home observation science solar system stars our galaxy cosmology astrobiology exoplanets astrophotography
Optically Identifying 20cm Radio Sources by Ricky Leon Murphy:

Back to Astrophotography

Click for poster

Radio Astronomy
The Very Large Array
The NRAO VLA Sky Survey
The VLA Faint Images of the Radio Sky at Twenty-Centimeters (FIRST)
Project Methods
Acquiring NVSS Radio Images
Acquiring FIRST Radio Images
NASA SkyView - NVSS Alternative
Telescope Use - CCD Image Acquisition
CCD Image Processing
Object Identification
Summary of Methods
Website References

“Optically Identifying 20 cm Radio Sources” 

1. Introduction: 

Radio astronomy gives us a unique view of the Universe. Many of the emission processes powering phenomenon like quasars, radio galaxies, Seyfert galaxies and other active galactic nuclei as well as supernova remnants are invisible using standard optical imaging. However, the object of emission is visible on film (or CCD). For this project, I have a list of eight radio objects that I will identify optically. Using coordinates provided by NED (NASA/IPAC Extragalactic Database), I will train a remotely controlled, 0.5 meter telescope to each object and capture on CCD the optical counterpart. The goal is to overlay the radio image over the CCD image in an attempt to identify the specific object of emission. In addition, I will summarize the object with a brief description as well as identifying the specific emission source responsible for the radio emission. The purpose of the project will be to demonstrate my effectiveness in identifying the optical counterpart of the radio source as well as justifying the need to capture these types of objects in both the optical and radio wavelengths. Note to the reader: computer and Internet commands and software menu items are capitalized for clarity.

Back to Top

2. Radio Astronomy: 

In 1931, Karl Jansky was tasked by the Bell Telephone Labs to determine the cause of static over long distance telephone lines. He constructed a large device in an attempt to detect any stray radio waves. 

(Figure 1

By 1932, the device was working and detecting anomalous radio emission, but Jansky determined the emission was coming from a specific region of the sky – traced along the Milky Way galaxy [W6]. Grote Reber, an amateur Ham radio operator, learned of Jansky’s discovery and wanted to follow up. After being denied a position at Bell Labs to work with Jansky, he decided to work alone, constructing the first ever radio dish in his back yard. In 1937, he began scanning the regions of the Milky Way at a variety of frequencies, finally settling on 160 MHz – successfully detecting and confirming the emissions initially detected by Jansky [W5]. 

(Figure 2

The image above shows the map of the Milky Way Galaxy as detected by Reber’s backyard dish [W5]. 

While Reber attempted to define the cause of the emission, it wasn’t until 1950 that V.L. Ginzburg, a Russian Physicist, theorized the radio emission was the result of synchrotron radiation – a non-thermal emission process which is the result of electrons traveling near the speed of light along magnetic field lines [W5]. 

Our atmosphere serves as a protector against high energy particles like gamma rays and UV radiation, but the lower energy of radio and optical waves (with just a dash of microwave) passes through with little or no resistance – in other words, transparent. 

(Figure 3)

While we must send probes above the atmosphere to examine phenomenon in the high energy wavelengths like X-ray and Gamma-Ray, we can safely and freely observe the Universe in radio and optical wavelengths from the ground using optical and radio telescopes.[1] 

(Figure 4

The radio wavelengths – as seen in the electromagnetic spectrum diagram above (cute, isn’t it) – are very long and requires a large “telescope” to detect them. Reber’s dish was 9.5 meters in diameter, but the larger the better – just like an optical telescope. A radio telescope is in the shape of a dish for good reason – radio emissions occur on Earth in the form of radio interference from telephones, radio stations, TV stations, cell phones, blenders, and so on and the shape of the dish helps to block any local emission. Radio observatories are usually in remote areas far from any possible interference. In addition, the dish shape prevents any interference from the ground and focuses the radio emission from the sky to the antenna. 

While describing the technology, design and functionality of a radio telescope is beyond the scope of this project, the image below demonstrates the basic parts and functions of a radio telescope. In simple terms, just as light is focused by a lens or mirror to an eyepiece, the radio dish focuses radio waves to an antenna which contains a receiver – sort of an electronic version of an eyepiece. 

(Figure 5

Because of structural limitations, the largest steerable radio telescope – that is a radio telescope that can move along both axes – is 100 meters. The famous 300 meter Arecibo dish in Puerto Rico (the one used by the SETI@Home folks) is also a single dish design, but is built into the ground. In order to improve resolution, a larger radio dish is required. To get around the inherent design flaws of a single large dish – an interferometer is created by using a series of two or more dishes. Interferometry uses the phase difference between signal arrival time between two or more radio dishes (Kutner, 2003). The phase difference combined with the distance between the dishes allows for a “virtual” radio dish with a diameter equal to the distance between the dishes. In other words, if the separation between two dishes is 500 meters, the effective diameter of the “virtual” dish is 500 meters. However, there are limitations which will require additional dishes inside the overall diameter to compensate.  One such limitation is angular resolution – the ability to discern small details. While the reasons and the maths involved are well beyond the scope of the project, it should be known that the Very Large Array in Socorro, New Mexico solves these limitations.  

Back to Top

3. The Very Large Array: 

The Very Large Array (VLA) is a member of the National Radio Astronomy Observatory. Its array of 27 dishes is located 50 miles west of Socorro, New Mexico – on the plains of San Agustin [W7]. 

(Figure 6

The VLA is an interferometer, meaning the dishes all work together to create a large dish capable of high resolution imaging. Each dish is 25 meters in diameter and weighs 230 tons. The Y-shaped array is capable of four different configurations [W7]: 

  • The A array – maximum dish separation of 36 kilometers
  • The B array – maximum dish separation of 10 kilometers
  • The C array – maximum dish separation of 3.6 kilometers
  • The D array – maximum dish separation of 1 kilometer

Changes in array configurations occur about every 4 months. In addition to a variety in dish configurations, a variety of receivers are also available. The receiver of a radio telescope determines the frequency, beam and resolution [W7].  

Table 1:


4 Band

P Band

L Band

C band

X band

U Band

K  band

Q band

Frequency (GHz)









Wavelength (cm)









Primary beam (arcmin)









Resolution (arcsec)









Back to Top

4. The NRAO VLA Sky Survey: 

The NRAO VLA Sky Survey, or the NVSS, used the VLA array from 1993 to September 1996. The survey covers the entire northern hemisphere from declination -40 degrees. While not the only radio sky survey, this survey made use of the lower noise L-Band receiver that was just installed prior to data acquisition (Condon et al., 1998). 

The purpose of the survey is two-fold: 

  • The data collected by the NVSS was made to service the Astronomical Community
  • To catalog the two major radio emissions – the strong and the weak

The strong radio emissions come from extra-galactic sources like active galactic nuclei (AGN) and radio loud galaxies. Weak sources of radio come from supernova remnants (Condon et al., 1998). 

All the data from the NVSS is available online, and for those who use a UNIX based computer, software is also available to search and download data.

Back to Top

5. The VLA Faint Images of the Radio Sky at Twenty-Centimeters (FIRST): 

The VLA FIRST survey is a project designed to map 10,000 square degrees over the north and south galactic poles. Using the B configuration of the VLA interferometer, image resolution is a sharp 5 arcseconds and is designed to pinpoint faint radio sources at 20cm. Even with 3 minute images, 50% of the FIRST survey catalog will have optical counterparts brighter than a magnitude of 23. It will be possible to detect the weaker portion of a radio signal even in the midst of “brighter” radio lobes (Becker et al, 1997). 

The FIRST survey will compliment other surveys – such as the NVSS – providing image resolution 10 times that of the other surveys currently available. The FIRST has been instrumental in the detection and continued study of quasars. The vast majority of published papers based on the FIRST survey ( indicate its valuable resource for quasar research. 

Back to Top

6. Project Methods: 

By determining the coordinates of a given list of eight radio objects, I will train a remotely controlled telescope using the determined coordinates to locate the optical counterpart of each radio object. The prerequisites of this project will require: 

  • A computer with Internet connection
  • A telescope with ample aperture and a CCD camera
  • Software to control the telescope and process the images

By taking this class, I have already demonstrated that I have a computer with Internet access. Such a tool will allow me to visit websites required to complete this project including the websites for the NVSS and FIRST survey.

The list of radio objects provided by my project supervisor: 


Other Name






Hydra A

09h18m05.7s -12d05m44s



12h29m06.7s +02d03m09s


NGC 4261                      

12h19m23.2s +05d49m31s



12h25m03.7s +12d53m13s


Coma A

12h54m11.7s +27d37m33s


Cygnus A

19h59m28.3s +40d44m02s


Hercules A

16h51m08.1s +04d59m33s


Cassiopeia A

23h23m25.4s +58d48m38s

Table 2:


List of objects, common names and coordinates in Right Ascension and Declination.

Using NED ( I was able to determine the coordinates of each target. In addition, I was provided with a common name. This is sometimes helpful when searching through the NVSS database (or any other) in the event the object number is not recognized. The left-hand set of coordinates is the Right Ascension while the right-hand set of numbers corresponds to the Declination. A visit to the NED website presents me with several options – one of which is “Objects – By Name” seen in Figure 7. 

(Figure 7)

With this page, I simply enter the object number, in this case 3C218, then click the SUBMIT QUERY button. 

I am now presented with a summary page with a wealth of data, but I am only interested in the first few lines of data – namely the coordinates of the object (figure 8). 

(Figure 8

The coordinates listed were used to complete Table 2

Back to Top

7. Acquiring NVSS Radio Images: 

A visit to the NVSS website gives me plenty of data and links, but for this project I am interested in the ‘Postage Stamp Server’ ( This is an online version of the image view tool that offers a JPEG or FITS version of the radio image. 

The search window is a bit complicated (figure 9) in that the Desired Image Size and Pixel Spacing cannot exceed an image size of 512 x 512. This is somewhat limiting as I have determined that my desired image size is to be 680 x 680 – discussed below.  

(Figure 9

The parameters I enter here are the object coordinates – in this case, the coordinates for 3C218 – a desired image size of 0.34 degrees and a pixel spacing value of 1.8 (having to enter “1.8 1.8” for a square image). These values will give me a desired image size of 680 x 680, but since this exceeds the allowed limit, I am presented with the dialog in figure 10. 

(Figure 10

A pixel spacing value of “2.5 2.5” would have given me an image, in which case the image would simply appear in the browser window as a JPEG. Of important note, figure 10 gives me an option to download the RAW IMAGE file, but this is a compressed FITS file that is 32 bit. Only software capable of opening 32 bit images (like Mira Pro) can view this image – and the image is not centered on the desired coordinates. 

There is an alternative to retrieve NVSS images in the desired image size using the NASA SkyView database, but I will cover that in a moment. 

Back to Top

8. Acquiring FIRST radio images: 

The FIRST website is much easier to navigate. A visit to the main site presents me with several options, one of which is IMAGES (pretty self-explanatory). The IMAGES page has some useful information about the project, and includes a link to the FIRST Cutout Server – the tool used to extract the radio images from this survey: 

(Figure 11

This page (figure 11) is rather simple to use – enter the coordinates and image size. The CCD camera I will use has a 19.6 x 29.8 arcminute field of view, so the parameter I chose for image size is 19.6 arcminutes:  

(Figure 12

The image that is extracted will be below the text “Extracted Image.” I did not include the entire screen shot (figure 12) as the image is large. It turns out the image itself (minus the legend data) is 680 x 680 which is how I determined the desired size for the NVSS search. 

Back to Top

9. NASA SkyView – an NVSS Alternative: 

As mentioned earlier, the NVSS Postage Stamp Server is limited in its image extraction size. I recently learned that the NASA SkyView service offers database access to almost all database images (optical, radio, gamma-ray, x-ray, and other images). The NASA SkyView homepage has a choice of interface; however, for our purpose the Advanced Interface is more than enough (figure 13). 

(Figure 13

The search is similar to the NVSS and FIRST database – entering the coordinates will retrieve the image. The SkyView database has the added option of using a target name instead of a set of coordinates – it confirms the target name through SIMBAD (an astronomical database) and extracts coordinates from that database. However, by using coordinates instead of a name I am guaranteed the center of the retrieved image will be the specified coordinates.  

In addition to the coordinates, I must select a survey. In this example (figure 13), I selected both the NVSS and FIRST surveys.  

Because I am concerned over image size parameters, this page has many options – available by scrolling down: 

(Figure 14

These optional parameters are used to constrain the image size – in our case, a 0.34 degree image with an image size of 680 x 680. Click on the SUBMIT REQUEST NOW button to retrieve the desired image. The results will display both images in a new window (if there is only an NVSS image, only the NVSS image will be displayed). 

Because my FIRST radio images are already 680 x 680 based, I acquired the NVSS radio images through SkyView instead of through the NVSS - 1.8 pixel space at 0.34 degree field of view - to ensure equal image size based on my CCD chip specifications. 

Back to Top

10. Telescope Use – CCD Image Acquisition: 

The capture of images is the meat and potatoes of the project – and the most fun. While I could most certainly acquire all my images through NASA’s SkyView, that would be cheating and no fun at all! 

The telescope I will use is an RC Optical Systems 20 inch Ritchey-Chrétien design (the specifications and field of view come from the observatory website and CCD calculator software, CCDCalc).  

  • 4300mm focal length
  • 8.3 f/ratio

This design is ideal for astrophotography as there is no coma (due to mirror shape) and no focus variations (due to open tube and tight primary mirror integration). The telescope is housed at the Blackbird Observatory in New Mexico, a private observatory owned and operated by Ron Wodaski (although there was a transfer of ownership during the course of the project). Through the use of my American Express Card, he has been gracious enough to allow me time on his telescope to complete this project. 

The telescope sits on a Software Bisque Paramount ME robotic mount which is capable of <5 arcseconds of periodic error. Tracking is performed via separate telescope and guide camera. The CCD camera is a Santa Barbara Instruments Group STL-11000M

  • 11 million pixels – 4008 x 2672
  • 36 x 24.7mm chip
  • 9 micron pixel size
  • Anti-bloom gate
  • 13 e- read noise
  • 50,000 e- full well capacity

The telescope system is controlled by Software Bisque’s Internet Astronomy powered by TheSky and CCDSoft.

The CCD camera and telescope provide an image of: 

  • Field of view 19.6 x 29.8 arcminutes, 0.34 x 0.50 degrees
  • Image scale = 0.44 arcseconds/pix

Images are acquired through the Browser Astronomy package using CCDSoft for camera control. 

(Figure 15)

The LOG IN page has important information. For one, it will list the current weather and observatory status. It will also list the schedule of users for the day. If no one is logged in, you will see a page like figure 16 – but if someone is logged in, a page like figure 17 will display.

(Figure 16)

(Figure 17)

After logging in…

(Figure 18)

After logging into to the Blackbird Observatory website, I am presented with the screen shown above. The interface is painfully simple as everything in regards to alignment is already taken care of. There are only three items on this page to be concerned about:

  1. The Slew to Object box
  2. The Take Image block
  3. The Autoguide link at the bottom next to Log Off

The first step is to enter the coordinates into the Slew to Object box, followed by clicking the SLEW TO button. 

(Figure 19

The program offers a real-time look into the slewing progress. You have an option to ABORT SLEW if you find that you entered the wrong coordinates. I did not have this problem. 

(Figure 20

After the slew is complete, the first page is refreshed, this time with the object of interest centered. You will know your target has been centered by the Take Image box – the coordinates will appear in the Prefix field (Figure 20) – if your object were M84 for example, the text "M84" would be in this box. 

Once the object is centered, autoguiding must be turned on and a guide star selected. The observatories default option is the autoguider is always running so a guide star must be selected. Once the AUTOGUIDING link is clicked (bottom of figure 20), a new window opens: 

(Figure 21

These are the default parameters for the autoguider settings. These settings are fine, but can be adjusted if a guide star is not located. Click the TAKE AUTOGUIDER IMAGE button to initiate autoguider capture. 

(Figure 22a)

Figure 22a shows a progress of the autoguider capture. When complete, a window opens with the result of the image capture. The smaller image window should display some white stars (figure 22b). Choose a star and click on it.  If there are no stars, you can CLICK HERE TO START OVER and change the parameters as seen on figure 21.

(Figure 22b)

Once the star is clicked, the autoguider analyzes the tracking errors and adjusts the mount to compensate. When the error reaches an acceptable level, you are given one option, CLICK TO CLOSE (figure 22c). This page is designed to inform you the mount is tracking properly. 

(Figure 22c)

The autoguider starts and the user is once again presented with the main page as seen in Figure 18

The final setting will be the Change LRGB Settings in the Take Image block: 

(Figure 23

These settings will define how long the image is captured in each filter. If you want an LRGB image, all the Active radio boxes should be selected. In this case, only the Luminance filter is active, and the timing of the image is 30 minutes.  

In addition to the timing, I decided to select 2x2 (figure 23 shows a 1x1 bin mode) binning on my images. Since the majority of the targets are faint, I prefer the increased sensitivity to a 2x2 bin mode (basically doubling the pixel size - analogous to film speed).

When satisfied with the settings, click the SAVE LRGB SETTINGS button. Again, the user is brought to the main page as seen in Figure 18

The final step to this process is to click the TAKE COLOR button. There is a status page when image capture is in progress: 

(Figure 24

This page shows the guide star as well as graphs that demonstrate the status of “seeing” conditions (DIMM) and guide star patterns. This information is really not that helpful so I have disabled them. 

Once image capture is complete, you are given some options: 

(Figure 25

You can view the image folder, go back to the main page or log off. 

There are steps to take if I want to take Luminance only images, or change the binning as well as capture reduction frames. I can also change some of the default actions of the guider – like planet guiding, comet or asteroid guiding, even turning the guiding feature off – but these steps will not be covered.  

Back to Top

11. CCD Image Processing: 

Once the images are captured, they are stored on the observatory’s local server – along with any image reduction frames that are provided (bias frames, flat frames and dark frames). Images are downloaded and run through MaxIm DL for calibration (more on calibration here): 

  • Open the raw image in MaxIm and save a copy of it

  • Use the calibration wizard to set the location of the Bias, Dark and Flat frames

  • Calibrate the image 

The Bias frame sets the black point of the image while the Dark frame compensates for electronic noise. The Dark frame should be equal in exposure as the light frame (light frame means the actual image). The Flat field is an image of the telescope optics, filters and camera protective plate and any CCD chip blemishes – this means that any blemishes on the chip, scratches and dust are imaged. This is divided from the light frame and provides a much cleaner and better looking image. It can also equalize the gradient in the background. Visit the Image Reduction page for more information.

Additional processing of pixel removal – any dead or hot pixels – is also performed through MaxIm DL. The SBIG STL-11000M chip does not include an overscan region so image reduction is a straight forward process (overscan is a feature used on professional CCD cameras that provide automatic Bias, among other things). 

Once calibration is complete, the image can be astrometrically calibrated using Mira Pro (the astrometry program in MaxIm DL is not perfect, and has crashed when attempting astrometry on the images I captured).

Back to Top

12. Astrometry:  

Astrometry determines where an object is in an image. When we determine the equatorial coordinates of a few known stars, we can calibrate the image so that when our cursor is located anywhere on the image, we will know the equatorial coordinates of the cursor location. This is helpful when looking for something that is not that obvious – like a galaxy that looks like a star, a supernova, or an asteroid. 

Using TheSky Professional (which uses both the GSC and USNO B catalogs), I can identify bright stars in the field of interest, but I need a field of view indicator to know where to look. In my version of TheSky, this is located under the VIEW menu. Select FIELD OF VIEW INDICATORS: 

(Figure 26

I can create an FOVI (Field Of View Indicator) by clicking the Create FOVIs tab (all I need is the focal length and aperture of the telescope). I am using the RCOS Carbon 20” telescope, so I name the FOVI accordingly. Make sure you select each telescope and make sure the radio box ON is not checked (the software manual can provide additional details) - only the RCOS Carbon 20” FOVI should be on: 

(Figure 27

Next, use the find tool and enter the equatorial coordinates (or common name if there is one). One of the obvious radio targets on our list of radio sources is M84 – an elliptical galaxy that is also a radio galaxy. In the Find box, I enter “M84”: 

(Figure 28

Click the CENTER & FRAME button to show the target. The next step is to orient the field of view to match the CCD image of the object area. Use the ROTATE TOOL for this: 

(Figure 29

Click and drag the N (N means North) upward: 

(Figure 30

The field of view is now oriented with the CCD image.  

Next, choose some bright stars in the field by clicking on it, like this one: 

(Figure 31

Note the OBJECT INFORMATION box has lots of information, including the coordinates. This is important: make sure you use the Equatorial 2000 coordinate set and not the “current” coordinate set as listed in TheSky’s OBJECT INFORMATION box. The NVSS and FIRST sky surveys (as well as a great many others) use the J2000 (Equatorial 2000) coordinate standard. I learned my lesson when I entered the “current” coordinate set and could not locate any targets – including M84! 

It is best to have the calibrated image open in Mira Pro and TheSky program running at the same time. I am lucky to have a widescreen monitor so I can have both programs side-by-side; otherwise you can ALT-TAB between programs. The process of astrometric calibration is pretty straightforward; with the image open in Mira Pro, notice the X and Y coordinates (figure 32): 

(Figure 32

The cursor will translate to pixel values in the X and Y axis. The Z axis remains the same as this is a 2D image. 

To begin calibration, activate the ASTROMETRIC CALIBRATION command (figure 33): 

(Figure 33

Clicking on the ASTROMETRIC CALIBRATION menu item will activate the function. 

Switch back to TheSky and select a star in the field that matches with the CCD image and note the equatorial 2000 coordinates. Once you have this, switch back to Mira Pro and click on the corresponding star, making sure the cursor is at the center of the star (note the small preview window). A box opens (figure 34): 

(Figure 34

This is where you enter the coordinates: Right Ascension first then Declination. In this example, the RA and Dec coordinates are: 12h54m38s and +27o30’49”. You would enter this as: 12 54 38, +27 30 49 (note the spaces). Once entered, click the MARK button and the box disappears. A mini spreadsheet opens confirming the point is set – click the image again to move this spreadsheet window behind the image.  

Repeat this for a total of at least 6 stars. Once all the points are marked, the calibration must be set (figure 35): 

(Figure 35

Click the CALCULATOR icon to the left of the image to set the calibration. The final step is to save the image – this saves the calibration to the FITS header in WCS format. 

(Figure 36

Once the calibration is set, notice in the image above that the numbers in the X and Y fields are replaced by the RA and Dec coordinates (figure 36). How cool is that! 

In order to be accurate, Mira Pro provides a summary once the calibration has been set. The summary looks like this: 

Astrometric Calibration for
E:\...\HET615\_3CR272.1 - M84\M84.FIT 

Field Center:
  RA:  12h24m50.944s at column 2004.000
  Dec: 12°50'51.054" at row 1336.000 

Plate Solution Errors:
  RA  rms: 0.4604 arcsec using 6 points
  Dec rms: 0.4136 arcsec using 6 points 

Image Scale:
  X scale: 0.4556 arcsec/pixel
  Y scale: 0.4560 arcsec/pixel 

Rotation of the sky on the image:
  RA Axis:  -9.3625 degrees.
  Dec Axis: 170.0964 degrees. 

WCS CD Matrix:
  x0=0.000830104005 C_11=0.000120997048 C_12=-0.000021125216
  y0=-0.000342961558 C_21=0.000019968335 C_22=0.000121110784

I am interested in the ‘Plate Solution Errors’ section. I am unclear as to how this works, but the Mira Pro manual suggests a value of 0.5 or less (the lower, the better). I did discover that if the selection of a star is off-center a bit, the error is greater – so make sure that the selected stars are centered in the cursor cross-hairs (this is what the 5x preview window is for – especially helpful for small stars). 

In testing this method, I used the coordinates of one of the radio sources – 3CR272.1, otherwise known as M84. This object is fairly obvious in the image so I moved the cursor around the image to locate 12h25m3.7s RA and +12d53m33s Dec: 

(Figure 37

That bright object in the 5x preview window is the (near) center of M84. Astrometric calibration is a success.  

Each of the 8 images has undergone astrometric calibration using the methods outlined above. Once the optical counterparts are identified (some where rather dim), I used Mira Pro to mark the target using the LABELS () button. This image will be screen saved[2] as a normal image to demonstrate the location of the radio source. Saving the image will not save the annotation.

Back to Top

13. Object Identification: 

With the coordinates of the radio object identified, the CCD image should be resized to match the radio image size. The overlay is a simple process, however there are a few pieces of data I need before resizing my image. Image resolution is not enough – I need to know the pixel scale. Knowing the pixel scale (0.44) for my camera alone will provide nothing outside of guessing – and that will take too long and will most likely be incorrect. The NVSS website does not list the pixel scale of their images; however, the FIRST website does list the pixel scale. At a value of 1.8 for the FIRST radio images, I can deduce two things: 

  • Since I set the SkyView image parameters based on the FIRST and NVSS images, they will be extracted with the same image scale
  • My images will need to be reduced to 25% (50% in size when using 2x2 binning) –using the equation: 0.44/1.8 = 0.2444444444

The NVSS and FIRST surveys oriented their data equatorially so that images are “extracted” from the virtual celestial sphere. This is significant as my images were captured on an equatorial mount. What does this mean? My images are oriented correctly with the radio data – at least in theory. I will proceed with this assumption. 

Knowing the orientation and scale of the images, overlay is simply a matter of creating a composite image. Here are the steps: 

  • Create a new RGB image (I use PhotoShop)
  • Select CCD Image and copy (CTL+A then CTL+C)
  • Select new image and paste (CTL+V)
  • With new image selected, create new layer, select layer
  • Switch to radio image and copy (CTL+A then CTL+C)
  • Select new image and paste (CTL+V)
  • In the layers palate, change the opacity of the selected layer (should be the radio image), change opacity to the CCD Image is visible underneath
  • Select the Move Tool and drag the radio image so the center of the image is over the discovered coordinate location on the CCD image - guides and rulers help here

Once the image overlay is complete, the image is now flattened (in the layers palate), cropped and saved. 

When I examined the result of the overlay, I discovered that the image quality was rather poor, and discernment of details is rarely possible. One solution to this issue is to do a simple side-by-side comparison with arrows marking the radio and optical sources. A novel idea suggested by Professor Pamela Gay is to create an RGB image using the three separate images: NVSS image as the Red channel, the CCD image as the Green channel and the FIRST image as the Blue channel – or a variety of combinations. 

The method I have chosen is a side-by-side comparison. Arguably this is the most simple method, but has the benefit of full view of each image; however, since the project defines an overlay, I will also include an image with an overlay using the composite method outlined above. This method provides better mixing of opacity through the various layers within PhotoShop. 

Back to Top

14. Summary of Methods: 

The details of methods used in this project are vital, but rather long. The steps above can be summarized in chronological order, so here is the project and methods in summary: 

  • Receive target list – names of known radio sources
  • Use NED to acquire the coordinates of the radio source names
  • Use the FIRST, NVSS and SkyView database to enter coordinates and extract the radio image
  • Determine the plate scale of the image – assumed to be 1.8 according to data from FIRST website and NVSS image scale
  • Acquire time and log into the Blackbird Observatory – the telescope I used to acquire the images (you may have your own telescope)
  • Enter the coordinates into the telescope SLEW TO command and acquire images
  • Download image and reduction frames, reduce the images
  • Use the GSC and USNO B database base within TheSky to acquire coordinates of bright stars in each of the images
  • Use Mira Pro to astrometrically calibrate the images using data from TheSky, save image
  • Use Mira Pro to locate the desired coordinates – locate optical counterpart of the radio source
  • When located, use Mira Pro to label the image – then screen save the image
  • Using PhotoShop, resize the image to match the scale of radio image – reduced to 25% for 1x1 bin images and reduced to 50% for 2x2 bin images
  • Create an overlay of the CCD image and radio image using PhotoShop, or
  • Create a side-by-side comparison of the optical and radio objects, or
  • Create an RGB image from the CCD, NVSS and FIRST images

Luckily I had the right tools for the job – that is I have in my arsenal an American Express card for purchase of telescope time, TheSky Professional, MaxIm DL, Mira Pro and PhotoShop CS2. There are probably methods available for those with limited resources, but this makes me nervous thinking about it. 

In the next section: objects. Each individual target will be discussed including basic information, facts and imaging notes will be covered. Following the next section will be a discussion on the mechanisms of the AGN (Active Galactic Nuclei) and the FR (Fanaroff-Riley) classification.

Back to Top

15. Objects: 

16. 3C218: 

(Figure 38)

(Figure 39) 

The object 3C218 carries the name Hydra A. The designation ‘A’ means it’s the first (or most prominent) radio source in the constellation Hydra. Hydra A is a very bright Fanaroff-Riley (FR) type I radio galaxy (Lane et al., 2004). While the NVSS data obscured the fine detail of this object, more recent observations with higher resolution has determined an “S” symmetry of the inner jets, possibly resulting from precession of the central engine (Lane et al., 2004). The central engine of this radio galaxy is believed to be a supermassive black hole. 

Optically, this galaxy is a faint cD type galaxy (elliptical galaxy) with an apparent magnitude of 14.8 (W8). In addition, this cD galaxy is the dominant member of A780, a poor Abell cluster of galaxies which lies at a distance corresponding to z=0.0549 (Lane et al., 2004). 

Imaging notes: this object was a bit of a challenge for imaging as the object sets early in the evening and westerly wind threatened the observatory resulting in a 30 to 50 degree closure of the west panel. There were only two days I was able to target this area, one of those days was poor weather resulting in loss of the guide star and trailing of the stars in the image. The second image attempt was successful; however, light contamination resulted in a gradient on the left side of the image (car, neighbors light, or something similar possibly). 

(Figure 40) 

Figure 40 is a composite of the CCD Image and NVSS Image. There is no FIRST radio image of this object.

Back to Top 

17. 3C273: 

(Figure 41)

(Figure 42) 

Object 3C273 is a well-known quasar discovered early during initial radio surveys. Its distance corresponds to a z=0.158 (Stawarz, 2004) and has a visual magnitude of 12.8 (W8). Radio occultation by the Parkes 64 meter dish identified this object as the first detected quasi-stellar object, or QSO (Burke and Graham-Smith, 2002). What is interesting about this object is its one-sided jet, demonstrated by super luminal motion with the counter-jet discovery being a controversial issue. Regardless, the core-dominated non-thermal emission of the one jet is still believed to be driven by a central engine, the AGN core. The one-sided view of this object is most likely due to its orientation to Earth. 

Imaging notes: a faint aircraft or some other faint object has streaked across the lower portion of the image. The area of interest is unaffected. 

(Figure 43) 

Figure 43 is a composite of the CCD Image, NVSS Image and FIRST Image. 

Back to Top

18. 3CR270.0: 

(Figure 44) 

(Figure 45

Object 3CR270.0 is also known as NGC 4261 (W8), an elliptical galaxy – one of the twelve bright members of the Virgo cluster of galaxies. The radio source of this galaxy is low-luminosity non-thermal emission from an AGN core. While categorized as an FR I BL Lac object, Hubble Space Telescope and Chandra data indicate a very hot and very large accretion disk surrounding a massive black hole (Chiaberge et al., 2003) with large jets. The radio jets are also symmetric on both axes. 

According to the NED database (W8), NGC 4261 has a visual magnitude of 10.4 and is at a distance corresponding to z=0.0075. Figure 45 demonstrates the extensive jets in the radio image – indicated by the arrows in the NVSS and FIRST images. 

Image notes: imaging of NGC 4261 was uneventful. Many smaller galaxies are visible in the field of view – the smaller members of the Virgo cluster. A much longer exposure would bring out the details of these smaller galaxies. 

(Figure 46) 

Figure 46 is a composite of the CCD Image, NVSS Image and FIRST Image.

Back to Top

19. 3CR272.1:

(Figure 47) 

(Figure 48) 

3CR272.1 is also known as 3C272. In addition, this object is the familiar elliptical galaxy M84 – member of the Virgo cluster of galaxies. This object is the closest galactic radio source of this group of targets, with a z=0.0354 and shines with a visual magnitude of 8.7 (W8). The radio source of this object is also non-thermal synchrotron radiation powered by a massive black hole. However, unlike other galaxies like this, the radio jets are much smaller and angular. It is suggested the material that is emitting the radio emission is also dense enough to affect the orientation of the jets as well as the orientation of the central engine causing a precession of the core (Quillen and Bower, 1999). 

Imaging notes: imaging of M84 was uneventful. Just like NGC 4261, M84 is a member is a galaxy cluster.  

(Figure 49) 

Figure 49 is a composite of the CCD Image, NVSS Image and FIRST Image.

Back to Top

20. 3C277.3: 

(Figure 50) 

(Figure 51)

Object 3C277.3 is also known as Coma A. This dominant radio source in the faint Coma cluster resides at a distance of z=0.085336 with a visual magnitude of a faint 15.9 (W8). Coma A is classified as an FRII type radio galaxy with the engine being typical of radio galaxies – a supermassive black hole (Morganti et al., 2002). While the core of a radio galaxy is thrilling enough, what makes this galaxy a bit more special is its proximity to us – meaning that it’s pretty close. This allows detailed views with spectroscopy and the Hubble Space Telescope. It’s not the core that is its fame, but its optical emission of the radio source – that is, emission lines have been detected in the optical wavelengths from the lobes of the radio jets (Miley et al., 1981). Also, absorption lines have been detected in the optical wavelengths as well, indicating that the jets are extending into a large gas disk surrounding the galaxy (Morganti et al., 2002). 

Imaging notes: Imaging of Coma A was uneventful. Compared to the other images, the star field is not as full and the galaxy distribution of this cluster is minimal. This object was at the lower edge of the main cluster.  

(Figure 52) 

Figure 52 is a composite of the CCD Image, NVSS Image and FIRST Image.

Back to Top

21. 3C405: 

(Figure 53) 

(Figure 54) 

Object 3C405 is also known as Cygnus A – a very powerful radio source at a relatively close distance, z=0.0561 (W8). Optically, the object has a dim visual magnitude of 16.2 (W8). The galaxy itself is very bright in the radio spectrum. In addition, the engine driving the emission is thought to be pointing near our line of site as indicated by the presence of the Broad-Line Region emission (BLR) – the same emission driving quasars (Thornton, Stockton and Ridgway, 1999). Cygnus A, being one of the brightest radio source in the sky, is classified as an FRII type radio galaxy (Burke and Graham-Smith, 2002). 

The NVSS image in figure 54 demonstrates the extent of the radio emission (red is bright, blue is not as bright) as indicated by the arrows.

Imaging notes: Cygnus A was a low target towards the east. Imaging of this target was uneventful, but had to be done late in the evening. This image is rather dramatic as the field is closer to the Milky Way band, resulting in a rather intense star field. 

(Figure 55) 

Figure 55 is a composite of the CCD Image and NVSS Image. There is no FIRST radio image of this object.

Back to Top

22. 3CR348: 

(Figure 56) 

(Figure 57) 

Object 3C348 is also known as Hercules A. This object is the faintest on our list with a visual magnitude of 18.25. It is also fairly distant compared to the other objects on our list, corresponding to a z=0.154 (W8). Hercules A is neither an FR I nor II type radio source. In fact, it is one of the few double-lobe radio sources and the only radio source that feature a double optical core (Sadun and Morrison, 2002). According to Sadun and Morrison (2002), the Hubble Space Telescope clearly identifies two galaxies, one is the AGN source – Hercules A and the other is a galaxy that does not emit in the radio. The presence of this second galaxy does not seem to affect, add or subtract anything from Hercules A. As such, the existence of this double-core object is a bit of a mystery, but it’s possible that this radio powerful, double-cored galaxy is the sum of a previous galaxy interaction. 

Imaging notes: imaging of Hercules A was uneventful. While the star field is somewhat rich, the area of interest is faint. 

(Figure 58) 

Figure 58 is a composite of the CCD Image and NVSS Image. There is no FIRST radio image of this object.

Back to Top

23. 3C461: 

(Figure 59) 

(Figure 60) 

3C461 is the only radio source on our list that is not a galaxy. Also known as Cassiopeia A, the optical counterpart was not noticed until bright radio emissions brought it to light (so to speak) – in fact, Cassiopeia A is the brightest radio object in the Northern Hemisphere (Fesen, 2001). Cassiopeia A is the youngest supernova remnant in our galaxy, which occurred some 300 years ago. The expanding shell is traveling at 6,000 km/second and colliding with the Interstellar Medium. It is this collision that emits synchrotron radiation visible at radio frequencies. If all goes well, emission from this object should continue for another 100,000 years with a gradual decrease in radio emission (Burke and Graham-Smith, 2002). 

The NVSS image in figure 60 demonstrates the extent of the radio emission (red is bright, blue is not as bright) as indicated by the arrows.

Imaging notes: This object was also difficult to image as the object was low in the horizon. This object was imaged in only one night – weather interfered with the rest of the imaging nights. In addition, the nebulosity is very faint resulting in 2x2 binning of the CCD to get good detail in a 15 minute exposure. 

(Figure 61) 

Figure 61 is a composite of the CCD Image and NVSS Image. There is no FIRST radio image of this object.

Back to Top

24. Discussion:

This project included eight radio targets – seven of them from galaxies and one from a supernova remnant. The radio emission from the galaxies has one thing in common, the central engine providing the source of the emission. The radio images that were acquired were from the 3C and 3CR (Third Cambridge Revised catalog) surveys, a radio survey that detected 320 extragalactic radio sources (Burke and Graham-Smith, 2002). 

These galaxies have been separated into two classes based on the presence of radio emission strength and orientation. This classification is called Fanaroff & Riley. An FR I galaxy is one that emits lower power radio emission and do not feature hot spots - strong central emission without prominent radio lobes. FR II type radio sources have prominent, supersonic jets that emit their radio emission primarily from the lobes “hot-spots” (Burke and Graham-Smith, 2002).  

Jets emanating from the core of the galaxy carry with it particles that are caught up into magnetic fields. This non-thermal radio emission is from these energetic particles moving through the magnetic field (Sparke and Gallagher, 2000), or synchrotron radiation. At first, it was thought there were several varieties of active galaxies: 

  • Quasars
  • BL Lac Objects
  • Seyfert Galaxies
  • Radio Galaxies

It is now widely accepted that all of these types of galaxies are really powered by the same engine – a supermassive black hole. The single group that covers all radio galaxies is Active Galactic Nuclei, or AGN (Sparke and Gallagher, 2000). The AGN model states that depending on the orientation of the galaxy core, the object will emit a different type of radiation (toward the observer) and therefore will appear to be a different object. 

(Figure 62

Figure 62 demonstrates the AGN of a galaxy. A BL Lac object, for example, will have the AGN facing us so we are looking directly at the jets. This will appear to make a BL Lac object a very strong radio source and one of the brightest phenomena in the Universe. As the orientation changes, the object will appear to look different – like a Seyfert galaxy or quasar (looking at an angle) or a “simple” radio galaxy (looking at the edge of the accretion disk). 

At the heart of an AGN is a very large black hole – a supermassive black hole. Material is caught in the black hole's intense gravitational field forming an accretion disk, generating intense heat. Some material is dragged along the intense magnetic fields, directed away from the black hole. As seen in figure 62, the magnetic field lines are perpendicular to the accretion disk. Energetic particles are funneled through the magnetic field and the result is the jets emanating from the core. Sometimes there are two sources of energy: the particles themselves emit synchrotron radiation, but supersonic particles can impact the ISM and create a “hot-spot” that also emit synchrotron radiation. 

The only non-galaxy object is Cassiopeia A, a supernova remnant. This used to be a high mass star; these large hot and bright stars do not last very long –only about 106 years. In these stars, hydrogen fusion gives way to helium fusion as the hydrogen runs out - then helium to carbon, carbon to oxygen, then oxygen to silicone. The final stage occurs when silicone fuses to iron. Iron cannot fuse so when it tries, the star blows itself apart in a supernova. All of the stellar debris is blasted out at about 6,000 - 10,000 km/s in an expanding shell of debris. This debris does two things, it seeds the surrounding interstellar medium (ISM) with heavy elements (W10) and impacts the ISM creating emissions of synchrotron radiation. The source of the magnetic field lines to create this emission is from the residual stellar core – a neutron star [W11]. 

Finally, a summary table: 


Common Name:

Visual Magnitude:

F-R Type

Special Feature:


Hydra A



‘S’ symmetry of jets, possible core precession





One of the first discovered quasars – one-sided jet


NGC 4261



Symmetric radio jets, large accretion disk





Smaller jets, possible core precession


Coma A



Emission lines from radio lobes detected in optical wavelength


Cygnus A



Brightest Radio Source, BLR Emission


Hercules A


I or II?

Double optical core, double-lobe radio source


Cassiopeia A


Not a galaxy

The only supernova remnant in our group

Back to Top

25. Summary:

Supernova remnants and AGN’s are strong emitters in the radio spectrum. Radio surveys like the 3CR (using the NVSS and FIRST survey) have identified a variety of radio galaxies. Tools like these radio surveys are vital in identifying some objects, for example Cassiopeia A was discovered by a radio survey as its optical emission is quite faint. In addition, some objects that look like stars are actually very distant galaxies called quasars. Studies of galaxies and supernova using radio and optical emission allow us to better understand the emission process as well. 

Radio galaxies are subset of active galaxies in a class known as Active Galactic Nuclei (AGN). Quasars, Seyfert galaxies, BL Lac objects and radio galaxies are all AGN’s. The source of radio emission is the acceleration of charged particles through magnetic fields generated by a supermassive black hole – called non-thermal synchrotron radiation. For FR II type galaxies, an additional radio source is the result of supersonic jet lobes impacting the Interstellar Medium (ISM). The supersonic velocity of charged particles in the supernova remnant also impacts the ISM emitting synchrotron radiation. 

This project demonstrates that amateur astronomers can do some real science with amateur class equipment. For example, a telescope with an aperture of 8 inches can reach a visual limiting magnitude of 14 – and a CCD camera attached to the telescope can go deeper than this. There is a host of data available online so all you need is a computer with an Internet connection and an object name. Sources like the NASA SkyView, FIRST or NVSS surveys (as well as MAST) can provide the rest.

This project was extremely fun. It should be noted that astrophotography can be extremely addicting so proceed with caution. I would like to thank Ron Wodaski for his tireless efforts in making sure the telescope was accessible. In addition, I would like to thank Josch Hambsh, R. Jay GaBany and Ken Levin (the new owners of the telescope) for adjusting their schedules to allow me to image my targets – or for helping me image when I had difficulties. And finally I must thank Professor Pamela Gay for her much appreciated guidance and patience.

Back to Top

26. References:

Becker, Robert; White, Richard and David Helfand. “The VLA’s FIRST Survey.” PASP, 1997. 

Burke, Bernard F. and Francis Graham-Smith. An Introduction to Radio Astronomy. Second Edition. Cambridge University Press, 2002. 

Chiaberge, Marco et al.”What do the Hubble Space Telescope and Chandra Tell Us About the Jet and the Nuclear Region of the Radio Galaxy 3C 270?” ApJ, 582:645-653. January 10, 2003. 

Condon, J.J., Cotton, W.D., Greisen, E.W., Yin, Q.F., Perley, R.A., Taylor, G.B. & Broderick, J.J. “The NRAO VLA Sky Survey.” 1998, AJ, 115, 1693. 

Fesen, Robert. “An Optical Survey of Outlying Ejecta in Cassiopeia A: Evidence for a Turbulent, Asymmetric Explosion.” ApJ, 133:161-186. March 2001. 

Kutner, Marc L. Astronomy. A Physical Perspective. Second Edition. Cambridge University Press, 2003. 

Kitchin, C.R. Astrophysical Techniques. Third Edition. Institute of Physics Publishing, Bristol. 2002. 

Lane, W.M. et al. “Hydra A at Low Radio Frequencies.” ApJ, 127:48-52. January 2004. 

Miley, George at al. “Optical Emission from the Extended Radio Source 3C 277.3 (Coma A).” AJ, 247:L5-L9. July 1, 1981. 

Morganti, R. et al. “Large-scale gas disk around the radio galaxy Coma A*.” Astronomy & Astrophysics, 387:830-837. 2002. 

Quillen, A.C. and Gary Bower. “M84: A Warp Based by Jet-Induced Pressure Gradients?” ApJ, 522:718-726. September 10, 1999. 

Sadun, Alberto and Philip Morrison. “Hercules A (3C 348): Phenomenology of an Unusual Active Galactic Nucleus.” ApJ, 123:2312-2320. May 2002. 

Sparke, Linda S. and John S. Gallagher. Galaxies in the Universe. An Introduction. Cambridge University Press, 2000. 

Stawarz, Lukasz. “On the Jet Activity in 3C 273.” ApJ. 613:119-128. September 20, 2004. 

Thornton, Robert; Stockton, Alan and Susan Ridgway. “Optical and Near-Infrared Spectroscopy of Cygnus A.” AJ, 118:1461-1467. October 1999. 

Back to Top

27. Website and Images References:

The NVSS Sky Survey Homepage

The VLA FIRST Survey Homepage

NASA SkyView Home Page

[W1] Keck Observatory: 

[W2] NASA: 

[W3] NASA:  

[W4] MIT Haystack:  

[W5] NRAO:  

[W6] NRAO:  

[W7] VLA:  

[W8] NED:  

[W9] AGN Model:  

[W10] Chandra:  

[W11] Chandra:  

Back to Top

28. Figures:

Figure 1:

Figure 2:

Figure 3:

Figure 4:

Figure 5:

Figure 6: NRAO / AUI / NSF

Figures 7 and 8:

Figures 9 and 10:

Figures 11 and 12:

Figures 13 and 14:

Figures 15 to 25:

Figures 26 to 31: TheSky 6 Professional

Figures 32 to 37: Mira Pro version 7.61

Figures 38 to 61: Core project images from FIRST, NVSS and me – processes with MaxIm DL, Mira Pro, PhotoShop CS2 and Illustrator CS2

Figure 62:


[1] From the highest elevation, we can observe in the higher end of the infrared – but only using near IR cameras that are super-cooled using liquid helium to about 3K (Kitchin, 2002)(W1).

[2] Screen save is done by the PRINT SCREEN button – but I used a program called SnagIt

Back to Top

Search | Site Map | Appendix
©2004 - 2024 Astronomy Online. All rights reserved. Contact Us. Legal. Creative Commons License
The works within is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.