46.8 PSFs
When dealing with the spherical aberration, point spread function (PSF) images are often an important element in the deconvolution of WF/PC-1 data. Characteristics of the WF/PC-1 PSF include:
- A sharp, 0.1 arcsecond radius core which contains about 15% of the energy.
- Broad, extended wings out to about 4".
- Strong position dependance due to camera vignetting.
- Wavelength dependance due to diffraction effects.
- Time variability due to dependance on the telescope's focus position.
- Shutter diffraction effects for exposures of less than 1/2 second.
- OTA focus variations and image motion on short time scale (< 1 hour).
- Spacecraft jitter.
Ideally, PSFs should be observed in a variety of positions on each chip, in all filters used and close in time to the observation, a time-consuming process. In addition, the shorter exposures needed to avoid saturating the PSF cores generally provide insufficient S/N in the wings. For these reasons, model PSFs will generally be needed to supplement the available observed PSFs. The Tiny Tim software package, which can be used to generate model PSFs, is available at:
http://scivax.stsci.edu/~krist/tinytim.html
Observed PSF Library
Deconvolution of WF/PC-1 observations is supported through a library of observed WF/PC-1 point spread function (PSF) images that can be found in the STScI Calibration DataBase (CDBS). Rather than storing entire WF/PC-1 datasets that already reside in the HST Archive, the library consists of smaller (typically 256 x 256) sub-sections, usually centered on the PSF star. Any PSF can be retrieved from StarView in the same manner as reference files (the file type is CDB). A complete PSF image name consists of the PSF rootname as provided in DATA_FILE column of the tables in the WWW memo plus the suffixes .r7h for the ASCII header and .r7d for the binary data file. The memo listing all available observed PSFs in the library is regularly updated on WWW.
Header keywords from a typical PSF image are described in Table 46.3. The secondary mirror actuator position keywords were populated using the results presented in the OTA Instrument Science Report #7.
| Keyword | Sample Value | Definition |
|---|---|---|
| Keywords Listed in WWW Memo | ||
| filtnam1 | F785LP | First STSCI filter name |
| camera | PC | WF or PC |
| detector | 6 | 1-4 for WF, 5-8 for PC |
| rootname | W0MN0A02T | Rootname of observation with PSF image |
| data_fil | C3U1428IW.R7H | Name of the PSF image file |
| targname | SAO107200 | Target name |
| spectral | unknown | Spectral type of source, if known |
| exptime | 0.23 | Exposure time in seconds |
| date_obs | 23/06/91 | UT date |
| xcorner | 502 | X pixel of (1,1) corner in PSF image |
| ycorner | 38 | Y pixel of (1,1) corner in PSF image |
| xcenter | 630 | X coordinate of PSF center on the chip |
| ycenter | 166 | Y coordinate of PSF center on the chip |
| Keywords in Headers, but not Listed in WWW Memo | ||
| filtnam2 | F122M | Second STScI filter name |
| mode | FULL | Full or area |
| origin | HST | Data source, e.g., HST, TIM or other |
| mjd | 48430. | Modified Julian date (JD-2400000.5) |
| calibrat | T | Was image calibrated (see Chapter 45)? |
| flatfile | c1916444w | Name of flatfield image (or INDEF) |
| psfscale | 1. | Divisor used to normalize PSF (1 if none) |
| obsmode | PC,6,F,DN,F785LP | Observation mode (for synphot) |
| refspec | CRCALSPEC: AGK_81D266_002.TAB | Reference spectrum |
| actuat25 | -1492 |
Position of secondary mirror actuator 25 (and actuator keywords 26 through 30) |
PSF Limitations and Effect of Jitter
Due to the limited dynamic range of the WF/PC-1, observed PSFs that are properly exposed in the core will have low signal-to-noise in the wings and will not reflect the true PSF. In addition, exposures shorter than 0.17 second (OV/SV Report, Chapter 9, Faber et al., 1992) suffer from diffraction effects which are caused by the shutter obscuring some of the light beam during flight time.
Except for very short exposures, both the observed and modeled PSFs will not perfectly match WF/PC-1 images due to the jitter in the HST pointing. The jitter is primarily caused by the motion of the telescope in response to the solar array oscillations, and possibly oscillations of the aperture door and the high gain antennas as well (SESD Report 92-15). The effect is greatest during earth terminator crossings, but is present at other times as well. On April 12, 1992, software called SAGA II was installed to correct for some of the jitter, effectively reducing it by about 50%. That is, the rms jitter dropped from 17 mas to 7 mas entering orbital day and fell from 12 mas to 6 mas entering orbital night.
The effects of the remaining 5-10 mas jitter, while not predictable, can be partially determined for observations taken in Fine Lock, via some of the associated engineering data. (Note that this is engineering data which is not part of the standard WF/PC-1 image set; users must submit a special request to help@stsci.edu to obtain more information on a particular observation from the Observation Monitoring System1.) For images obtained on Coarse Track, the effects of jitter can only be estimated.
1This information became part of the standard pipeline products delivered with science data after June 1994.
Copyright © 1997, Association of Universities for Research in Astronomy. All rights reserved.