Telescopes in XML

Earlier I described my idea for an RSS-like XML feed for telescopes. The idea was to allow anyone to keep up with what particular telescopes were doing. In this post I will try to describe my current idea. First off, before I give a more detailed description of each part, let me show an example of what the finished thing may look like:

<?xml version="1.0" encoding="UTF-8"?>
<telescope xmlns=""
<desc>The 42ft telescope at Jodrell Bank Observatory</desc>
<link href="" />
<image href="" />
<object type="pulsar" name="Stuart's fake pulsar" />
<ra h="0" m="34" s="56.89" />
<dec d="+50" m="12" s="3.7" />
<az act="170.06" dem="169.5" />
<el act="70.06" dem="70.23" />

OK, now for a bit of a description. The first line is a standard start to an XML file. The second line defines the start of a telescope tag. I've defined a namespace at a currently non-existent URL but as far as I can tell, that doesn't actually matter. The remaining lines define aspects of the telescope. First of all we need to define a telescope in a way that would give remote applications a bit of useful background information:

  • name - the given name of the telescope
  • desc - a short description of the telescope for use in information balloons, short biographies etc.
  • link - a relevant web link. I chose this over <a href=""> because having link text didn't seem to make sense
  • image - this should be a picture of the telescope. Perhaps <img src=""> makes more sense but I was keeping away from HTML tags.
  • Point - this follows the Geography Markup Language (GML) format for marking positions after a suggestion from Dave P. The coordinates are the latitude and longitude (in that order) of the telescope on the Earth and are given in degrees. This should probably be optional as it makes no sense for telescopes in space!
  • height - the height above mean sea level in metres. Probably should be optional to space-based 'scopes.

This information would not be expected to change often. Next we define the time that this file was generated:

  • time - the time that this file was generated in ISO 8601 date format. Dave P suggested this rather than the time format used in RSS files.
Now that we've got the preliminary stuff out of the way, define what we are looking at right now.

  • status - a flag to specify things like "slewing", "observing" and"parked/closed/stopped". Should these words be defined?
  • object - the aim here is to provide both a public-friendly description (the name attribute) and a computer readable flag (the type attribute) for the type of object. My suggested types are: asteroid, comet, dwarf planet, planet, star, nebula, ISM, galaxy, galaxy cluster, calibration object. However, the full A&A keywords list of objects may be better to work from as it is accepted within the astronomical community.
  • position - this is where we finally get around to defining the position that the telescope is observing. The most important coordinates to report are the ra (Right Ascension) and dec (declination). I thought it would be sensible to give these broken into hours/degrees, (arc)minutes and (arc)seconds for maximum flexibility. It would also be possible for a telescope to omit the (arc)minutes/(arc)seconds attributes if they wanted to degrade the reported resolution for some reason. A given ra/dec only make sense if an epoch is also defined. I expect this will normally be J2000 but sometimes telescope systems still work in B1950. I also thought it would be good to have an azimuth (az) and elevation (el) specified in degrees, although these could be optional.
So there you have it. It is fairly basic but then the resulting file should be straightforward and easy to use. The only thing this file doesn't do is store a history of observing targets but I'm not sure if that is necessary. A nice acronym for this XML format is also required, so suggestions for that are welcome below e.g. can anyone contrive an acronym for STAR? Please add any problems, suggestions for improvements, additions or previously established standards of relevance in the comments.

Tags: | |
Posted in astro blog by Stuart on Tuesday 04th Sep 2007 (20:33 BST) | 11 Comments | Permalink

Comments: Telescopes in XML

Maybe add some kind of optional subcategory to the object tag? This way the pedantic world of astronomy can be kept happy and the overarching object tag can be kept simple.

e.g. Nebula could cover a plethora of subcategories of planetary nebula, stellar nursery, reflection nebula etc.

Posted by Robert Simpson on Tuesday 04th Sep 2007 (20:27 UTC)

That looks like an excellent start Stuart. Nothing really obvious strikes me as being missing and it encapsulates enough information to meet the aims stated in your previous posting.

I definitely agree with Robert that object type sub-categories make sense as many objects can be sub-divided: clusters (open, globular, galaxy); planet (dwarf, exo) etc. unless you want to have the specific object type defined in that element to keep it short. The A&A document you refer to already gives you the broad categories and sub-categories to make life easier.

As to STAR, how about: Simple Telescope Access and Retrieval

Posted by Nick Bramhall on Tuesday 04th Sep 2007 (21:21 UTC)

Looks like an excellent start. I'm sure this would also be great if used with astronomical databases and telescope hardware - you could track something along with dozens of other telescopes around the world through an RSS feed!

And, for the name, so long as it doesn't end up as "TeleXML" or "StarXML". :D

Posted by Jordn on Wednesday 05th Sep 2007 (10:29 UTC)

For object names, I now notice that STSCI (HST) use a list very similar (possibly identical) to that used by A&A. It is probably best to use that list to make things easier in the future.

Posted by Stuart on Friday 07th Sep 2007 (10:56 UTC)

Thinking about the RA/Dec, perhaps it is better to provide these as strings with the hours/minutes/seconds (degrees/arcmin/arcsec) separated with spaces. Does anyone have a preference?

Posted by Stuart on Friday 07th Sep 2007 (11:04 UTC)

List of telescopes currently providing their pointing information in some form:

Jodrell Bank Observatory telescopes, Torun 32m Radio Telescope and Mt Pleasant 26m Radio Telescope.

Posted by Stuart on Friday 07th Sep 2007 (11:28 UTC)

Regarding the status, and keeping in mind that I know nothing about how pro observatories work, I'd be tempted to say that there's little point in defining a set of status messages. I'd have thought it would probably make more sense to keep it as "free text"?

As for the location on Earth: personally I think I prefer the idea that lat/lon is broken down into two bits of information (as I did with my own observing logs (I was going to link to my example here, but your blog thinks I'm a spammer for doing that)) but I'm guessing there's something to be said for using existing (micro-?)formats.

For for the general problem of location, perhaps build in some future-proofing? Like, specify planet, or in space, which body or L point it's in orbit around, or.... Okay, I'm being silly there. ;)

Posted by Dave Pearson on Friday 07th Sep 2007 (13:38 UTC)

Stuart - this sort of thing obviously needs to be standardised, preferably through the IVOA. The place you'd want such a description of a telescope and its facilities is a resource entry in a VO Registry. Check out the wiki page for IVOA working groups and maybe get involved. If you don't, I might send somebody an email anyway...

The initial standards from IVOA have been mostly about data access - as well as how you set out the metadata for resources in general, how you specify access to images, spectra, databases etc, and standardising the meaning of table columns. But as you already know, the VO Event group have defined how to describe an event like a transient or burst, and there is also a group working on how to describe theoretical services.

But it has always been obvious that some time we would have to describe access to facilities - imagine how nice it would it be if you trawled the VO for data on your favourite source, found there was no X-ray data, and lo ! the XMM proposal form pops up, half filled in. The other reason is the one Alasdair Allan is interested in - running networks of automated telescopes, that send each other messages.

Oh and the other obvious thing is - go down the corridor and speak to Anita Richards !

Posted by Andy Lawrence on Monday 15th Oct 2007 (07:40 UTC)

You should probably look at RTML

which is already being used to describe observations and make observing requests to telescopes. See


for details. Also VOEvent, which is already providing real-time information on telescope pointing (e.g. SWIFT utility messages) and notification of events via RSS feeds, see,


Posted by Alasdair Allan on Monday 15th Oct 2007 (09:29 UTC)

Andy and Alasdair, I think you are making more of my plan than I intended. The idea is to have a very simple format for straightforward reporting. The plan is not to control telescopes or do real astronomical research. It is simple reporting so that those who aren't professional astronomers can answer the question "what is it looking at?". I'm hoping it might be useful for other things too, but this basic idea was the primary motivation.

RTML (and VOEvent etc) looks great (I had actually looked at it before writing this entry) but just too complicated for the purposes I mentioned.

Posted by Stuart on Monday 15th Oct 2007 (17:26 UTC)

gravatarYour Site Are Best Thanks for you Build

Posted by Celestron 21041 60mm PowerSeeker Telescope on Tuesday 24th May 2011 (09:49 UTC)


Don't provide an email/URL unless really necessary as your comment may get caught in the spam filter. No URLs get turned into links so don't bother. The ground rules for commenting are:
  1. No profanity or personal attacks please. Keep it clean.
  2. Restrict comments to subjects relevant to the post.
  3. Don't mention Pluto. If you do it'll be replaced by Goofy.
  4. No spam i.e. anything commercial unrelated to astronomy.
  5. If you think you've discovered a Theory of Everything, a replacement to Relativity, or something similar then please publish it in a journal rather than in my comments.
Comments against the spirit of these ground rules may be removed.

* required fields