PHAS3332 (Practical Astronomy 3: Field Trip) 2008


Use only data files which have been produced by reading the CCD in "bin" mode (these files are small: about 12 kilobytes in size). Identify the files which contain the observations which you intend to reduce, along with bias (offset) and flat field frames. You can read the FITS header of each file with command fitshead; you will need to enable this command by entering kappa, once.

Create a `Starlink Data Format' version of each of your observation and calibration files by using the Figaro command rdfits. For example:

rdfits p25983.fits p25983 swap=true float=true

Average the bias frames. For example, if you had four bias frames p25983.sdf, p25984.sdf, p25985.sdf and p25986.sdf, you would average them as follows:

iadd p25983 p25984 biasum
iadd p25985 biasum biasum
iadd p25986 biasum biasum

Divide the summed bias frame by the number of frames used to produce it, so that you are left with the mean bias level. Continuing the previous example:

icdiv biasum 4.0 biasmean

Subtract the mean bias frame from each of your observation and flat field frames, e.g:

isub p25988 biasmean p25988

Combine the flat field files into an average flat field in a manner similar to that used to combine the bias frames. For example, if you have three flat field files p25988.sdf, p25989.sdf, p25990.sdf (from which the bias has already been subtracted), you would do the following:

iadd p25988 p25989 flatsum
iadd p25990 flatsum flatsum
icdiv flatsum 3.0 flatmean

Now divide the flat field file into each file that contains the spectrum of a star. For example:

idiv p26031 flatmean p26031
Do not flat field any of the calibration arcs.

You should calibrate an arc file for wavelength by using the Figaro command arc. An appropriate line list (needed for the "type of arc" parameter) may be found here. Use a third-order polynomial (i.e. enter 3 in response to "order") and a line width ("sigma") of 1.4 pixels.

Copy the wavelength scale from the calibrated arc file by using the Figaro command xcopy.

Finally, read the spectrum into dipso and push it onto the data stack. Rectify the spectrum: divide (adiv) it by a continuum function which you have previously fitted to it using the command cdraw. Don't forget to push the rectified spectrum onto the data stack, and to save the stack before you quit dipso.

The reduction of your stellar spectrum is now complete. You should next proceed to analyse it in whatever manner is appropriate for the spectroscopic project that you have chosen. Please consult the course tutors if you need advice with this.

Stephen J. Boyle
2008 February 6