Research:Angst
From Astronomy Facility Wiki
This page is dedicated to keeping track of the photometric data processing for the ACS Nearby Galaxy Survey Team, led at the UW by Julianne Dalcanton. The entries should help to keep track of issues and solutions that are studied throughout the process of creating the best photometry catalogs possible.
Contents |
Dolphot Problems and Solutions
Problem 1: FLT VS CRJ IMAGES
Dolphot CMDs are deeper when inputing crj files instead of the individual flt files that were used to create the crj files.
Solution: The Readnoise values in the headers of the crj images after running splitgroups are the same as those in the flt headers, so the noise level is wrong in the S/N calculations in dolphot. Since the flt readnoise values are correct, the S/N values output by dolphot when using crj images are systematically high!
Problem 2: 3 FILTER ARTIFICIAL STARS
Dolphot's acsfakelist task only accepts a range of magnitudes and colors in 2 filters, but many angst fields have B,V, and I data.
Solution: If one feeds the task the .phot output with 3 filters of data, but designates the 2 deepest bands (V and I for the Angst data) as for defining the color and magnitude ranges, acsfakelist will automatically generate the numbers for the third band, giving the stars appropriate colors in both B-V and V-I.
Problem 3: IMAGE ALIGNMENT
acsfitdistort is a pain to use, as it requires matching stars in flt and drz images by hand in order to get an approximation of the shift from the central pixel of the reference image to the location of the central pixel of each chip of the flt image on the reference image.
Solution: The world coordinate system of the flt image is not very accurate in the corners of the chips because of the ACS distortion, but near the center it is accurate. Therefore we can grab the RA and DEC of the center of each chip in the flt image, find the pixel coordinates of that RA and DEC in the reference image, subtract from the pixel coordinates of the center of the reference image, and input those as the X and Y shift estimates for dolphot. they aren't perfect, but they get close enough that the dolphot fine tuning can determine the correct shift to within 0.01 pixels.
Problem 4: ARTIFICIAL STAR PLACEMENT
Dolphot's artificial star lists include the corners of the reference image, which are not actually covered by the ACS chips.
Solution: We are always going to grab subsets of artificial stars based on their background surface brightness, so the stars in these areas will never be used because their background surface brightness will be undefined!
Useful Theoretical Models -- text from Leo Girardi's Nov 29 email
Web tools
In http://stev.oapd.inaf.it/~lgirardi/cgi-bin/cmd_1.0 there's a web interface for ordering isochrones for arbitrary age-Z values, including ACS and WFPC2 filters.
In http://trilegal.ster.kuleuven.be/cgi-bin/trilegal, you can simulate the Milky Way foreground (this is not that interesting now, but could become later when we have data for many fields with sizeable numbers of faint foreground stars.)
Transformations to HST filters
What you find in the web pages are still 3-year old tables of transformations, that could be updated using latest spectral libraries and revised filter throughputs. Lep Girardi is collecting the latest spectral libraries. Currently there are WFPC2 curves provided by Jon Holtzman in 2000, and ACS curves by Marco Sirianni in 2003. What is needed are throughputs including filters+camera+CCD, and separately for each chip if available. A resolution of Delta(lambda) ~ 10 angstrom is good enough.
Stellar models
Leo Girardi and Paola Marigo are now fixing the AGB part of the isochrones, to obtain more realistic AGB numbers and RGB/AGB ratios (see e.g. latest Ben Williams' astro-ph on Virgo intracluster CMDs for a description of the problem in old pops.), and the right numbers/colours/magnitudes for C stars. There are some relatively-minor details pending - regarding mostly the bolometric corrections for C stars, and stars with high mass-loss at all ages - but there's a good hope it will be ready for the team meeting in January.
