3
$\begingroup$

@Jack's answer mentions 29 separate spectral bands for the Worldview-3 satellite, and that these are implemented using filters.

Is there one giant 29+ position filter wheel incorporated into a single telescope, or are there multiple imaging systems each with a subset of the filters?

I know that sometimes spacecraft will need to have more than one optical system in order to cover all the bands. For example in the case of SOHO (also) that's because the short UV and EUV are involved and special multilayer coatings are needed which are reflective only for narrow ranges. In this case, are all 29 filter positions on one giant filter wheel?

Improve this question
$\endgroup$
2
  • 2
    $\begingroup$ 29 filters is quite possible. HIRS, which dates from the 1970s, has 20 filters on its filter wheel. $\endgroup$
    – gerrit
    Commented Jul 20, 2018 at 10:50
  • $\begingroup$ @gerrit Now I want to ask the question "What's the largest number of positions on a filter wheel in space?" now, but probably it's a bit too far on the edge. That's great to know! If you happen to run across a photo or link about such a large wheel, that would certainly be interesting to see! $\endgroup$
    – uhoh
    Commented Jul 20, 2018 at 11:27

2 Answers 2

Reset to default
7
$\begingroup$

The WorldView-3 instrument is a pushbroom imager, which constructs an image one row at a time as the focused image of the Earth through the telescope moves across the linear detector arrays, which are located on the focal plane.

(From "Radiometric Use of WorldView-3 Imagery", DigiGlobe technical note 2016-02-22)

Monochromatic pushbroom imagers just sweep a 1D image sensor along the track, making up the 2D image as the vehicle moves.

enter image description here

Simple multispectral cameras do the same, using filters over each subsequent sensor bar. That's cheap, but reduces the exposure time for each band. It appears that's what WorldView-3 does for the brighter bands.

enter image description here

For faster (or dimmer light) multispectral imaging, the light from the telescope is separated by the optics to form the separate bands. One way to do this is to use a diffraction grating to do the spectral separation, and then have side-by-side high-efficiency detectors.

enter image description here

(images from G. Petrie "Airborne Pushbroom Line Cameras")

The side-to-side extent sets the bandpass, because the grating has correlated position and frequency. You end up with something like this:

enter image description here

where each peak is a detector bar. The use of parallel detectors gives you complete image coverage (you don't miss parts of the Earth as you swap detectors or filters), some redundancy, good optical efficiency (i.e. you use every possible photon), and sharp edges in wavelength and position.

Improve this answer
$\endgroup$
4
  • $\begingroup$ Very nice! Other examples of "pushbrooming" here and here. $\endgroup$
    – uhoh
    Commented Jul 19, 2018 at 23:31
  • 1
    $\begingroup$ Careful with the use of the word "pushbroom". In its purest sense a pushbroom satellite uses the satellite's motion around the earth to move the line scanner. Thus your image "rectangle" is always aligned with the ground track. WorldViews, Ikonos, and OrbViews do not rely on this technique. They actively rotate the spacecraft, allowing them to scan in any direction (E-W, W-E, N-S, S-N, diagonal). Otherwise, and excellent answer. A combination of different sensors and diffraction gratings is indeed how it is done. $\endgroup$
    – Carlos N
    Commented Jul 23, 2018 at 3:26
  • 1
    $\begingroup$ @CarlosN note that the pushbroom quote is from Worldview $\endgroup$
    – Bob Jacobsen
    Commented Jul 23, 2018 at 3:28
  • 2
    $\begingroup$ @BobJacobsen - Yup. Even those in the know misuse the word. I'll admit it is a loose definition. And then there is "whiskbroom" to add to the confusion. Hence I prefer "scanning sensor". $\endgroup$
    – Carlos N
    Commented Jul 23, 2018 at 3:30
2
$\begingroup$

Rather than a filter wheel the typical approach to multi-spectral reception is to split the beam in a prism and place detectors at several places, i.e. each split angle corresponds to a different wavelength.

Some satellites do use two telescopes to cover the wavelength range, I've seen one design with six parallel imagers in different bands. I don't know the ins and outs of what engineering choices lead that way though it could be that it means you can use a different kind of glass for the thermal infra-red end of the spectrum.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.