Practice Example: reduction / calibration with SMS


2 images of eps Aur are taken, each with 600s exposure time, with a Sigma 1603ME on Lhires III (grating 2400 g / mm) + C14 telescope. After that was taken at the same grid position, a neon calibration spectrum. A day earlier the masterflat was created.

The analysis was conducted with the SMS package of GŁnter Gebhard, an excellent MIDAS script that makes the data reduction very quickly. The evaluation of a series of recordings is performed in less than 3 minutes. Each step can be reconstructed and retrospectively reviewed, if the user something arouses suspicions.

I concentrate here on the idea of some very useful monitoring tools, which allow the user to check the quality of the data reduction. All of these control instruments are automatically computed and displayed only if the user so wishes.

The use of SMS is described in detail in the documention of GŁnter Gebhard.

The screenshot shows the two superimposed spectra (which is why we see only 1 horizontal strip of total). The neon-frame is superimposed. 7 perpendicular slit images can be seen in the light of the neon emission (neon glow lamp).

The yellow markers indicate the neon emission wavelelengths, which has found the routine on the basis of a implemented catalog of neon lines. With these 7 lines, the star spectrum is then calibrated with a calibration function of third degree.

The horizontal window defines the user via a mouse click.This is the area used for evaluation. The yellow marked neon line sections be bent automatically in the course of the routine now.

It was used in this case no darks. The dark correction happens automatically when the internal step of the sky background correction is done.

The used calibration neon spectrum is shown on the right for illustration in the form of a graph.
A routine in SMS find automatically the lines in calibration Neon spectrum and assigns them to the exact pixel numbers (by Gaussian fitting, subpixel accuracy).

2 of the found calibration lines are identified by the user. This is done by marking the 2 lines with the mouse and entering the approximate wavelength. The exact wavelength then selects SMS from the implemented neon wavelength catalog. The so-identified 2 lines are marked in blue in the graph.

 

According to this "guidance", the program automatically finds the other lines in the catalog. In this case, the calibration spectrum contains 7 lines, so that the calculation of a calibration of the third degree is useful. With the determined calibration function Wavelength = f (pixel number) is calibrated the extracted sum spectrum of the two star images.

The non-linear part of calibration function is shown here as graph. The absolute deviations (in angstroms) of the catalog wavelengths (crosses) from the calculated calibration function (solid line) are given.
Here the relative deviations of the calibrated spectrum of neon are represented (residuals). The standard error is 0.007 = 0.7%.

And if the user wants to know the resolution R of the spectrograph, it is the width (the neon FWMH's) in the calibration spectrum calculated and R displayed. R is in this case in the center approximately 10,800 and at the edges less.

This is caused by the field curvature of the 40 mm diameter achromatic lense noticeable (CCD chip edge length = 13.8 mm for the KAF 1603ME). It is for all grating positions as similar.

The final, normalized spectrum as a graph.

back to content