Telescope ASPH 611 Term Project
A University of Calgary Department of Physics and Astronomy
Graduate Course in Radio Astronomy

Brief Intro to Radio Astronomy
FT Theory
Telescope Information
Telescope Pointing
First Light
Radio Sources of Interest
Acquisition Software
Electronics Characterisation
Temperature Conversion
Noise Investigation
Preliminary Observations
RFI Problems
PRIMARY Observation
ASPH 611 Team

Primary Observation

Jeff Dever, Julie Grant, December 14, 2005

While there were numerous technical problems with the project, there was one dataset taken that was chosen to apply our knowledge and produce a final, processed image. The galactic region on a DEC Countour of 41o ranging from approximately -2 to 9 Hours of RA was observed on December 5th and 6th, 2005. This scan is a sweep through the Galacitc plane at the constellation Auriga.

Figure 1 shows the raw data as it was aquired from the telescope over the channels that contain the HI line. This data has many issues with gain compression, gain drift and RFI, much of which is not visible in this band, but was very strong across many channels. The same data processed as described below is shown in Figure 21
Raw image Processed Data
Figure 1: Raw data as produced by the Spectrometer
Figure 2: Processed and calibrated data
FITS Downloads
Raw Fits File Processed Fits File Processed Fits Subset

Processing Steps

  1. Divide out the bandshape using data from an absorber measurement.
  2. Convert from arbitrairy units to brightness temperature in Kelvins using data from the hot load/cold load measurements.
  3. Shift the frequency scale to correct for a systematic offset.
  4. Convert from frequency to velocity scale in km/s.
  5. Subtract off the continuum by averaging the clean channels on either side of the HI line, and subtracting from every frequency.
  6. Temporal RFI rejection and time averaging. RFI in channel 1893 was observed to pollute the entire spectrum, including harmonics right on top of the HI line. Time smoothing of 30 time steps was done to achieve 30 second integrations spaced about 2 minutes apart.
  7. Spectral RFI rejection and frequency averaging. There were numerious persistant in time low amplitude peaks in frequency. The spectra of all time points were averaged together and a distribution of differences built to detect the worst of these peaks. Frequency smoothing of 4 channels was done to achieve a per channel frequency resolution of 6KHz.
  8. An additional step that was not completed is to flatten the baseline by fitting a curve to the background pixels on either side of the HI line and then subtract it out.

A sanity check was made with the Liden-Dwingeloo Survey in order to solidify our data. The observation we made was located in the Liden-Dwingeloo survey. Initially an attempt was made to convolve the survey to a 4 degree beam which was successful, but the resulting scaling factor was not well understood, and therefore was not used. But the problem of making a direct comparison was still needed, so a simple descrete method was developed. A single spectra from our observation could be compared with a region of the Liden-Dwingeloo map by averaging a box that has the dimension sqrt(1.133) * 4o = 4.26.

A set of coordinates to makes the comparisions was selected. These coordinates was chosen along the observed DEC, and on integral RA hour angles. The following image shows the Liden-Dwingeloo Survey with annotations of where the comparisions were made. The circle annotation shows the 4o beam of our telescope. The Square shows the area that is averaged together to emulate the gaussian of the beam. The numbers represent the time on the ASPH611 scan that corresponds to the comparison position. Constellation Auriga is shown in blue2.

Liden-Dwingeloo comparision positions ASPH611 comparision positions
Figure 3: Sky pointing positions used for comparisions.

The following table shows the comparsions between the ASPH611 data and the Liden-Dwingeloo survey. It is clear that the baseline for the ASPH611 data needs to be flattened to be directly comparable. However, the agreement between the two datasets is surprisingly good. The band shape is quite consistent showing similar features in both cases. A velocity alignment was made arbitrairily to match the start of our observation with Liden-Dwingeloo, but does diverge in the later part of the observation. This is due to the frame of reference not being taken into account in our data. The effect is largely due to the rotation of the Earth changing the direction the telescope is pointed in relation revolution around the Sun.

Liden Dwingeloo SurveyASPH611 Scan

After reviewing the data from this ovservation, it is clear that as we pass through the Galactic plane, the HI line gets broader and is blueshifted to a maximum of approximately -50 km/s. This is due to the higher temperatures and the thermal motions in the cloud as well as the speed at which it is moving towards us. The finer detail in the line can be attributed to the internal motions of the cloud, and other clouds on the same line of sight, which is why we see structure in the spectrum even though the HI emission is emitted at a precise frequency.

This observation is very encouraging for several reasons. First, as far as we know, this is the first HI line radio detection done from the University of Calgary grounds. Second, not only did the spectrometer detect the HI line, it has sufficient spectral resolution to to show line peaks at varying frequencies. Third, there are other regions of the Galactic plane that are stronger and more dispersed than the one thin scan analyzed in this section. We hope that in the future we can make an entire HI map of the visible sky from the telescopes location.

  1. Images produced in kvis written by Richard Gooch.
  2. Annotations were made available by Steven Gibson.

Last modified: 10:22 am July 17, 2014

Valid HTML 4.01 Transitional