Astrometry Tools and the Aladin Sky Atlas by Peter Campbell-Burns

I have recently completed an Advanced University Certificate in ‘Astrophysics of Galaxies’, a challenging two-year distance learning course run by the University of Central Lancaster.  One of the course assignments required students to research the characteristics of a given peculiar galaxy using astronomical databases and other information sources available on the Internet.  Whilst gathering the information I needed, I came across some powerful yet easy to use Virtual Observatory software tools for the retrieval and analysis astronomy images and data.  I was able to use these with very good effect to complete my assignment.

The following article presents a brief overview of three of the more interesting tools.  If there is sufficient interest amongst members then perhaps a presentation on this subject can be arranged for a future meeting.  In the mean time, there are a number of references and links at the end of this article should you want to find out more. 

The Aladin Sky Atlas

The Aladin Sky Atlasis perhaps the most exciting and useful tool for the amateur astronomer.  Aladin was developed and maintained by the Centre de Données Astronomiques de Strasbourg (CDS) and is one of several tools that comprise the European Virtual Observatory (VO) toolset.  It is available to download as Java program which can be run as stand-alone software on the PC, or as a java applet which will run in your Internet browser.

With Aladin you can perform tasks such as:

• Visualising digitized astronomical images;
• Locating data of interest;
• Visualising multi-wavelength data by superimposing entries from astronomical catalogues or databases;
• Photometric analysis.

In just a few clicks you can access and download related datasets and information from many sources including the SIMBAD database, the VizieR service and many other astronomy data archives.

Crab Nebula visualised using Aladin and showing reference points and proper motions (click to enlarge)

With Aladin you can select individual data points and drill down into the data itself to view data such as distances, spectra, proper motions, etc.

SIMBAD data for the Crab Nebula progenitor pulsar (click to enlarge)

Aladin gives unprecedented access to contextualised multi-wavelength data and is a valuable tool for any amateur astronomer looking to understand more about the universe.  However, for astronomers who spend their nights acquiring beautiful deep-sky images Aladin enables these to be combined with scientific data acquired by the large ground based and space observatories.  This is where Aladin really comes into its own.

Astrometric Calibration

To use Aladin with your own images (or an image acquired from the Internet) you need to first perform an astrometric calibration to create a calibrated image in FITS format.  A major feature of the FITS format is that data about the image is stored in a human readable ASCII header.  A calibrated FITS image header holds information on location in the sky and image orientation; this information is held in WCS (world co-ordinate system) format.

There are a number of ways to perform astrometric calibration – or plate solving as it is usually called (a hangover from the days of photographic plates).  Plate solving consists of using star pattern matching against a given stellar catalogue to calculate the coordinates of the centre of an image. 

Some commonly used image acquisition tools such as Maxim DL and Astroart include a plate solving function – these are effective but require more than a few hints about image scale and the approximate location in the sky of the area imaged.  Also, both Maxim DL and Astroart are not cheap – and are likely to be of interest only to imaging astronomers.

An alternative is to use astrometry.net.

Plate Solving with astrometry.net

Astrometry.netis plate solving software but unlike Maxim-DL and Astroart it performs ‘blind astrometry’ on a FITS or jpeg image. – in other words it does not need to know anything in advance about the location, orientation or image scale to calibrate an image.  Solving blind can take a lot of CPU time (there’s a lot of stars out there), but if you are able to provide the software with some information about the image – e.g. an approximate image scale or location then it will solve images much more quickly.  When it has solved an image astrometry.net returns a number of useful outputs which include:

• A calibrated FITS image;
• An annotated image identifying the main objects in the field (constellations, NGCs, Messier Objects);
• Overlays which can be imported into Google Sky.

Astrometry.net is offered as a web service to alpha testers but the software is offered under an open source license and can be run on your own PC (requires Linux – it is not available for Microsoft Windows).  Alternatively, you can submit your astronomy images to the Flikr which runs an astrometry.net ‘bot’ to calibrate all images submitted to its astrometry group (though this does put your images in the public domain).

I’ve processed over 100 images using astrometry.net (wide field and deep sky) and it has calibrated all but a few images – invariably, these failed images were of poor quality, featured noisy backgrounds and perhaps some false artefacts as a result of processing.

A small word of warning however: if you plan to run a local version of astrometry.net – as well as having some understanding of the Linux operating system you will need to be prepared to download a lot of data.  The indexes are huge (almost 30Gb)!

Plate Solving With Elbrus

Elbrus is another interesting plate solving application which was brought to my attention recently by Donal McDonnell.  What differentiates Elbrus is that its stated objective is to improve telescope pointing accuracy and it does this first by determining the image centre (i.e. plate solving).  Elbrus can then feed the image centre back to the mount to improve pointing accuracy.   

However, unlike astrometry.net it cannot blind solve; to solve a particular image it needs to know:

• The pixel scale of the frame measured in arcseconds per pixel in both axis, and
• The approximate location of the centre of the frame.

Pixel scale is the angular dimensions of the area of the sky covered by each pixel – as opposed to plate scale which is the angular dimensions of the area of the sky covered by the entire frame.   Pixel scale is calculated as the angular distance between two points in arcseconds (in the x or y dimension) divided by the number of pixels (in the x or y direction).  Alternatively it can be calculated as the overall plate scale in arcseconds divided by the CCD dimension measured in number of pixels. 

Determining the plate scale requires an initial one-off calibration step for the particular telescope and CCD setup that you are using at the time.  Elbrus can be calibrated by identifying stars in a calibration image and providing their GSC catalogue numbers so that Elbrus can match these to positions held within its database – or if you know the image scale already the required parameters can be simply entered directly.  Accurate calibration is essential.  Almost any image taken with the same telescope and CCD configuration will serve as a calibration image so long as you can identify the stars and obtain their GSC catalogue numbers.  For my own calibration I used an Image of M31 and SkyMap Pro to identify the stars and their GSC numbers.

Elbrus has many potential uses and is a valuable tool before, during and after imaging.  Its uses include:

• Determining the precise location and angle of a frame;
• Analyzing pointing errors;
• Creating a multipoint star map for alignment;
• Measuring field rotation (and therefore as a tool for analysing polar alignment);
• Re-aligning to a previous frame so new frames can be captured at the same precise location;
• Adding WCS positional data to a fits header (e.g. for post image analysis with Aladin).

Having some information to work with makes Elbus fast compared with an astrometry.net blind solve and, potentially, a more useful tool to have with you beside your telescope.  astrometry can be fast if pixel / plate scale and location information is provided but being Windows based and easy to install Elbrus will almost certainly be better suited to the needs of most amateur astronomers as an ‘at the telescope’ tool.

My own interest in Elbrus lays in being able to more quickly find objects and centre them on the CCD.  Without a permanently mounted telescope, the need to realign at every session means that my GOTO’s are sometimes just off target.   If after an initial alignment my intended target object is out of view I should be able to use Elbrus to determine the actual image centre and work out in which direction and by how much I need to move the telescope.  (I gave up on the idea of using setting circles and sky maps the moment I connected my EQ6 to my laptop and installed the EQMOD software.)

Elbrus is a COM server; if you are a whiz with a computer then it should be possible to pick up the frame centre from the solved plate in Elbrus and then work out automatically where to slew to centre a target. 
Whilst the possibilities seem almost endless – I’m just relieved that I no longer need to be ‘lost in space’ and wasting time and a precious clear sky whilst trying to find and centre my target!

As mentioned already, similar plate-solve functionality is provided by the likes of Maxim DL and Astroart but my own experience is that Elbrus will find solutions more readily than Astroart (though this may be down to the operator)  and Maxim DL is beyond the software budget of some amateurs – including me.

I’m not an expert on this subject but would expect there to be a limit beyond which field curvature (e.g. a large format CCD used without a field flattener) to cause generic plate solving algorithms to fail in finding a solution unless there are algorithms to somehow model or compensate for curvature.  Whether or not there is any practical limitation for this software with typical telescope / CCD combinations is something that I would like to understand better and I’d welcome any comments on this.

Elbrus can be downloaded from the StarLocatorElbrus group on Yahoogroups.com. 

Summary

Initiatives such as the European Virtual Observatory are making more tools and distributed data accessible to the professional and amateur alike.  Now even the amateur has the opportunity to undertake original research using resources that were once available only to professional astronomers.  There is no denying that this is a very exciting time to be an amateur astronomer.

If you find any more useful tools out there on the Internet I would be glad to hear from you.   John Murrell of the Croydon Astronomical Society has posted on his web site some excellent material on the AmateurVirtual Observatory (including Aladin and astrometry.net) and links to this material are also provided at the end of this article.  John has already published some updates so it is well worth keeping a regular look-out on his site.

A key message is that Virtual Observatory tools are not something of relevance to only imaging astronomers they can be used by anyone with a deep interest in astronomy, an Internet connection and just a little computer knowledge. 

Post Script: DASCH (Digital Access of a Sky Centuary @ Harvard

Astrometry.net is currently being used by the DASCH to digitise the Harvard Astronomical Plate Collection; There are over 500,000 glass photographic plates in the collection, exposed in both the northern and southern hemispheres between 1885 and 1993 making it a unique astronomy resource.  This project moves the collection from the 19^thCentury into to the 21^stCentury and extends the virtual observatory with over 100 years of historical data.

Further Research:

Some useful references are as follows:

• Aladin download is available here
• The astrometry.net web site
• A paper on astrometry.net and blind astrometric calibration is available here
• Astrometry.net on Flikr
• An article on astrometry.net and Flikr is available here
• Elbrus on yahoogroups.com
• John Murrell’s website:
  • Article on the Amateur Virtual Observatory
  • John’s presentation to The Astronomer annual meeting 2009
  • Amateur Virtual Observatory links
• Information about the FITS image format is available here
• Information on the European Virtual Observatory is available here
• The SIMBAD astronomical database can be found here
• The VizieR astronomical database can be found here