Landmarks and Changes to the Asteroid Orbital Elements Database

Asteroid Orbits File Landmarks

Wednesday 9 April 2008
Today, we have implemented a new roster of perturbers in our integration scheme. Previously, the perturbers were Mercury through Pluto (Eath and Moon separately) plus 1 Ceres, 2 Pallas, and 4 Vesta. We have now added perturbations from 10 Hygiea, 15 Eunomia, 52 Europa, 511 Davida, and 704 Interamnia. Most of the asteroid masses have been adopted from JPL's DE-405 ephemeris.

Wednesday 9 April 2008
The switch to the new server is partially complete. Our software appears to run correctly on the new server, but WWW URLs do not work. Please be patient while we try to fix the web interfaces, a nontrivial task.

Tuesday 1 April 2008
Today, we will switch to a new server. Therefore, our asteroid services URLs will be down for a while---perhaps for as much as a day or two. We will issue an "all clear" when services appear to have been restored.

Saturday 12 May 2007
Several coorespondents have recently pointed out errors in asteroid absolute magnitudes (H). Larry Wasserman and I (Ted Bowell) have identified a couple of coding errors that have led to the problem, and I have recomputed H values (and orbits) for the offending asteroids---about 3000 of them. Please let me know if you spot any more errors in H values.

Friday 3 September 2004
The problem mentioned on Wednesday 25 August has been dealt with.

Wednesday 25 August 2004
We have corrected a small error in the computation of the geocentric-to-topocentric positions of asteroids. The correction affects orbits and ephemerides (using our software) computed on or before Tuesday 24 August 2004. The correction is generally small (< 0.01 arcsec in ephemerides), except for some Earth-approaching asteroids, where it may amount to several tenths of an arcsecond. We will be recomputing orbits for the affected NEAs in the near future.

Thursday 12 February 2004
The problem reported on Monday 9 February appears to have been corrected.

Monday 9 February 2004
The corruption problem should have been corrected, though we will be monitoring the orbit files for a few days to verify. Full daily updates should now proceed as normal, beginning tomorrow morning.

Monday 9 February 2004
Corruption of two asteroid orbital element files has occurred. Full daily updates have been suspended pending correction.

Wednesday 3 July 2002
Full daily updates of astorb.dat have restarted following inclusion of data from the 24 June 2002 Minor Planet Circulars. The shortcuts referred to in the note of Monday 3 June 2002 will be applied till further notice.

Monday 3 June 2002
Full daily updates of astorb.dat will be resumed on Tuesday 4 June 2002 UT following inclusion of data from the 26 May 2002 Minor Planet Circulars. Note that we have taken some short cuts in the preparation of astorb.dat this month. In particular, ephemeris and similar programs may not work for a while for all asteroids if dates far from current are requested. Please let know of any problems you encounter.

Monday 17 December 2001
The problem noted on 25 November has now been fixed, and a new astorb.dat, with related files, is available.

Saturday 15 December 2001
The problem noted on 25 November seems not to have gone away. For the time being we will not be able to post all the usual files generated by our automatic nightly update. astorb.dat will be posted, however.

Tuesday 11 December 2001
The problem noted on 25 November has been isolated but not fixed. We expect it will go away when data from the 30 November 2001 MPCs have been assimilated, which should occur by this weekend.

Sunday 25 November 2001
One of the component programs of the nightly update of astorb.dat is hanging. Unfortunately, I (Ted Bowell) and my knowledgeable colleagues are attending a meeting in New Orleans, and we cannot fix the problem remotely. Expect a fix in early December. Meanwhile, orbits from the daily MPECs are being incorporated into astorb.dat as usual.

Tuesday 31 July 2001
Henceforth we will be including some MPC-published orbits in astorb.dat. On a day-to-day basis, these will comprise MPEC orbits, and will be so identified by the MPEC reference in the "orbit computer" field. During times of transition, when MPC monthly update files have been published but orbits have not been computed at Lowell, MPO orbits will be found in astorb.dat. The presence of an MPEC or MPO orbit implies that it is preferred to the orbit we can calculate at Lowell. This is because MPEC and MPO orbits are based on more observations than are available to us. For the same reason, MPEC and MPO orbits contain no ephemeris uncertainty data. Likewise, the arclengths of multi-apparition MPEC and MPO orbits are guesstimates.

Friday 23 February 2001
The problem reported yesterday should now have been corrected.

Thursday 22 February 2001
We've been having a problem with one of the component programs of the full nightly update procedure. (It's a program that generates a current file of ephemerides, which, in turn, is used by utilities such as plast). Should be fixed for tonight's run, which we hope, therefore, will produce uncorrupted output.

Thursday 28 September 2000
Full daily updates of astorb.dat will be resumed on 29 September UT following inclusion of data from the 13 September 2000 MPCs. We apologize for the delay, which has been caused by file dimension problems (internal files got too large for our operating system!). Please let us know if you see any problems.

Sunday 6 August 2000
The residual problems with astorb.dat, noted on Thursday 3 August 2000, have now been corrected.

Thursday 3 August 2000
Owing to lightning damage to part of Lowell Observatory's internal network, we have been unable to complete the orbit database update today. A full daily update will be carried out tonight, but users should note that there will be known errors, such as a lack of new names for numbered asteroids, that will persist for a day or two. Watch for an "all clear" message.

Monday 10 July 2000
The problem of the corruption of orbits by duplicate observations, reported on Saturday 8 July, should now have been corrected.

Saturday 8 July 2000
Full daily updates have been resumed following inclusion of data from the 21 June 2000 Minor Planet Circulars. Due to the publication of the original PLS, T1S, T2S, and T3S in those MPCs, and the way we process new observational data, there are about 1000 orbits in astorb.dat that contain duplicate observations (and have thus been slightly corrupted). Those orbits will be recomputed soon.

Saturday 22 April 2000
From this date on, orbit computations will make use of perturbing body ephemerides from DE-405. Old orbits, which use DE-403, will gradually be replaced.

Thursday 20 January 2000
astorb.dat should be in good shape again (see the note of Monday 3 January). Thanks to Tiziano Magni and Zdenek Moravec for pointing out some y2k bugs (we thought, in our innocence, that there would be few if any). Please let Ted Bowell ( know if you see any additional bugs.

Monday 3 January 2000
Owing to my (Ted Bowell's) vacation, there will be no full updates of astorb.dat till at least Friday 14 January 2000.

Saturday 25 September 1999
Most of the upgrade process, referred to on Friday 17 September, went smoothly. One prolonged snag has been reinstituting full nightly updates of astorb.dat. However, we think that all the Lowell asteroid services are functioning normally. Let us know if you find otherwise.

Friday 17 September 1999
We will be upgrading our server for Lowell Observatory's asteroid services, including the provision of astorb.dat, on Monday 20 September. We hope that the process will take less than a day, but there may be snags... Also note that asorb.dat will not be fully updated from tomorrow until at least Wednesday 22 September.

Saturday 11 September 1999
Full nightly updates of astorb.dat will be resumed tonight following incorporation of data from the 1999 August 31 MPCs. However, there may be one or two strange bugs. We are working on them, but do let us know if you find anything unusual with astorb.dat or the various asteroid services.

Tuesday 31 August 1999
Further to the note of Thursday 26 November 1998: Orbits hitherto uncomputable due to a relatively new method of computing covariance matrices have now been reinserted in astorb.dat following a change to our computational methods. Mostly, the covariance matrices remain indeterminate, which results in erroneous ephemeris uncertainties. The ephemeris uncertainties should, however, be large, so users will be alerted that the objects affected cannot easily be recovered.

Monday 5 July 1999
Peter Schlatter has pointed out that the high inclinations of 1999 LD31 and 1999 LE31---both larger than 100 deg---caused overflow in their records in astorb.dat. The problem has been fixed with a format change, as also noted in astorb.html.

Tuesday 4 May 1999
Mark Sykes has pointed out that asteroid diameters given in astorb.dat are from a preliminary IRAS database. Accordingly, we have replaced them with values from the final IRAS database, as provided by the Small Bodies Node of the Planetary Data System.

Friday 2 April 1999
Peter Schlatter has pointed an error in the epoch of orbital elements in astorb.dat: rather than being 19990502 it is 19990002. The error has now been fixed, but users who have downloaded astorb.dat during the past three or four days should check to see that the epoch is indeed 19990502. Otherwise they should download astorb.dat again.

Monday 1 February 1999
The LONEOS web service failures referred to on Sunday 17 January should now have been fixed.

Sunday 17 January 1999
Following the format change to astorb.dat (Thursday 31 December 1998), certain LONEOS asteroid web services do not work: obs (asteroid observability chart), astfinder (asteroid finder chart). The following works for some asteroids and not others: astephem (asteroid ephemeris), The underlying problems should be fixed over the coming days.

Friday 8 January 1999
We have modified the meanings of values of integer code 1. Specifically, value = 1 now pertains to Aten asteroids, value = 2 to Apollos, and value = 4 to Amors. Other values have been corrected where errors were detected. For details, refer to the table of integer-code values in astorb.html.

Wednesday 6 January 1999
We have heavily modified our "critical list" of asteroids in need of astrometric observation. In particular, there are now two lists: one containing all appropriate asteroids, the other containing just those that come to opposition during the current lunation. Click on here.

Thursday 31 December 1998
The format of astorb.dat has been changed in accordance with the remarks (below) dated Saturday 26 December 1998. The text describing astorb.dat (astorb.html, accessible from the web) has been modified to describe the change.

Saturday 26 December 1998
We will be changing the format of astorb.dat to accommodate the impending numbering of asteroids > 10000. The changeover, which will most likely occur during Tuesday 29 through Thursday 31 December 1998, will comprise an extra column at the left of each record. Stand by for further information.

Monday 7 December 1998
We are modifying the procedure for generating our critical list of asteroids in need of astrometric observation. Things may go wrong in the next few days, so please bear with us.

Thursday 26 November 1998
Our new method of computing covariance matrices (see note of Friday 9 October) results in some quasi-indeterminate orbits being rejected. The objects concerned (and therefore missing from astorb.dat) are 1998 SM35, 1998 SL45, 1998 TN1, 1998 UZ14, 1998UD16, and 1998 UQ16.

Monday 9 November 1998
We have now more or less recovered from last Friday's disk crash. Please report any oddities to

Saturday 7 November 1998
We've had a serious disk crash, which has affected astorb.dat. Efectively, we are at present unable to do the daily updates, so the ephemeris uncertainties and some integer codes are stale. We may not be able to resolve the problems till next workweek.

Friday 9 October 1998
We have been improving the computation of ephemeris uncertainties. Data for multiple-apparition asteroids should be largely unchanged, but ephemeris uncertainties for many single-apparition asteroids will increase. We hope that the new results will better reflect reality.

Tuesday 3 February 1998
Integer code 2:128 has been implemented. It pertains to asteroids that have been sought but not found. Several possible reasons can be adduced: (1) Orbit and/or current ephemeris uncertainty in error. (2) Search not deep enough. (3) Asteroid image blended with that of a star. (4) Observer error.

Wednesday 24 December 1997
We have developed a "critical list" of asteroids, containing numbered asteroids whose orbits are unsatisfactory or suspect and unnumbered asteroids that are likely to be numberable after a modest astrometric effort. Start at

Friday 12 December 1997
We think we have cleared up the problem mentioned yesterday. But, as usual, please do not place much reliance on orbits and current ephemeris uncertainties derived from small numbers of observations spanning short arcs.

Thursday 11 December 1997
Re the note on Monday 1 December: In changing our protocol for computing covariance matrices, a problem has arisen that affects mainly short-arc, single-apparition asteroids: a few CEUs and PEU10s have zero value. Clearly, these are erroenous and should be ignored.

Monday 1 December 1997
We are making some major changes to the structure of some files that are used to generate astorb.dat. The changes will be transparent to most users, except that we will be inserting new orbital elements of numbered asteroids over the next few days. These will show up in the *.add files. Most of the new elements will differ very little, if at all, from those already in astorb.dat.

Thursday 23 October 1997
Re the note on Sunday 5 October 1997: P/1996N2 = 1979 OW7 has now been numbered as an asteroid (7968). Therefore, the orbit for 1979 OW7 has been deleted.

Sunday 5 October 1997
Seeing no reason why the Themis-family object P/1996 N2 = 1979 OW7 is not classified as an asteroid, we have added an orbit for it to astorb.dat under the designation 1979 OW7.

Monday 22 September 1997
Orbits of asteroids resulting from Hipparcos observations have been incorporated in astorb.dat.

Tuesday 2 September 1997
The problem mentioned on 13 and 14 August should now have been solved.

Thursday 14 August 1997
The problem mentioned yesterday has recurred in today's automatic update of astorb.dat. But now the symptoms are clear: affected are data for asteroids whose ephemeris uncertainty parameters have become outdated, and are therefore in need of recomputation. We are uncertain when the problem will be corrected, but will try to put things to rights on a day-to-day basis for the next few days.

Wednesday 13 August 1997
Yesterday, we made some changes to our internal asteroid files that are used to create astorb.dat. After the daily automatic update, we noticed that future ephemeris uncertainties and their dates of occurrence, along with corresponding integer codes number 6, had been corrupted in the case of more than 100 asteroids (both numbered and unnumbered---no obvious pattern). Because of vacation time and upcoming travel, it is possible that we will not be able to correct the problem till early September. It might be that the corruption is cumulative (i.e., data for an additional hundred or more asteroids will become corrupted each night), in which case users should be aware that very large values of future ephemeris uncertainties (+/- 10^15 arcsec, say) are erroneous. Also, note that we do not expect to incorporate orbits deriving from the 18 August 1997 Minor Planet Circulars until early September.

Monday 2 June 1997
The known errors in astorb.dat, referred to on Sunday 1 June 1997, have been corrected.

Monday 2 June 1997
Ladislav Nemec has pointed out some typos in designations [for asteroids (7196) and (7590)] and duplicate orbits [for 1997 EM6 and 1997 GR9]. These errors have been corrected.

Monday 12 May 1997
Milos Tichy and Francois Ochsenbein have pointed out that, for some asteroids, the second integer codes in astorb.dat have values of -128 or -126. These are erroneous, and should be 64 or 66, respectively. Please contact if you notice a recurrence of the problem.

Tuesday 18 March 1997
Thanks to Albert Timashev, St. Petersburg, Russia, we have corrected 35 mis-spelled or missing asteroid names. We now believe the names of numbered asteroids to be correct (less diacritical marks, of course). Please let us know if you find additional errors. (Note that the principal preliminary designations of numbered asteroids are entered by hand, so errors undoubtedly exist among them.)

Saturday 1 March 1997
Owing to a procedural error, the file of deleted orbits created on 27 February (970227.del) has been lost. To avoid duplication of orbits, users are urged to download a new version of astorb.dat.

Tuesday 4 February 1997
Tim Spahr (University of Florida) has pointed out an error in the absolute magnitude (H) of 1977 TD. Investigation revealed that a software bug, which has apparently persisted for several months, caused values of H for unnumbered asteroids not to be updated when new observations containing magnitude estimates were incorporated in new orbits. Concerned users of astorb.dat are urged to re-download the file.

Sunday 29 December 1996
The error remarked on Sunday 29 December 1996 has been corrected, so daily updates have been resumed.

Friday 3 January 1997
The epoch of orbits in astorb.dat has been changed from 19961113 to 19970221.

Sunday 29 December 1996
Due to a change in operating system on one of our computers involved in the daily automatic updating of astorb.dat, the compressed version of the file (astorb.dat.gz) and the *.add files cannot be generated. As a temporary fix, we will post current versions of astorb.dat (uncompressed). The fault may persist for another week (that is, until year-end revelries have come to an end).

Tuesday 3 December 1996
New values of magnitude parameters H and G for numbered asteroids have been incorporated into astorb.dat in accordance with those in the 25 November 1996 Minor Planet Circulars.

Friday 11 October 1996
Some users have reported difficulty reading the absolute magnitude H because of variable format (values such as 14, 14.5, 14.56). In FORTRAN, a fix is as follows

C---  At some point you read the data such that the 5 characters (at most
C---  DD.DD of the H value are in HCHAR, then proceed as follows:
      IF(HCHAR(3:3).EQ.' ') HCHAR(3:3)='.'
1     FORMAT(F5.2)

Wednesday 9 October 1996
In response to users who have difficulty downloading the complete astorb.dat, we are instituting a method of logging changes in the orbital elements therein. Users will now find files such as yymmdd.add and yymmdd.del in our public ftp area accessible from the WWW. These cover activities for the past 30 days. Thus, 961007.add comprises records for asteroids whose orbital elements were either added to or replaced in astorb.dat on 7 October 1996; and 960930.del contains records for asteroids whose orbits were deleted from astorb.dat on 30 September 1996. Note the following: (1) The *.add files are refreshed every day. Thus, an orbit that was replaced on, say, 961001 and then again on 961002 will have been found only in the 961002.add file on or after the latter date. (2) The *.add files reflect changes in orbital elements only. Changes to other parameters (such as H, G, and integer codes), as well as daily updates of ephemeris uncertainties, are not reported in the *.add files. Users who are interested in the latter will need to download the entire astorb.dat.

Monday 7 October 1996
Along with automating the routine daily updating of astorb.dat, we are now updating the computation of the minimum orbital intersection distances (MOIDs) for planet-approaching asteroids. They are available starting at

Thursday 3 October 1996
We have automated the routine daily updating of astorb.dat. The update will start at 08:00 UT, and should take about 1 hour. It consists of: (1) Recomputing current ephemeris uncertainties of date; (2) Recomputing peak ephemeris uncertainties that have "expired" (that is, have peaks at past dates); (3) Recomputing integer codes; (4) Transferring updated files to the publicly accessible ftp area for access via ftp or the WWW. The files will be unavailable during step (4), which should take a few minutes at about 09:00 UT.

Thursday 22 August 1996
To solve the problem encountered on 17 August (q.v.), the format for semimajor axis has been changed from F11.8 to F12.8, with the result that the length of records in astorb.dat has increased from 264 to 265 columns. See astorb.html or astorb.txt for new record examples and FORTRAN format statement.

Saturday 17 August 1996
The semimajor axis of 1996 PW overflows the available field size, being more than 100 AU. The value should be a = 141.3007235273188 AU. The error will be corrected in a day or two.

Sunday 11 August 1996
All DE205 orbits have now been recomputed and replaced by DE403 orbits (see note of 15 November 1995).

Wednesday 19 June 1996
The few known errors in PEU10, resulting from corruption of the observational database and remarked on Friday 14 June, have been corrected. We hope that similar remaining unknown errors are not abundant.

Friday 14 June 1996
The errors in PEU10, remarked on Thursday 30 May, have largely been corrected. A few errors may remain, due to corruption of our observational database. We will be checking the database next week.

Tuesday 11 June 1996
Asteroids having unknown absolute magnitude H (about 200 of them) have been flagged by setting integer code 2 = integer code 2 + 64. H has been set to 14 mag (a guess).

Friday 31 May 1996
Values of integer code 6 (see astorb.html) have been redefined. The new definitions will be given in astorb.html.

Thursday 30 May 1996
The error in CEUs, remarked on Tuesday 7 May, should have been corrected. However, a new error, in which the 10-year peak ephemeris uncertainties (PEU10) of some well-observed unnumbered asteroids are much larger than they should be, has come to light. We hope to correct this latter error within a few days.

Tuesday 7 May 1996
An error has been detected in the method of computing current ephemeris uncertainties (CEUs) for 3-observation asteroid orbits in astorb.dat. The error, which has persisted since the provision of CEUs on 16 April 1996, leads to grossly underestimated CEUs. We are working on a fix.

Tuesday 16 April 1996
Ephemeris uncertainty parameters have been incorporated in astorb.dat. See astorb.html or astorb.txt for details of the resulting format change and for notes on applications of the uncertainty parameters.

Tuesday 2 Aril 1996
Most orbits having assumed eccentricities (integer code 2:2) have been recomputed.

Saturday 30 March 1996
A software improvement has resulted in (statistically) more reliable assumed-eccentricity orbits computed on and after this date.

Friday 8 March 1996
Brian Marsden, Director of the Minor Planet Center, has recently discontinued the availability of batches of astrometric observations of asteroids and comets in conjunction with the monthly publication of the Minor Planet Circulars (though observations of individual asteroids are still available as previously to those who subscribe to the MPC's extended computer service). Consequently, we will not, until further notice, be able to acquire the observations published in the MPCs very rapidly, and will not be able to update astorb.dat soon after publication of the MPCs.

Wednesday 14 February 1996
The format of records in astorb.dat has been changed.

Wednesday 15 November 1995
The planetary ephemeris used to compute asteroids was switched from DE245 to DE403. DE245 orbits will gradually be replaced by DE403 orbits.

Last updated 16 Apr 1996 at 22:47:47 U.T.
Contact: Bruce Koehn (
Web Curators: Ted Bowell and Bruce Koehn
Icon Go (Back) to the Lowell Observatory Home Page
Icon Go (Back) to previous page