======================================================== 
Content description for the Asteroid Occultations bundle 
======================================================== 
This data set is intended to include all reported timings of observed asteroid, planet, and planetary satellite occultation events, as well as dimensions and other data derived from those timings. The dataset includes observations made over a period of more than 50 years, and has grown rapidly in recent years. Most of these timings are otherwise unpublished. This version is complete through to about September 2024, with partial observations through to [LastDate]. 

In relatively recent years light curve inversion 3D models (Shape Models) have become available for a large number of asteroids via the DAMIT and ISAM web services [https://astro.troja.mff.cuni.cz/projects/asteroids3D/ and http://isam.astro.amu.edu.pl/]. Wherever possible, the asteroidal occultation observations are matched to available Shape Models.

Since the last version (PDS4 V4.0), an extra file has been added to include details on occultation events where either (i) a known asteroid satellite has been detected, (ii) an asteroid satellite has been discovered [20 discoveries as of late 2025], (iii) the observation has been recorded and the light curve suggests there might be a satellite but other explanations cannot be reasonably excluded, and (iv) visual observations where the observer reported a second light drop that might have been caused by a satellite, but there is no other evidence to confirm or support that conclusion. 


==============
Data 
==============
The data set is made up of 11 files

Four files contain the complete set of observation data. 
* [Asteroid]
* [AsteroidTimes]
* [Planet]
* [PlanetTimes]

[Asteroid] contains all data elements relating to an an asteroid occultation event as a whole, while [AsteroidTimes] contains the details of the individual observations of an event. The files are linked by a sequence number in the [Asteroid] file. Comets are included in these files.
[Planet] and [PlanetTimes] contain observation of occultations by the major planets and their satellites.

Two files are summary files, limited to the object size/shape derived for each individual observation
* [AsteroidSummary]
* [PlanetSummary]

Five files contain the primary results derived from an analysis of the observations
* [AsteroidAstrometry]
* [PlanetAstrometry]
* [AsteroidDiameters]
* [DoubleStars]
* [Satellites]

[AsteroidAstrometry] lists the astrometric position of asteroids & comets, while [PlanetAstrometry] contains astrometry for the major planets and their satellites. The data content is that set by the IAU Astrometry Data Exchange Standard (ADES) standard for small solar system objects, and includes all columns required for submission to the Minor Planet Center under observatory code 275. This includes the columns headed deltaRA and deltaDec, which are always set to 0.
* [AsteroidDiameters] provides volume-equivalent diameter of the asteroid derived from the fit of all observations against an available shape model.
* [DoubleStars] provides the double star solution(s) for double stars discovered in an occultation, together with a relevant reference to Journal of Double Star Observations
* [Satellites] provides observations of asteroidal satellites

One file contains images of the 'best' occultation observations.
* occultations.pdf 
This provides a graphic of all events with fit quality code 3 or 4. There are 385 images, which is almost 9% of all events. The scale of the images varies according to the object. If a satellite is involved, the scale is reduced so as to show both the primary object and the satellite.  The file is located in the document directory.


================
Acknowledgments
================
This dataset contains [ObserverCount] observations made at [EventCount] events by over 4,300 individuals from around the world, made over a period of more than 55 years. Observers have generally made these observations at their own expense, including occasions when they have traveled significant distances. Users of this database are requested to acknowledge their contributions with a statement like:
--------
We acknowledge the contributions of the 4300 observers who have provided the observations in the dataset. Most of those observers are affiliated with one of more of:
* European Asteroid Occultation Network (EAON)
* International Occultation Timing Association (IOTA)
* International Occultation Timing Association - East Asia (IOTA-EA)
* Trans-Tasman Occultation Alliance (The Occultation section of the Royal Astronomical Society of New Zealand)
--------
The occultation observations would not have been possible without the efforts of those who make, or have facilitated the making of (by way of software), predictions of occultation events over the duration of the observations in this dataset. Special mention must be made to the coordination software Occult Watcher and OccultWatcher Cloud which facilitates coordination of observers wherever they are located around the world. Final mention is made to the various software packages specifically designed to measure occultation events on video or CCD recordings.


=============================
Occultations Analysis Program 
============================= 
The observations are reduced using the software package Occult that is freely available for download at: 
https://occultations.org/sw/occult/occult4.htm 
with tutorials and tips at https://occultations.org/observing/software/occult/
The basic methodology is given in 'Precise astrometry and diameters of asteroids from occultations - a data set of observations and their interpretation', Herald et al, MNRAS 499, 4570-4590 (2020)

Processes for reporting observations have evolved over the duration of the dataset. For many years now observers submit an observation report that provides disappearance and reappearance timings (in UTC), observing location coordinates (WGS84) and elevation (Mean Sea Level), observing conditions and circumstances. The observations are reduced using the Besselian fundamental plane. The methodology is as follows. 

- The apparent position of the asteroid is computed at four 1-hour steps; the time of the 2nd step is rounded to 0.1 hrs, and is close to the mean time of the observed occultation events. These positions are combined with the apparent positions of the star computed for the same four times (corrected for stellar parallax, proper motion) to derive coordinates of the asteroid's shadow on the fundamental plane. Those coordinates are used to derive a cubic expression for the _motion_ of the asteroid's shadow on the fundamental plane. 

- For each reported time, the (x,y) coordinate of the observer on the fundamental plane is computed. The time of the observation is increased by the light travel time from the observer to the fundamental plane.

- The reference time for the event is set as the unweighted mean time of all reported Disappearance or Reappearance event used to derive the position/shape of the shadow. All observer positions on the fundamental plane as at the observed time are transposed to their position on the fundamental plane as at the reference time, using the cubic expression for the motion of the asteroid's shadow. This creates a set of (x,y) observer coordinates on a reference frame that is fixed relative to the moving asteroid's shadow. 

- For fitting purposes, the origin of the (x,y) coordinates on the reference frame is set at the mean location of the (x,y) observer coordinates. Most commonly, the events used to define the origin of the reference frame are all Disappearance and Reappearance observations. However:

  * Events given a zero weight, and Miss events, are excluded;
  
  * If the star is a double star, only events involving the primary component are used to define the origin of the reference frame. 

  * If a satellite is involved, the relative motion may be unknown. To this end, the reference frame (including the reference time) is defined using only events involving the satellite. This results in the position of the main body being transposed in that reference frame in accordance with its (presumably) dominant motion, to the mean time of the satellite events - thereby ensuring correct evaluation of the separation and position angle at the reference time. For a small number of satellites an ephemeris is available from the IMCCE Miriade system [https://ssp.imcce.fr/webservices/miriade/api/ephemsys/], and the 'known' relative motion of the satellite is incorporated.

- Fitting a shape model, ellipse, double star solution, and satellite solution, occurs on this reference frame.

- The axis of the occultation shadow corresponds to the center of the fitted ellipse, or the center of a fitted shape model. The astrometric position of the asteroid is set to be the position of the star as seen from an observer 'located' at the point of intersection on the fundamental plane, of that shadow axis. This location is converted to geocentric equatorial GCRS J2000 coordinates, with the time of observation being the reference time. 

- If the event involves a double star, there may be a single solution, two solutions, or four solutions (depending upon the number of observed chords). The solution provides a Separation and Position Angle on the Apparent reference frame. For reporting purposes the Position Angle is converted to the BCRS reference frame. The separation of almost all double stars detected in an occultation is smaller than the resolution of the Gaia catalogue; to provide an astrometric position relative to the Gaia star position, the photo-center of the double star components is used. 

- For Miss events, the listed time is generally (but, by omission, not always) set to be the time when the miss chord passes closest to the ellipse fitted to the event.


============
Deriving data from the observations
============
For each event in the dataset a quality code is provided. That code has the following definitions:

0 = 'No reliable position or size'. Such events have been assessed as being unreliable. 
1 = 'Astrometry only. No reliable size'. Such events typically involve only one, a small number of closely spaced, observed chords. The event can be used to report astrometry, but nothing meaningful about the asteroid's size.
2 = 'Limits on size, but no shape'. The chords provide a poor coverage around the profile, but a diameter of the asteroid is determinable (often with a fit to a shape model)
3 = 'Reliable size. Can fit to shape models'. There is good coverage around much of the profile enabling a reliable determination of the asteroid's size - even without a shape model 
4 = 'Resolution better than shape models'. The profile of the asteroid well covered with highly consistent results - even without a shape model.
5 = 'No astrometry. AOTA2B = Possible'. For events where there is only one positive chord, and the duration of the potential occultation is for one video frame or camera exposure. The AOTA2B test compares the light drop of a suspected event to other light drops in the light curve, and allocates a rating of Probable, Possible, or Unlikely, This setting is for when the Rating for such an event is 'Possible'. Such events are not used for any reporting process.
6 = 'No astrometry. Only 1 star of a double'. For events where the observed light drop was clearly less than the expected drop, and could not be explained by variation in the brightness of the asteroid - leaving the only explanation being the occultation was one component of a double star, with the other component not detected. No astrometry because the location of the photo-center to match the Gaia star position cannot be determined.

===========================
Astrometric uncertainties - Error Code 
==========================
The astrometry derived from occultations includes a Error Code from the error model used for the observations. 

The error model is somewhat complex, covering the range of practical situations that occur. Because the observation is of event time, the dominant sources of uncertainty are best resolved in the Along-path / Across-path directions, rather than Right Ascension/Declination. Some example of issues that are dealt with are (i) single chord observations, where the Along-path uncertainty is dominated by uncertainty in the event times, whereas the Across-path uncertainty is associated with the diameter of the asteroid, and (ii) the presence of a near-by Miss chord to a single-chord observation can result in the across-path uncertainty being greatly reduced. For the final astrometric report, the Along-Path and Across-Path uncertainties are rotated to provide uncertainties Right Ascension and Declination, with an associated correlation coefficient.

Parts of the Error Code use the following 4 parameters, which relate to the distance of certain chords on either side of the center of the asteroid. In these parameters, the side of the path is indicated as Plus, or Minus - with Plus referring to the north-side of the path. Hit refers to positive occultation events, while Miss refers to Miss events. The Hit parameters specify the distance of the northern-most and southern-most positive chords. The Miss parameters specify the distance of the closest Miss chord on either side of the path, with a value of +9 or -9 if there is no such chord on the particular side. The distances are expressed in units of the asteroid's assumed radius.

The 4 parameters are:
  Well-located   : PlusHit >= 0.3 _and_ MinusHit <= -0.3
  Poorly-Located : (PlusHit - MinusHit) > 0.5, with either PlusHit < 0.3 or MinusHit > -0.3 
  Constrained    : either PlusMiss < 1.3 or MinusMiss > -1.3
  Unconstrained  : none of the above
  
Two additional parameters apply to the code (d) values (that is, to major planets and their large moons) - ScaleFactor_AlongPath & ScaleFactor_Across_Path
  The default values for both these factors is 1.
  If the absolute value of (PlusHit - MinusHit) <0.3:
  - the MEAN value is Absolute(PlusHit + MinusHit)/2, limited to being no greater than 0.8, and no less than 0.3
  - ScaleFactor_AlongPath = 1/Sqrt(1 - MEAN*MEAN)
  - ScaleFactor_AcrossPath = 1/MEAN

The Error Codes are as follows:

a - the uncertainty is that from a least squares fit of an unrestricted ellipse to the observed chords. This provides the minimal value of the uncertainties.  (a) is usually replaced by one of the following codes. However it will appear for occultations by the major planets and larger planetary satellites, when they have a known diameter and are known to be spherical.
         
b - For asteroids only. If the astrometric position is set to match the _center_ of a shape model (center of mass solution) - the uncertainties for both the Along-Path and Across-Path directions are set at 2% of the assumed diameter.

c - For asteroids, if (b) does not apply. If the observation has been fitted to a shape model, with the quality of that fit being either 'poor' or 'good'
  c1 - If the overall event quality is 'Reliable Size...' or 'Resolution better than Shape models': uncertainties for both the Along-Path and Across-Path directions are set at 4% of the assumed diameter.
  c2 - If the overall event quality is 'Limits on size, but no shape': uncertainties for both the Along-Path and Across-Path directions are set at 8% of the assumed diameter.

d - for major planets & their large moons. This relies on they being essentially circular in profile, and their diameters being accurately known - to provide good astrometry in circumstances which would not apply to asteroids.
  d1 - If Well-located:   uncertainties of (a), unchanged
  d2 - If Poorly-located: three times the uncertainties of (a)
  d3 - if Constrained:
  (i) If only 1 chord, or the spread of multiple chords [absolute value of (PlusHit-MinusHit)] is less than 5% of the assumed diameter:
     * if the displacement in the time-uncertainty-period is greater than the Along-Path uncertainty, the Along-Path and Across-Path uncertainties are set as that displacement multiplied by ScaleFactor_AlongPath, and ScaleFactor_AcrossPath, respectively.
     * otherwise the Along-Path and Across-Path uncertainties are the uncertainties from the fit, multiplied by ScaleFactor_AlongPath, and ScaleFactor_AcrossPath, respectively.
  (ii) if not (i), the Along-Path and Across-Path are the uncertainties from the fit, multiplied by ScaleFactor_AlongPath, and ScaleFactor_AcrossPath, respectively.
  d4 - if Not-constrained:
  (i) If only one chord, 
     * the Along-Path uncertainty (but not the Across-Path uncertainty) is computed on the same basis as for 'd3 - if Constrained', (i), first dot point, but multiplied by ScaleFactor_AlongPath
     * the Across-Path uncertainty is:
       - if the chord is located more than 2 arc-seconds from the center of the object, the Across-Path uncertainty is computed on the same basis as for 'd3 - if Constrained', (i), first dot point - but multiplied by ScaleFactor_AlongPath 
       - otherwise the Across-Path uncertainty is set as 25% of the object's assumed diameter.
  (ii) If more than one chord,
     * the Along-Path uncertainty is computed on the same basis as for 'd3 - if Constrained', (i),second dot point, but multiplied by ScaleFactor_AlongPath
     * if 
       - the MeanLocation value is greater than 25% of its assumed radius, AND (PlusHit - MinusHit) >0.05, OR
       - the MeanLocation value is greater than 15% of its assumed radius, AND (PlusHit - MinusHit) >0.10 
       the Across-Path uncertainty is the uncertainty from the fit, multiplied by ScaleFactor_AcrossPath
     * otherwise the Across-Path uncertainty is set as 25% of the object's assumed diameter.   

e - if (b), (c) and (d) do not apply: 
  for events where quality is better than 'Astrometry only', and either 
  (i) the Major + Minor axes are included in the Least Squares solution, or 
  (ii) the solution is for a circle. 
     The uncertainties are separately considered in the Along-Path and Across-Path directions, and are set as the larger of the uncertainty from (a), and the uncertainty specified in the relevant one of e1 to e8:
    
     If the overall event quality is 'Limits on size, but no shape'
      e1 - Well-located:   uncertainty must be at least 8% of assumed diameter
      e2 - Poorly-located: uncertainty must be at least 12% of assumed diameter
      e3 - Constrained:    uncertainty must be at least 16% of assumed diameter
      e4 - Unconstrained:  uncertainty must be at least 20% of assumed diameter
    
    If the overall event quality is 'Reliable size', or 'Resolution better than Shape models'
      e5 - Well-located:   uncertainty must be at least 5% of assumed diameter
      e6 - Poorly-located: uncertainty must be at least 8% of assumed diameter
      e7 - Constrained:    uncertainty must be at least 12% of assumed diameter
      e8 - Unconstrained:  uncertainty must be at least 16% of assumed diameter

f - if none of (b) to (e) apply:
    The Across-Path uncertainty set on following basis:
    f1 - Well-located: the uncertainty in the assumed diameter of the asteroid.
    f2 - Poorly-located: twice the uncertainty in the assumed diameter of the asteroid.
    f3 - Constrained: twice the uncertainty in the assumed diameter of the asteroid (same as f2).
    f4 - Unconstrained: 40% of the assumed diameter of the asteroid.

    For each of f1 to f4, the Along-path uncertainty set on following basis:
       For each positive chord:
         if chord length > 0.8 of assumed diameter - the Along-Path uncertainty for that chord is set at 5% of assumed diameter of the asteroid.
         else if chord length > 0.6 of assumed diameter - the Along-Path uncertainty for that chord is set at 10% of assumed diameter of the asteroid.
         Otherwise - the Along-Path uncertainty for that chord is set at 20% of assumed diameter of the asteroid.
     Along-Path uncertainty set as the mean of the individual Along-Path uncertainties, assessed in inverse quadrature (to give greater significance to smaller uncertainty values).


==============================
Star identifiers and positions; gravitational deflection
==============================
The star identifiers used in this dataset are (in order of priority) HIP (for Hipparcos2), Tycho2, UCAC4, USNO-B1 and NOMAD; 83 stars are identified by Jhhmmss.s-ddmmss, or Jhhmmss.ss-ddmmss.s. The identifiers are merely for identification purposes. The listed GCRS position of the star for the date of the event (used as the reference point for reporting astrometry) incorporates stellar parallax, proper motion and foreshortening. 

An occultation observation relates to the exact coincidence in the apparent position of the star and asteroid. Gravitational light deflection displaces the apparent position the star and asteroid differently; the deflection of the asteroid is always less than that for the star, by an amount dependent on the geocentric distance to the asteroid. As that distance depends upon any orbit solution derived using the observations, it is inappropriate to incorporate allowance for gravitational deflection outside of that orbit solution methodology. Hence the position of the star is not corrected for gravitational deflection by the Sun or any of the planets.

The star positions for 99.8% of the events are from Gaia EDR3 (VizieR catalogue I/350). The following is a list of the 31 stars and associated events where the source of the star position is not Gaia EDR3. The majority of these are too bright for Gaia EDR3, and their position has been taken from the USNO Bright Star Catalog (UBSC, VizieR catalogue J/AJ/164/36) which provides high accuracy recent epoch position for such stars. It can be assumed that there is a problem with the position of the remaining remaining 5 listed stars. 

10 stars from USNO Bright Star Catalog (2022).
HIP  27989 =           : 2023 Dec 12  (3199) Leona
HIP  28380 = theta Aur : 2022 Apr 13  (2826) Ahti
HIP  31681 = gamma Gem : 1991 Jan 13  (381) Myrrha
HIP  37740 = kappa Gem : 1975 Jan 24  (433) Eros, 2023 Nov 2 (109657) 2001 RQ10
HIP  43103 = iota Cnc  : 2007 Apr 18  (411) Xanthe
HIP  49669 = alpha Leo : 1959 Jul  7  (P2M00) Venus, 2005 Oct 19 (166) Rhodope, 2015 May 24 (1669) Dagmar, 2016 Oct 13 (268) Adorea
HIP  56647 = nu Leo    : 1999 Mar  4  (748) Simeisa
HIP  78821 = beta2 Sco : 1971 May 14  (P5M01) Io
HIP  92855 = sigma Sgr : 1981 Nov 17  (P2M00) Venus
HIP 112961 = lambda Aqr: 2014 Apr 16  (P2M00) Venus

3 stars from Hipparcos2  (VizieR catalogue I/311)
HIP  13702: 1989 Feb 17  (P4M00) Mars
HIP  20719: 1991 Dec 31  (50) Virginia
HIP  28416: 1997 Jan  6  (363) Padua

1 stars from Gaia DR2 (VizieR catalogue I/345)
UCAC4 521-018804: 2005 Sep  8  (814) Tauris

1 star from Gaia EDR3, with no proper motion available. The UCAC4 position (VizieR catalogue I/322A) at the UCAC4 epoch was used to derive and add proper motion values
Tycho2 1299-00981-1: 2012 Oct 5 (232) Russia

Additionally, there is one occultation observed on 2017 May 15 by a radio telescope (the Very Long Baseline Array), of an object identified as J0141+268. This is the QSO WISEA J014433.55+270503.0 which is present in Gaia DR3 as Gaia id = 298674156068220672. 
============
Shape Models
============
The data set includes Shape model fits wherever possible. There can be up to 6 shape models for an event in this data set. Two sources of Shape Models have been used:

* DAMIT [Database of Asteroid Models from Inversion Techniques] https://astro.troja.mff.cuni.cz/projects/asteroids3D/. [See also Durech et al. (2010), DAMIT: a database of asteroid models  https://ui.adsabs.harvard.edu/abs/2010A%26A...513A..46D/abstract]

* ISAM [Interactive Service for Asteroid Models]  http://isam.astro.amu.edu.pl/
In 2023 the shape models in ISAM were renumbered, with model numbers less than 100 being duplicates of models in DAMIT. Models numbered greater than 100 are not present DAMIT. This data set only includes ISAM models with numbers greater than 100.

A shape model quality setting is provided, with the following definitions:
0 = Not fitted. Too few chords to make a fit to the shape model.
1 = Bad occn data. The occultation observations are either inconsistent or unreliable, with no ready way of distinguishing between reliable and unreliable observations
2 = Model wrong. The shape model is clearly inconsistent with the observed chords
3 = Minimum Dia. Usually associated with a single chord observation, with that chord being matched to the largest dimension of the shape model to give its minimum diameter
4 = Diameter but no fit. The chords do not sensibly match the shape model, yet it is possible to derive minimum and maximum possible diameters. [NOTE: flag 1 - Model wrong - should be used unless there is a degree of confidence that the derived diameters are 'in the right ball-park']
5 = Poor fit. The shape model is generally consistent with the observed chords - but has some broad deviations
6 = Good fit. The shape model has a high degree of correspondence with the observed chords.
7 = Not constrained. Is used when ever the chords are insufficient to place any reasonable constraints on the size of the asteroid. Typical situations are: chords with very large event uncertainties compared to the nominal diameter, chords which can be fitted to the shape model over a large range of diameters, and chords which are shorter than about 60% of the asteroid's nominal diameter,

The Shape model fit does not redetermine the shape of the shape model. It is simply a matching of the shape model (with possible rotational variation) to obtain the best fit of the chords to the shape model. The diameter of the asteroid is obtained by deriving the length of the unitary scale of the shape model that gives the best visual fit to the occultation chords.
 
 
========================
Light curve availability
========================
Over 13000 asteroid occultation light curves from recent events are available at:
https://vizier.cds.unistra.fr/viz-bin/VizieR-3?-source=B/occ
The number of available light curves is rapidly growing, with them being a requirement for observations from the start of 2024.


=====================
Modification History 
==================== 
The first version of this data set, introduced in 2003, included occultations only through to 1998. The update of 2004 not only added occultations through to March 1, 2004, but also provided a more systematic arrangement of the data. The data set has been updated annually since then. From 2019 to 2023 the dataset has been extensively reviewed and modernized to cater for more recent observational techniques, improve the error modeling, incorporate the results of shape model fitting, and to detect and correct data errors. 

The number of asteroid occultations included in each successive version is as follows: 
Year: Version: Number of occultations: 
2003 V1.0 183 
2004 V2.0 524 
2005 V3.0 680 
2006 V4.0 865 
2007 V5.0 1055 
2008 V6.0 1203 
2009 V7.0 1417 
2010 V8.0 1662 
2011 V9.0 1935 
2012 V10.0 2102 
2013 V11.0 2275 
2014 V12.0 2469 
2015 V13.0 2717 
2016 V14.0 2933 
2017 PDS4 V1.0 3224 
2018 PDS4 V2.0 3598 
2019 PDS4 V3.0 4342
2024 PDS4 V4.0 9729
2025 PDS4 V5.0 [EventCount]
