Re: Identifying observed satellite

Bjoern Gimle (
Thu, 26 Jun 1997 10:21:18 +0200

Atul Sowani wrote:

>>...While observing the satellites (known to be visible at that time),
>> I came across a few "unknown" satellites as well. I have recorded their
>> paths roughly, WRT stars, i.e., I know rough RA, Dec. Is there any
>> software avaible which will read these RA, Dec, and will read the TLE
>> database I have, and then tell me the possible satellites following
>> that path? Does anybody know a similar software? Where can I get it?

Bruno Tilgner answered:
>I am in the process of finishing such a software myself. If you (or anybody
>else for that matter) give me the following information I will identify the
>object with a high degree of confidence, provided it is in the TLE database
>at all:
>1. Geographical coordinates of observer.
>2. UTC date.
>3. Time of observation in UTC plus estimated error.
>4. RA and declination or Greek letter and constellation of the star where
>   the unknown object passed.
>5. When it did not pass exactly over that star: approximate distance and
>   direction.
>6. Approximate direction of flight (very important to distinguish between
>   several potential candidates).
>7. Any other useful information such as magnitude, flashes etc.
>Items 3 to 6 may be repeated for several stars in the path of the same object.
>1. 48.85N, 2.02E
>2. 22 June 1997
>3. 20:55:15 +/- 15 sec
>4. eps Uma
>5. less than 2 deg to the east
>6. NNW to SSE
>7. Moving rather slowly
>Using this data with a recent MOLCZAN.TLE file of 1235 objects would yield
>122 objects which came into view during a time window of +/- 5 minutes 
>before and after the time of observation, rounded to the nearest minute.
>The search would take less than a minute on a machine with a Pentium 100.
>A larger database, such as Mike McCant's ALLDAT.TLE, would take proportionally
>longer to search.
>Of these 122 objects, 4 would pass closer than 6 degrees from eps Uma. Two
>would satisfy the time of observation but only 1 would fit the direction of
>flight. This object would be identified as Cosmos 588 r. ...

I shouldn't evaluate software only based on a description, but it sounds like
a fine piece of artificial intelligence. For me, the main advantages seem to
be for observations done by others, where either:
a. The only direction information is in the form 6. above
b. The time/position/magnitude/speed uncertainty is so large that a lot of
   irrelevant objects are shown by programs like QuickSat and SkyMap.

For my own observations, I find it more accurate to memorize and/or draw
the track among the stars, and note the local direction of the track
(like in QuickSat "Dir" column, in Bruno's example 325 degrees,
or as "5 o'clock to 11 o'clock") and speed
(e.g. time to cross the binocular field, or the distance between two stars)

0. If I have RA/dec of observations, I may want to edit them into a copy
   of a deepsky.txt file, to have them appear on the map. The coordinates
   can be read from a SkyMap cursor position. 

Then I load a SkyMap configuration file, which always contain max.perigee/
range/magnitude/prediction uncertainty, field-of-view, latitude/longitude,
elset file name, map grid/projection, star magnitude limit etc.

1. Change any of the old values, if necessary.
2. Set the date/time, then press F10 on 'Satellite Info' to copy to 'Search 
3. Move the map center by cursor, star/planet/galaxy name or RA/dec, or alt/az.
4. On the main menu, save the updated .cfg if you wish, press F10 to show 

On my 50 MHz 486, Bruno's example takes 60-90 seconds, depending on if I have
used the elset file already, or not.

The direction/position of tracks, time tags/ticks, and USSPACECOM ID along
the tracks makes it easy to see which object matches the observation.

A screen dump of the example, with a 12 deg. FOV, showing the four
close tracks, is posted on
The config file I used is
The HP-GL plot file is, and
the text report (listing also satellites outside the selected area is

A printout from SkyMap's HPGL file to a plotter, a LaserJet 4 or similar
printer, or a HPGL emulator program, has a 6000*8000 resolution, compared
to the 480*640 of the screen dump.

Skymap 5.11 can be loaded from Neil Clifford's ftp site as two files and or one at

and mail and donations can be sent to author Rob Matson,, who can tell you when/where 6.0
or later versions are available.

--              +46-8-7428086  (home)  --
-- 59.22371 N, 18.22857 E           AND member of :       --
--   SeeSat-L              --
--                +46-8-59095783 (office)--

--              +46-8-7428086  (home)  --
-- 59.22371 N, 18.22857 E           AND member of :       --
--   SeeSat-L              --
--                +46-8-59095783 (office)--