Data Reduction and Image Processing

First Group Project

Last class we successfully acquired multi-color images of several clusters. In one good night, we got enough data to complete this project. Now the hard work begins!

1) Collect and Organize the data. To speed things up, I already copied the data from Lestrade to Sherlock. It currently resides at /sherlock/altau6/

a) make an ASTR377 directory in your home directory

b) make a 11sep12 subdirectory in this directory

c) cp /sherlock/altau6/*.FIT ~yourusername/ASTR377/11sep12 (this might take a while; you only need the darks and the data for your star, at least to get started)

d) Decide how you want to arrange the data into subdirectories. For example, you might want to put all the darks in one directory. You might want to put all the B data in one directory and the V in another. At this stage of the game, you never know the most optimal way of arranging your data, but you've got to get started now and be flexible later (i.e. rearrange things as necessary).

2) cd ~yourusername/Iraf and cl then change to your working directory cd ../ASTR377/11sep12 (you always need to start iraf from your Iraf directory so it can find your login.cl and uparm directory, but you probably don't want to use it as a working directory)

3) Screen the data. Using display, imexamin, implot, imstat, imhead, and any other tools you've learned, have a look at the data. Make sure it's what you think it is, the headers are right, nothing is corrupted, etc. Delete anything that is obviously junk. You will be using the "thermal" images as a combined dark and bias. You have enough of them that you can toss any that have a huge cosmic ray hit. Maybe some of your data files are corrupted, too. You probably don't have time to individually inspect each image, although you should. Look at enough that you know what's going on and that you are comfortable with these screening tools. You might end up re-arranging your subdirectories as you go along.

4) BIAS/DARK average. You will be using noao/imred/ccdproc. You could treat these as bias frames (using zerocombine), but you need to make sure that iraf doesn't get hung up on the header keywords. You might have to do a massive hedit to insert the right keyword.

5) FLAT averages. You'll be using flatcombine, but there are a lot of parameters. We'll look at them in real-time and decide how to proceed.

6) Ideally, we would next average and cosmic-ray filter your "object" images. But we might want to register (shift and add) them AFTER pixel-by-pixel corrections to the flat and bias are applied. So we'll process everything at once, but we'll have to look at the ccdproc parameters very carefully before we do. You'll then apply the average bias, average flat, and trim off the edges. This step will OVERWRITE your data, so you have to do it right (ALWAYS keep a copy of the original data in a directory well away from your working directory).

7) Finally, we want to register, cosmic-ray filter, and average the reduced cluster images. You should end up with one image for each filter for your cluster.

Before moving on to the next step, go back and review what you've done. Fill in the gaps in your understanding. Re-do things if you think you messed them up. Try setting parameters differently and see what effect it has. Practice makes adequate. So practice some more, and read over the iraf documentation, and ask questions.

Next time, we'll learn some iraf tools to take our images and turn them into a table of RA, Dec, star id (hopefully), B, V, R, I. Then we'll wrap this up by learning some plotting tools (in IDL) to analyze our results.