Site of Jakob Selbing

[Main]
[Astronomy]
[Electronics]
[Music]
[Gallery]
[Scuba diving]

New Swedish Solar Telescope

On the island of La Palma, Canary Island, lies an international observation site called ORM - Observatorio del Roque de los Muchachos. It is situated 2,300 m above sea level, on the rim of a volcanic crater which makes up almost the entire island.

The New Swedish Solar Telescope is the only solar telescope at ORM. The NSST is run by the ISP - the institute for solar physics in Stockholm, Sweden, and it is a 1-meter vaccum telescope. Currently (2005) it is the highest resolving solar telescope in the world, mainly due to the advanced adaptive optics, which can, under good conditions, compensate for the atmosphere's blurring effect. It was the first solar telescope in the world to reach the "magical" 0.1 arcsecond resolution, at which point new structures - dark cores in the penumbra filaments of Sun spots - begin to show.

I came in contact with Mats Löfdahl at the ISP when he held a lecture for the astronomy club ÖAS in the autumn of 2003. I was offered both a Master thesis assignment and a temporary job at the observatory. In June 2004 I went to La Palma, to work as observatory assistant for 3 months. When I got back to Sweden I started working on my Master's thesis at the ISP in Stockholm. It was finished not sooner than after 9 months! There were several problems involved in the process, which is why it took so long to complete.

Photo gallery

Please visit my photo gallery for some photos from the beautiful island of La Palma. Below follows a description of my Master thesis. If you're not a scientist you might find it excruciatingly boring.

My Master thesis

The title of my Master thesis is "SST polarization model and polarimeter calibration".  The task was to develop methods of calibration for a new instrument, a polarimeter, and to make a model of the telescope's polarizing properties. You can download it in PDF here.

A polarimeter is a device that can measure the polarization state of incoming light. The polarization state is described using a Stokes vector, which has 4 components: I, Q, U and V. The I component is the intensity of the light, Q and U are measures of the degree and orientation of linear polarization, and V is a measure of the degree of circularly polarized light. In this case, we use a polarimeter that can image the Stokes vector across the Sun's surface. It is also possible to use the polarimeter in conjunction with a spectrograph (which will give us a spectropolarimeter), which gives us polarization and wavelength information across a narrow stripe of the Sun's surface. The light we receive is polarized because of the strong magnetic fields around Sun spots. If the magnetic field is along the line of sight, we get more V signal; if the magnetic field is perpendicular to the line of sight the Q and U signals are strong. This effect is very strong in the 630.2 nm absorption line (Fe-I).

The polarimeter I worked with is nothing more than one CCD camera, one linear polarizer and two liquid crystals. The liquid crystals have a polarizing property which changes when a voltage is applied. In order to use this type of polarimeter it needs to be calibrated. This was not as easy as first believed, because we also need to measure properties of some optical elements that are used in the calibration. This turned out to be quite difficult (at least with the required accuracy), due to non-linear effects of the CCD cameras. In the end I came up with a better way to measure the properties of the calibrational elements.

The next step is to make a model of the telescope's polarizing properties. Unfortunately the polarization effects in this type of telescope are quite large, especially for the linear polarization because these components are usually not larger than approximately 10% (compared to I). Further, the effect depends on the pointing of the telescope. Therefore we need to record the pointing throughout the observation, in order to find out the telescope properties for each polarimetric observation. This is done automatically by the telescope control PC.

The telescope is modelled using 10 parameters. The telescope's 1-meter lens accounts for 5 of these, because we used a very general model. The two folding mirrors account for 2 parameters, and the rest of the optics (which has quite weak polarization effects) account for 3 parameters. The parameters were fitted to polarization data aquired by putting a very large linear polarizer in front of the telescope, setting it to certain angles, and measuring the polarization at the telescope's output (where the polarimeter sits). In order to determine the local properties of the 1-meter lens, we used a small lens in front of the polarimeter to re-image the 1-meter lens onto the CCD camera. The polarimeter data was recorded during one day, giving around 5000 images. When the parameters were fitted to the data, it turned out that the model seemed to be very well-behaved; in principle there was only one unique best solution, even with randomly initialized starting points.

There was only one opportunity for me to test the polarimeter and the telescope model. Unfortunately the observed Sun spot (AR0772, June 3, 2005) was very irregular, making it almost impossible to make any assumptions about its magnetic field. However, in some cases we could. And on the whole, the calibration methods and the telescope model seemed to work! I was very relieved when I saw this, since I had put so much effort in this. The actual images are shown below, courtsey of O. Khomenko and M. Collados of Instituto de Astrofisica de Canarias, Spain. Notice how different Q and U appear with compensation. This is due to the strong cross-talk from the V signal. Literally, the Q image has about 24% V, while the U image has -33% V. It is easy to see the effect near the left edge, where the faculae from the V signal appears. The faculae is completely removed by the telescope compensation, indicating an accurate model.

	Telescope-compensated I
Raw Q	Telescope-compensated Q
Raw U	Telescope-compensated U
Raw V	Telescope-compensated V