Near-Earth objects
Tens of thousands of rocks share the inner Solar System with us, and every one with an orbit worked out can be projected forward — which is what the agencies that track them are doing continuously. This is what they are reporting right now.
The state of the sky
43
close approaches
within 0.05 au over 60 days
2
closer than the Moon
in that same window
2,204
objects on Sentry
monitored for any possible impact
0
above Torino 0
on the scale's own definition
Those last two numbers are the ones worth dwelling on. Sentry monitors every object whose orbit leaves any possibility of an impact open, however remote — 2,204 of them — and not one currently rates above zero on the Torino scale, which is the ordinary state of affairs. 2 sit above the Palermo scale's −2 threshold, and all of them remain below zero — that is, below the background risk from all objects of comparable size.
Approaching next
Live| Object | Closest approach (TDB) | Distance | Relative velocity | Estimated size | Risk monitoring |
|---|---|---|---|---|---|
| 2026 RF39Not yet catalogued in AsteriaStar | 2026-09-16 05:39 TDB±< 00:01 (3σ) | 5.68 LD3σ range 5.65–5.71 LD | 7.6 km/s | 17 m–37 mestimated from H 26.0 | Not on the Sentry table |
| 2026 RM39Not yet catalogued in AsteriaStar | 2026-09-16 05:49 TDB±< 00:01 (3σ) | 4.24 LD3σ range 4.22–4.26 LD | 9.8 km/s | 8 m–17 mestimated from H 27.7 | Not on the Sentry table |
| 2023 SZ3Not yet catalogued in AsteriaStar | 2026-09-16 06:10 TDB±< 00:01 (3σ) | 6.70 LD3σ range 6.70–6.70 LD | 10.3 km/s | 43 m–96 mestimated from H 23.9 | Not on the Sentry table |
| 2026 RA4Not yet catalogued in AsteriaStar | 2026-09-16 07:40 TDB±< 00:01 (3σ) | 3.35 LD3σ range 3.34–3.36 LD | 11.7 km/s | 15 m–34 mestimated from H 26.2 | On the Sentry table → |
| 2026 RB39Not yet catalogued in AsteriaStar | 2026-09-17 02:29 TDB±< 00:01 (3σ) | 1.00 LD3σ range 1.00–1.00 LD | 9.3 km/s | 5 m–10 mestimated from H 28.8 | On the Sentry table → |
| 2026 NT2Not yet catalogued in AsteriaStar | 2026-09-17 05:07 TDB±< 00:01 (3σ) | 16.1 LD3σ range 16.08–16.09 LD | 12.5 km/s | 235 m–525 mestimated from H 20.3 | Not on the Sentry table |
| 2026 RN4Not yet catalogued in AsteriaStar | 2026-09-17 07:36 TDB±< 00:01 (3σ) | 16.1 LD3σ range 16.02–16.24 LD | 8.9 km/s | 14 m–32 mestimated from H 26.3 | Not on the Sentry table |
| 2026 RD38Not yet catalogued in AsteriaStar | 2026-09-17 22:21 TDB±00:01 (3σ) | 15.2 LD3σ range 15.13–15.23 LD | 8.2 km/s | 26 m–58 mestimated from H 25.0 | Not on the Sentry table |
The section
- Close approaches
Every object passing within 0.05 au over the next 60 days, with its uncertainty.
- Objects
The objects behind the feeds, and whether AsteriaStar has catalogued them.
- Impact risk
JPL's Sentry table: what it monitors, and what its numbers actually mean.
- Recently discovered
New entries in the small-body database, and the MPC's unconfirmed candidates.
- Planetary defence
How objects are found, tracked and — in one demonstrated case — deflected.
Live records and catalogued objects
Of the 43 objects approaching in this window, 0 match something AsteriaStar has catalogued. That ratio is normal and it is not a gap to be closed: CNEOS tracks every object whose orbit is known well enough to project, and most of them are metres-wide rocks with provisional designations that will never warrant an encyclopedia entry.
A live record that matches a catalogued object is linked to it. One that does not is shown as what it is — a provider record — and labelled not yet catalogued in AsteriaStar. It is never quietly turned into a permanent entity: minting encyclopedia entries from a feed is how an encyclopedia fills with things nobody wrote.
Reading these numbers
Between −2 and 0 on the Palermo scale: the hazard is smaller than the background risk from comparable objects, and warrants monitoring rather than concern.
An impact probability of “1 in 300,000” is not a forecast that something will happen; it is the width of the remaining uncertainty in an orbit. JPL states plainly that these probabilities “can easily be inaccurate by a factor of a few, and occasionally by a factor of ten or more”.
Objects leavethis table. That is what normally happens: more observations shrink the orbit's uncertainty until the possible impact is ruled out entirely. A disappearance from Sentry is the system working, not data going missing.
Where the data comes from
JPL Solar System Dynamics / CNEOS
NASA Jet Propulsion Laboratory, California Institute of Technology
JPL publishes a Fair Use Policy rather than a numeric limit: requests must be “reasonably necessary”, automated processes must not be “unnecessarily frequent, repetitive, or redundant”, only ONE request may be open at a time, and processes must back off on errors. AsteriaStar serialises JPL requests and stops asking after three consecutive failures — within a server instance, which is the honest scope: this deployment is serverless, so a second instance has its own queue and its own failure count. What bounds the rate across instances is the response cache in front of these pages, which is measured in hours: a page view is not a provider request, and the products are re-fetched when a cached page revalidates, not when someone reads it.
JPL states of Sentry impact probabilities: “The probability computation is complex and depends on a number of assumptions that are difficult to verify. For these reasons the stated probability can easily be inaccurate by a factor of a few, and occasionally by a factor of ten or more.” It also notes there is no guarantee any particular API remains available.
IAU Minor Planet Center
International Astronomical Union / Center for Astrophysics, Harvard & Smithsonian
The MPC publishes machine-readable data files rather than an API, with no registration, no credentials and no documented rate limit. AsteriaStar reads one of them — the NEO Confirmation Page, about six kilobytes — and caches it for ten minutes.
The NEO Confirmation Page lists CANDIDATES, not discoveries. An entry may turn out to be an already-known object, a main-belt asteroid rather than a near-Earth one, or an artefact; most entries leave the page within days. The score is the MPC's estimate of how likely the object is a NEO, not a probability that it exists.
Every figure here is JPL's or the Minor Planet Centre's. AsteriaStar computes no impact probability and applies no danger rating of its own: where a hazard is named, it is on a scale the issuing agency defined, quoted as they define it. A close approach is described by its distance and its uncertainty, because that is what the data says. Objects pass within a few lunar distances routinely, and the word for that is “routine”.
Provenance
Close approaches
Live- Status
- Live
- Kind
- Model product
- Provider
- JPL Solar System Dynamics / CNEOS
- Organisation
- NASA Jet Propulsion Laboratory, California Institute of Technology
- Fetched
- 2026-09-16 22:08 UTC
- Provider cadence
- 1 d
- Cached for
- 1 h
- Treated as stale after
- 1 d
- Licence
- Public domain (US Government work), NASA/JPL-Caltech.
Source file: https://ssd-api.jpl.nasa.gov/cad.api?date-min=now&date-max=%2B60&dist-max=0.05&sort=date&fullname=true&diameter=true
Near-Earth object close approaches from JPL Solar System Dynamics / CNEOS, fetched from https://ssd-api.jpl.nasa.gov/cad.api?date-min=now&date-max=%2B60&dist-max=0.05&sort=date&fullname=true&diameter=true. Close-approach solutions change only when an object's orbit is refitted from new observations, which happens on a timescale of days. An hour's cache is far finer than that, and JPL's Fair Use Policy asks that automated requests not be unnecessarily frequent.
Times are TDB (barycentric dynamical time), not UTC — the provider publishes them that way and AsteriaStar does not silently convert them. Distances are the nominal solution with 3-sigma minimum and maximum bounds alongside; a close approach is a computed prediction from a fitted orbit, not an observation. The window is the next 60 days within 0.05 au (about 20 lunar distances).
Sentry risk table
Live- Status
- Live
- Kind
- Model product
- Provider
- JPL Solar System Dynamics / CNEOS
- Organisation
- NASA Jet Propulsion Laboratory, California Institute of Technology
- Fetched
- 2026-09-16 22:08 UTC
- Provider cadence
- 1 d
- Cached for
- 6 h
- Treated as stale after
- 1 d
- Licence
- Public domain (US Government work), NASA/JPL-Caltech.
Source file: https://ssd-api.jpl.nasa.gov/sentry.api
Sentry impact-risk table from JPL Solar System Dynamics / CNEOS, fetched from https://ssd-api.jpl.nasa.gov/sentry.api. Sentry is re-run when new observations arrive for a listed object; entries persist for months or years and change slowly. Six hours is well inside that, and keeps a page view from becoming a request to a research service.
Impact probabilities are JPL's own, with JPL's own caveat that they can be wrong by a factor of a few and occasionally by ten or more. Diameters are estimated from absolute magnitude assuming an albedo of 0.154 unless a measurement exists. Objects LEAVE this table when further observations eliminate their potential impacts — a disappearance is good news, not missing data. The Torino scale is defined only for potential impacts less than a century away.
Recent database entries
Live- Status
- Live
- Kind
- Observation
- Provider
- JPL Solar System Dynamics / CNEOS
- Organisation
- NASA Jet Propulsion Laboratory, California Institute of Technology
- Fetched
- 2026-09-16 22:08 UTC
- Provider cadence
- 1 d
- Cached for
- 6 h
- Treated as stale after
- 1 d
- Licence
- Public domain (US Government work), NASA/JPL-Caltech.
Recently observed near-Earth objects from JPL Solar System Dynamics / CNEOS, fetched from https://ssd-api.jpl.nasa.gov/sbdb_query.api?fields=full_name,pdes,neo,pha,H,diameter,first_obs,class,moid&sb-class=IEO,ATE,APO,AMO&sb-cdata=%7B%22AND%22%3A%5B%22first_obs%7CGE%7C2026-07-18%22%5D%7D&sort=-first_obs&limit=1000. A new NEO enters the database once its orbit is computed, which happens over days. Six hours cannot delay one meaningfully, and the date window only advances once per day.
`first_obs` is the date of the FIRST OBSERVATION used in the orbit solution — not the date the object was announced, and not the date anyone recognised it as new. Absolute magnitude H is a brightness, not a size; a diameter appears only where one has actually been measured or modelled.
Unconfirmed candidates
Live- Status
- Live
- Kind
- Observation
- Provider
- IAU Minor Planet Center
- Organisation
- International Astronomical Union / Center for Astrophysics, Harvard & Smithsonian
- Fetched
- 2026-09-16 22:08 UTC
- Provider cadence
- 15 min
- Cached for
- 10 min
- Treated as stale after
- 12 h
- Licence
- Public data of the IAU Minor Planet Center.
Source file: https://minorplanetcenter.net/Extended_Files/neocp.json
NEO Confirmation Page (unconfirmed candidates) from IAU Minor Planet Center, fetched from https://minorplanetcenter.net/Extended_Files/neocp.json. The page is regenerated as observations arrive, and entries live on it for hours to days. Ten minutes bounds how long a new candidate is missing while keeping the request rate to a small file trivial.
These are CANDIDATES awaiting confirmation, not discoveries. An entry may prove to be an already-known object, not a near-Earth object at all, or an artefact, and most leave the page within days. The score is the MPC's estimate of how likely the object is a NEO — it is not a probability that the object exists, and it is not an impact risk.
Sources & references
The primary and reference sources this topic draws on.
- NASA JPLJet Propulsion Laboratory / Solar System Dynamics
Orbital data, ephemerides, and small-body parameters for planets, asteroids, and comets.
- NASANational Aeronautics and Space Administration
Mission data, planetary science, space telescopes, and public-domain imagery.
Most NASA-produced imagery is in the public domain; individual items are checked for usage terms before publication.
- Minor Planet CenterIAU Minor Planet Center
Designations and orbits of asteroids, comets, and minor bodies.