|
 |
On Tuesday, March 9 at 10am EST, the UDF Team has released the reduced, calibrated, stacked and mosaiced images acquired with NICMOS in parallel to the main ACS field. The released images have been fully processed using the best calibration and reference files available at the time of release. More general information on the NICMOS Parallel UDF fields may be found on the parallels site of the UDF program.
The release consists of the complete, multi-epoch stacked mosaics of the UDF NICMOS Parallel data, for the two passbands and for the two fields covered during the ACS observations of the main field (made with two different roll angles of the telescope).
Other data products include the weight map
and a photometric catalog of the sources.
The data set is composed by a set of 4 images, i.e. 2 passband (F110W and F160W) images for each NICMOS Parallel field (hereafter referred to as Field 1 and Field 2).
To optimize PSF sampling, the original, pipeline processed NICMOS images have been drizzled from the original scale of 0.2 arcsec per pixel onto a larger virtual grid with scale ~0.09 arcsec/pixel.
The file naming convention contain information on the program ('udf'), the instrument ('nic'), the fact that it is a parallel observation ('p'), the passband ("f110w" or "f160w"), the field (whether "f1" for Field 1 or "f2" for Field 2), and the image type (whether it is a science image 'img' or a weight map 'wht'). For example, the FITS file containing the final drizzle combined F110W science data for Field 1 is called
h_udfnicpf110wf1_img.fits
where the prefix "h" is adopted to comply with the general naming rules of the STScI data archive. The file containing the corresponding weight map is called
h_udfnicpf110wf1_wht.fits.
The pixel values of the science images report the flux count rate calibrated in DN/second. This complies with the standard output units for calibrated NICMOS data. The zero points needed to convert the count rate into AB magnitudes for the two passbands are:
Z0_F110W = 23.4033
Z0_F160W = 23.2146and can be derived from the following equations (See also the NICMOS Data Handbook ):
Zpt (AB) = -2.5 log (PHOTFNU × Count Rate × 10-23) - 48.6
Zpt (AB) = -2.5 log (PHOTFNU × Count Rate) + 8.9by putting Count Rate = 1DN/second. The value of PHOTFNU (Jy s DN-1) is given in the fits header. Multiplying your image by the PHOTFNU value yields fluxes in Jansky. An approximate Vega normalized flux may be computed using the equation
m = ZP(Vega) - 2.5 log (PHOTFNU × Count Rate x fnu(vega)-1)where ZP(Vega) is the magnitude of Vega, around 0.0 (depending on the photometric system). To obtain fluxes in units of erg cm-2 s-1 A-1, simply multiply the images by the value of PHOTFLAM (erg cm-2 A-1) values in their headers. More details of the photometric calibration for the NICMOS cameras, as well as other information on NICMOS photometric properties can be found in the NICMOS Data Handbook .
The weight maps are images produced as part of the data reduction process, and give a measure of the background + instrumental nominal noise per unit area (pixel) in the science data. Because the NICMOS images are mosaics consisting of different numbers of overlapping images , the total exposure time varies as a function of position. In addition, pixels may have been masked for a variety of reasons, including rejection of occasional cosmic rays or of persistently bad pixels, rejection of diffraction spikes and spurious illumination resulting from bright stars falling on the edge of the detector, etc. Also, because of the nature of the parallel program, the depth of the field depends on the number of images at that particular pointing. Finally, the sky background showed some variation from exposure to exposure, and even within single exposures during the Field1 observations. All of these effects, directly or through the data processing needed to mitigate them, contribute to variable "depth" across the image mosaics.
A noise model has been used to calculate the expected noise per pixel at the background level, resulting from the combination of sky background (modulated by the flat field), readout noise, dark current, and amplifier glow. The effects of Poisson shot noise due to signal from objects in the image have not been included in the noise model. The noise model has been used to build weight maps equal to the expected inverse variance per pixel. The weight maps have been combined with masks that excluded (i.e., set to zero weight) pixels for all various reasons outlined above, and then used to weight the combination of images in the drizzling process.
The resulting output weight map should be equal to the expected inverse variance (i.e., 1/RMS^2) per pixel. The interpolations introduced by drizzling the images (shifting, rotating, correcting distortion, and sub-sampling pixels onto a finer grid) result in correlation between pixels in the drizzled science images. Therefore, the apparent RMS background noise that one measures in the image is smaller than that given by the (inverse) weight maps, because the apparent RMS is suppressed by the effects of correlation. The weight maps are normalized to show the expected noise per pixel that the images would have in the absence of correlation. Or, put in another way, the sum of the variances (inverse weight values) over some aperture larger than the correlation scale (a few pixels) should accurately reflect the measurement uncertainty due to the background + instrument noise. (We note again that no attempt is made to include Poisson uncertainty due to signal from objects.) For a more detailed discussion of weight map conventions and noise correlation in drizzling, please see Casertano et al. 2000, AJ, 120, 2747, especially Section 3.5 and Appendix A.
The scaling of the weight maps has been validated by
comparing their values to
the measured image noise, after a correction for the measured
autocorrelation
of background pixels.
2.5 File Size and Data Set Size
The files for each field, both the science and the weight map images, are listed below along with the file size, image size, exposure time and the GEMS tile which overlaps the observations. Note that the actual observed area is smaller than the image size of the fits files.
|
|
|
|
|
GEMS Tile |
| h_udfnicpf110wf1_img.fits | 6.8MB | 1400 x 1200 |
|
Tiles 49/42 |
| h_udfnicpf110wf1_wht.fits | 6.8MB | 1400 x 1200 | ||
| h_udfnicpf160wf1_img.fits | 6.8MB | 1400 x 1200 |
|
Tiles 49/42 |
| h_udfnicpf160wf1_wht.fits | 6.8MB | 1400 x 1200 | ||
|
|
||||
| h_udfnicpf110wf2_img.fits | 7.1MB | 1400 x 1300 |
|
Tiles 25/19 |
| h_udfnicpf110wf2_wht.fits | 7.1MB | 1400 x 1300 | ||
| h_udfnicpf160wf2_img.fits | 7.1MB | 1400 x 1300 |
|
Tiles 25/19 |
| h_udfnicpf160wf2_wht.fits | 7.1MB | 1400 x 1300 | ||
3.0 Data Reduction and Calibration
Raw data have been processed using the standard NICMOS pipeline
(CALNICA), which provides the basic reduction steps of dark and bias
subtraction, linearity correction and flat-fielding. It also provides
data quality files
that flag known hot pixels, bad-columns and other cosmetic defects.
Since many of the images were taken shortly after an SAA passage, and
both fields contained bright objects, the raw images
were also processed to remove residual signal from pixels and columns
containing
ghost signal. This is known as "Mr. Staypuft" in NICMOS images and is
due to the
pull down of the power supply after reading a large value before it's
asked to
read another (in another quadrant for example). CALNICA processing can
continue as normal after
this signal has been removed. The final reduced
images consist of the individual exposures, or "dithers", taken in
each band for each NICMOS pointing, where an exposure is defined as the
final combined
multiaccum sequence.
The dark reference frame is the combination of darks acquired through the NICMOS UDF Treasury observations (proposal ID 9803, P.I. Thompson). These darks are close in detector temperature to our parallel field observations and represent a more accurate measure of the actual dark current than the currently available NICMOS temperature-dependent darks which are generated by the calibration pipeline.
Standard NICMOS flat fields were used to calibrate each image. The file used are referenced in the fits headers, and can be found through the NICMOS Instrument web site, or retrieved through Starview.
An extremely variable and bright sky was observed during Field 1 observations. In order to remove the extra signal from the images, a sky image was created using object frames from the proposal 9803 observations. The pointing in this case was close enough to provide a good estimate of the zodiacal and local background, while the observing strategy was very similar except for the dithering pattern, large enough to remove unwanted objects from the frame. This sky was then normalized, scaled and subtracted from each individual read in the affected images. Since the sky was variable on a time-scale shorter than the average exposure length of a single integration ramp (image set), a special procedure was developed to exploit the non-destructive readout mode used by NICMOS. This ad-hoc sky processing was not used for Field 2, since in that case the sky remained constant within each ramp.
4.1 Geometrical Distortion Coefficients
The geometric distortion applied to the individual input images uses a value for the x/y scale ratio which has been updated to best reflect the most recent plate-scale measurements for post-NCS NICMOS data (Cycle 11 and beyond). The full distortion solution is currently being re-evaluated and appears to show less than ~ 1% change from the Cycle 7 values. A future NICMOS Instrument Science Report will be released detailing the work done on recalculating the NICMOS distortion coefficients, however the scale ratio is the most important term in the correction and any change to the other coefficients is expected to have minimal impact on the solution. The values below may be copied into a new drizzle coefficients file to replace the drizzle$coeffs/nic-3 file which is referenced in the drizzle package when drizzling NICMOS images:
# cycle7 coeffs but with x/y scale ratio
adjusted
0.0 1.0014705 0.0 8.0317971e-06 1.3219373e-05
5.8285553e-06 0.0 0.0 0.0 0.0
0.0 -0.00089368516 0.99853067 -1.8073393e-05 5.9911861e-07
-1.1582927e-05 0.0 0.0 0.0 0.0
Drizzling has been done in two phases. In the first phase, images of the same tile taken in the same filters are identified, sky subtracted and drizzled onto a common pixel grid with the same scale as the input images (0.2 arcsec/pixel). Cosmic rays and deviant pixels are identified during this process and flagged in mask files specifically created for this purpose. Information included in post-pipeline masks (which have also been drizzled onto the same grid) is included in the new masks at this time.
During the second phase, the images and the mask files are blotted back to the original positions, drizzled again onto a common astrometric grid with scale 0.09 arcsec/pixel, and stacked together. During this process corrections for the NICMOS geometrical distortion are applied, cosmic rays flagged during the previous processing block are masked out from the stack, and additional, low-level cosmic rays and defects are identified and masked, too.
Initially a set of shallow, single-exposure-depth mosaics are created which are then combined to create a clean "median" mosaic, using signal-to-noise thresholds to reject cosmic rays and any remaining bad pixels. This technique is extremely robust at producing a clean median mosaic, with the primary requirement being accurate astrometric alignment between epochs. The relevant pieces of this median mosaic are then transformed back to the frame of each of the original input files using blot.
Next, the standard dither package tasks of "deriv" and "driz_cr" are used to compare this blotted image and its derivative image with the original input file, and generate a cosmic ray mask based on the comparison. Finally, all the files, together with their newly created cosmic ray masks, are drizzled onto a single output mosaic, which has units of count-rate in DN/second in each pixel.
An advantage of NICMOS over other HST instruments is that data are taken in MULTIACCUM sampling mode, where the image can be read out non-destuctively over the course of the full exposure. In this way, cosmic rays can easily be identified and removed inside the CALNICA pipeline reduction process,allowing for cleaner calibrated images. It is possible to keep track of which pixels were affected by cosmic rays by sampling the, all TIME array image in the final _cal.fits CALNICA product.
A source
catalog has been produced for each NICMOS field. The file names are
h_udfnicp_f1.cat and h_udfnicp_f2.cat, and follow the same naming
convention adopted for the images. The catalogs have been compiled
using SExtractor.
Sources have been identified on the F160W images and individually
inspected to reject spurious detections, mostly at the edge of the imaged
fields. The basic parameters used by SExtractor are similar to those used
for the HDF-S
NICMOS observations. With the exception of the magnitudes and SNR, the
source parameters reported in the catalog files are relative to the F160W
filter.
The following table lists the parameters listed in the two catalog files. The semi-major and semi-minor axis of the source ellipse are indicated with a and b, respectively.
Parameter units comment Source ID x pixel baricenter y pixel baricenter RA (J2000) deg DEC (J2000) deg Theta deg P.A. betwen a and RA (CCW) Ellipticity 1-b/a R50 pixel 50% flux radius FWHM pixel Stellarity SExctractor CLASS_STAR (0: galaxy -> 1: star) mag[J] AB mag isophotal Dmag[J] AB mag isophotal SNR from isophotal flux mag[H] AB mag isophotal Dmag[H] AB mag isophotal SNR from isophotal flux
The NICMOS arrays are especially susceptible to orbits which pass through the South Atlantic Anomaly (SAA). Although the detectors are turned off during each passage, cosmic ray persistence can still have a non negligible affect on data taken up to 80 minutes after a SAA passage (this is variable and depends on the strength and duration of the SAA contour). A post-SAA dark observing program has been implemented in the post-NCS NICMOS era to help alleviate these affects. After each passage through an SAA, ACCUM darks are taken and stored in the archive. They can be used to create a residual image of the cosmic ray impacts accumulated during the passage. The residual image can then be used to remove some of the noise in the following science images which were impacted. For more details on the SAA and it's affect on NICMOS data, see NICMOS-ISR-2003-10. About half of the observations taken during the NICMOS parallel observations were affected by the SAA. These datasets were cleaned using the process described in the ISR mentioned above, before final inclusion in the drizzled mosaic.
The nature of the parallel program for the NICMOS observations created large discrepancies in the exposure depth across the field (this applies to both Field 1 and Field 2). Each field is made up of basically 2 pointings (each of which contain small sub-dithers) which have vastly different cumulative exposure times, approximately by a factor of 6. The center of the drizzled mosaic always achieves the longest exposure time. The exposure times listed in the headers of the final drizzled mosaics represent the largest possible exposure time in the image, i.e. the exposure time achieved by a pixel at the center of the field which had never been rejected in any of the samples that were drizzled together. The sensitivity (SNR) across the image is strongly non-uniform, and this should be kept in mind when doing relative photometry across the field.