Correcting Hubble images

[tweetmeme only_single=false source=allinthegutter]

I’m going to be upfront here: this post is about CCDs and readout electronics. Wait, come back, it’s going to be interesting I promise*. It involves the Hubble Space Telescope. Everyone likes that, don’t they?

Many astronomical instruments, including the ones on Hubble, use Charge Coupled Devices or CCDs. These detectors are like containers for electrons, and they’re divided up into grids of boxes, sort of like ice-cube trays. When a photon of light falls on the a particular part of the CCD it drops an electron into the corresponding box on the grid. To make an image the number of electrons in each grid section need to be counted as they represent the distribution of the detected light. To do this each row in the tray is emptied in turn, with the electrons shifted box-by-box towards the edge of the grid, as illustrated by this handy animation:

I’m glossing over a lot of the details here, but that’s basically how a CCD works (for more details try this How Stuff Works page on digital cameras, which use the same technology).

The CCDs onboard the Hubble Telescope are continually bombarded with cosmic rays which gradually damage them. In particular pockets can form which trap electrons and prevent them from being read out at the correct time. When they’re eventually released, they’re erroneously counted and included wrongly in the image, leaving distinctive trails behind bright objects.

This is a major problem for people who need to accurately measure the shape of galaxies in their Hubble images. One such person is Richard Massey who studies weak lensing – the small distortion in the shapes of galaxies caused by their light being bent by a massive foreground object, like a galaxy cluster. Obviously, the presence of little trails behind these galaxies is going to mess things up. Good news though, Richard, and his collaborators have come up with a way to fix this issue. They model the CCD system and manage to figure out where all the trapped electrons have come from and therefore can put them back where they belong at the initial stage of the data processing. This means that degraded images can be corrected back to how they would have been if they’d been taken years earlier, before the cosmic radiation had damaged the CCDs. As you can see in the images below, this will hopefully make things easier for the weak lensing people.

The before-correction (left) and after-correction (right) images from Massey et al 2010, demonstrating how successful this technique is at removing the trails seen behind many objects on the left

*interest levels not guaranteed

ResearchBlogging.orgMassey, R., Stoughton, C., Leauthaud, A., Rhodes, J., Koekemoer, A., Ellis, R., & Shaghoulian, E. (2010). Pixel-based correction for Charge Transfer Inefficiency in the Advanced Camera for SurveysMonthly Notices of the Royal Astronomical Society, 401 (1), 371-384 DOI: 10.1111/j.1365-2966.2009.15638.x

ResearchBlogging.orgRichard Massey (2010). Charge Transfer Inefficiency in the Hubble Space Telescope since Servicing Mission 4 MNRAS arXiv: 1009.4335v1

4 Comments on “Correcting Hubble images”

  1. Nick Cross says:

    This is not just a problem for Hubble, or just a problem for the Weak Lensing community. For the European Space Agency mission Gaia, trapping and subsequent rerelease is a mission critical problem. The point-spread function position can be modified by several tenths of a pixel, but for the main astrometric mission it is necessary to measure the centroid to ~1/1000th of a pixel. The offset depends on history of charge transfer along the same CCD column, but the constraints on the data transfer rate mean that most of the information about this must be thrown away and only a fraction of the most useful data is kept. Using this data, it should be possible to correct the data through self calibration and periodic resetting the trap levels, but the methods that Richard has developed for HST data cannot be used as they require the full charge history. Future space missions that require high levels of positional/shape accuracy will also need to worry about CTI.

  2. Emma says:

    Good point, Nick, thanks.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s