|
|||||||||
|
|||||||||
Cas-A |
| 1. Create a region, centered on the pulsar, that is about the same same size as the entire remnant. |
|
| 2. In the analysis drop down menu, select "light curve plot". Place a value of 32.4104 in the upper part of the dialog box, and check both other boxes. Does the intensity versus time appear to be constant? It looks like the intensity is varying wildly! But is this real? Now we come up against one of the most difficult of all scientific problems; deciding whether data is telling us that something unusual is happening to the source, or whether we are just looking at a statistical fluctuation that you would expect to see whenever you examine the object. |
|
Notice that although the light curve seems to change, the value of the intensity is always 265 counts/sec, plus or minus 10 counts/sec. Let's get this replotted so we can get a better idea of the true variation.
a. In the light curve plot, select view--> axis range... and change the Y axis to read from 0 to 300.; Deselect the auto box. Hit OK. Now look at the same data!!; Quite a bit smoother, isn't it? What do you believe? There are sophisticated statistical tests that we make to show that indeed, the light curve does NOT vary in time, at least within the limitations that we have with the Chandra satellite.


3. Now, make some smaller regions so you can see what happens to the light from little knots of material. Make light curves of these regions. Do they look similar?
4. Let's examine the background (regions that do not contain the remnant). For example, the region surrounding Physical X=4080 and Physical Y=4600 is within the field of view of Chandra, but is looking at an empty part of space. What does this light curve look like?
5. Now, do a background subtraction. Select a region inside the remnant, and then make several regions of background, by pressing the "b" key within each background region. This will cause the selected region to become dotted. Now do a light curve. What happened?
Since any instrument has
"noise" or a level of signal that occurs even when
you are not looking at anything, we use the above technique to
get a truer picture of what the actual source is doing. What
you have done here is to take the data from the remnant (which
really consists of data PLUS background noise) and subtract off
the background component by selecting regions of blank sky.
Note: It is very important that the light curve
"tool" have a time bin that is an exact multiple of
3.24104 sec, so we can avoid spurious aliasing of the intensity.
Try this without this restriction (use, for example, 30 instead
of 32.4104 seconds, and see what happens. Can you explain why?)
Last updated: 7/22/04
|
|
|
|
Harvard-Smithsonian Center for Astrophysics 60 Garden Street, Cambridge, MA 02138 USA Phone: 617.496.7941 Fax: 617.495.7356 |
Comments
& Questions? cxcedu@cfa.harvard.edu |
|
This site was developed with funding from NASA under Contract NAS8-39073. |
|