Occultations¶

The Occultations tool at /tools/occultations lists upcoming stellar occultations by asteroids as seen from each Skynet site, recomputed every day for the next 60 days, and lets you plan a high-cadence observation of any event with one click. Nobody maintains an object list: every numbered asteroid whose size and orbit quality clear a floor is screened against the Gaia DR3 star catalog, and the geometry at every Skynet observatory is worked out in advance.

When an asteroid passes in front of a star, the star disappears for the few seconds it takes the asteroid's shadow to sweep over the observer. The shadow is as wide as the asteroid, so from a fixed site most predicted events are near misses, and the tool's job is to say which ones are not: when the event happens at your site, how likely the site is to be inside the shadow, how long the star will be gone, and whether a camera at that site can time it.

What the tool shows¶

The event list¶

The list page shows one row per site event, an event as seen from one observatory, sorted by time. Each row names the object, the star, the site, the time of closest approach in UTC and local time, the probability that the site is inside the shadow (drawn as a bar, so a 2% event reads as a long shot), the expected duration at the site, and the magnitude drop.

The filters on the left narrow the list, and their defaults are aimed at the CCD cameras Skynet runs today:

Sites: any set of public observatories, default all of them.
Window: a date range, default the next 30 days.
Star: faintest Gaia G magnitude, a RUWE ceiling, and a switch to exclude stars Gaia flags as duplicated or variable.
Event: minimum probability, minimum duration, minimum magnitude drop, and a switch to keep only objects with a measured diameter rather than one estimated from H.
Distance from the shadow: a ceiling on how far the site may sit from the shadow, in kilometers or in shadow widths.
Sky at the site: minimum star altitude and the Sun-altitude cut, with a switch to include daytime events.

Filters persist in your browser. "Reset to defaults" puts them back.

The event page¶

Every site event has its own page at https://skynetgo.org/tools/occultations/<eventUid>?site=<slug>, where <eventUid> is the event's identifier and <slug> is the observatory's slug. Open it from a list row, or share the link. Without the site parameter the page shows the geocentric event with the site table only.

The page has four parts:

The map. A world map drawn for the event time: the shadow's center line in white with a dot every 10 minutes UTC, the nominal shadow edges in red, the 1σ band dashed, the chosen site's horizon as a circle, and the other Skynet telescopes as gray markers. Night and the Sun and Moon are as they are at the event time.
At the site. The site's closest approach in UTC and local time, the miss distance from the center line (and whether the site is inside the nominal shadow or how far outside its edge), the probability that the site is inside the shadow, the expected chord duration, the timing window (plus or minus three along-track sigmas plus two minutes), the star's altitude and azimuth, the Sun's altitude, and the Moon's separation and illumination. Below it, an Other sites table lists every other Skynet site the shadow passes near, with links to their pages.
The event. The object with its diameter (marked when estimated from H), apparent V magnitude, MPC uncertainty code U, and distance; the star's Gaia DR3 source id, G, BP−RP, RUWE and position; the magnitude drop; the shadow's width, speed and central duration; the cross-track uncertainty in km and the along-track uncertainty in seconds; the geocentric closest approach in milliarcseconds with its position angle; and the Orbit and ephemeris line, which names the orbit solution the prediction was propagated from (for example assist sbdb 123 2026-10-30, meaning the ASSIST integrator started from JPL Small-Body Database orbit solution 123 dated 2026-10-30, or assist astdys 2026-10-28 for an AstDyS solution) and when it was computed.
Observe it. The planning panel described next, or a notice that the event has passed.

Observing an event¶

Sign in, open the event page for the site you want to observe from, and look at the Observe it panel. Skynet sizes an observation for the event on every optical imager at that site, with the same exposure solver the scheduler uses, and reports which of them can time it. An imager is compatible when it reaches the target signal-to-noise ratio on the star in an image short enough that the expected chord holds at least three samples (ingress, bottom and egress). Slower cameras are listed with the reason they fall short and left out of the plan.

Choose which observing account the observation belongs to and press Plan an observation at . The observation editor opens prefilled:

Target: a fixed position at the star's Gaia coordinates at the event epoch, named after the event, for example "(168) Sibylla occults Gaia DR3 3399145974004627072".
Window: the observation may start 300 seconds before the site's closest approach (slew, focus and settle time) less three along-track sigmas, and ends 120 seconds plus three sigmas after it, with two further minutes of slack so a late start still covers the event.
Instruments: the site's compatible imagers, selected; you can add or remove instruments.
One series request: a series runs a single camera continuously for the whole window and delivers as many images as fit back to back at that camera's cadence, each image reaching the target depth. The target signal-to-noise ratio comes from the magnitude drop (deep enough to see the dip at five sigma, between 5 and 50), and each image may last at most a third of the expected chord. The filter is the site's highest-throughput option.

Review the observation and create it as you would any other. The scheduler treats it as a normal observation with a hard window: it sizes exposures per camera at promotion and realizes as many images as fit the series. Nothing is reserved automatically, so a higher-priority observation inside the window can still displace it; for a one-shot event it is worth checking the observation's status the day before.

What comes back¶

The observation produces one series result per camera: a contiguous run of images on that camera, each image a stack at the requested depth, never stacked across images. The camera keeps capturing until the series clock runs out, so the frame count you see is what the camera managed, not the planned number. Frames appear on the observation detail page as they land, and the download page's shutter-time window lets you pull just the minutes around the event.

To get a light curve, open the observation in Afterglow, run photometry on the target star and a few comparison stars across the frames, and plot the result in Afterglow's Light Curve tool. An occultation shows as a sharp drop in the star's flux to the asteroid's brightness alone (the magnitude drop the event page predicts), lasting the chord duration, with the asteroid and star blended into one source throughout. The ingress and egress times from the two edges of the dip, together with the frame timestamps, are the measurement; the FITS header's TIMESRC card records whether each frame was timed by the camera or by the SkyNode clock.

The first real event observed this way was (168) Sibylla occulting a Gaia DR3 star on 2026-11-06, planned from the event page and observed from CTIO with two telescopes; the dip is visible in both light curves.

If you observe a chord, consider reporting it to the Occultation Portal (OPOP), the community site that collects occultation observations and the campaign predictions of the Lucky Star project and its partners. It is not a dependency of this tool, but it is where the wider occultation community looks for results.

How the predictions are made¶

A daily job runs the predictor at 12:00 UTC against the catalogs below. In outline:

The screening set is read from Skynet's copy of the Minor Planet Center's orbit catalog, enriched with JPL's physical and fit-quality data. It takes every numbered asteroid with uncertainty code U ≤ 3 and diameter ≥ 5 km, every outer-system body (semimajor axis ≥ 4.6 au) with diameter ≥ 50 km regardless of U, and every unnumbered object with H ≤ 5 and U ≤ 3 that clears the same diameter floors (the large trans-Neptunian objects now being found). Comets and Pluto are excluded. Diameters are NEOWISE, IRAS, AKARI, occultation or radar measurements where JPL has one, otherwise estimated from H with an assumed albedo of 0.15 and marked as estimated.
One orbit solution with covariance per object comes from the JPL Small-Body Database, cached and re-fetched only when JPL refits the orbit. A numbered object that has no cached JPL solution yet is propagated from the AstDyS catalog instead, and the event page says which source was used.
Propagation is numerical, with the ASSIST integrator: the planets and Moon from JPL's DE440 ephemeris, the sixteen most massive main-belt asteroids, Earth and Sun oblateness, and relativistic corrections. The orbit's covariance is carried as twelve sigma-point clones integrated beside the nominal orbit, so the uncertainty at the event comes out of the dynamics rather than a linearization.
The screen samples each object's track hourly across the horizon and keeps every Gaia DR3 star (G ≤ 15) inside a corridor of Earth radius plus object radius plus three sigma around it, after moving the star to the epoch by proper motion and parallax, then drops any hit whose shadow does not pass within ten diameters plus three sigma of at least one Skynet site.
Refinement resamples the survivors at 60-second steps around each candidate and computes the closest approach, the shadow path, the per-site miss distance, the probability the site is inside the shadow, the expected chord, the magnitude drop from the IAU H-G phase law, and the local circumstances. Events within 24 hours are frozen rather than recomputed.

The full design, with the measured yields, is in the repository at docs/internal/design/catalogs/OCCULTATION_PREDICTIONS_DESIGN.md.

Data sources and what to cite¶

The predictions rest on catalogs, ephemerides and software maintained by others. If you publish a result that started from this tool, the providers below ask to be credited, and several of them specify the wording. Where a provider states a citation or acknowledgement request, it is quoted here exactly as written, with a link to where it is stated. Where a provider states none, this page says so rather than suggesting one.

Resource	What the tool takes from it	Citation or acknowledgement requested
JPL Small-Body Database (SBDB)	Orbit solutions with covariance; diameters, albedos, H, U code, fit quality	Yes, a citation form for data downloaded from the SSD site
JPL planetary and small-body ephemerides (DE440, SB441-N16, de440s)	Planet, Moon, Sun and asteroid positions in the integrator; the Earth's position	A reference paper for DE440; crediting the kernel producers is encouraged
Minor Planet Center (MPCORB)	Designations, names, elements, H, G, U code for every asteroid	Yes, an acknowledgement sentence
AstDyS-2	Fallback orbit solutions with covariance for numbered asteroids	None found on the site; the literature cites one paper
ASSIST	The integrator	A preferred citation in the package's citation file
REBOUND and IAS15	The N-body framework and integrator under ASSIST	Yes, a citation request
Gaia DR3 (ESA / DPAC)	The star catalog	Yes, an acknowledgement paragraph and two papers
SORA	Three occultation-geometry functions, ported	Yes, a citation request
Astropy and astropy-healpix	Time scales, frames, Earth position, HEALPix search	Yes, an acknowledgement sentence and three papers

JPL Small-Body Database¶

The Small-Body Database of NASA's Jet Propulsion Laboratory Solar System Dynamics group is the tool's source of orbits. For every object in the screening set the predictor fetches one solution document from the SBDB API: the osculating elements, the 6×6 covariance at the solution epoch, the solution date, and H and G. A weekly bulk pull from the SBDB Query API supplies every numbered asteroid's measured diameter and its uncertainty, albedo, H, uncertainty code, fit RMS, number of observations, data arc and SPK-ID, and the same for the brightest unnumbered objects. The event page's "Orbit and ephemeris" line records the SBDB orbit id and solution date behind each prediction.

The SSD group's About page states, under Acknowledgements:

For referencing the data downloaded from the SSD website, please use the following citation:

"NASA/JPL Solar System Dynamics. Data retrieved on (Downloaded Year, Month, Date). (Title of the Page). https://ssd.jpl.nasa.gov"

For an SBDB orbit, the download date is the solution's fetch date, which is at or after the solution date shown on the event page, and the page title is "Small-Body Database (SBDB) API". The SBDB API documentation itself states no separate citation requirement.

JPL planetary and small-body ephemerides¶

Three JPL ephemeris files sit under every prediction. The ASSIST integrator reads the DE440 planetary and lunar ephemeris (linux_p1550p2650.440) and the SB441-N16 small-body perturber file (sb441-n16.bsp, the sixteen most massive main-belt asteroids), both downloaded from the SSD ephemeris export area. The Earth's own position, which the shadow geometry needs to a few kilometers, comes from the de440s.bsp SPICE kernel published by NASA's Navigation and Ancillary Information Facility (NAIF), read through Astropy.

The SSD ephemeris page lists the DE440 reference paper, and the SSD citation form quoted above covers these downloads as well:

Park, R. S., Folkner, W. M., Williams, J. G., and Boggs, D. H. 2021, "The JPL Planetary and Lunar Ephemerides DE440 and DE441", The Astronomical Journal, 161, 105. DOI 10.3847/1538-3881/abd414.
The SB441-N16 perturber file is described in Farnocchia, D. 2021, "Small-Body Perturber Files SB441-N16 and SB441-N343", JPL Interoffice Memorandum IOM 392R-21-005, March 30, 2021 (PDF).

NAIF's Credit and References page gives no fixed wording but asks:

In your publications and presentations consider giving appropriate credit to the teams or persons who have provided you the ancillary data and/or services used in producing your results.

and

Referencing your use of SPICE or NAIF/PDS resources in publications, tools and findings is encouraged; such community support will help ensure NASA will continue to fund the maintenance and further development of the SPICE system.

Minor Planet Center¶

Skynet downloads the MPCORB orbit file from the International Astronomical Union's Minor Planet Center every night. It is the catalog the screening set is drawn from and the source of each object's designation, number, name, H, G, osculating elements and uncertainty code U. Where a prediction rests on an AstDyS orbit, the MPC's H is the one used for the object's brightness.

The MPC's data page asks:

If using these or any other data files, or a webtool from the MPC, for your research publication, we would appreciate it if you could thank us in the acknowledgements section. This will help us track the use of MPC data and services in publications, which will be of interest to funding agencies. Please use the following snippet:

This research has made use of data and/or services provided by the International Astronomical Union's Minor Planet Center.

AstDyS-2¶

AstDyS-2, the Asteroids Dynamic Site maintained by SpaceDyS with the University of Pisa, publishes a catalog of orbit solutions with covariances for every numbered asteroid (allnum.ctc). The predictor uses it as the fallback orbit source for a numbered asteroid that does not yet have a cached JPL solution, since the JPL solutions are fetched slowly to respect JPL's request limits, and as the fallback if the JPL service is unavailable. The event page's "Orbit and ephemeris" line reads assist astdys <date> for such a prediction.

No citation or acknowledgement request was found on the AstDyS site, which could not be read in full for this page because it sits behind an automated-access gate; if you used an AstDyS-based prediction, check the site for any current request. Publications that use AstDyS data conventionally cite the paper that describes the service:

Knežević, Z. and Milani, A. 2003, "Proper element catalogs and asteroid families", Astronomy & Astrophysics, 403, 1165–1173. DOI 10.1051/0004-6361:20030475.

ASSIST¶

ASSIST integrates each object and its twelve covariance clones as massless test particles under JPL's small-body force model. Its README states no citation request in words, but the repository's citation file names a preferred citation, and the package carries a Zenodo DOI, 10.5281/zenodo.7778016:

Holman, M. J., Akmal, A., Farnocchia, D., Rein, H., Payne, M. J., Weryk, R., Tamayo, D., and Hernandez, D. M. 2023, "ASSIST: An Ephemeris-quality Test-particle Integrator", The Planetary Science Journal, 4, 69. DOI 10.3847/PSJ/acc9a9.

ASSIST is distributed under the GNU General Public License, version 3.

REBOUND and IAS15¶

ASSIST is built on REBOUND and uses its IAS15 integrator. REBOUND's README states, under Acknowledgments:

If you use this code or parts of this code for results presented in a scientific publication, we would greatly appreciate a citation. The simplest way to find the citations relevant to the specific setup of your REBOUND simulation is: sim.cite()

The papers that setup resolves to for this tool are the code paper and the integrator paper:

Rein, H. and Liu, S.-F. 2012, "REBOUND: an open-source multi-purpose N-body code for collisional dynamics", Astronomy & Astrophysics, 537, A128. DOI 10.1051/0004-6361/201118085.
Rein, H. and Spiegel, D. S. 2015, "IAS15: a fast, adaptive, high-order integrator for gravitational dynamics, accurate to machine precision over a billion orbits", Monthly Notices of the Royal Astronomical Society, 446, 1424–1437. DOI 10.1093/mnras/stu2164.

REBOUND is distributed under the GNU General Public License, version 3 or later.

Gaia DR3¶

The stars are the Gaia Data Release 3 sources brighter than G = 15 (about 37 million), with positions at epoch J2016.0, proper motions, parallaxes, G and BP−RP magnitudes, RUWE, and the duplicated-source and photometric-variability flags. Skynet downloaded them from the ESA Gaia Archive and keeps them in its own catalog store (the in-repository skycat package, which is not a published package and has no citation of its own). Every event names the star by its Gaia DR3 source id.

The Gaia DR3 documentation's Credit and citation instructions ask:

If you have used Gaia DR3 data in your research, please use the following acknowledgement:

This work has made use of data from the European Space Agency (ESA) mission Gaia (https://www.cosmos.esa.int/gaia), processed by the Gaia Data Processing and Analysis Consortium (DPAC, https://www.cosmos.esa.int/web/gaia/dpac/consortium). Funding for the DPAC has been provided by national institutions, in particular the institutions participating in the Gaia Multilateral Agreement.

and

If you have used Gaia DR3 data in your research, please cite both the Gaia mission paper and the Gaia DR3 release paper:

Gaia Collaboration et al. (2016b): The Gaia mission (provides a description of the Gaia mission including spacecraft, instruments, survey and measurement principles, and operations);

Gaia Collaboration et al. (2023j): Gaia DR3: Summary of the contents and survey properties.

The same page states that "The Gaia data are open and free to use, provided credit is given to 'ESA/Gaia/DPAC'." The two papers in full:

Gaia Collaboration, Prusti, T., et al. 2016, "The Gaia mission", Astronomy & Astrophysics, 595, A1. DOI 10.1051/0004-6361/201629272.
Gaia Collaboration, Vallenari, A., et al. 2023, "Gaia Data Release 3: Summary of the content and survey properties", Astronomy & Astrophysics, 674, A1. DOI 10.1051/0004-6361/202243940.

SORA¶

The per-event geometry (the closest-approach search, the projection of each site into the shadow plane, and the ephemeris sampling) is a port of three functions from SORA, the Stellar Occultation Reduction and Analysis package, which is MIT-licensed. SORA's Using and Citing page asks:

SORA is a free software, and we kindly ask users that make use of it (or of part of it), especially in projects that results in scientific publications, to include the following citation:

Gomes-Júnior et al. (2022). SORA: Stellar occultation reduction and analysis. MNRAS, Volume 511, Issue 1, March 2022, Pages 1167–1181 DOI: 10.1093/mnras/stac032

The same page sets out terms for co-authorship that apply to members of the Rio, Granada and Paris collaboration who use SORA on objects within the collaboration; read it if that describes you.

Astropy and astropy-healpix¶

Astropy supplies the time scales, the Earth's position from the SPICE kernel, the site frames, the altitude and azimuth of the star, and the Sun and Moon circumstances. astropy-healpix supplies the HEALPix cone search the screen uses to find the stars along each track. Astropy's Acknowledging or Citing Astropy page provides this acknowledgement text and asks that all three Astropy papers be cited (or the most recent one where space is limited):

This work made use of Astropy: a community-developed core Python package and an ecosystem of tools and resources for astronomy (Astropy Collaboration 2013, 2018, 2022).

Astropy Collaboration, Robitaille, T. P., et al. 2013, "Astropy: A community Python package for astronomy", Astronomy & Astrophysics, 558, A33. DOI 10.1051/0004-6361/201322068.
Astropy Collaboration, Price-Whelan, A. M., et al. 2018, "The Astropy Project: Building an Open-science Project and Status of the v2.0 Core Package", The Astronomical Journal, 156, 123. DOI 10.3847/1538-3881/aabc4f.
Astropy Collaboration, Price-Whelan, A. M., et al. 2022, "The Astropy Project: Sustaining and Growing a Community-oriented Open-source Project and the Latest Major Release (v5.0) of the Core Package", The Astrophysical Journal, 935, 167. DOI 10.3847/1538-4357/ac7c74.

The astropy-healpix README states no citation request of its own.

Methods without a citation request¶

Two standard relations are used as published, and neither source asks for an acknowledgement:

The object's apparent magnitude at the event, and therefore the magnitude drop, follows the IAU H-G phase law of Bowell, E., Hapke, B., Domingue, D., Lumme, K., Peltoniemi, J., and Harris, A. W. 1989, "Application of photometric models to asteroids", in Asteroids II (University of Arizona Press), 524–556.
A diameter estimated from H uses D = 1329 km × 10^(−H/5) / √p with an assumed geometric albedo p = 0.15, after Harris, A. W. and Harris, A. W. 1997, "On the Revision of Radiometric Albedos and Diameters of Asteroids", Icarus, 126, 450–454. DOI 10.1006/icar.1996.5664.

Used for validation only¶

Two external services shaped the tool but are not used to produce a prediction. JPL Horizons ephemerides and uncertainty ellipses are the reference the integrator, geometry and brightness are tested against; the SSD citation form above applies if you quote those comparisons. The Occultation Portal is the external prediction and reporting site the predictions were compared with during development, and the place to report an observed chord.

Reference¶

OpticalImagingRequest — the request record; a series request carries series_duration_sec.
MpcOrbit — the catalog row an event's object refers to, with the SBDB diameter and fit fields.
EphemerisRecord — ephemeris records as the rest of the platform uses them.