Exercises with IRSA tools

From CoolWiki
Revision as of 17:07, 30 July 2024 by Rebull (talk | contribs) (Created page with "=Multi-wavelength images, the simplest version= IRSA home page → Finder Chart → put in any target you want. Turn off catalog searching; just search for images. Look at re...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search

Multi-wavelength images, the simplest version

IRSA home page → Finder Chart → put in any target you want. Turn off catalog searching; just search for images. Look at results and compare images across wavelengths. Turn on 3-color images if you want. Why do things look like they do as a function of wavelength?

Multi-wavelength images, a more complex version

Based on this idea: https://vmcoolwiki.ipac.caltech.edu/index.php/Dustier,_Messier_Messier_Marathon Goal : explore differences between optical and IR properties of images of various types of objects

Go here: https://en.wikipedia.org/wiki/Messier_object pick your favorite Messier object and/or pick one of each broad type (globular cluster, galaxy, star-forming region, planetary nebula, etc.).

Go to IRSA Finder Chart (for small objects) or IRSA Viewer (for larger objects, or objects where you want to explore images beyond what is available in Finder Chart). You may want to just start in Finder Chart and see what it looks like, then move to IRSA Viewer once you learn a little about the object(s), like size, or how different it looks in visible vs. IR.

Finder Chart has DSS, SDSS (both optical); 2MASS, WISE, Spitzer (cryo only), AKARI, IRAS (all IR). IRSA Viewer has much more data, but of the stuff that covers a large enough fraction of the sky that any given Messier object might be in it, in addition to the stuff also in Finder Chart, try SINGS (for galaxies), GLIMPSE (for galactic plane, e.g., star-forming regions), ZTF/PTF, Herschel (several versions, incl HHLI, *HPDP), MSX (for galactic plane, e.g., star-forming regions).

Skill building: play with the color stretch. Why does this matter? Why would you need to play with the stretch? What details does it bring out in any given image you’ve selected?

Science: does any given object type look the same or different in optical vs. IR? Why? Does it look the same in NIR and FIR? Why? What are the images you have loaded telling you about the spatial resolution across the wavelengths of your target?

Extension: make color images. In Finder Chart, it’s a single click that makes a 3-color image at the end of each row of images, and you don’t get to control which band is which color plane. In IRSA Viewer, you control what image is in each plane. Conventionally, red is the longest wavelength, but do what you wish. Note that IRSA Viewer will downsample images to the red plane, so if you choose an IRAS image to be the red plane, all the images will have the same enormous pixels. What is the 3-color image telling you about the science in your image? What is bright in which wavelength? What is the 3-color image you have created telling you about the spatial resolution across the wavelengths of your target? Can you create a 3-color image where the colors enhance the spatial resolution differences?

Possibly relevant IRSA videos (not all yet updated for new look):

Optional extension to coding: Visualizing images (and making 3-color images) is something that really does need to be done interactively, e.g., it’s not necessarily something that can be easily done in a lights-out “let me write code to do this” kind of way. Finder Chart does have a “batch mode” that can be used to make thumbnails or color images for hundreds of targets at once. You can also interact with IRSA’s holdings to pull FITS images (large or small) from our holdings, but then it’s on you to change the stretch and color table. At this point, I’d recommend using IRSA tools (or ds9) to work with the images interactively. In the longer term, you can use python to either just use the astropy tools to visualize the images, or invoke Firefly from your notebook.

Relevant links (I will be of little help):

Quick, help me make a CMD and don’t make me think too hard!

Make an Gaia absolute color-magnitude diagram of nearby stars, and find white dwarfs and giants among the nearby stars. Go here: https://caltech.box.com/s/uq8a92vyyq1m4bgyqo152lb2ppo88oqg Quickest path to success: download gj_gaia_culledcolumns.tbl from that link. Go here: https://irsa.ipac.caltech.edu Click on the big “IRSA Viewer” link. Drag-and-drop that gj_gaia_culledcolumns.tbl into the IRSA Viewer link (or the “upload” tab). After it loads, in the plot tab in the upper right, click on the gears to change what’s plotted. Put “bp_rp” (which is B-R in “Gaia database” parlance) on the x-axis and “gmag” (or “phot_g_mean_mag”; both are the same, it’s just the latter is what the Gaia database calls the G magnitude measurement) on the y-axis; under “chart options”, click on ‘reverse’ for the y-axis to put the bright objects at the top. “Apply.” But! You have distances from Gaia, so you can do a better job: m-M=5*log(d)-5 with distance in parsecs and distance in parsecs = 1/parallax in arcseconds. Get the plot options back by clicking on the gears (if you don’t still have that window up), and for the y-axis, use gmag-(5*log10(1000/parallax) - 5) because the parallax as retrieved from Gaia is in millarcsec. Make sure that you reverse the y-axis to put the bright objects at the top. Where are the giants? Where are the white dwarfs? Pick any object in the plot that you think is a giant or a white dwarf and click on it. It’s highlighted in the table at the bottom. Go to the microscope in the top right of the table, and pick “Go to and search Simbad at row with 5” radius” to see what this object is. Were you right, is it what you thought it was?

This catalog I had you use is a version of the “Gliese-Jareiss catalog of bright stars” which was at one point the most complete catalog of nearby stars, since ‘bright’ often also means ‘nearby’... it has since been surpassed by more complete, better catalogs of truly nearby stars, but this is sufficient for our purposes. For completeness, let me acknowledge that (a) the GJ catalog is of bright stars so some of the stars actually turn out to be rather far away (you can find them in the list!); (b) simply inverting the Gaia-provided parallax is ‘good enough’ for these purposes, but technically, you need to do lots more sophisticated things to get good distances. See 2021A&A...649A...6G for both a more recent/complete list of nearby stars AND a discussion of what that group did to get reliable distances from Gaia in this context. (Note the link under “Related materials” on the upper right of the ADS page that goes to all sorts of online data tables associated with this paper.)

Hm, actually, I’d rather you make me think harder to make a CMD…

Based on this idea: https://vmcoolwiki.ipac.caltech.edu/index.php/Gliese_Catalog_Explorations Goal : make Gaia absolute CMD and find white dwarfs and giants among the nearby stars

Find the gj.tbl file here: https://caltech.box.com/s/jrz5cxv77pivbjdxkja78r570satebuh and download it, renaming it if need be. This is the Gliese-Jareiss catalog of nearby stars. All of them should be pretty good coordinates; the positions come from https://ui.adsabs.harvard.edu/abs/2010PASP..122..885S/abstract

Aside for completeness: This catalog is a version of the “Gliese-Jareiss catalog of bright stars” which was at one point the most complete catalog of nearby stars, since ‘bright’ often also means ‘nearby’... it has since been surpassed by more complete, better catalogs of truly nearby stars, but this is sufficient for our purposes. For completeness, let me acknowledge that (a) the GJ catalog is of bright stars so some of the stars actually turn out to be rather far away (you can find them in the list!); (b) simply inverting the Gaia-provided parallax is ‘good enough’ for these purposes, but technically, you need to do lots more sophisticated things to get good distances. See 2021A&A...649A...6G for both a more recent/complete list of nearby stars AND a discussion of what that group did to get reliable distances from Gaia in this context. (Note the link under “Related materials” on the upper right of the ADS page that goes to all sorts of online data tables associated with this paper.)

Go to IRSA Catalog Search Tool, pick Gaia DR3, do a multi-object search, 1-to-1 matching with a 2 arcsecond radius, and upload this list of sources. Save the catalog to your disk and upload it back to IRSA Viewer, or make the plot entirely within the IRSA Catalog tool. (I think the IRSA Viewer interface is moderately easier to use and more powerful.)

Make a color-magnitude diagram! You can put G (phot_g_mean_mag) on the y-axis (don’t forget to reverse the y axis to put bright objects at the top) and either the field bp_rp (which is B-R) or explicitly “phot_bp_mean_mag-phot_rp_mean_mag” on the x-axis. Look at that CMD. But wait! You can do better.

Because you have matched to Gaia data, you now have distances to these stars, so you can do more than just make a color-magnitude diagram; you can make a color-absolute magnitude diagram! However, parallax is tabulated, not distance. Note that the parallax is tabulated in units of milliarcsec (mas). Because the IRSA plotting can do simple mathematical manipulations including logarithms, you can use the information there to make an absolute color-magnitude diagram. STOP HERE AND DON’T READ FURTHER UNTIL YOU HAVE AT LEAST TRIED BY YOURSELF TO FIGURE OUT HOW TO DO THIS.


(Use phot_g_mean_mag- (5*log10(1000/parallax) - 5) for the y axis, and don’t forget to reverse the y axis to put bright objects at the top, and for the x axis, use either the field bp_rp or explicitly phot_bp_mean_mag-phot_rp_mean_mag.) Look at how much better your diagram looks when you take distances into account! The scatter goes way down on the main sequence, and the giants and white dwarfs differentiate themselves much more clearly.

Science 1: Which stars are white dwarfs in your diagram? Which stars are giants in your diagram? Click on any source you think is a white dwarf in the plot. The star corresponding to the point in the plot is highlighted in the table. Once that row is highlighted, especially if your table is in IRSA Viewer, you can use the native Firefly tools to search Simbad. (Go to the binoculars and “Go to and Search Simbad at row” to spawn another window or tab at Simbad with the source in question loaded (or all the sources within 5 arc sec, sorted by distance). Were you right? Is the source you picked a white dwarf?

Extension 1: Scroll down and find the references on the object. Find the most recent paper that mentions this object. Is it a paper about Gaia observations of white dwarfs? What data are the paper using? Learn something you didn’t know about this white dwarf. Repeat this again for another white dwarf. Can you find something different about this white dwarf compared to the first white dwarf you found?

Science 2: Repeat for any red giant.

Skill building: Sort the catalog. Which is the closest/furthest star from us in the list? Where does Alpha Cen end up in the CMD? Filter the catalog. How many of the stars don’t have matches in Gaia? What happens to the fraction that is matched if you change the 1-to-1 matching radius to something larger or smaller? What are the risks of just setting the matching radius to 15 or 20 arcsec?

Challenge (MUCH bigger challenge than I thought it might be): How might you find main sequence binaries in this catalog?

Possibly relevant IRSA videos (not all yet updated for new look):

Extension to coding: I would try this in IRSA Viewer first so that you “get the idea” of what you’re trying to do. But then, once you do that, then you should be able to do this entirely within the routines you can find in astropy and pyvo. You can pull catalogs (https://pyvo.readthedocs.io/en/latest/dal/index.html#pyvo-tap), merge them by position (https://docs.astropy.org/en/stable/coordinates/matchsep.html#matching-catalogs) to create a master merged catalog, and make plots like the ones above. Again, though, I will be of minimal help. I just know generally that it’s possible within the astronomical python universe of code.

Do you have something with YSOs for me to do?

Goal: Compare color-mag and color-color diagrams for the young stars in Taurus and the Gliese-Jareiss catalog of nearby stars to see how different they are. (Hint: they are very different in many cases!)

Do the prior activity to get Gliese-Jareiss catalog matched to Gaia and loaded into one IRSA Catalog search window. Go here https://caltech.box.com/s/jntjw6wo7zqvk9e55evwb4k8865s3sfd , find the taurus.tbl file and download it. This is a catalog from Taurus (a star-forming region), already in IRSA table file format. Start another browser window, but this time, Luhman (an author from an article in the literature, I think this one if my notes are correct: https://ui.adsabs.harvard.edu/abs/2019AJ....158...54E/abstract ) has already done the catalog matching for us, and this catalog has all the Gaia, 2MASS, Spitzer, and WISE matches included. Start IRSA Viewer, click on the catalogs tab, and upload this taurus.tbl file into IRSA Viewer. It should recognize it as a tbl file and interpret all the columns correctly.

Make a color- absolute magnitude diagram for Taurus. Now the columns are named differently (sorry), so you need bmag-rmag for the x-axis and, for the y-axis, gmag - (5*log10(1000/par) - 5).

Science: why does this Taurus CMD look so different than the Gliese-Jareiss one? Why are there points below the main sequence in Taurus?

Go back to your Gliese-Jareiss browser window. Go back to the catalog search. This time, do a 2MASS point source catalog search, again a multi-object search (on the Gliese-Jareiss catalog), 1-to-1 matching, 3 arcsecond radius. Change the plot to be J-H on the y-axis (j_m-h_m) and H-K on the x-axis (h_m-k_m). Note that there are some clearly not-real data points that are large outliers in this plot when you first make it. In order to get rid of them, you will need to filter down the table to get rid of the limits. The best way to do this is to filter on j_snr>0, h_snr>0, and k_snr>0. Note that this immediately makes the plot much better behaved. Pin the plot so that you can keep it.

Go back to your Taurus browser window, and make the same JHK plot there. (jmag-hmag and hmag-kmag). In this catalog, there are no upper limits, so the plot is better behaved. Pin the plot so that you can see it next to the Gliese-Jareiss one. Does it look like the Gliese-Jareiss one? Why or why not?

Repeat this for Gliese-Jareiss and AllWISE, [W1] vs. [W1]-[W4] (w1mpro vs. w1mpro-w4mpro). And for Taurus (w1mag vs w1mag-w4mag). These plots look HUGELY different from each other (between catalogs, not just between wavelengths). Why?

Challenge: What is the deal with the things brighter than [W1]~4 in Gliese-Jareiss? Why does this plot do that?

Challenge (HARD!): Why haven’t I asked you to do this for IRAC?

Challenge: Try any other color-mag or color-color combination you want and compare Gliese-Jareiss with Taurus. Do the two look the same or different in your chosen parameter space?

Extension to coding: If you were able to do idea #2 above, then you can certainly do this one. 😊 You need to merge catalogs to get the Gliese-Jareiss catalog, but I’ve given you the entire catalog you need for Taurus, so that part is easy; you’re pretty much just making plots at this point.

Retrieved from "https://vmcoolwiki.ipac.caltech.edu/index.php?title=Exercises_with_IRSA_tools&oldid=13925"