IRTF: Data Reduction

How the data is reduced.
In short, we use a program developed specifically for reducing spectra from SPEX called SPEXTOOL.  It is written in a commonly used programming language for Astronomers called IDL.  I'm going to skip over a lot of the details, but in SPEXTOOL there are basically two steps:

1. Construct Calibration Frames (for wavelength calibration, flat-field subtraction).
2. Extract the spectral orders.

As you may recall from my last post, we took a series of calibration frames called arcs.  I've extracted a section of the image in my last post as a reminder:

The bright lines in the image (here shown in red) are spectral lines from the Argon arc lamp.  The software automatically finds the spectral features, whose wavelengths are known, and generates a linear regression to determine the wavelength as a function of pixel value (i.e. a wavelength vs. pixel value graph) for each order in the image. Then, this data is cross-correlated with a stored model for where the spectral lines should appear and the pixel differences are recorded a wavecal file.  These data look like the below image:

I've scaled this image logarithmically and added false color to make things easier to see.  (Also note that these may not be the same order or even the same mode as I don't recall what image I used for the sample spectra in the last post, this is just for visualization).  After this is done, we extract the spectral orders using the calibration information we have just obtained.  We used the telescope in AB mode in which the star was moved up and down the slit, so our data is reduced in AB mode.  In terms of data reduction, this means that we feed in two files and use both to construct the spectra.  An example of this extraction is shown in the image below:

Now we have to remove atmospheric features from the data.  This is done with another IDL program distributed with SPEXTOOL called Xtellcor.  Sadly these steps take a long time and I forgot to take screenshots as I was going along so for now if you want to know the full details I'm going to refer you to the paper on Xtellcor. Xtellcor has a built-in model of the throughput of the instrument and a model of an ideal A0V star.  Our first step is to convolve our comparison (A0V) star with the model A0V star to account for any residual wavelength shifts in the data.  Then we fix the continuum levels of the spectra by scaling atmospheric absorption lines of our comparison star to match the built-in model.  Finally we apply these corrections to our target star spectra and we're (almost entirely) done reducing the spectra.  We can still do more (i.e. normalize the spectra), but for now we can at least look at the data and see spectral lines:

What happens next?

From here, where do we proceed?  It's mostly measuring equivalent widths and seeing if anything has changed and of that which has changed, trying to explain what is going on in terms of the eclipse and the hydrodynamics/composition/structure of the disk.  You might be wondering "did we see any CO in the disk" as we originally discussed?  The answer is, not yet.  As expected, CO shouldn't appear until after mid-eclipse.  Did we see anything else interesting in the spectra?  Yes, but we can't talk about it yet as we need additional confirmation.

IRTF: Getting Observing Time

Before I entered graduate school, I had no idea about the complexity of research proposals, especially those in astronomy/astrophysics.

All telescope time for which I have applied has been peer reviewed.  This means that my proposal is read by fellow astronomers/astrophysicists and evaluated against some scoring method.  It is my responsibility, as the proposer, to convince the committee that not only is my research interesting, but also doable from a technical perspective. 

The basic outline of the proposal process is thus:

1. Determine a facility that can do the type of science you need.  This often means locating a facility that first covers the wavelengths in which you are interested.  As I mentioned in yesterday's post, we hope to observe the appearance of CO lines in the infrared so we obviously need an infrared telescope.  Although it goes without saying, the facility needs to be in the correct hemisphere to observe your target.  Being quite new to research (presently I'm in the third year of my Ph.D.), I often go ask my adviser, Dr. Bob, or another professor in our department about instruments they have used in the past that might cover the wavelength region of intrest.
2. See if the research facility has an instrument that can take the data you need.  The next step is to look at the facility's instrument list.  IRTF has a nice list of infrared instruments that can be used for spectroscopy and photometry.  We wanted to observe the CO spectral lines in the ~2.5 micron region so SpeX, and CSHELL are good options (there are separate observations looking in the mid-infrared with MIRSI and BASS).
3. Next, we need to verify that the instrument can observe our target.  This involves looking at sensitivities, exposure times, and estimating signal-to-noise ratios based upon previously observed targets with similar properties (or a model of the machine's performance characteristics).  With eps Aur, we have no problems getting enough light and in the case of some instruments, eps is too bright to be observed  without very special pixel placement.
4. After we have the basics down, we need to start looking at the proposal schedule.  Some observatories offer yearly, bi-yearly, or quarterly proposal opportunities.  We need to choose a proposal time that permits our target to be observed (in our case, any time in the winter is good).
5. Now comes the hard part: paperwork.  There's a series of forms and check lists that must be filled out detailing our observations.  It is in this step that we start filling out a description and scientific justification for our observations for peer review.  Some telescope time comes with funds so you might have to fill out a budget and then run it by your institution's business department.  These budgets can be tricky as your hosting institution can get a considerable cut of your incoming funds.
6. Then comes the easy part: submit and wait.  Most research facilities are good about getting proposals processed in a timely fashion, taking a month or two to get back to you.  Some of them can take six or more months.  During this time the reviewers are processing hundreds, possibly thousands of applications in order to pick the ones that offer the most interesting science.
7. And the best/worse part.  Acceptance or Rejection of your proposal.  If your proposal was accepted, great!  If not... well, try again I suppose.  At the big research facilities, telescope time is very competitive and for the space-based telescopes, it is extremely competitive, especially if the observing time comes with funds.  I can't give you any details statistics, but getting a rejection letter is common.  Fortunately, the reviewers provide feedback which describes why your proposal wasn't accepted which can often help when you re-propose.

Be sure to stop by tomorrow in which I'll discuss how we prepared for our IRTF observing run.

IRTF: Preparing and Conducting Observations

Picking up where I left off, in this blog post I will cover two topics: how we prepare for an observing session and how we conduct the observations.

How we prepare for an observing session.
The answer is short: meetings.  Lots and lots of meetings.  As part of our proposal we already specified our primary target (eps Aur) and verified that the instrument is capable of observing the star.  We also specified at least one A0V star to use as a calibration star for telluric (atmospheric) line removal.  In our final meetings, we double-check exposure times, calibration stars, and coordinates.  During this time we finalize our observing plan by deciding the order in which we will observe the stars on our list and insert flats, darks, and arcs (all required for calibration) so that we can maximize our observing time.

How the observations are conducted.
Although I have repeatedly tried to convince Dr. Bob to send me out to Hawaii to conduct the observations, we can't justify the airfare for only four hours of observing time.  Therefore we will be observing remotely over the Internet.  At IRTF there is a telescope operator (TO) who, among several things, controls the telescope.  There is also a support astronomer who is on duty to help out if we have a problem.

While the telescope operator takes care of the telescope, we control SPEX (the spectrograph).  To do so, we fire up two VNC sessions to control Big Dog, the spectrograph's user interface, and Guide Dog, the guider's interface.  Both interfaces are very similar.  They both consist of two main windows, the left hand panel is the main array that controls the camera, lamps, mirrors, dichroics, slit selection, slit rotation, nodding mode, and filters.  I've included a screenshot of GuideDog's main interface below:

The second panel on Guide Dog show the slit and star.  In the upper left panel of the image below, you can see back scattered light from the target star and it's position on the slit (the darker grey band going vertically).  In the lower right you see a zoomed-up view of the green box in the upper left.  This shows the telescope's current location (green crosshairs), current computed centroid of the star (the purple box), and the vector offset that is sent to the telescope control system (TCS) for tracking (the small purple line).  We observe in and AB-nodding mode in which the star is moved from position A (inside the green box) to position B (not shown, but about the same distance away from the bottom edge of the slit as the green box is from the top of the slit.

Assuming this object is our first star, the TO will slew to the target and have us set the slits and filters.  The TO will then adjust focus for us before giving us control of the camera and the fine control portion of the TCS.  From here we center up the star in the 'A' position and engage autoguiding.  After waiting a few moments to ensure autoguiding is working properly, we switch over to Big Dog.  Here we set our exposure times, Coadds, Guide Mode, Slow Count and Non-Destructive Reads (NDRs).  After setting up everything, we simply click a button and let the macros take the data for us.

Stay tuned for my next blog post in which I will discuss what the data looks like and how we reduce it.

IRTF: What the data looks like

As I have mentioned before, the instrument we used at IRTF is called SPEX.  It is a medium resolution spectrograph.  Specifically, it's a cross-dispersing spectrograph.  What does this mean?  Well, instead of dispersing light like a prism where the colors are all in one line, SPEX breaks the spectrally dispersed light into several orders that are displayed along side each other.  I've included a copy of a figure I made for a previous presentation that includes SPEX data in LXD1.9 mode to give you an idea about the dispersion:

Please note that the camera takes only black and white images.  I have tweaked the intensities and applied a rainbow map to the intensities using SAO DS9.  During our observations we actually use two different dispersive modes.  One is called SXD and the other is LXD2.1 and we merge the two together to form the final spectra.  The array is a 1024x1024 Aladdin 3 InSb array.  The camera is a very excellent device, but it is not without small defects that are common in all instruments.  For example, check out this image below:

There you see what appears to be a pit.  More interesting is the line that goes from top to bottom in the image that even show up the flat-field images:

Those lines, if I recall correctly, are actually defects in the structure of the camera's detector (I want to say that these are crystalline defects, but I'm not for certain).  The above image is actually a flat from a light source inside of the camera.  In the lower-left side, you can see that there appear to be overlapping orders.  This effect is called persistence and is literally one of the previous exposures.  In the lower-right, you can see some leftover spectra (the vertical bands) from a previous observation.  To calibrate our data, we need to take spectra of a known object that has well-defined spectral lines.  To do so, we expose the camera to an argon arc lamp and obtain calibration frames (called "arcs") like what follows:

In this image the most intense regions (shown in red-orange).  These are the arc lines from the arc lamp.  You can also see that the background isn't as clean as it was in previous images, this is simply background noise that will need to be removed during the processing steps. 

Spectra from stars looks different.  In the image below you can see the spectral bands from the atmosphere (the most obvious is the vertical red-orange line in the third order from the bottom), as well as the star's spectrum (red-orange bands that follow the orders).  You can also see the spectra from a previous exposure (the nearly horizontal line that crosses the second order from the bottom).

In my next post I'll discuss how the spectra is reduce from it's image form to the spectral form which we are accustomed to seeing.

IRTF: What we hope to observe

During the 1983 eclipse something funny happened in the infrared spectrum.  In addition to the continuum from the F-star, additional absorption lines started to appear.  This doesn't seem that odd until one considers that some of the spectral lines, carbon monoxide (CO) in particular, did not appear until after mid-eclipse.  The high resolution spectra obtained during the ecipse also provided some other interesting details, but I'll let you read the paper to find out those tidbits of information.  (As an interesting sidenote, the integration times for these spectra were very long.  One spectra i particular took 670 minutes!  That's a little over 11 hours to get a single spectra!)

Now here comes the big question: where did the CO come from?  The F-star is too hot for CO to form so we can't find it there.  The leading edge of the disk doesn't cause CO emission so it isn't there either, right?  In our proposal we argued that the CO is present in the disk and is frozen down to grains of dust in the disk.  When the disk rotates past the F-star, it's intense radiation causes the CO to sublimate off of the grains and become a gaseious cloud like I've drawn below:

(yes, this really is a sketch in my notes.)

With such a hypthosis going into our observation, and data that already proves the CO is present only in the trailing edge of the disk, how are we going to get observing time?  The experiment has been done and proven to be true.  Fortunatly we are trying to prove a few things in addition to the apperance of CO and the sublimation off of grains.  Because this is an active research topic (which may result in a publication), I can't openly discuss it, but what I can talk about is how we go from having a need for data to applying for telescope time.  That's coming up in tomorrow's blog posting.