Photometry (doing it)

From CoolWiki
Jump to navigationJump to search

If you've done photometry before and expect to do it the same way on Spitzer or Herschel images, BE CAREFUL because IT WILL NOT WORK. The units of Spitzer/Herschel images can be tricky.

Most coherent, developed, tested materials

  • APT, the Aperture Photometry Tool, was developed by Dr. Russ Laher in response to very early NITARP team needs (when it was still the Spitzer program)! It has since taken on a life of its own. Here are some NITARP-developed resources on APT:

Other important CoolWiki pages

Somewhat less coherent (or less standalone) materials

  • (this is cross-posted in 'concept' and 'skill' for Photometry) In 2019, an alumni team (Olivia Kuper and Tom Rutherford) were trying to carry our original photometry on Herschel images further. We had sources clumped close together on the sky, and even though aperture photometry was ok for our original project, they were trying to learn how to do PSF photometry in Python. They were struggling with vast array of free parameters. I wrote this Photometry Scaffolding document in an effort to help them understand, in words, what the free parameters were and why they mattered. This may or may not be useful for anyone else, but as I put a lot of time into it, I thought I'd post it here. -- Dr. Luisa Rebull, 2019

Other sources of interest

  • The IRAF manuals are old but still very, very valid. If you can pay attention to the concepts and ignore the IRAF-specific instructions, you can learn an awful lot from these manuals. Oldies but goodies.
  • "Magnitudes, distance moduli, bolometric corrections, and so much more" by David W. Hogg
  • Software you can use to do photometry:
    • Aperture Photometry Tool (APT) - just does aperture photometry, obvs. Developed specifically for NITARP!
    • MOPEX - is the tool developed for Spitzer data. It's free and not too horrible to install. It depends on having a bunch of keywords in the FITS header. Getting it to work on non-Spitzer data is theoretically possible, but I've never managed. MOPEX lets you tell it, hey, I know I have 27 sources of about this brightness at these positions; go force a fit at those locations with those fluxes as initial guesses and tell me what you get, and give me back the image where all the point sources have been subtracted off so that i can check on how well it fit the sources as distinct from each other and the nebulosity. This is incredibly helpful for testing your photometry.
    • HIPE - is what we're supposed to use on Herschel data. It grabs data from the archive directly, and it knows natively how to compensate for scan directions. This is free software, but it is NOT practical for a NITARP team because: (A) it took me, on my desktop with a Caltech-quality network connection, more than a day to install it. (B) it is NOT easy to use. (C) it needs to work on the whole set of images from a given observation.
    • IDL - this is what I mostly program in, and it costs a LOT of money to buy a license, so also not practical for NITARP. It's been used by the astronomy community since the late 1980s. there is a well-tested suite of astro-related IDL software: Note it has a DAOPhot section. I
    • starfinder and xstarfinder - - works in IDL.
    • IRAF - depends on DAOPhot. Has routines from there incorporated into it. But IRAF is no longer supported! See Pyraf below.
    • DAOphot - really old workhorse. Website seems last updated in early 2000s. the core algorithms are still used in many places.
    • SExtractor - and yes it has an unfortunate name (would rather not work with HS students using this software). This is largely used by the extragalactic community.
    • astropy because most of the community is shifting/has shifted to python, the astropy library has lots of code and is updated all the time, all free. IRAF was migrated to Pyraf. If you are looking to learn something, this is likely your best, most flexible bet.
    • tractor - - a suite of code in python that was described as modeling (meaning PSF fitting) across multiple bands at once.
    • There are other smaller-scale PSF fitting routines that i know of, but they're all intimately tied to the instrument for which they're designed. Like, for Chandra, the images are really a table of photon energies and arrival times; it renders in ds9 like a regular image, but it's not. The one for Kepler/K2 is intimately tied to the very large pixels it has and the super complicated instrumental effects they have to cope with on that telescope.