Overview

Your task is to use Spitzer Space Telescope observations to determine how much infrared is being emitted from the center of certain galaxies. You will use images of galaxies in 4 different wavelengths from the Infrared Array Camera (IRAC); 3.6, 4.5, 5.8 and 8.0 microns. The galaxies we have targeted are called Active Galactic Nuclei, or AGN. They are called "active" because they emit large amounts of energy at many wavelengths above and beyond what we would calculate based on the total numbers of stars at their center. We believe the excess energy is generated by material flowing into a huge black hole at their centers. Your measurements of the infrared emission from these objects will be combined with your (or other students’) measurements of the radio emission from these same objects, as well as estimates made by professional astronomers of the of the black holes at their centers. We hope that we will learn more about how AGNs by determining if and how their infrared and radio emissions depend on the of the black hole.

To determine the infrared energy emitted from the surrounding region of the black hole, you will do these things:

1) Log onto a Lewis Center computer, and run two computer programs (called IDL and ATV) that are used for visualizing and manipulating data.

2) Load your Spitzer data into ATV and adjust the image of the AGN so you can more easily see it.

3) Determine how much energy was received from the center of the AGN. This is a two step process. First, you measure how much total energy is coming from the central region. , you must subtract off any background emission that is present. (This background is emission from many other things that are in the , in addition to the nucleus of the galaxy.) As you will see, the ATV program is set-up to make this very easy for you to do!

4) Answer some questions about what you have learned and relate them to your radio work.

At the end of this document we list the masses of the black holes you have observed. You will use this information in the analysis of the data.

1. Start the Data Analysis Software

To start data analysis you must log into: http://idl2.gavrt.org/

You will get a page with a password requirement.

The password is: g4vrt4u

Then you will get the following screen that may also require a password. Put in the same password.

You will have an image of another desktop. On that desktop is an icon called “IDL.” Double click on it to open

This will start two screens, one is running a program called IDL:

and the other has a program called ATV which is generated by IDL. It is the tool you will use to display and analyze the data from the Spitzer Space Telescope. Right now it should be displaying a cross, which is the default image:

IMPORTANT NOTE: At the cursor can leave a trail that can cover the text (see figure on left). If you need to refresh the screen DO NOT use the “Page Reload” option of the browser (the one with the circular arrow on top)!!! Either use the “Refresh” button at the top center, or minimize and then maximize the screen. If you use the “Page Reload” button you will be taken to the login screen again, though your work will not be destroyed.

2. Loading Your Data and Adjusting the Image

Spitzer files are saved in FITS format which is the most common format that astronomical images are saved. It stands for Flexible Image Transport System.

Go to the “File” menu and open the image that you wish to analyze using “Readfits.” These images should have been uploaded to the Lewis Center and placed in a directory where you should be able to open them. If the images you are looking for do not show up when you use the “Readfits” function, contact the Lewis Center.

As an example you can load in the image n4945-24um.fits. It should look like this:

Once you have opened an image you can change the display in two forms to get a better look at the data. This is called changing the “stretch” of the image.

First you can press the left mouse button and hold it down and move the mouse. This will change the relative brightness on the screen. If you only have one mouse button, which is typical on Apple computers, don't worry! You can jump straight to the second way of stretching the image. You also could try pressing either the "option" key, or the down- arrow key (if you have one) at the same as pressing your mouse button.

Second you can apply a mathematical function to the image. This is the most useful method since often the main problem is that there are very bright and very faint things in an image so trying to display both simultaneously is difficult. But if you apply a logarithm function to the images where each pixel value is replaced by the log of that value then it becomes easier to see bright things next to faint things.

For example if one pixel has a value of 10 and the pixel next to it is 10 times brighter and has a value of 100, it is difficult to analyze. But if you took the log of both pixel values then their pixel values change from 10 and 100 to 1 and 2 (log(10)=1, log(100)=2) so it become easier to see the brighter value next to the fainter one.

Go to the “Scaling” menu and select “log” for the image you have loaded. You can still use the left mouse button to change the relative brightness on the screen.

A final way to change the display is to change the minimum (Min) and maximum (Max) values displayed. If the image has a value higher than the maximum you specify, all pixels higher than that maximum value are displayed as if they had that maximum value. Similarly, if you specify a minimum value to plot, image values lower than that are treated as if they have the minimum value. You specify the minimum and maximum values in the upper middle of the ATV window, just to the left of the two small images. Change those numbers to see how the image changes.

Now you are ready to do .

3. Determining the From the AGN

Photometry is the measurement of the brightness of an object. The brightness that is recorded on an electronic detector (or any kind of detector) is a combination of the brightness of the source plus the brightness of the background that the source is on.

In the case of the center of the galaxy above, a great deal of the "background" is in fact from the surrounding host galaxy. But what we are interested in is only the light from the center and not from the rest of the galaxy. To account for this, we will determine how much light is coming from where the center of the AGN is, and then compare it to how much light is coming from near the center of the AGN. We assume that the "background" near the AGN is the same as the background right on the AGN, so by subtracting the two, we are left with only the brightness of the center of the AGN.

To start the photometry process, place the cursor on the center of the galaxy (or in general on whatever you want to do photometry on) and press the “P” key. Notice that on the upper right of ATV, there is a window that magnifies the region your cursor is on, or you can simply zoom in using the “ZoomIn” function just above the image. Once you press “P,” then a new window opens with a magnified region around where your cursor was with 3 circles drawn around your photometry target.

The green circle is around the target whose light you want to measure and it is called your “photometry aperture.” The blue and purple circles define a ring in which the sky measurement will be made which will give the average background value per pixel that will be subtracted from each pixel within the aperture value. Then after all the pixels within the photometry aperture have this average per pixel sky value subtracted from them, they are totaled up and are displayed on the right as “Object Counts: 53770.5”

The next step revolves around the radius of the photometry aperture. How do you know if it is big enough or maybe if it is too big? To see if you have included the light mainly from the center and not mostly from the background, press the “Show Radial Profile” key:

This is a one dimensional representation of the image. The brightest part of the image is at zero radius on the x-axis (the numbers are partially hidden at the bottom of the screen) and has the most counts (2300 counts on the y-axis). But as you go away from the center of the image and to the right on the plot you see that the counts come down and then they are almost flat but they don’t actually reach zero counts. This is because there is background or sky contribution which in this case comes from the rest of the host galaxy and is noted in red as “sky level.” The sky level is measured within a ring between the inner sky measurement radius (insky) and the outer sky measurment radius (outsky). Unfortunately our aperture radius (aprad) chosen above is too small and we cut off some of the light from the source before it reaches the red line of the “sky level.”. So we need to change the radius from 5 pixels to 9 pixels where the image goes flat. You must also make sure that the aprad does not cross into the insky which is at 10 pixels in the above image. We usually also like to put some space between the aprad and insky to make sure that the sky and point source values do not mix, but in this case the background is not uniform and comes right up to the source we are measuring so we need to make the sky measurements right outside of the aprad. So the new values for the aprad and insky radii will be 9 and 9. There is no need to move the outsky radius since there is plenty of space between insky and outsky. The ring for measuring the sky background should not be smaller than 3 pixels and we have 11 pixels between the insky radius of 9 and the outsky radius of 20. Once the new radii are put in then the Object Counts is updated with the new value being 65661.6.

The reason that this is a higher number is that we have all the central region in the aprad. As an exercise you can change the aprad and see how much the final counts change. You have now measured the light coming from the center of this galaxy!

To convert the Object Counts into a more commonly used physical value, the conversion value is 3.38x10-5 (explained in Appendix A). The counts measured in this case are 65661.6 which when multiplied by 3.38x10-5 give 2.2 Jy.

Here is another example from the data you will be using. Load all 4 images one after the other of the AGN 3C78 (3c78-3.6um.fits, 3C78-4.5um.fits, 3C78-5.8um.fits, 3C78- 8.0um.fits). Display it as shown previously. Notice how different each image looks at the different wavelengths:

Notice how the fuzzy part of the host galaxy gets smaller as the wavelength gets longer? That is because that fuzz is about 100 billion stars and stars emit more light at shorter wavelengths (3.6 micron) than at longer ones (8 micron). Notice how there are fewer stars in the entire image at 8.0 microns vs. 3.6 microns. Those are foreground stars in our own Milky Way galaxy that, like all stars, emit less 8 micron radiation than 3.6 micron radiation and hence are not as prominent in the longer wavelength image as they are in the shorter wavelength one.

But why does the center of the galaxy not get dimmer at longer wavelengths as well?

That is where the black hole at the center of the galaxy comes in. There is a lot of gas that is flowing into the black hole and it is heating up as it goes in. This process, called accretion, causes a large amount of light in optical and UV to be generated by the hot gas (no light comes from the black hole itself). The black hole is surrounded by dust and that dust absorbs the UV and optical photons and reradiates them at longer wavelengths thus making the nucleus brighter than the stars in the longer wavelengths.

Since our main concern is the infrared light coming from around the black hole and not the light coming from the stars, we will choose the photometry aperture based on the wavelength least affected by starlight: the 8 micron image. So load the 8micron image and display it. Use the log value and set the Maximum to a high number like 100. Then do the steps for photometry by pressing “P” on the center and displaying a radial profile and choosing the appropriate aprad, insky, and outsky:

In this case to contain all the light from the center we have chosen aprad=12, insky=12, and outsky=15 for total Object Counts of 1295.66 which by using the earlier conversion is 1295.66 x 3.38x10-5 = 0.04 Jy.

Now use the same aperture values for the three other wavelengths for this AGN (3.6, 4.5, and 5.8 microns). Notice how important it is to do the background subtraction right outside the photometry aperture for the 3.6 micron image where there is a lot of contamination from the starlight of the AGN host galaxy. This is also the reason why we have chosen a small gap between insky and outsky (3 pixels) so that primarily we remove the local from the host galaxy. You can notice the level of host galaxy starlight contamination by changing the insky and outsky values and seeing how much the final Object Counts change for the 3.6 microns image and then comparing the same insky and outsky changes for the 8.0 micron image. But for your final results the Object Counts you measure for all four images should use the same aperture radius and sky annulus that you determined for the 8 micron image.

You can then convert them to Jys based on the conversion number given and plot the ratio of the IR to radio flux to the black hole masses given at the end of this document.

4. Some Questions to Consider

1. How does the process to eliminate background emission from our Spitzer measurement compare to the cross-scan technique used on the DSS-12 radio telescope?

2. Why do we calculate the average brightness per pixel instead of the total brightness in the sky region? (Hint, when you are doing the photometry using the colored circles on the image, is the area within the green circle the same as the area between the blue and purple ones?)

5. Known Black Hole Masses for 3C radio Galaxies from Woo & Urry (2002)

Name Black Hole masses kg 3C 29 3.2x1038 3C 31 6.3x1038 3C 33 4.8x1038 3C 40 1.4x1038 3C 62 9.4x1038 3C 76.1 2.7x1038 3C 78 8.0x1038 3C 84 6.2x1038 3C 88 2.1x1038 3C 89 6.6x1038 3C 98 1.5x1038 3C 120 2.7x1038 3C 192 3.2x1038 3C 196.1 3.2x1038 3C 223 2.8x1038 3C 293 2.0x1038 3C 305 1.7x1038 3C 338 1.2x1039 (Note that the exponent here is 39 NOT 38) 3C 388 3.0x1039 `` 3C 444 9.6x1037 (Note that the exponent here is 37 NOT 38) 3C 449 4.3x1038

APPENDIX A

What exactly are the counts that you have measured? The units of the image are MegaJanskys per Steradian per Pixel. That is a flux density (MegaJansky) per unit area of the sky (steradian) striking each pixel (per pixel). Janskys are a unit of flux density equal to 10-26 /m2/Hz, which is a flux (in this case measured in W/m2 which is the number of Watts that pass though a square meter) per unit (in this case measured in Hz). So, by analogy, flux is how many cars are going through the entrance of a tunnel, and a flux density is how many cars of a particular color, say red cars, are going through the entrance of the tunnel. Instead of cars, Spitzer is measuring infrared photons, and the Jansky is a measure of how many photons of a particular frequency are impacting a unit area of our detector.

The rest of the units are the steradians per pixel, which tells us how much of the sky is seen by each pixel of Spitzer’s detector. A steradian is a measure of area on the sky. It is an angular area, not the area we typically talk about. To understand steradians, you first have to know what a is. A radian is a measure of an . For example, some of you may know that a right triangle is a triangle with one angle equal to 90 degrees. If we measured that 90- angle in , we would say it is about 1.57 radians. The ratio of a degree to a radian is Pi/180, so 90o x 3.1415926 radian /180o = 1.57 radians. Similarly, I could say that one radian is about 0.017 degrees. A steradian is a patch of sky one radian on a side (or about 0.017 degrees on a side). In the case of the Spitzer camera that we are using which is the Infrared Array Camera, each pixel sees 0.000333 degrees by 0.000333 degress on the sky. Converting the degrees to radians gives 5.8177584x10-6 radians by 5.8177584x10-6 radians . Since our pixels are square, the area of the pixel is just the length of one of its sides squared so by squaring 5.8177584x10-6 radians we get an area of 3.3846331x10-11 steradians or we can round to 3.38x10-11.

Putting all of the above together, the counts are a measure of how many W/m2 are hitting each pixel at a particular frequency from a small patch of the sky. Sometimes, however, we just want to talk about the total amount of energy/sec entering each pixel of our detector, rather than the amount of energy/second from a unit area on the sky. To do this, we multiply the MegaJansky per steradian per pixel which we’ve measured above by the number of steradians in each pixel. So to convert the counts into simply MegaJanskys, multiply the MegaJansky/Steradian/pixel by the Steradians/pixel (3.38x10-11 steradians per pixel) to get the value in Megajanskys. To get the value in Jy instead of MegaJy you multiply 3.38x10-11 by 1x106 which gives the final conversion value of 3.38x10-5.