This describes the Hubble Ultra Deep Field 2009 (HUDF09) program observations, obtained as HST program 11563 (PI: Garth Illingworth) in Cycle 17. The program uses WFC3/IR as the prime instrument for 192 orbits to image the deep ACS fields that were obtained in the original HUDF (PI: Steven Beckwith) program and in the HUDF05 (PI: Massimo Stiavelli) program. The three principal fields are:
Observations for the HUDF09 program were taken from August 2009 to February 2011 and images presented here make use of the full two-year WFC3/IR datasets over the HUDF09, HUDF09-1 and HUDF09-2 fields.
The HUDF09 WFC3/IR observations were obtained using a similar dither strategy to the original HUDF. The dither strategy for the WFC3/IR primary and ACS/WFC parallel observations were devised to include
Observations taken in visits 46, 49, 4A, 4B and 4D failed due to a Science Instrument Command & Data Handler (SIC&DH) lockup and visits 5C and 5F failed due to guide star issues. These visits were rescheduled and data were retaken in visits 4I, 4J, 4K, 4L, 4H, 5H, 5I.
View the HUDF09 MAST HST Search
|Exposure Time (s)
The final science and weight images where generated by the WFC3 data reduction pipeline WFC3RED.
The WFC3RED processing of the WFC3/IR observations in the HUDF09 fields follow a very similar procedure used in reductions of ultra-deep NICMOS observations or other deep near-IR data. The process begins with individual flt files requested from the MAST HST archive. These frames have already been subject to bias correction, dark subtraction, flat-field correction, and cosmic ray rejection. The flt files obtained from the MAST are reasonably well calibrated, given that the on-the-fly pipeline processing by STScI already takes advantage of darks and flat fields constructed from latest on-orbit calibration data.
To improve the pixel-by-pixel S/N (and correct for any imperfections in the flats or darks), we median stacked all of the observations in the F105W, F125W, and F160W bands to create supermedian frames — which were subsequently subtracted from the individual images. Compact, bright point-like sources are then identified in each WFC3/IR image and in the ACS data covering the HUDF09 fields. These source lists were used to align the WFC3/IR images with the ACS data using WFC3RED module superalign.
The WFC3/IR observations were then drizzled onto the appropriate ACS reference image frame using the MultiDrizzle software while clipping 4σ outliers. As a result of this clipping procedure, small artifacts in individual exposures — such as hot pixels — will not significantly affect our reductions. We also mask out pixels on the WFC3/IR camera that were affected by source persistence. This masking was performed by remapping our initial reductions of the data (or more precisely blotting median stacks of the data) back to the original frames, subtracting them from the original frames, coadding these subtracted frames for all exposures within a visit, smoothing, and then flagging all pixels above a 3σ threshold. This together with our dither strategy and CR clipping procedure ensures that our final reductions are not significantly affected by source persistence.
In the drizzling process not only were latest distortion solutions (as of 10/11/2010) utilized, but each source in the WFC3/IR reductions was cross-correlated with the corresponding source in the v1.0 HUDF ACS F850LP band observations. Small corrections to the distortion solution were required to obtain robust registration (i.e., RMS differences of <0.01″) for the F105W band data (no WFC3/IR distortion solution has been explicitly derived for this filter by STScI). Also to further refine the registration of images, for each visit we used MultiDrizzle to create image stacks for each visit and then cross-correlated the sources in each image stack with the HUDF and HUDF05 ACS F850LP-band images. Minor adjustments were made to the coordinates of each visit to obtain a more optimal alignment.
To further optimize the pixel-by-pixel S/N in our reductions, we repeated our median stacks of the exposures in each band — but now masking out the sources apparent in our final stacks (rather than just those apparent in the individual exposures). We then subtracted these supermedian images off of each of the individual exposures and drizzled the data together to generate our final reductions.
The data are organized into sets of images by HUDF09 pointing and by passband (WFC3/IR F105W, F125W & F160W). Each image is approximately 3k x 3k pixels in size and a scale of 0.06 arcsec/pixel. All three pointings reside in the GOODS/Chandra South field and for each pointing we provide the drizzled science image and a weight image.
The files name contains information relative to the pointing, the passband (F105W, F125W or F160W), the data release version (v1.0 in this case) and the image type (science image or a weight map). For example, the FITS file containing the F160W science data for the HUDF09-2 pointing is called:
These data may also be downloaded with anonymous ftp (ftp archive.stsci.edu then cd /pub/hlsp/hudf09). The data are viewable from a browser (and therefore wget or curl) at http://archive.stsci.edu/pub/hlsp/hudf09.
The data may be displayed with a HLA-like display.
The pixel values of the science images report the flux count rate calibrated in electron/second. The zero points to convert the count rate into an AB or Vega magnitudes for the three HUDF09 WFC3/IR passbands are the following:
|Zero Point ABMAG
|Zero Point VEGAMAG
The weight map image is generated by MultiDrizzle and is equal to the inverse variance (i.e., 1/rms^2) per pixel. A detailed discussion of weight map conventions and noise correlation in drizzling, can be found in Casertano et al. 2000, AJ, 120, 2747, especially Section 3.5 and Appendix A.
If you use this data in work that you publish, we’d appreciate it if you could acknowledge the fact by including the following citation:
Last update: 2011-04-04