Let's all take a moment to think ourselves back in time. We're going all the way back to Ancient Greece, so it might have to be a long moment.
Here's a picture of the Parthenon to help you out.
Are we all there? Good. Back then, the greatest minds in the world, like Aristotle, thought that stars, planets, and other heavenly bodies were all perfect spheres. That changed when people discovered sunspots. If the sun, the most perfect heavenly body of all, had dark marks--one might even go so far as to call them bruises--then surely the rest of the universe was filled with imperfections as well.
That was basically a really long (and maybe unnecessary) way for me to say that data isn't perfect. Telescopes have flaws. Nights will be cloudy. You'll decide to punish yourself by trying to do optical observations in Boston. How do we deal with this? Calibration.
There are three basic effects that one would need to calibrate for
- Instrumental Imperfections
- Changes in the sky or atmosphere
- Fluxes of nearby stars (how much light they're producing per unit area per unit time)
Each one of these calibrations is really important when it comes to collecting accurate data.
Let's start at the beginning (a very good place to start) and talk about how to account for imperfections in your image due to instrumental errors. To do this, you take what astronomers call dark images. That means you take a picture in complete darkness (see what they did there?). When you do that, you get a baseline measurement of how each pixel in the CCD (the part of the camera that actually collects and records the photons being received) responds when it takes a picture. But the most important instrumental effect that needs to be accounted for in dark imaging is thermal noise.
Electronic devices create heat, which excites electrons. CCDs record images by keeping track of how many electrons get excited to new energy levels. They don't care whether those electrons were excited by heat from the device or photons from the imaged source--it blends them all together. So it's important to keep the device as cool as possible. Still, it's not like we're going to reach Absolute Zero, so there will be some thermal noise. By taking dark images, you determine how much thermal noise your device inherently has, and then you can subtract that from your final image.
Some cameras automatically take dark images by snapping a picture with the shutter closed before taking the actual photo, but this is still a nice thing to keep in mind.
Now onto atmospheric influences.
Electronic devices create heat, which excites electrons. CCDs record images by keeping track of how many electrons get excited to new energy levels. They don't care whether those electrons were excited by heat from the device or photons from the imaged source--it blends them all together. So it's important to keep the device as cool as possible. Still, it's not like we're going to reach Absolute Zero, so there will be some thermal noise. By taking dark images, you determine how much thermal noise your device inherently has, and then you can subtract that from your final image.
Some cameras automatically take dark images by snapping a picture with the shutter closed before taking the actual photo, but this is still a nice thing to keep in mind.
Now onto atmospheric influences.
Isn't that picture beautiful? Yes, but it's also a pain for astronomers. See how the light is uneven--the top of the picture is darker than the bottom? That doesn't make really useful images for people who actually want to do high-level science with them. And it all has to do with differences in the atmosphere. But there's a way around it, and it's called Flat Fielding.
To take a flat field image, you have to take a picture of a completely evenly-illuminated field. You could do this at dawn or dusk (when the sky is all one color and there are no stars) or you could illuminate a piece of paper or a T-shirt and take a picture of that. Get creative! Just make sure the field is evenly-illuminated.
This basically lets you know how your camera responds to a blank sheet of light, so when your image ends up (almost inevitably) having an uneven field, you can normalize it to your flat.
And, last, but certainly not least, it's important to calibrate to the fluxes of well-studied nearby sources of light. This is especially important in a project like ours where photometry (measuring how much light we're receiving from the target field) is so key. Astronomers have worked hard to put together catalogs of stars and their fluxes so that others can use them as references when trying to determine the fluxes of their target.
Think of it like this: You're looking at two of your friends standing a football field away. You can't remember how tall friend A is, but you know friend B is 6 feet tall. You can then figure out how tall friend A is based on the height difference you observe between A and B.
Since our entire project depends on being able to catch slight dips in one star's flux relative to its neighbors, this calibration is particularly important.
Hopefully you now understand why calibration is so important. Or maybe you already knew that and the only thing you really learned by reading this post is what the Parthenon looks like. Either way, calibration is both awesome and necessary! If only we could just calibrate the imperfections out of our everyday lives...
