NRAO Skynet 20-meter telescope
Details for various kinds of observing projects
Making Images
Preamble -- For those not familiar with radio astronomy, note that the 20-meter telescope is like a
camera with one pixel, which we call a "beam". To make an image, one must scan the telescope back and forth
across the object and build up an image, pixel by pixel, a process called mapping.
- Raster scans
- The 20-meter can do two types of mapping: raster (i.e., back and forth), or daisy patterns.
- The raster scans come in 2 flavors: "RALongMap" or "DecLatMap".
- Mapping is in either RA/DEC or Galactic Latitude/Longitude coordinates depending on the coordinate type selected in the Target screen.
- RaLongMap: sweep along the Right Ascension (or Longitude) direction, and step sucessive sweeps in Declination (or Latitude).
- DecLatMap: sweep along the Declination (or Latitude) direction, and step sucessive sweeps in Right Ascension (or Longitude).
- Daisy: scan in a flower-petal pattern with the object of interest at the center.
Beamwidth -- The beamwidth of a telescope is given by Bw = 1.2λ/D (in radians).
in which λ is the observing wavelength, and D is the telescope diameter.
If you use the frequency in GHz (Fghz) instead of wavelength, then the
beamwidth in arcminutes is:
Bw = 62/Fghz (for a 20-meter telescope)
- The 8-10 GHz receiver has a beam size of about 7 arcmin.
- The 1.4 GHz receiver has a beam size of about 44 arcmin.
- The 1.4 GHz receiver is presently on the telescope.
General Advice -- Maps should be at least 5 beamwidths across, otherwise you cannot
distinguish the object from the background.
- For the 1.4 GHz receiver, the beam is about 44 arcminutes, so raster maps should be no
smaller than 4x4 degrees; and daisy maps should have a radius of 2 degrees or more.
Cautions:
- Blocking near the horizon: Use a minimum elevation somewhat higher than the default 5; there is blockage in certain directions near the horizon. The horizon is less blocked to the south; for far south sources such as the galactic center, the minimum elevation of 3 degrees will probably be ok.
- Also, due to the radiation from the atmosphere, mapping near the horizon shows the variations in system temperature due to the atmosphere, which may drown out the brightness from weaker sources. Best to use a minimum elevation of 25 degrees when doing maps.
- The 20-meter can pick up radiation from the Sun in its far sidelobes. To avoid this try to keep all parts of a map at least 15 degrees away from the Sun. In the Skynet interface, set the "Solar Separation" to 15, unless you really want to observe the Sun.
Quick Advice
For the 20cm receiver (installed as of Aug 2013)
Here are some rules of thumb that will get you pretty good maps.
Set the minimum elevation to 25 degrees
Set the solar separation to 15 degrees
Daisy Pattern:
-----------------
1. use a radius of at least 90 arcminutes.
2. number of petals: 4 for a quick look,
8 or 12 petals for an reasonable image of the central part of the field.
3. For Radius:
up to 180 arcmin -- use 45 sec per petal
200-300 -- use 75 sec per petal
300-400 -- use 100 sec per petal
(above that use radius divided by 10)
Multiply these times per petal by the number of petals
to get the total duration.
4. use integration time of 0.2 seconds in all cases.
Map pattern:
--------------
1. "Map Size" : Use a size of 6x6 beam widths or larger.
2. "Sampling Density" :
Set the gap between sweeps to 1/4 for average maps
Set the gap to 1/5 for higher quality.
3. Under "Map Depth"
Select "integration time"
Look at the Slew Speed: this should be 0.6 or less.
Put integration time = 0.3 seconds.
Then check the Slew Speed, if greater than 0.6,
increase the integration time in steps of 0.1
until slew speed is less than 0.6
Usually you can just set the map size and the gap between sweeps. The interface will
fill in reasonable values for the rest.
Happy mapping!
|
At the bottom of the page, note the number of sweeps, the sweep duration, and the total duration of the map. Generally you are discouraged from making maps that take more than 30 minutes or so to complete. If you want to map a very big area, it is more efficient to make several smaller maps.
[Frank D. Ghigo, NRAO-Green Bank, Nov 2013, rev Jan 2014, rev March 2015, rev June 2020]