check: rate × residence time

For the 30-AU sphere:

\[ 1{,}907\ \mathrm{year}^{-1}\times6.21\ \mathrm{years}\approx11{,}851. \]

For the 1-AU sphere:

\[ 5.91\ \mathrm{year}^{-1}\times0.139\ \mathrm{year}\approx0.823. \]

That agreement is expected: in steady state, the mean inventory equals the incoming rate multiplied by the mean time spent in the region. This is an application of a general counting principle, not independent evidence validating the assumed population density.

6.7 An interesting scale: average separation and mass

If the far-field density really were 0.1 objects/AU³, a characteristic spacing would be \(n^{-1/3}\approx2.15\) AU. The objects would still almost never be close to one another, because their individual sizes are only hundreds of meters while their separations are hundreds of millions of kilometers.

As another illustration only, a spherical 200-meter-diameter object with bulk density 1,000 kg/m³ would have mass about \(4.19\times10^9\) kg. If all 11,851 modeled bodies were of that mass, their combined instantaneous mass inside 30 AU would be about \(4.96\times10^{13}\) kg, only about \(8.3\times10^{-12}\) of Earth's mass. Their collective gravity would be negligible. This is not a composition, size-distribution, or shape measurement of the ISO population.

6A. What “one detected object” can and cannot statistically prove

The original density estimate was enabled by the appearance of one bright enough, fast enough, correctly positioned and linkable visitor in a survey with limited sensitivity. The relevant statistical quantity is not simply “objects per year”; it is

\\[ \lambda(\Theta)=\int dD\,d^3v\,d^3x\,dt\;f_{\rm pop}(D,\mathbf x,\mathbf v,t\mid\Theta)\,S(D,\mathbf x,\mathbf v,t), \\]

where \(S\) is the telescope-and-pipeline selection function and \(\Theta\) describes the unknown population. For \(k\) detected events, a simple Poisson discovery likelihood is \(P(k\mid\lambda)=e^{-\lambda}\lambda^k/k!\). If \(\lambda\) is directly proportional to a density normalization \(n\) and exactly one object was discovered, the maximum-likelihood \(\lambda\) is one—but its statistical uncertainty is enormous.

As a pedagogical Poisson-only exercise, the classical exact 95% interval on the expected event count after one event is approximately 0.025 to 5.57 events. This alone spans a factor of more than 200 between endpoints; survey-calibration and physical-model uncertainties can widen it further. Do not apply this interval as an official confidence interval on Do et al.'s density: their inference depends on a richer detectability model, and later ISO discoveries do not share an identical survey selection function. [R01] [R04]

How detections can remain rare even if entries are common

For example, if a region has a model-based passage rate \(\sim6\) objects per year and each passage has an independent 1% chance of satisfying a particular telescope-and-pipeline detection requirement, the expected detections in ten years are \(6\times10\times0.01=0.6\). In a Poisson approximation, the probability of zero detections is \(e^{-0.6}\approx55\%\). This example does not assert that the real efficiency is 1% or that all past surveys are equivalent. It simply demonstrates why physical passage rate and published discovery rate cannot be interchanged.

Magnitudes are not diameters

For an inactive reflective body of fixed phase behavior, absolute optical magnitude has a relationship of the form \(D\propto p^{-1/2}10^{-H/5}\), where \(p\) is geometric albedo and \(H\) is the absolute magnitude. A factor of four uncertainty in assumed albedo corresponds to a factor of two in inferred diameter. For an active comet, coma brightness further undermines the conversion from observed magnitude to nucleus size.

A properly inferred size-frequency distribution would model brightness, colors, activity and albedo jointly, rather than assign every ISO one diameter and fit a slope to the resulting three numbers. [R01] [R02] [R07] [R08]

6B. Cross-checks, assumptions, and dimensions

An extremely useful quantity is the gravitational length \(\ell=GM_\odot/v_\infty^2\), which is about 0.986 AU at 30 km/s. With \(\sigma=\pi(R^2+2\ell R)\), the dimensions work out explicitly: \(n\) has units AU⁻³, \(v\) has AU/year, \(\sigma\) has AU², so \(nv\sigma\) is per year. All reported “entries/year” numbers are for one inbound boundary crossing. In the Sun-only unbound model, the same body later makes one outbound crossing, but that exit is not a new arrival.

The density formula used for instantaneous inventory follows from Liouville's theorem for a uniform, isotropic, mono-speed distribution at infinity. At distance \(r\), trajectories have speed \(\sqrt{v_\infty^2+2GM_\odot/r}\). Mapping the initially thin, isotropic velocity-shell distribution to the local shell gives

\\[ \frac{n(r)}{n_\infty}=\frac{\sqrt{v_\infty^2+2GM_\odot/r}}{v_\infty}. \\]

This is not a general result for a directional stellar stream, and it requires a stationary ensemble with inward and outward unbound populations, no loss at the Sun, no planet scatterings and no collisions. The integral for population inventory is well behaved at the origin in the point-mass approximation; physically, the Sun has a finite radius and some extremely close trajectories would strike it, a negligible correction to the 30-AU results.

Why ~1,900/year differs from the review's ~1,000/year: Jewitt & Seligman's rough estimate uses a \(\sim10\)-year crossing time and \(\sim10^4\) objects inside Neptune. The present report uses an exact stated mono-speed gravitational cross section for inward crossings, with a model-derived residence time \(\sim6.2\) years, giving \(\sim1.9\times10^3\) annual entries. These are compatible order-of-magnitude calculations with different averaging approximations—not competing measured rates. [R02]

A numerical consistency check for the outer-planet region

A sphere of radius 30 AU has volume \((4\pi/3)30^3\approx113{,}097\;\mathrm{AU}^3\). Multiplying by the far-field density 0.1 AU⁻³ produces roughly 11,310 objects if gravity is ignored. Solar focusing adds about 541 to the computed steady-state inventory, for ~11,851 objects total. This modest enhancement of a vast volume is much more intuitive than imagining a giant inward funnel.


Part IV — The three confirmed macroscopic visitors

7. Actual observed orbital geometries

The following parameters are rounded descriptions based on published orbital determinations; osculating orbital elements shift slightly with data arc, epoch, heliocentric versus barycentric reference frame, and fitted nongravitational forces. [R04] [R06] [R07] [R19]

ObjectDiscoverySpeed at infinityPerihelion \(q\)Ecliptic inclination \(i\)Least angle between orbital planesNature
1I/ʻOumuamuaOct. 2017~26 km/s~0.256 AU~122.7°~57.3°Inactive-looking; nongravitational acceleration
2I/BorisovAug. 2019~32 km/s~2.007 AU~44.05°~44.05°Active comet
3I/ATLASJuly 2025~58 km/s~1.356 AU~175.1°~4.9°Active comet, retrograde

Conclusion: Two out of three orbital planes are nowhere near coplanar with the ecliptic. The premise that all three confirmed visitors have traveled along the Solar System's disk is demonstrably incorrect. 3I/ATLAS is the unusual close-to-coplanar example. [R05] [R06] [R07] [R19]

7.1 1I/ʻOumuamua

Its trajectory is securely hyperbolic, and its discovery by Pan-STARRS after perihelion emphasized the limited observability window for faint objects. Its inbound heliocentric excess speed was roughly 26 km/s. Despite its asteroid-like appearance, nongravitational acceleration complicated simple physical interpretations; a primary astrometric analysis by Micheli et al. (2018) directly documents that signal [R31]. The object illustrated how strongly detection volume depends on brightness and timing. [R04] [R05]

7.2 2I/Borisov

Its very hyperbolic orbit and obvious coma made it a different discovery class: a cometary object that brightens with gas and dust activity. It approached on a plane tilted by about 44° to the ecliptic and never approached closer than about 2 AU to the Sun. This argues against using one simple asteroid-detection efficiency for every interstellar visitor. [R06]

7.3 3I/ATLAS

3I was discovered July 1, 2025, and moved on an almost ecliptic but retrograde orbit. Its perihelion was about 1.356 AU, outside Earth's orbital radius. NASA reports that its closest approach to Earth was approximately 1.8 AU; it never represented an Earth impact threat. Its high hyperbolic excess speed is ~58 km/s. [R07] [R18] [R19] [R32]

A March 2026 Hubble-based nucleus extraction estimated an effective radius of 1.3 ± 0.2 km, conditional on assumed geometric albedo 0.04. The implied size should not be confused with an assumption-free physical measurement or with the diameter. [R08] A separate 2026 analysis of nongravitational acceleration and sublimation arrived at a smaller inferred radius, about 0.45–0.65 km. The difference illustrates how assumptions about albedo, activity and mass-loss physics remain consequential. [R26]

7.4 Don't infer a common origin from three unusual trajectories

The three objects have different velocities and orbital geometries. Galactic kinematic modeling of 3I suggests its pre-encounter motion is consistent with a population older than the likely source population of 1I, but such age inferences are model-dependent and do not identify a particular parent star. [R11]


Part V — How “random directions” really work

8. Incoming sky directions are not equivalent to random orbital planes

Consider two different probability questions under a simplified ideal population.

Question A: For an incoming velocity direction uniformly distributed over a celestial sphere, what fraction of radiants lie within \(\pm\theta\) of the ecliptic great circle?

\[ \boxed{P_{\rm radiant}(|\beta|<\theta)=\sin\theta}. \]

At 5°, this is 8.72%.

Question B: If an object's orbital angular-momentum direction is isotropic (so its orbital plane is random), what fraction of orbital planes lie within \(\theta\) of the ecliptic plane, counting both prograde and retrograde?

\[ \boxed{P_{\rm plane}(<\theta)=1-\cos\theta}. \]

At 5°, this is 0.3805%.

These are different random variables. An incoming radiant near the ecliptic does not imply an orbit near the ecliptic, and vice versa.

The probability of finding at least one orbital plane within 5° among three independently and isotropically distributed planes is

\[ 1-(\cos5^\circ)^3\approx\boxed{1.14\%}. \]

This is the appropriate simple-null-model probability for “one or more of three,” not 2.3%, and not a calibrated significance for our actual survey-selected three-object sample.

Why 3I's alignment cannot be assigned a clean surprise probability

The isotropic-plane model is pedagogical, not a realistic end-to-end survey model. A valid probability for detected objects must incorporate:

  1. the Galactic velocity distribution;
  2. the offset (impact-parameter) distribution and orbital orientation;
  3. Sun-centered gravitational focusing;
  4. sky brightness, solar elongation, Earth position and seasonal visibility;
  5. survey footprints and cadence;
  6. cometary activity and photometric behavior;
  7. streaking, detection and orbit-linking efficiency.

Moreover, a small sample invites a post-hoc interpretation problem: the particular “striking” alignment may be noticed after examining several possible peculiarities. One rare event in three does not demonstrate a preferred physical plane.

9. Solar apex and Galactic anisotropy: the correct sign

Imagine the Sun moving through a comparatively stationary local reservoir. In the Sun's rest frame, that reservoir appears as a headwind. An object ahead of the Sun arrives from the direction toward which the Sun is moving, called the solar apex.

  • The object's incoming radiant (sky position it comes from) tends toward the apex.
  • The object's inbound velocity vector points approximately away from the apex, toward the antapex.
  • Its outbound path after flyby is a different question because the Sun has bent its trajectory.

Confusing the radiant with the velocity vector reverses the conclusion. In published Galactic-population modeling, the solar apex is the expected enhanced incoming radiant, though details of the distribution depend on the adopted stellar and ISO kinematics. [R09] [R10]

This preference is not a narrow beam. The local population includes multiple stellar velocity dispersions and Galactic dynamical structures. The Ōtautahi–Oxford model draws on Gaia-derived stellar kinematics to predict a complex distribution in velocity and arrival direction; survey coverage further filters what observers see. [R10] [R11]

A key practical result from Dorsey et al. (2025) is that a realistic simulated Rubin sample does not preserve all sky directions with equal detection efficiency. Their modeled northern flux can interact with Rubin's predominantly southern footprint, so a raw discovered-sky map is not the true underlying distribution. [R10]

9A. Why “random” can look like a stream without implying anything is aimed at the Sun

A more general incoming population is described by a velocity-distribution function \(f_\infty(\mathbf v)\) in the Sun's frame. The annual entry rate is then

\\[ \dot N(R)=n_\infty\int d^3v\,f_\infty(\mathbf v)\,|\mathbf v|\,\pi R^2\left(1+\frac{2GM_\odot}{R|\mathbf v|^2}\right), \\]

with \(\int f_\infty d^3v=1\). Distinct components matter: older Galactic disk stars tend to have a different velocity dispersion from young stars; ejection speeds add to parent-system velocities; nearby streams can make the true radiant map uneven. The Sun's reference frame matters: the mean Galactic disk may produce an apparent headwind, but individual ISO velocities are not identical, and the solar apex itself depends on the reference population used. [R09] [R10] [R11]

A crucial fact sometimes missed in popular illustrations is that a spatially homogeneous wind consists of almost parallel trajectories with very different offsets, not of arrows that converge. A sea of particles with entirely random velocity directions likewise consists of broadly offset trajectories. Only those with small enough impact parameters for a chosen target sphere are mathematically included in its entry statistics.

Conditional samples are not an unbiased map of space

As soon as we say “objects inside Neptune's orbit,” we have already chosen a target region. The selection discards the vastly larger background population that misses that sphere. When we say “objects discovered by a particular telescope,” we apply further filters for brightness, timing, angular rate, field of view, weather and identification. Consequently, a chart showing only discovered flybys is highly conditioned; it is not a random diagram of all ISO tracks through a large surrounding cube.

One especially instructive discovery detail is 3I/ATLAS's position near the Galactic plane, where stellar crowding hinders some surveys. Seligman et al. (2025) reported that the ATLAS survey's coverage made that particular discovery possible, while certain deeper programs tend to avoid crowded Galactic-plane fields. That is a different bias from ecliptic-orbit preference. [R07]


Part VI — Does “many interstellar objects” include tiny rocks and dust?

10. Size determines the answer far more than people expect

Important correction about physical size: In early observations, ʻOumuamua's inferred mean/effective radius was roughly 80–100 m, given assumed reflectivity and shape; its effective diameter is therefore nearer 160–200 m in those models, not necessarily 100 m. Its unusual lightcurve also makes “the diameter” an oversimplification. The published \(0.1\)–\(0.2\,\mathrm{AU}^{-3}\) number is an inference for ʻOumuamua-like detectability/absolute brightness, not a hard measured census of every object bigger than exactly 100 m. This matters enormously when multiplying a steep size-frequency law. See the original observational analysis [R04] and the density modeling [R01–R02].

The benchmark \(0.1\) AU⁻³ density is not an abundance for all sizes; it is a normalization for roughly ʻOumuamua-scale objects under specific assumptions. Extrapolating it to 1-meter rocks, pebbles or dust requires a size-frequency distribution, which remains poorly constrained.

One common model specifies differential counts

\[ \frac{dN}{dD}\propto D^{-q_s}, \]

which implies cumulative counts

\[ N(\geq D)\propto D^{1-q_s}\quad(q_s>1). \]

For a lower size threshold ten times smaller, a slope \(q_s=3\) predicts 100 times as many bodies, not one-hundredth as many. It is essential to keep this sign correct.

Illustrative extrapolation from an explicitly hypothetical 200-m diameter threshold

Assumptions: all counts are cumulative (\(D\ge D_{\rm min}\)), we hypothetically set \(n(D\ge200\,\mathrm m)=0.1\) AU⁻³ (as an approximate proxy for ʻOumuamua-scale brightness), fix velocities, and extrapolate a single power law without a break. **This normalization is assumed for illustration and is not directly established by observations.** None of these small-size rates is a direct observation.

Differential index \(q_s\)Inside 1 AU/year, ≥200 m≥20 m≥2 m
2.55.91875,908
3.05.959159,076
3.55.91,868590,759

The orders-of-magnitude spread is the scientific point. This table should not be used to claim approximately 590,000 bodies of diameter at least 2 m definitely arrive each year; the power-law may break, and different surveys constrain different size ranges and compositions.

What about kilometer-scale comets?

Extending the revised illustrative model from a 200-m reference threshold upward to \(D\ge1\) km would give approximately 0.53, 0.24, or 0.106 inner-1-AU crossings per year for the three slope choices above. Those figures result from multiplying 5.91/year by \((1000/200)^{1-q_s}\). These are only mathematical examples and are not a substitute for active-comet population analyses. 2I/Borisov and 3I/ATLAS are much more readily detectable when active; their selection functions and nucleus sizes differ from those of inactive ʻOumuamua-like bodies. [R03] [R07] [R08]

11. The measured dust population is not a smooth continuation of the large-body estimate

Spacecraft such as Ulysses directly observed microscopic interstellar dust. Dust behaves differently because solar radiation pressure, Lorentz forces on electrically charged grains, and the heliospheric magnetic environment can all alter trajectories. Large rocks are dominated by gravity; micrometer grains need not be. [R14] [R15]

A 2025 paper by Peña-Asensio & Seligman, The interstellar flux gap: From dust to kilometer-scale objects, compared spacecraft dust fluxes, meteor limits, and macroscopic objects. It found that a simple single-slope distribution across these sizes conflicts with the data, with differences of approximately 2–7 orders of magnitude between dust measurements and some extrapolated predictions. [R16]

Consequently, phrases such as “there must be millions of interstellar rocks passing Earth daily” are not supported by the three large-object discoveries alone. The number of tiny particles may be very high, but particles and large bodies cannot be counted by blindly extending one inferred power law.

12. What meteor surveys add

Ground-based radar and optical meteor surveys provide constraints on the intermediate-size population. Establishing an object as interstellar from a meteor is challenging because Earth's motion, atmospheric deceleration, measurement errors, and borderline solar escape speeds can produce apparently hyperbolic solutions. A 2020 Canadian Meteor Orbit Radar analysis found only a handful of candidates in a large data set and carefully assessed uncertainties; they are not equivalent in confidence to the three tracked telescopic visitors. [R17]

The 2025 flux-gap analysis argues that existing meteor limits make some simple large-to-small size extrapolations untenable. Improving the intermediate-size census is therefore critical to understanding where the large-object and dust populations connect—if they form one population at all. [R16]


Part VII — Why the detection count is so tiny

13. Observing a visitor is not the same as its being physically present

Imagine 12,000 benchmark objects scattered over a sphere 60 AU across. Most are small, dark, and much too distant for a survey to detect. Even an object inside 1 AU of the Sun may be far from Earth, poorly illuminated, near the Sun in Earth's sky, or detectable for only a short window.

For an inactive reflective body, the received optical flux approximately scales with

\[ F\propto\frac{p D^2\,\Phi(\alpha)}{r^2\Delta^2}, \]

where \(p\) is reflectivity, \(D\) is diameter, \(\Phi(\alpha)\) is a phase function, \(r\) is Sun-object distance, and \(\Delta\) is observer-object distance. This is schematic, not a full thermophysical or survey photometric model.

The detection probability is therefore a strong function of many variables, not a constant applied to every ISO:

\[ P_{\rm detect}=P_{\rm detect}(D,p,r,\Delta,\alpha,\mathbf v,\Omega,t,\mathrm{activity},\mathrm{survey}). \]

Important survey filters

Faintness: especially severe for dark inactive bodies. Solar elongation and daylight: most telescopes cannot observe close to the Sun's apparent sky position. Sky coverage: a survey covers particular declinations and sky regions with particular cadence. Seasonal geometry: Earth moves while the object passes through. Rapid apparent motion: the object can streak across detector pixels, reducing signal and linking effectiveness. Orbit linking: one moving dot is insufficient to identify a secure hyperbolic orbit. Cometary activity: gas/dust increases apparent brightness without necessarily increasing nucleus size. Weather and downtime: further reduce effective visibility.

Do et al. explicitly modeled detection cross-section and trailing losses for Pan-STARRS. Walker & Seligman (2026) investigated apparent sky-plane motion distributions and showed why extremely fast-moving ISOs may evade existing detection pipelines. [R01] [R13] [R34]

Why “near ecliptic in a survey” does not automatically mean “near-ecliptic orbit”

Many asteroid searches concentrate on regions where ordinary Solar System asteroids are most likely found. But a high-inclination ISO can cross the ecliptic at its nodes and be discovered there. A snapshot of an object's location near the ecliptic does not establish a near-zero orbital inclination. Proper bias correction must simulate whole orbital trajectories and actual survey observations.

14. What Rubin Observatory can clarify

As of the November 2026 research cutoff, Rubin's published observing-strategy tracking indicates that its ten-year Legacy Survey of Space and Time began June 29, 2026. That matters: older forecasts describing the survey as “upcoming” should not be mistaken for its current operational status. [R20] [R33]

Forecasts still differ widely:

  • Marčeta & Seligman (2023): about 0–70 asteroidal ISOs per year across their examined assumptions, with large sensitivity to size-frequency slopes, albedos and trailing losses. [R12]
  • Dorsey et al. (2025): about 6–51 ISOs across the ten-year LSST under their different population and realistic discovery-criteria assumptions. [R10]
  • Hoover et al. (2022): roughly 0.9–1.9 detections/year under specific conservative and optimistic cases. [R09]

These are not three measurements and not necessarily contradictory: they model different physical populations, brightness functions, survey decisions and linking criteria. The disagreement is a useful warning that selection effects matter enormously.

The major gain will not merely be a larger number of objects. A more systematically selected sample can measure the joint distribution of sizes, speeds, arrival directions, activity and inclination—and may test whether the local incoming flux follows Galactic stellar kinematics.

14A. The intermediate-sized population: an especially valuable negative result

In the previous version the text focused mainly on a positive report of a few radar candidates. The deeper comparison needs a crucial non-detection study: Wiegert et al. (2025) searched high-quality Global Meteor Network data and found no conclusively hyperbolic millimeter-class meteoroid in the analyzed sample. They reported a model- and sensitivity-dependent upper limit of fewer than roughly one interstellar meteor per million interplanetary meteor impacts at millimeter scales, with further analysis needed. [R27]

This observation is not evidence that interstellar dust or macroscopic planetesimals do not exist: those are independently observed. Instead, it challenges simplistic assumptions connecting those populations across many orders of magnitude in diameter. The 2025 Peña-Asensio & Seligman analysis makes the same point explicitly: the microscopic, magnetic-field-sensitive interstellar-dust population and the macroscopic, gravitationally dominated planetesimal population should not automatically be fitted by one universal power law. [R16] [R27]

For a future census, the most informative observation could be several well-measured 1-cm to 10-m bodies, bridging the huge size gap between spacecraft dust detections and telescopic interstellar asteroids. Such detections would need credible pre-atmospheric heliocentric speeds, carefully propagated measurement errors and corrections for Earth's gravitational acceleration.


Part VIII — Earth encounters, capture, and physical consequences

15. Crossing Earth's orbital radius is not an Earth encounter

The 1-AU sphere used in the annual-rate model is centered on the Sun, not Earth. A body may cross that boundary on the other side of the Sun while Earth is elsewhere on its orbit. Even reaching a perihelion below 1 AU does not guarantee the object's path comes near Earth's location at the appropriate time.

A genuine Earth close-approach calculation requires the relative three-dimensional trajectory and the timing of Earth and object at their nearest separation. A collision adds the much smaller cross section of Earth itself, together with Earth's own gravitational focusing. These probabilities are far lower than the Sun-centered spherical entry rates. [R02]

For perspective, the 2023 review quotes an estimated 100-m-scale interstellar impact rate of order \(5\)–\(10\times10^{-9}\) per year (roughly one in 100–200 million years) from a particular population model. Earlier papers have given different estimates depending on the assumed abundance and impact cross section. These are theoretical order-of-magnitude rates, not a forecast that an impact is due. [R02] [R01]

15A. A transparent but highly idealized near-Earth encounter calculation

To quantify the geometry without pretending to predict impact risk, we can build an independent local illustration. From the baseline isotropic model, at 1 AU the Sun-focused number density is

\\[ n(1\,\mathrm{AU})=0.1\sqrt{1+\frac{2GM_\odot}{(1\,\mathrm{AU})v_\infty^2}}\approx0.172\;\mathrm{AU}^{-3}. \\]

Suppose, only for this example, the relative Earth–object speed is fixed at 60 km/s, or about 12.66 AU/year, and the local relative motion is adequately approximated by a straight-line flux across a small target. For a small encounter distance \(d\),

\\[ \Gamma(d)\approx n_{\rm local}\,v_{\rm rel}\,\pi d^2. \\]

Earth-centered encounter radiusIllustrative mean rateEquivalent time between events
0.01 AU (~1.50 million km)\(6.85\times10^{-4}\)/year~1,460 years
Mean lunar distance (~384,400 km)\(4.53\times10^{-5}\)/year~22,100 years
Earth-radius collision cross section, including Earth focusing\(1.29\times10^{-8}\)/year~78 million years

For the collision example, the extra factor is \(1+(v_{\rm esc,Earth}/v_{\rm rel})^2\), about 1.035 for 11.2 versus 60 km/s. These estimates are illustrative calculations for the same highly uncertain ʻOumuamua-like normalization; they are not a full Earth-impact model. In particular, the actual relative-velocity distribution near Earth is not isotropic/mono-speed, and Earth-bound collision cross sections cannot be inferred from simple Sun-centered rates by a single universal factor. The approximate Earth-impact timescale being tens of millions of years is at least consistent in order of magnitude with independent literature models [R01] [R02].

Interpretation: many thousands of objects inside a 30-AU Sun-centered sphere can coexist with very infrequent passages within a lunar distance of Earth. The difference is almost entirely geometry, not a mysterious repulsive effect.

16. What happens to them after the flyby?

For a purely two-body Sun–ISO encounter, an object arriving with positive specific orbital energy

\[ \epsilon=\frac12v_\infty^2>0 \]

leaves on another hyperbolic asymptote with the same \(v_\infty\) magnitude it had before the encounter. The Sun changes the direction of motion but does not permanently capture the object.

Planetary encounters are different. A close approach to a moving planet can exchange energy in the Sun-centered frame, in some cases temporarily or permanently trapping an interstellar body or ejecting it on a different trajectory. Such captures are a special, low-efficiency subset and are not included in our simplest single-Sun model. [R02] [R21]

Physical transformations are also possible. Ice sublimation, fragmentation, nongravitational outgassing, solar heating, and tidal interactions can change an individual object's brightness or motion. For small dust grains, magnetic fields and radiation pressure also matter. [R02] [R15]

16A. Could we have a hidden population of captured interstellar bodies?

An ISO can be unbound when it enters the planetary region and later become bound through three-body energy exchange—especially through an encounter with Jupiter. This is qualitatively different from solar gravitational focusing, which alone does not capture an initially unbound body in a conservative two-body orbit. A captured object also ceases to obey the simple “single inbound crossing, one outbound crossing” residence-time model.

A 2025 numerical investigation by Marquardt-Demen, Li & Wray, Capture of Interstellar Objects and Binaries, reported model-dependent possible captures and candidates; its prediction of approximately ten bound ISOs is not an observationally verified number. A related 2026 study argued that Jupiter-assisted captures can leave long-lived, difficult-to-identify bound bodies with particular orbital distributions. [R28] [R29]

A separate 2026 preprint by Raymond, Kaib & Hopkins investigated capture in stellar encounters: a passing star can weakly change the gravitational binding of very distant ISOs and leave some in the outermost Oort-cloud region. Such captures are extremely different from the brief flyby traffic inside Neptune's orbit, and estimates of retained objects are contingent on a long Galactic history model. [R30]

This distinction is worth maintaining in any census:

  • Free interstellar traffic: positive energy relative to the Sun, mostly one-pass hyperbolae.
  • Temporarily captured material: altered orbital energy after planetary or stellar interactions; residence time can be much longer.
  • Native long-period comets: formed in our own planetary system but may visit the same regions on nearly parabolic trajectories.

Without an unambiguous dynamical history or a unique chemical signature, some capture candidates cannot be classified by present-day orbital elements alone.


Part IX — What does all this imply about other planetary systems?

17. The Milky Way may contain an immense reservoir of ejected planetesimals

If the local density inferred from ʻOumuamua were representative of much of the Galactic disk, a vast number of similar bodies would exist throughout the Milky Way. Jewitt & Seligman (2023) discuss a population on the scale of 10²⁵–10²⁶ objects, with substantial uncertainty. [R02]

Where could they come from? Planet-forming disks produce many planetesimals. Interactions with giant planets, passing stars, binary stars, and stellar evolution can eject some of those bodies from their original systems. Over billions of years, multiple stellar generations can populate interstellar space with debris. The number and physical composition depend on formation and ejection efficiency, so the interstellar abundance can test planet-formation models. [R02] [R22]

There is a mass-budget puzzle: a very high density of 100-meter-scale bodies, when extrapolated over an entire Galaxy and a steep size distribution, can imply substantial mass ejected per star. But we should not assume the local density or single power law applies everywhere or at all sizes. [R01] [R02] [R16]

18. Arrival directions can encode Galactic history

In principle, the heliocentric speed and incoming radiant of an ISO contain information about how its velocity relates to local stellar populations. Models using Gaia stellar kinematics predict that the velocity distribution can reflect the ages and dynamics of the source populations. [R10] [R11]

But simply drawing a line backward through today's Solar System generally cannot identify the parent star: Galactic gravitational forces, long flight times, unknown ejection velocities and incomplete stellar phase-space histories accumulate uncertainty. An object's arrival direction is not a direct address for its birthplace. [R02] [R11]

19. Why three objects are simultaneously exciting and statistically inadequate

Three discoveries provide proof of existence and a preview of diversity: inactive-looking versus active, different velocities, different orbital inclinations, and different sizes. They do not provide a robust population-level velocity distribution, orbital-plane distribution, size-frequency slope, physical-activity fraction or precise yearly entry rate.

An apparent anomaly can seem statistically unlikely under one simplistic null model but become less surprising once one considers the actual selection function and multiple possible patterns. The most valuable next step is not to multiply improbabilities from post-hoc coincidences; it is to compare forward models of discovered samples with the real observations as survey data accumulate.


Part X — Audit of earlier claims and illustrations

20. Corrections that materially change the interpretation

Error 1: Drawing every trajectory pointed at the Sun. Correction: Object velocities and impact parameters both vary; most objects in a large Solar System volume have relatively distant flybys. For the baseline model, just 0.310% of inward crossings at 30 AU get inside 1 AU.

Error 2: Claiming that all three confirmed objects are aligned with the ecliptic. Correction: Their orbital inclinations are approximately 122.7°, 44.1°, and 175.1°. Only 3I's orbital plane is within 5° of the ecliptic (and its motion is retrograde).

**Error 3: Confusing near-ecliptic incoming radiants with near-ecliptic orbital planes. Correction: Their ideal isotropic probabilities are different: within 5° of the ecliptic great circle is 8.72%** for incoming radiants; within 5° of the ecliptic plane is 0.3805% for isotropic orbital-plane normals.

Error 4: Getting the three-object alignment probability wrong. Correction: The chance that at least one of three isotropically oriented orbit planes lies within 5° of the ecliptic is ~1.14%, not ~2.3%. This is not a survey-corrected probability.

Error 5: Reversing the solar-apex arrival direction. Correction: The incoming radiant enhancement is toward the solar apex; approaching objects' velocity vectors point generally the opposite way. These are two different directions.

Error 6: Reversing the size-distribution scaling. Correction: With \(N(\ge D)\propto D^{1-q_s}\) and \(q_s>1\), making the threshold smaller always increases cumulative counts. For \(q_s=3\), reducing the threshold from 100 m to 10 m multiplies predicted counts by 100, not 0.01.

Error 7: Reporting entries into a 1-AU sphere as near-Earth flybys. Correction: The former is a Sun-centered crossing criterion; the latter requires a much smaller Earth-centered distance and appropriate timing.

Error 8: Reporting extrapolated model results as measured counts. Correction: Approximately 12,000 inside 30 AU and approximately 1,900 entries per year are conditional on a particular density and speed model. We have only a tiny number of confirmed macroscopic objects, not a direct observed census.

Error 9: Treating small dust and km comets as one measured population. Correction: Distinct detection methods and a probable flux gap prevent a single robust all-size extrapolation.


Part XI — How to draw a physically faithful schematic

21. Detailed visual-design specification

A trustworthy scientific visualization should have at least six distinct panels with explicit assumptions and accurate direction conventions.

Panel A — Far-field interstellar flux (3D): Draw a large volume centered on the Sun containing many straight or nearly straight trajectories with various 3D velocity directions. Vary their offsets from the Sun. Most should neither converge on nor pass close to it. Do not use a spherical set of arrows aimed radially toward the Sun.

Panel B — Two-body gravitational scattering: Pick three incoming parallel tracks with different impact parameters as an illustrative controlled experiment: distant miss (almost straight), moderate flyby (bent), close perihelion (strongly bent). Mark \(b\), \(q\), \(\delta\), inbound and outbound asymptotes, and the Sun's location. Explain that this panel holds incoming direction fixed only to isolate the effect of impact parameter.

Panel C — Encounter cross sections: Draw concentric spheres for 1, 5.2 and 30 AU, not concentric filled planar rings as if all orbits lie in the ecliptic. Mark sample trajectories entering 30 AU but missing 1 AU. Clearly state the benchmark fraction reaching inside 1 AU: 0.31%.

Panel D — 3D orientation versus plane: Show the ecliptic as a thin transparent plane and multiple inclined orbital planes. Label the confirmed inclination angles; distinguish 3I's nearly coplanar retrograde orbit. Include the formula \(P_{\rm plane}(<\theta)=1-\cos\theta\), separate from \(P_{\rm radiant}(|\beta|<\theta)=\sin\theta\).

Panel E — Galactic anisotropy: A celestial-sphere map of incoming radiants, with the solar apex marked as an expected upwind enhancement. Draw velocity vectors in the physically opposite direction from the incoming sky positions. Label the map illustrative model, not a measured observed histogram.

Panel F — True population versus detected subset: Show the large, mostly undetected population throughout a huge volume and a narrow survey selection window favoring brightness, observational geometry, and acceptable apparent motion. Separate inactive asteroids from active comets. Make counts in the illustration proportional only if actual sampling is quantitatively defined.

Panel G — Numbers by radius and size: Present the annual influx, instantaneous populations, residence time, and model assumptions alongside uncertainty/sensitivity tables. Avoid labeling any of these as observed counts.

Panel H — Evidence ledger: Three confirmed large ISO discoveries, interstellar dust measurements, survey non-detections/limits, and planned survey gains, each visually distinguished by confidence level.

A graphic made using this specification should not show a broad annulus of arrows all converging toward the Sun. If particles all happen to be moving toward a central point, that is a different physical boundary condition than the interstellar background discussed here.


Part XII — Reproducibility and methods

22. Minimal numerical implementation

The following self-contained Python 3 example uses only the standard library. It reproduces the single-speed benchmark rates and inventories using a composite Simpson integration for the inventory. To avoid the integrable singularity at \(r=0\), the integration is performed in \(x=\sqrt r\), with \(r=x^2\).

from math import pi, sqrt, atan, degrees, cos, radians

# Units: AU and Julian year
N_INF = 0.1                     # AU^-3; hypothetical Oumuamua-like brightness/size class
V_INF = 6.3284857971           # AU/year = 30 km/s
MU = 39.4769264143             # AU^3/year^2 (GM_sun)


def focusing_factor(R):
    return 1 + 2 * MU / (R * V_INF**2)


def annual_entry_rate(R):
    return N_INF * V_INF * pi * R**2 * focusing_factor(R)


def inventory(R, n_steps=20000):
    # Transform r = x^2, dr = 2*x*dx.
    xmax = sqrt(R)
    step = xmax / n_steps

    def integrand(x):
        if x == 0:
            return 0.0
        r = x*x
        return 8*pi*N_INF*x*r*r*sqrt(1 + 2*MU/(r*V_INF**2))

    total = integrand(0) + integrand(xmax)
    for j in range(1, n_steps):
        total += (4 if j % 2 else 2) * integrand(j * step)
    return total * step / 3


for R in [0.1, 1.0, 5.2, 30.0]:
    rate = annual_entry_rate(R)
    count = inventory(R)
    print(f"R={R:>4g} AU: focus={focusing_factor(R):7.4f}, "
          f"entries/yr={rate:10.4f}, present={count:11.4f}, "
          f"residence years={count/rate:.5f}")

p_radiant_5 = __import__('math').sin(radians(5))
p_plane_5 = 1 - cos(radians(5))
p_one_of_three = 1 - cos(radians(5))**3
print(p_radiant_5, p_plane_5, p_one_of_three)

Expected benchmark output (rounded):

R= 0.1 AU: focus=20.7140, entries/yr=    0.4118, present=     0.0023, residence years=0.00552
R=   1 AU: focus= 2.9714, entries/yr=    5.9076, present=     0.8231, residence years=0.13933
R= 5.2 AU: focus= 1.3791, entries/yr=   74.1407, present=    73.5130, residence years=0.99153
R=  30 AU: focus= 1.0657, entries/yr= 1906.9202, present= 11850.8133, residence years=6.21464
0.0871557427 0.0038053019 0.0113725199

Limits of this implementation

This is not an N-body simulation, a complete survey simulator, or a Monte Carlo reconstruction of the Galactic ISO population. It deliberately assumes:

  • one incoming speed rather than the velocity distribution from Gaia;
  • an ideal isotropic far-field population;
  • no planetary scattering or binary stellar encounters;
  • no radiation pressure or electromagnetic grain forces;
  • no physical destruction, volatile activity, detection effects, or size evolution;
  • no local streams, seasonal visibility, or covariance between size and speed.

For a defensible population inference, a next-level program would sample \(\mathbf v_\infty\), an isotropic or data-driven impact parameter vector, a size/albedo/activity model, integrate heliocentric trajectories with planets, simulate Earth viewing geometry and the actual telescope visits, and fit the observed sample with a survey selection function. [R10] [R12] [R13]


Part XIII — Direct answers to the original questions

23. “Shouldn't these objects approach from almost any direction?”

Yes, in a broad geometric sense. No Solar System disk constrains their incoming directions. Their far-field velocity vectors are distributed in 3D, with some Galactic anisotropy. Crucially, each object also has an offset from the Sun's position. Most tracks are not directed at the center.

24. “Why would any be close to the Solar System's disk?”

There is no geometric prohibition. A particular orbital plane may by chance be within a few degrees of the ecliptic. In a simple isotropic-orbit orientation model, one within 5° has probability 0.38%; real detection probabilities differ because incoming velocities and survey selection are not isotropic.

25. “Why did the three visitors seem aligned?”

They were not all aligned. ʻOumuamua's plane was tilted by ~57° from the ecliptic as an unoriented plane, Borisov's by ~44°, and 3I's by ~4.9° (retrograde). A top-down 2D projection can hide substantial inclination differences. In addition, discovery near the ecliptic is not the same as having a near-ecliptic orbital plane.

26. “Does this mean immensely more stuff is coming through than we see?”

Very probably, yes. Survey-derived density estimates imply a much larger invisible macroscopic population. For a deliberately simple normalization, ~12,000 are present within 30 AU at any moment and ~1,900 per year cross inward through that boundary, while ~6 per year pass inside Earth's orbital radius. Actual numbers remain model-dependent. Most of these flybys are not near Earth.

27. “Would everything get pulled toward the Sun eventually?”

No. In an unbound Sun-only flyby, an ISO approaches, is gravitationally deflected, and escapes again. The Sun focuses encounter cross sections, not the eventual fate of the whole interstellar population. Most trajectories never approach sufficiently close for significant bending.

28. “What single observation would improve the answer most?”

A statistically useful, well-characterized sample from Rubin and other surveys, with their detection efficiencies modeled in detail, would constrain the population density and size-frequency distribution. Distributing discoveries across incoming radiants and orbital inclinations would also test the Galactic-kinematic predictions.


Part XIV — Questions that a better survey or simulation can actually answer

29. A prioritized research roadmap

A useful project should treat the interstellar population as a forward-inference problem, not simply count found comets. The following design separates independently testable hypotheses from assumptions:

QuestionMinimal observation/model requiredFailure mode to avoid
What is \(n(D,p,\text{activity})\)?Calibrated search completeness by size/brightness, including non-detectionsUsing three discoveries as three identical trials
Are incoming radiants isotropic?Galactic velocity prior plus full-sky, epoch-dependent selectionConfusing sky discovery coordinates with inbound radiant
Are orbital planes preferentially near the ecliptic?Debiased three-dimensional orbital sampleUsing flat 2D diagrams or post-hoc small probabilities
How many pass inside 1/5.2/30 AU?\(\int f(\mathbf v) v\sigma(v,R)\,d^3v\) with sampled impact parametersTreating all passing tracks as solar-directed
What happens at submeter scales?Confirmed hyperbolic meteor sample with accurate error modelContinuing a size law across dust/meteor/asteroid regimes
Are some visitors captured?Long-term N-body integrations and candidate dynamical historiesCounting captures as ordinary inbound hyperbolae
How many close-Earth encounters occur?Earth phase, relative geometry and gravitational cross sectionsConfusing a Sun-centered 1-AU sphere with a near-Earth tube

30. A reproducible forward-model architecture

A physically correct computational experiment would:

  1. Define a very large injection sphere with uniformly sampled positions and a chosen Galactic velocity distribution, not just arrows aimed at the Sun. A flux-weighted prescription must be used for boundary crossings; uniform angular coordinates on a sphere are not by themselves a correct injection measure.
  2. Sample impact parameters with area weight: conditional on \(b\le b_{\rm max}\), draw \(b=b_{\rm max}\sqrt U\) for \(U\sim\mathcal U(0,1)\), and a uniform azimuth around each velocity vector.
  3. Integrate trajectories under solar gravity; optionally add planetary N-body forces, nongravitational acceleration and physical fading. Confirm energy/angular-momentum conservation in the Sun-only limit.
  4. Record \(v_\infty\), incoming radiant, \(b\), perihelion \(q\), ecliptic inclination \(i\), velocity deflection \(\delta\), entry and exit epochs, and residence time for every region independently.
  5. Forward-project the population into the Earth's moving sky and apply a real observing footprint, cadence, photometric threshold, solar-elongation mask, streak losses and tracklet/orbit-linking requirements.
  6. Fit the simulated detected population to data while preserving the number of undetected injected tracks. Compare competing size laws and Galactic kinematic prescriptions, reporting posterior predictive checks instead of a single unqualified count.

The difference between this and the inaccurate infographic is fundamental: the point of departure is a 6D phase-space distribution, not vectors selected to intersect the Sun. [R09] [R10] [R12] [R13]

31. Worked inverse problem: what would 30 new Rubin detections tell us?

Suppose, purely as an illustration, that an accurately modeled Rubin simulation predicts 10 detections for a reference abundance \(n_0\), while the real survey finds 30 objects with matching distributions of sizes and trajectories. Under the simplest fixed-shape Poisson model the abundance maximum-likelihood estimate is \(3n_0\). But if instead the discoveries are systematically brighter and slower than the synthetic sample, simply increasing \(n_0\) may not fix the mismatch: the size law, activity fraction or speed distribution may also be wrong.

That's why characterizing the recovered orbit and brightness distributions can be more informative than multiplying a raw discovery count by a guessed correction factor. Dorsey et al.'s simulations explicitly find selection effects in velocity, perihelion and sky direction [R10]. The result of any actual survey needs an analysis of the detection pipeline as operated at the time, not merely a theoretical telescope limiting magnitude.

32. Decision-quality statements for future public explanations

Strongly supported: Interstellar small bodies exist; they approach on unbound three-dimensional trajectories; most far-field tracks miss the inner Solar System; gravity focuses selected passages; our macroscopic detections are severely incomplete; the three confirmed visitors have very different orbital-plane orientations. [R02] [R04] [R06] [R07]

Reasonable but model dependent: A population of order ten thousand ʻOumuamua-like bodies inside 30 AU at once; of order thousands of inward crossings per year at 30 AU; of order several/year inside 1 AU; Galactic solar-apex asymmetry; a substantial Rubin discovery yield. [R01] [R02] [R09] [R10] [R12]

Not yet settled: Precise rates for 1-m and 10-m ISOs, physical size slope across many orders of magnitude, the fraction of active vs inactive objects, a calibrated ecliptic-plane alignment probability for the detected sample, and the number of captured interstellar planetesimals. [R16] [R27] [R28] [R30]

Not supported: All ISO tracks pointing at the Sun; every visitor entering the 1-AU sphere passing near Earth; three objects proving ecliptic alignment; a precise all-size ISO census; a single dust-to-kilometer-size power law already established by observation.

Sources and annotated bibliography

Source key: Entries below identify the evidence used for observational claims. Model calculations in this report were recomputed from stated assumptions, and their results are not attributed as measured facts to these publications. When a paper is a simulation or forecast, that limitation is stated explicitly.

R01 — Do, A., Tucker, M. A., & Tonry, J. (2018). Interstellar Interlopers: Number Density and Origins of ʻOumuamua-like Objects. Astrophysical Journal Letters. Pan-STARRS detection-volume model; inferred ~0.2 AU⁻³; explains survey assumptions and mass budget. <https://arxiv.org/abs/1801.02821>

R02 — Jewitt, D., & Seligman, D. Z. (2023). The Interstellar Interlopers. Annual Review of Astronomy and Astrophysics, 61, 197–236. Broad peer-reviewed review of observations, dynamics, physical properties, capture, meteor constraints, and inferred ~0.1 AU⁻³ scale. <https://doi.org/10.1146/annurev-astro-071221-054221>

R03 — Engelhardt, T., et al. (2017). An Observational Upper Limit on the Interstellar Number Density of Asteroids and Comets. Earlier survey non-detection constraints for modeled active/inactive populations; different size/brightness assumptions from R01. <https://arxiv.org/abs/1702.02237>

R04 — Meech, K. J., et al. (2017). A brief visit from a red and extremely elongated interstellar asteroid. Nature, 552, 378–381. Primary discovery and early orbital measurements for 1I; the linked manuscript repository uses a slightly different manuscript title. <https://doi.org/10.1038/nature25020> (published article); <https://pmc.ncbi.nlm.nih.gov/articles/PMC8979573/> (author manuscript).

R05 — NASA Science (2017; maintained). ʻOumuamua. Public discovery summary and explanation of the first confirmed interstellar body. <https://science.nasa.gov/solar-system/comets/oumuamua/>

R06 — Guzik, P., Drahus, M., Rusek, K., et al. (2020; published online 2019). Initial characterization of interstellar comet 2I/Borisov. Nature Astronomy, 4, 53–57. Hyperbolic orbit, ~32 km/s excess speed, and coma observations. <https://www.nature.com/articles/s41550-019-0931-8>

R07 — Seligman, D. Z., et al. (2025). Discovery and Preliminary Characterization of a Third Interstellar Object: 3I/ATLAS. Astrophysical Journal Letters. Discovery, eccentricity, ~58 km/s, inclination and perihelion. <https://doi.org/10.3847/2041-8213/adf49a>

R08 — Hui, M.-T., Jewitt, D., Mutchler, M. J., Agarwal, J., & Kim, Y. (2026). Nucleus and Postperihelion Activity of Interstellar Object 3I/ATLAS Observed by the Hubble Space Telescope. Astrophysical Journal Letters. Effective radius 1.3 ± 0.2 km conditional on an assumed albedo. <https://doi.org/10.3847/2041-8213/ae471c>

R09 — Hoover, D. J., Seligman, D. Z., & Payne, M. J. (2022). The Population of Interstellar Objects Detectable with the LSST and Accessible for In Situ Rendezvous with Various Mission Designs. Planetary Science Journal. Galactic kinematic radiants, incoming solar-apex preference, and survey-yield estimates. <https://doi.org/10.3847/PSJ/ac58fe>

R10 — Dorsey, R. C., et al. (2025). The Visibility of the Ōtautahi–Oxford Interstellar Object Population Model in LSST. Planetary Science Journal, 6, 214. Realistic survey/kinematic modeling; 6–51 predicted ISO discoveries over LSST; directional selection effects. <https://doi.org/10.3847/PSJ/adf8ca>

R11 — Hopkins, M. J., et al. (2025). From a Different Star: 3I/ATLAS in the Context of the Ōtautahi–Oxford Interstellar Object Population Model. Astrophysical Journal Letters. Gaia-related kinematics, 3I's velocity and Galactic context; model-dependent age implications. <https://arxiv.org/abs/2507.05318>

R12 — Marčeta, D., & Seligman, D. Z. (2023). Synthetic Detections of Interstellar Objects with the Rubin Observatory Legacy Survey of Space and Time. Planetary Science Journal, 4, 230. Survey simulation predicting 0–70 detections/year across differing assumptions; trailing-loss limitations. <https://doi.org/10.3847/PSJ/ad08c1>

R13 — Walker, C. E., & Seligman, D. Z. (2026). Sky-Plane Velocity Distributions of Interstellar Objects and Implications for Their Detection. Survey sensitivity to apparent angular velocity and orbit linking. <https://arxiv.org/abs/2606.05344>

R14 — Grün, E., et al. (1993). Discovery of Jovian dust streams and interstellar grains by the Ulysses spacecraft. Nature, 362, 428–430. Early direct evidence for interstellar dust in the planetary region. <https://www.nature.com/articles/362428a0>

R15 — Baalmann, L. R., et al. (2025). Interstellar dust measured in situ by Ulysses: New aspects of the particle size distribution and its modulation by the heliosheath. Astronomy & Astrophysics, 698, A138. Dust measurements and heliospheric filtering. <https://doi.org/10.1051/0004-6361/202453532>

R16 — Peña-Asensio, E., & Seligman, D. Z. (2025). The interstellar flux gap: From dust to kilometer-scale objects. Astronomy & Astrophysics, 704, L1. Why a universal power law linking dust and large interlopers is unsupported. <https://doi.org/10.1051/0004-6361/202557337>

R17 — Froncisz, M., Brown, P., & Weryk, R. (2020). Possible Interstellar Meteoroids Detected by the Canadian Meteor Orbit Radar. Meteor candidate analysis and error assessment. Candidates should not be equated to secure telescopic detections. <https://arxiv.org/abs/2005.10896>

R18 — NASA Science (2025–2026; maintained). 3I/ATLAS Facts and FAQs. Discovery dates, speed evolution, comet properties, and closest Earth distance. <https://science.nasa.gov/solar-system/comets/3i-atlas/3i-atlas-facts-and-faqs/>

R19 — International Asteroid Warning Network / JPL (2025; orbit solution shown). 3I/ATLAS Orbit. Published q, e, i, perihelion epoch and visualization; orbit solutions update with additional astrometry. <https://iawn.net/obscamp/3I_ATLAS/3I_orbit.shtml>

R20 — Rubin Observatory Survey Strategy (2026; living documentation). Survey Progress. Records LSST start date June 29, 2026 and ongoing baseline strategy. <https://survey-strategy.lsst.io/progress/index.html>

R21 — Jewitt & Seligman (2023), section on interstellar capture. Within R02; summarizes Galactic tides, planetary captures and low capture efficiencies. <https://doi.org/10.1146/annurev-astro-071221-054221>

R22 — Moro-Martín, A., Turner, E. L., & Loeb, A. (2009). Will the Large Synoptic Survey Telescope Detect Extra-Solar Planetesimals Entering the Solar System? Astrophysical Journal, 704, 733. Pre-discovery theoretical context linking interstellar planetesimals to planetary-system ejection. <https://doi.org/10.1088/0004-637X/704/1/733>

R23 — NASA JPL Solar System Dynamics (maintained). Small-Body Orbits & Ephemerides. Official reference describing orbital solutions, uncertainties and Horizons ephemerides. <https://ssd.jpl.nasa.gov/sb/orbits.html>

R24 — de León, J., Licandro, J., de la Fuente Marcos, C., et al. (2020). Visible and near-infrared observations of interstellar comet 2I/Borisov ... Monthly Notices of the Royal Astronomical Society, 495, 2053. Detailed orbital element table including \(i\approx44.05^\circ\) and \(q\approx2.007\) AU. <https://academic.oup.com/mnras/article/495/2/2053/5827545>

R25 — Pomazan, A., et al. (2026). The Trajectory of 3I/ATLAS Through the Inner Solar System: Close Encounters and Possible Collisions. Astronomical Journal, 171, 96. Detailed implications of 3I's nearly coplanar retrograde orbit for interactions with native small bodies. <https://doi.org/10.3847/1538-3881/ae2ea6>

R26 — 2026 independent post-perihelion analysis of 3I/ATLAS. Post-perihelion photometric and spectroscopic monitoring of the interstellar comet 3I/ATLAS. Monthly Notices of the Royal Astronomical Society, 552(3). Nongravitational force modeling yields radius estimates of approximately 0.45–0.65 km, illustrating unresolved differences between physical inference methods. <https://academic.oup.com/mnras/article/552/3/stag1800/8832565>

R27 — Wiegert, P., Tran, V., Gregg, C., Vida, D., & Brown, P. (2025). An Upper Limit on the Interstellar Meteoroid Flux at Video Sizes from the Global Meteor Network. Searches high-quality meteor orbits; no conclusively hyperbolic detections; upper limit at ~5-mm scale. <https://arxiv.org/abs/2503.19042>

R28 — Marquardt-Demen, R., Li, G., & Wray, J. J. (2025 preprint). Capture of Interstellar Objects and Binaries. Simulation of Jupiter gravitational braking and capture; numerical candidate counts are model results, not confirmed extrasolar identities. <https://arxiv.org/abs/2509.05905>

R29 — Marquardt-Demen, R., Li, G., & Wray, J. J. (2026 manuscript / institutional copy). Candidate Captured Interstellar Objects in the Solar System. 2026 framework for capture and survival; conclusions depend on assumed ISO flux and N-body evolution. <https://repository.gatech.edu/entities/publication/85feecd4-3813-4ad8-9ad4-19f290625c87>

R30 — Raymond, S. N., Kaib, N. A., & Hopkins, M. J. (2026 preprint). Capture of interstellar objects during stellar encounters. Models possible capture at extremely large distances during stellar flybys. <https://arxiv.org/abs/2607.18551>

R31 — Micheli, M., et al. (2018). Non-gravitational acceleration in the trajectory of 1I/2017 U1 (ʻOumuamua). Nature, 559, 223–226. Original astrometric evidence for nongravitational acceleration. <https://doi.org/10.1038/s41586-018-0254-4>

R32 — NASA Science (maintained). Comet 3I/ATLAS: Overview. Institutional account of its detection, inbound trajectory, solar passage and safe Earth distance; not a source for the Galactic population density. <https://science.nasa.gov/solar-system/comets/3i-atlas/>

R33 — Rubin Observatory (2026; maintained). Legacy Survey of Space and Time. Primary institutional statement that LSST began in June 2026, distinct from older papers' predictions of a future start. <https://rubinobservatory.org/explore/how-rubin-works/lsst>

R34 — Walker, C. E., & Seligman, D. Z. (2026). Sky-plane velocity distributions of interstellar objects and implications for their detection. Monthly Notices of the Royal Astronomical Society, 553, 1, article stag1866; journal publication of [R13]. <https://doi.org/10.1093/mnras/stag1866>

Citation usage notes

  • Density estimates [R01–R03] describe inferences under survey models, not directly enumerated objects.
  • The trajectory/scattering and rate equations are standard mechanics derived in this report; astronomical input assumptions are linked to [R01–R02].
  • Forecasts [R09], [R10], and [R12] have distinct assumptions and must not be averaged as if they were measurements.
  • The 2026 Hubble size estimate [R08] depends on albedo. Different physical methods, including [R26], can yield substantially different size estimates.
  • Publications and observatory pages provide an evolving picture: archived orbital elements and forecasts are not necessarily the latest fit.

Final synthesis

What the physics says: An interstellar population fills a huge three-dimensional volume with trajectories characterized by many arrival directions and many impact parameters. The Sun alters those paths according to their distance and speed; it does not turn the population into a radial inflow toward its center.

What the statistical model says: If the literature's order-of-magnitude density estimate for roughly ʻOumuamua-scale bodies is representative, the outer planetary region could contain thousands of passing interstellar bodies at a given time. Very few reach the inner 1 AU. The absolute numbers remain uncertain because of small discovery statistics and poorly known size/velocity distributions.

What the observations say: The confirmed large visitors are three objects with widely different properties and orbital inclinations. Microscopic interstellar dust is also known directly. We have not yet measured a reliable, continuous census of all intermediate sizes.

What this means scientifically: The Solar System may be continually traversed by debris from numerous planetary systems. A far better survey sample could reveal the history of planetesimal ejection across the Galaxy—without implying that objects are mysteriously targeting the Sun, tracking the ecliptic, or frequently approaching Earth.

<!-- Markdown reference definitions make every [Rxx] mention above clickable. --> [R01]: https://arxiv.org/abs/1801.02821 [R02]: https://doi.org/10.1146/annurev-astro-071221-054221 [R03]: https://arxiv.org/abs/1702.02237 [R04]: https://doi.org/10.1038/nature25020 [R05]: https://science.nasa.gov/solar-system/comets/oumuamua/ [R06]: https://www.nature.com/articles/s41550-019-0931-8 [R07]: https://doi.org/10.3847/2041-8213/adf49a [R08]: https://doi.org/10.3847/2041-8213/ae471c [R09]: https://doi.org/10.3847/PSJ/ac58fe [R10]: https://doi.org/10.3847/PSJ/adf8ca [R11]: https://arxiv.org/abs/2507.05318 [R12]: https://doi.org/10.3847/PSJ/ad08c1 [R13]: https://arxiv.org/abs/2606.05344 [R14]: https://www.nature.com/articles/362428a0 [R15]: https://doi.org/10.1051/0004-6361/202453532 [R16]: https://doi.org/10.1051/0004-6361/202557337 [R17]: https://arxiv.org/abs/2005.10896 [R18]: https://science.nasa.gov/solar-system/comets/3i-atlas/3i-atlas-facts-and-faqs/ [R19]: https://iawn.net/obscamp/3I_ATLAS/3I_orbit.shtml [R20]: https://survey-strategy.lsst.io/progress/index.html [R21]: https://doi.org/10.1146/annurev-astro-071221-054221 [R22]: https://doi.org/10.1088/0004-637X/704/1/733 [R23]: https://ssd.jpl.nasa.gov/sb/orbits.html [R24]: https://doi.org/10.1093/mnras/staa1190 [R25]: https://doi.org/10.3847/1538-3881/ae2ea6 [R26]: https://academic.oup.com/mnras/article/552/3/stag1800/8832565

[R27]: https://arxiv.org/abs/2503.19042 [R28]: https://arxiv.org/abs/2509.05905 [R29]: https://repository.gatech.edu/entities/publication/85feecd4-3813-4ad8-9ad4-19f290625c87 [R30]: https://arxiv.org/abs/2607.18551 [R31]: https://doi.org/10.1038/s41586-018-0254-4 [R32]: https://science.nasa.gov/solar-system/comets/3i-atlas/ [R33]: https://rubinobservatory.org/explore/how-rubin-works/lsst [R34]: https://doi.org/10.1093/mnras/stag1866