Difference between revisions of "C-CWEL Bigger Picture and Goals"

From CoolWiki
Jump to navigationJump to search
m
m
 
(3 intermediate revisions by the same user not shown)
Line 14: Line 14:
  
 
=The overall "story arc"=
 
=The overall "story arc"=
 +
 +
(this is a very high level description)
  
 
OK, so I've done this a few times before. :*)  There are four potentially useful pages on the wiki with the "overall story arc" or "to-do list" describing the major tasks we have to accomplish towards actually reducing our data and analyzing it. The [[Working with L1688]] page (from 2008) tries to explain the 'story arc' by using a cluster similar to the ones we were studying to demonstrate the tasks. The [[Working with CG4+SA101]] page (from 2010) goes through the actual region that that group studied.  The more recent [[Working with the BRCs]] page (from 2011) goes through those two regions that that group studied (this was John's year). Finally, last year's [[Working with the C-WAYS data]] page (from 2012) started out trying to help that team make it through all three regions, but we decided mid-summer-visit to limit our work to just BRC27.  I found that being concrete was better than being abstract, so there will be a [[Working with the C-CWEL data]] page that will be fully updated when we get to that point.  
 
OK, so I've done this a few times before. :*)  There are four potentially useful pages on the wiki with the "overall story arc" or "to-do list" describing the major tasks we have to accomplish towards actually reducing our data and analyzing it. The [[Working with L1688]] page (from 2008) tries to explain the 'story arc' by using a cluster similar to the ones we were studying to demonstrate the tasks. The [[Working with CG4+SA101]] page (from 2010) goes through the actual region that that group studied.  The more recent [[Working with the BRCs]] page (from 2011) goes through those two regions that that group studied (this was John's year). Finally, last year's [[Working with the C-WAYS data]] page (from 2012) started out trying to help that team make it through all three regions, but we decided mid-summer-visit to limit our work to just BRC27.  I found that being concrete was better than being abstract, so there will be a [[Working with the C-CWEL data]] page that will be fully updated when we get to that point.  
  
 
The important part is that there are lots of exit ramps off this particular highway.  If we get through only a few of those tasks, but you really understand them, that's fine!  If we get through them all, that's fantastic. I expect we as a group will get most of the way through them before we have to write the poster.
 
The important part is that there are lots of exit ramps off this particular highway.  If we get through only a few of those tasks, but you really understand them, that's fine!  If we get through them all, that's fantastic. I expect we as a group will get most of the way through them before we have to write the poster.
 +
 +
 +
=The overall process=
 +
 +
(from email 27 feb 2013 during the process of proposal writing)
 +
 +
here is a better, clearer analysis chain of "we will do, a, b, c, d, and e"...
 +
 +
OK. Try this. This has more words (and more informal words), for your benefit, than go into the proposal.
 +
 +
*We will start with the WISE+2MASS catalogs in the region of interest.
 +
*We will (we have) run Xavier Koenig's selection algorithm on the WISE+2MASS catalog, which picks things based on colors.  What information I currently have to explain this selection algorithm is here: http://coolwiki.ipac.caltech.edu/index.php/Gutermuth_color_selection
 +
This process will identify (has identified) a subset of objects in which we will be more closely interested, e.g., a set of IR-selected YSO candidates.  (I asked Xavier to run this stuff a year ago, and he has since improved his algorithm, so there may be another pass through the selection process.)
 +
 +
Somewhat separately, on somewhat of a parallel track:
 +
*We will start with the catalogs obtained from the literature and archives, including:
 +
**X-rays from Getman+; smallish region
 +
**Optical from Choudhury+, Chauhan+ (incl spectra), Ogura+; small region
 +
**Optical from IPHAS, Nakano+, Barentsen+
 +
**NIR from 2MASS
 +
**NIR from Beltran+ (small region)
 +
**MIR from WISE
 +
**MIR from Akari
 +
**MIR from Spitzer (small region, and for now all we have is an incomplete list of objects because we have the Choudhury+, Chauhan+ catalog but that is just for some objects, and just for those in the central 4-band IRAC region)
 +
*We will assemble a big catalog of "everything we know" about every point source in this region that we can identify. We will identify a subset of objects in this catalog that any of these folks have ever identified as a high-confidence (e.g., confirmed) YSO or a YSO candidate, e.g., a set of literature-identified YSOs and YSO candidates.  But we will also have, at this point in the process, a merged catalog of everything (not just YSOs) at multiple wavelengths from prior studies, assuming they published a list of more than just the YSO candidates (and some of them have).
 +
 +
*We will merge these two catalogs ("everything in the literature" and "things Xavier picked as YSO candidates") such that we can do, at the very least, the following:
 +
** identify which of the YSOs/candidates from the literature were also recovered by Xavier Koenig's selection algorithm
 +
** identify which of the YSOs/candidates from the literature were NOT recovered by Xavier Koenig's selection algorithm
 +
** identify which objects (not necessarily YSOs) from the literature that were recovered by Xavier Koenig's selection algorithm (we want to identify additional bands beyond the 2MASS+WISE data that we can add to the SED to help us decide if each of the YSO candidates identified by Xavier's method are legitimate and worth keeping as YSO candidates.)
 +
 +
Portions of this process, for reasons I only sort of understand (otherwise I would fix it), is perennially confusing. Just about everyone, every year, has a terrible time keeping straight what is a "complete" catalog of "everything" in the region, what is a catalog of just literature YSOs/candidates, what is a catalog of just IR-selected YSO candidates, and the fact that "objects of interest" consist of the IR-selected YSO candidates PLUS the literature YSOs/candidates. I will keep repeating this again and again and some of you will still be confused. So please just keep asking until it makes sense!
 +
 +
OK, so now at this point in the process, we will have a catalog of "everything we know" about "everything of interest" in this region. 
 +
 +
*Now, we need to inspect each of the objects of interest to see if we can understand more about the nature of these objects. This process is necessarily recursive and will probably not end up being as linear as I portrayed on the phone today or here.
 +
 +
*For each of these objects, we need to:
 +
**inspect the object in as many images, from as many bands, as we can access. Does it look like a point source in all of them? If yes, keep; if no, drop.
 +
**inspect where each object falls in as many CMDs/CCDs as feasible to check. Does it fall in the right place to be a YSO in this region, at this distance? If yes, keep; if no, drop.
 +
**construct an SED for the object using as many bands as we can access. Does it look like a YSO SED? If yes, keep; if no, drop.
 +
 +
There will be shades of grey here. There will be, most likely, at least the following categories for each of the "if/then" statements above: keep!, keep, keep?, drop?, drop, drop!  We will discuss many if not all of them until we agree. There is an easy way to get POSS+2MASS+WISE images all at once for each object. We will have to separately inspect the images at other bands (including Spitzer) because there is no unified tool for this.
 +
Constructing SEDs will take some effort for at least the first few, but I will help turn the crank and do all of them for you.
 +
 +
We will have to make sure that we include in this process images & points from Spitzer where we have them. In order to get the Spitzer photometry (for the CMDs/CCDs or SEDs) for at least some of the sources, we will have to do photometry ourselves.  For that we will use APT on the pipeline-produced mosaics. And this will mean some recursion here -- we will check all the objects we can with all the bands we can and only go through the effort with doing photometry on Spitzer for the objects that still survive, and then reassess those based on Spitzer images and photometry. 
 +
 +
APT could potentially eat half a day of our summer visit, but it is easier to learn in person rather than over the phone. It does have benefits for understanding what the guts of a photometry program does, and how much to trust photometry from *any* catalog, and what decisions need to be made (and, e.g., why WISE's catalog gets fooled). But for the relatively few Spitzer points that we will get out of it, I am unsure if that much time during our visit is a wise investment. It will sort of depend on what you guys think, what "life skills" you want to get out of this project, and how fast we make it through some of the other "must do" tasks.  I can always churn through it with IDL or MOPEX as a backup.
  
 
=The timeline=
 
=The timeline=

Latest revision as of 16:11, 28 February 2013

The big goal

We have WISE data for a patches of sky likely to harbor young stars around BRC 38.

One of the signatures of young stars is that they have "more infrared than you'd expect" (e.g., they are redder than you expect) because of their circumstellar disk. We will use this property, as seen in the WISE data (as combined with data from some other archives), to identify new CANDIDATE young stars. The word "candidate" is important, because there is likely to be contamination in our sample from things that have colors that make them look like young stars, but they are actually not young stars. Most likely, the contaminants will be active galactic nuclei (AGN) in the distant background. The word "new" is also important -- there have been previous searches for young stars in these regions, so we need to make sure that we understand what has been done before so that we can compare what we did to what other people did, and make sure that we are not, say, announcing "OMG 30 new young stars!!!1!" when in reality 25 of them were found before by someone else, and we are rediscovering them -- rediscovering them independently, mind you, but rediscovering them nonetheless.

We will attempt to "compare and contrast" our results in this BRC with the results from prior NITARP teams (which worked with Spitzer and WISE data in other similar regions). Ideally, we will also extend this discussion to the rest of the literature. We benefit considerably from the C-WAYS team, who did a lot of the ground work for us in BRC 38 before they decided to focus just on BRC 27.

The big question driving all of this is - Why do certain stars like our Sun 'choose' to form planets? What makes that happen? Because we can't watch a single star from start to finish, and then set it up again, perturb it in a different way, and watch it go from start to finish again, we have to assemble as many stars as possible in as many different environments in the hopes that we can statistically unravel what is going on. This is why we are looking for new young stars.

The concrete goal

We have to come up with a science poster (and an education one) for the AAS in Jan 2014. BUT because the posters can be simple or complex, this goal is a little squishy, perhaps squishier than you might be comfortable with. What I describe here (and elsewhere on the wiki) is the kind of goal I would give a grad student. But getting through even a part of it (rather than all of it) is still a success!! This may be hard to really internalize, but it's true.

The overall "story arc"

(this is a very high level description)

OK, so I've done this a few times before. :*) There are four potentially useful pages on the wiki with the "overall story arc" or "to-do list" describing the major tasks we have to accomplish towards actually reducing our data and analyzing it. The Working with L1688 page (from 2008) tries to explain the 'story arc' by using a cluster similar to the ones we were studying to demonstrate the tasks. The Working with CG4+SA101 page (from 2010) goes through the actual region that that group studied. The more recent Working with the BRCs page (from 2011) goes through those two regions that that group studied (this was John's year). Finally, last year's Working with the C-WAYS data page (from 2012) started out trying to help that team make it through all three regions, but we decided mid-summer-visit to limit our work to just BRC27. I found that being concrete was better than being abstract, so there will be a Working with the C-CWEL data page that will be fully updated when we get to that point.

The important part is that there are lots of exit ramps off this particular highway. If we get through only a few of those tasks, but you really understand them, that's fine! If we get through them all, that's fantastic. I expect we as a group will get most of the way through them before we have to write the poster.


The overall process

(from email 27 feb 2013 during the process of proposal writing)

here is a better, clearer analysis chain of "we will do, a, b, c, d, and e"...

OK. Try this. This has more words (and more informal words), for your benefit, than go into the proposal.

This process will identify (has identified) a subset of objects in which we will be more closely interested, e.g., a set of IR-selected YSO candidates. (I asked Xavier to run this stuff a year ago, and he has since improved his algorithm, so there may be another pass through the selection process.)

Somewhat separately, on somewhat of a parallel track:

  • We will start with the catalogs obtained from the literature and archives, including:
    • X-rays from Getman+; smallish region
    • Optical from Choudhury+, Chauhan+ (incl spectra), Ogura+; small region
    • Optical from IPHAS, Nakano+, Barentsen+
    • NIR from 2MASS
    • NIR from Beltran+ (small region)
    • MIR from WISE
    • MIR from Akari
    • MIR from Spitzer (small region, and for now all we have is an incomplete list of objects because we have the Choudhury+, Chauhan+ catalog but that is just for some objects, and just for those in the central 4-band IRAC region)
  • We will assemble a big catalog of "everything we know" about every point source in this region that we can identify. We will identify a subset of objects in this catalog that any of these folks have ever identified as a high-confidence (e.g., confirmed) YSO or a YSO candidate, e.g., a set of literature-identified YSOs and YSO candidates. But we will also have, at this point in the process, a merged catalog of everything (not just YSOs) at multiple wavelengths from prior studies, assuming they published a list of more than just the YSO candidates (and some of them have).
  • We will merge these two catalogs ("everything in the literature" and "things Xavier picked as YSO candidates") such that we can do, at the very least, the following:
    • identify which of the YSOs/candidates from the literature were also recovered by Xavier Koenig's selection algorithm
    • identify which of the YSOs/candidates from the literature were NOT recovered by Xavier Koenig's selection algorithm
    • identify which objects (not necessarily YSOs) from the literature that were recovered by Xavier Koenig's selection algorithm (we want to identify additional bands beyond the 2MASS+WISE data that we can add to the SED to help us decide if each of the YSO candidates identified by Xavier's method are legitimate and worth keeping as YSO candidates.)

Portions of this process, for reasons I only sort of understand (otherwise I would fix it), is perennially confusing. Just about everyone, every year, has a terrible time keeping straight what is a "complete" catalog of "everything" in the region, what is a catalog of just literature YSOs/candidates, what is a catalog of just IR-selected YSO candidates, and the fact that "objects of interest" consist of the IR-selected YSO candidates PLUS the literature YSOs/candidates. I will keep repeating this again and again and some of you will still be confused. So please just keep asking until it makes sense!

OK, so now at this point in the process, we will have a catalog of "everything we know" about "everything of interest" in this region.

  • Now, we need to inspect each of the objects of interest to see if we can understand more about the nature of these objects. This process is necessarily recursive and will probably not end up being as linear as I portrayed on the phone today or here.
  • For each of these objects, we need to:
    • inspect the object in as many images, from as many bands, as we can access. Does it look like a point source in all of them? If yes, keep; if no, drop.
    • inspect where each object falls in as many CMDs/CCDs as feasible to check. Does it fall in the right place to be a YSO in this region, at this distance? If yes, keep; if no, drop.
    • construct an SED for the object using as many bands as we can access. Does it look like a YSO SED? If yes, keep; if no, drop.

There will be shades of grey here. There will be, most likely, at least the following categories for each of the "if/then" statements above: keep!, keep, keep?, drop?, drop, drop! We will discuss many if not all of them until we agree. There is an easy way to get POSS+2MASS+WISE images all at once for each object. We will have to separately inspect the images at other bands (including Spitzer) because there is no unified tool for this. Constructing SEDs will take some effort for at least the first few, but I will help turn the crank and do all of them for you.

We will have to make sure that we include in this process images & points from Spitzer where we have them. In order to get the Spitzer photometry (for the CMDs/CCDs or SEDs) for at least some of the sources, we will have to do photometry ourselves. For that we will use APT on the pipeline-produced mosaics. And this will mean some recursion here -- we will check all the objects we can with all the bands we can and only go through the effort with doing photometry on Spitzer for the objects that still survive, and then reassess those based on Spitzer images and photometry.

APT could potentially eat half a day of our summer visit, but it is easier to learn in person rather than over the phone. It does have benefits for understanding what the guts of a photometry program does, and how much to trust photometry from *any* catalog, and what decisions need to be made (and, e.g., why WISE's catalog gets fooled). But for the relatively few Spitzer points that we will get out of it, I am unsure if that much time during our visit is a wise investment. It will sort of depend on what you guys think, what "life skills" you want to get out of this project, and how fast we make it through some of the other "must do" tasks. I can always churn through it with IDL or MOPEX as a backup.

The timeline

At our workshop in January, I talked about a global timeline for an entire NITARP project. Look here (link to the CoolCosmos site) for the document I distributed at the January workshop. IMPORTANT NOTES: (1) This is a schedule in the ideal case. Your mileage may vary. (2) This includes everything for the entire cycle, including applications, so that I can give it to other people and have the entire thing make sense. (We're fundraising and recruiting here too.)

The important dates are:

  • Jan-Mar - proposal prep
  • Apr-June - background work
  • June-Aug - summer visits, learn and start analysis
  • Sep-Oct - continue analysis
  • Oct? - AAS abstracts due
  • Oct-Nov - finish analysis
  • Dec - write posters
  • Jan - present them!