The Polarimeter

A dual-beam polarimeter suitable for measuring linear polarization usually contains the following optical components:

1. A focal-plane mask.
2. A half-wave plate.
3. An analyser.
4. A detector.

The light collected by the telescope passes through these components in the order listed (see the next figure ). Each component is described more fully below.

The main optical components in a typical dual-beam imaging polarimeter.

The heart of the polarimeter is the analyser, which splits incoming partially plane polarized light up into two beams; one (called the ordinary, or $O$ ray) contains the component of the incoming light which is polarized parallel to the axis of the analyser, and the other (called the extraordinary, or $E$ ray) contains the component of the incoming light which is polarized orthogonally to the axis of the analyser. These two beams are recorded simultaneously on a suitable detector such as a CCD. The advantage of this system over a single-beam instrument (in which only one state of polarization is recorded on a given exposure), is that variations in sky background between exposures affect both states of polarization equally, and so can be eliminated.

In an imaging polarimeter, the two beams form two images on the detector, displaced by some distance determined by the design of the instrument; both images representing the same area of the sky. A masking system is used to prevent any overlap between the two images. In some instruments this takes the form of a series of parallel, equally spaced bars in the focal plane of the telescope (see the next figure). In this case, the instrument is designed so that the displacement between the two images formed by the $O$ and $E$ rays is perpendicular to the bars, and equal in size to the width of a bar. Thus, the two images form two inter-leaving sets of bars. There are several other systems (such as a mask containing only a single aperture), but the principle is the same.

An example of a masking system used in a dual-beam imaging polarimeter.

If the incoming light is only partially polarized, then at least two exposures are required to estimate both the degree and the orientation of the polarization, each exposure recording the intensity in two orthogonal states of polarization as described above. The analyser axis is rotated in steps of 45 degrees between these exposures. In practice, physically rotating the analyser would result in the displacement between the $O$ and the $E$ ray images also rotating. This would cause the images to overlap on the detector and would make the data reduction process much harder (if not impossible). For this reason, the analyser is usually left in a fixed position, and the plane of polarization of the incoming light is rotated instead. This is achieved by placing a half-wave plate in front of the analyser, and rotating it in steps of 22.5 degrees, resulting in a rotation of the plane of polarization of 45 degrees. Using this scheme the positions of the $O$ and $E$ ray images on the detector are unchanged.

The orientation of the plane of polarization of the incoming light is measured relative to a fixed ``reference'' direction. The analyser axis and the 0 degrees position of the half-wave plate are usually parallel to this direction.