Chapter 5:
ImagingIn this chapter . . .
5.2 Important Considerations for ACS Imaging / 64
5.3 Wide Field Optical CCD Imaging / 68
5.4 High-Resolution Optical and UV Imaging / 70
5.5 Ultraviolet Imaging with the SBC / 72
5.1 Imaging Overview
ACS can be used to obtain images through a variety of optical and ultraviolet filters. When the selected ACS camera is the WFC or the HRC, the appropriate filter in one of the two filter wheels is rotated into position, and a clear aperture is automatically selected on the other filter wheel. For SBC imaging the single filter wheel is rotated to the required position. A number of apertures are defined for each ACS camera. In general, these refer to different target positions on the detector.
Table 5.1, Table 5.2, and Table 5.3 provide a complete summary of the filters available for imaging with each detector. Figures 5.1 through 5.6 show the filter transmission curves, and Figure 5.7 shows the integrated system throughputs.
The CCD filter wheels contain filters of two different sizes. Some filters (F435W, F475W, F502N, F550M, F555W, F606W, F625W, F658N, F660N, F775W, F814W, F850LP, and, G800L) are full-sized filters that can be used with both WFC and HRC. Others (F220W, F250W, F330W, F344N, F892N, POL0UV, POL60UV, POL120UV, POL0V, POL60V, POL120V, and PR200L) are smaller, giving a full unvignetted field of view when used with the HRC, but a vignetted field of view of only 72 ¥ 72 arcseconds when used with the WFC. Use of the small UV filters is not supported with the WFC due to the unpredictable behavior of the silver coatings shortward of 4000 Å.
The ramp filters are designed to allow narrow or medium band imaging centered at an arbitrary wavelength. Each ramp filter is divided into three segments, of which only the middle segment may be used with the HRC. More information is available at Section 7.7.2.
Note that although the CLEAR filters are specified in the filter wheel tables, users do not need to specify these filters in their HST proposals; they are added automatically in conjunction with the desired filter in the complementary wheel. In the SBC filter wheel, every third slot (#1, 4, 7, 10) is blocked off, so that in the case of a bright object limit violation, it is only necessary to rotate the filter wheel to an adjacent slot to block the incoming light.
With either the WFC or HRC, it is possible to select a filterless observation by specifying CLEAR (this is an “available-but-unsupported” filter) as the filter name, although the image will be of degraded quality. Rough wavelengths and widths, when used with the WFC or HRC, are listed in Table 5.1 under CLEAR entries. Use of CLEAR will provide slightly degraded PSFs with the HRC and seriously degraded PSFs for the WFC. More details on PSFs with use of CLEAR are provided in ACS ISR 2003-03. Applications are expected to be rare, but a valid use could be astrometry of extremely faint targets with the HRC when color information is not required.
Table 5.1: ACS WFC/HRC filters in Filter Wheel #1.[/WFC] indicates that polarizer filters, designed for the HRC field of view, induces vignetting when used with the WFC, producing a 72 by 72 arcsecond field of view.Table 5.2: ACS WFC/HRC filters in Filter Wheel #2.[/WFC] indicates that polarizer filters, designed for the HRC field of view, induces vignetting when used with the WFC, producing a 72 by 72 arcsecond field of view.Table 5.3: ACS SBC filter complement.Figure 5.1: ACS broad-band filters.Figure 5.2: ACS SDSS filters.Figure 5.3: ACS UV and medium-band filters.Figure 5.4: ACS narrow-band filters.Figure 5.5: ACS SBC filters.Figure 5.6: Comparison between the ACS and WFPC2 ramp filters.The crosses and open circles are for the ACS narrow and medium band ramps. The open squares are for the 4 WFPC2 ramps. For each of the ACS ramps the peak throughput that was calculated for eleven central wavelength values is plotted. For the WFPC2 ramps, the peak throughput calculated every 100 Å within the field of view of any of the 4 chips, and a 0 filter rotation angle (as mapped in Figures 3.4 and 3.5 of the WFPC2 Instrument Handbook, version 3.0), are plotted.5.2 Important Considerations for ACS Imaging
There are a few characteristics of ACS that should be taken into account when imaging with ACS:
- The HRC and WFC filters are housed in two filter wheels shared by both cameras. As a consequence, when a filter is chosen for the primary camera, the filter used in the parallel camera is not independently selectable.
- The ACS cameras are designed to be used with a single filter, and for this reason unfiltered imaging or imaging through two filters leads to significantly degraded imaging quality (particularly in the WFC) and is not normally used except for polarization observations, or bright target acquisitions with the HRC. The polarizer filters were designed with a zero optical thickness so that they can and should be used with another filter.
- The geometric distortion of the WFC is significant and causes the projected pixel area to vary by ±9% over the field of view. This distortion affects both the photometric accuracy and the astrometric precision, and must be accounted for when the required accuracy is better than 10%.
- The ratio of in-band vs. out-of-band transmission for the ACS CCD UV filters is similar to that of WFPC2, once the two detector QE curves are taken into account (the red leak on ACS F330W is very small). This implies that the effect of filter red leaks needs to be calibrated for UV imaging of intrinsically red objects.
- The cosmic ray fluxes for HRC and WFC are comparable, respectively, to those of the STIS CCD and WFPC2. As with these instruments, typical imaging observations will need to be split or dithered for cosmic ray rejection.
- Section 4.3.5 provides further details and a recommendation that separate exposures with small dithers be used as a means of helping to remove residual hot pixels.
- The large format of the WFC requires significantly more shifts to read out data compared to STIS or WFPC2. Therefore, the impact of decreasing charge transfer efficiency will be encountered earlier. Section 4.3.7 details current expectations.
- The default GAIN setting for WFC primary observations is GAIN=2. This allows for good sampling of the readout noise and allows one to reach the full well counts of WFC. For the HRC primary observations, the default gain is GAIN=2. To sample the detector full well depth, GAIN=4 is needed, but this results in modest undersampling of the readout noise. For HRC ACQ data, the default setting is GAIN=4. Users may select the GAIN they wish to use for their ACS observations by using the GAIN optional parameter in their Phase II proposal. However, not all GAIN settings are supported (see Section 7.6).
- At wavelengths longward of ~8000 Å, internal scattering in the HRC CCD produces an extended PSF halo. This should affect only a small number of observations since the WFC camera is normally preferred at these wavelengths. The WFC CCDs include a front-side metallization that eliminates the large angle, long wavelength halo problem for l < 9000 Å. (For observations of red targets with the F850LP refer to Section 9.3.2).
- The ACS filter complement is not as rich as that in WFPC2. In particular, the Strömgren filter set and several narrow band filters available in WFPC2 (F375N, F390N, F437N, F469N, F487N, F588N, F631N, F656N, F673N, F953N) are not available on ACS. In general, these filters were not heavily used by the GO community. For most applications they can be replaced with the ACS medium and narrow ramps.
5.2.1 Optical Performance
Testing of the WFC and HRC cameras, following fine alignment activities on-orbit, has shown that the optical quality objectives of the cameras are met. The encircled energy values obtained from observations made in SMOV are given in Table 5.4.
Table 5.4: Encircled energy measurements for the ACS channels.5.2.2 CCD Throughput Comparison
Figure 5.7 shows the throughput of the two unfiltered ACS CCD cameras: WFC and HRC. Superposed on this plot are unfiltered WFPC2 (WF4) and the clear STIS throughputs.
5.2.3 Limiting Magnitudes
Table 5.5 contains Johnson-Cousins V magnitudes for unreddened O5 V, A0 V, and G2 V stars, generated using the Exposure Time Calculator. WFC and HRC values used the parameters CR-SPLIT=2, GAIN=2, and a 0.2 arcsecond circular aperture. For the SBC, a 0.5 arcsecond circular aperture was used. An average sky background was used in these examples. However, limiting magnitudes are sensitive to the background levels; for instance, the magnitude of an A0 V in the WFC using the F606W filter changes by ±0.4 magnitudes at the background extremes.
Figure 5.7: ACS CCD system throughputs + OTA versus those of STIS and WFPC2.Table 5.5: V detection limits for ACS, HRC, and SBC direct imaging.5.2.4 Signal-To-Noise Ratios
In Chapter 10, we present, for each imaging mode, plots of exposure time versus magnitude to achieve a desired signal-to-noise ratio. These plots, which are referenced in the individual imaging-mode sections below, are useful for getting an idea of the exposure time you need to accomplish your scientific objectives. More accurate estimates require the use of the ACS Exposure Time Calculator (http://www.stsci.edu/hst/acs/software).
5.2.5 Saturation
Both CCD and SBC imaging observations are subject to saturation at high total accumulated counts per pixel. For the CCDs, this is due either to the depth of the full well or to the 16 bit data format. For the SBC, this is due to the 16 bit format of the buffer memory (see Section 4.3.1 and Section 4.5.2).
5.3 Wide Field Optical CCD Imaging
The Wide Field Channel of ACS was designed primarily for high throughput observations at visible wavelengths. The use of protected silver mirror coatings, the small number of reflections, and the use of a red sensitive CCD have provided the high throughput required for this camera at the expense of a 3700 Å blue cutoff. The WFC detectors are two butted 2K by 4K thinned, backside-illuminated, SITe CCDs with a red optimized coating and long-l halo fix. The plate scale is 0.050 arcseconds per pixel, which provides a good compromise between adequately sampling the PSF and a wide field of view. The WFC PSF is critically sampled at 11,600 Å and undersampled by a factor 3 at the blue end of the WFC sensitivity range (3700 Å). We expect that it will be possible to achieve a final reconstructed FWHM of 0.100 to 0.140 arcseconds for well-dithered observations. Because the WFC PSF FWHM is largely dependent on the blurring caused by CCD charge diffusion, dithering will not be able to recover the full resolution of the optical system. See Section 7.4 for more discussion of how to use dithered observations to optimally sample the PSF.
The optical design of the camera introduces a two-component geometric distortion. The detectors themselves are at an angle with respect to the optical axis. This produces an 8% stretching of one pixel diagonal compared to the other. As a result, WFC pixels project on the sky as rhombuses rather than squares. These effects are purely geometrical and are routinely corrected in the ACS data reduction pipeline. The second component of geometric distortion is more complex. This distortion causes up to ±9% variation in effective pixel area and needs to be taken into account when doing accurate photometry or astrometry as the effective area of the detector pixels varies nonlinearly with field position.
5.3.1 Filter Set
WFPC2 and Johnson-Cousins filters
All of the most commonly used WFPC2 filters are included in the ACS filter set. In addition to a medium and a broad V band filter (F550M and F606W), there is a complete Johnson-Cousins BVI set (F435W, F555W, F814W).
Sloan Digital Sky Survey filters
The Sloan Digital Sky Survey (SDSS) griz filter set (F475W, F625W, F775W, F850LP) is designed to provide high throughput for the wavelengths of interest and excellent rejection of out-of-band wavelengths. The filters were designed to provide wide, non-overlapping filter bands that cover the entire range of CCD sensitivity from blue to near-IR wavelengths.
Narrow Band filters
The Ha (F658N), [OIII] (F502N), and [NII] (F660N) narrow band filters are full-size, and can be used with both the WFC and HRC.
Ramp filters
ACS includes a complete set of ramp filters that provide full coverage of the WFC wavelength range at 2% and 9% bandwidth. Each ramp filter consists of three segments. The inner and outer filter segments can be used with the WFC only, while the middle segments can be used by both WFC and HRC. Unlike the WFPC2, where the desired wavelength is achieved by offsetting the telescope, the wavelength of ACS ramps is selected by rotating the filter while the target is positioned in one of the pre-defined apertures. The monochromatic field of view of the ramp filters is approximately 40 by 80 arcseconds. Details of how to use the ramp filters are given in Section 7.7.2.
Polarizer filters
The WFC/HRC filter wheels contain polarizers with pass directions spaced by 60×, optimized for both the UV (POL0UV, POL60UV, and POL120UV) and the visible (POL0V, POL60V, and POL120V). All the polarizer filters are sized for the HRC field of view. They induce vignetting when used with the WFC, for which the FOV will be about 72 by 72 arcseconds. More information on the use of the polarizers is given in Chapter 6.
Grism and Prism
The CCD channels also have a grism (G800L) for use with both WFC and HRC from 5500 Å to 11,000 Å, and a prism (PR200L) for use with the HRC from 1600 Å to 3500 Å. These are described more fully in Chapter 6.
5.4 High-Resolution Optical and UV Imaging
The High Resolution Channel of ACS is the prime ACS camera for near-UV imaging. HRC provides high throughput in the blue and a better sampling of the PSF than either the WFC or other CCD cameras on HST. The HRC pixel size critically samples the PSF at 6300 Å and is undersampled by a factor 3.0 at the blue end of its sensitivity range (2000 Å). With this capability, the HRC functionally replaces the Faint Object Camera as the instrument able to critically sample the PSF in the V band. For this reason, most of the usage of HRC will be for UV and blue imaging. HRC can also be convenient for imaging in the red when the PSF sampling is important. As an example, better PSF sampling is probably important for accurate stellar photometry in crowded fields. We expect that the photometric accuracy achievable by the HRC will be higher than that achievable with the WFC. Well-dithered observations with the HRC should lead to a reconstructed PSF FWHM of 0.03 arcsec at ~4000 Å, increasing towards longer wavelengths. HRC also includes a coronagraph that is discussed in Chapter 6. The HRC CCD presents a long wavelength halo problem similar to the STIS CCD since the front-side metallization correcting the halo problem for the WFC CCDs was implemented only after the HRC CCD had been procured. Although most of the HRC imaging is likely to occur in the UV, users should be cautioned to take into account the effects of the long wavelength halo when using the HRC in combination with near-IR filters (See Section 5.6.5).
5.4.1 Filter Set
The HRC-specific filters are mostly UV and blue. The set includes UV and visible polarizers (discussed in Chapter 6), a prism (PR200L, discussed in Chapter 6), three medium-broad UV filters (F330W, F250W, and F220W) and two narrow band filters (F344N and F892N). Use of the UV filters with the WFC is not supported because of the uncertainty of the WFC silver coating transmission below 4000 Å.
All broad, medium and narrow band WFC filters can be used with the HRC whenever a better PSF sampling is required. In general, the throughput of WFC is higher than that of HRC where their sensitivity overlaps. Only some of the WFC ramp segments can be used with the HRC since only the middle segment overlaps with the HRC FOV. In particular, HRC can use the FR459M and FR914M broad ramps, and the FR505N [OIII], FR388N [OII], and FR656N (Ha) narrow ramps.
5.4.2 Multiple Electron Events
Like the STIS CCD but unlike WFPC2, the HRC CCD is directly sensitive to UV photons and for this reason is much more effective in detecting them. However, whenever a detector has non-negligible sensitivity over more than a factor two in wavelength, it becomes energetically possible for a UV photon to generate more than one electron and be counted more than once. This effect has been seen for STIS and also during the ground testing of the HRC detector. The effect is only important shortward of 3200 Å, and reaches a magnitude of approximately 1.7 e–/photon at 2000 Å. Multiple counting of photons has to be taken into account when estimating the detector QE and the noise level of a UV observation since multiple photons cause a distortion in the Poisson distribution of electrons.
5.4.3 Red Leaks
When designing a UV filter, a high suppression of off-band transmission, particularly in the red, has to be traded with overall in-band transmission. The very high blue quantum efficiency of the HRC compared to WFPC2 makes it possible to obtain an overall red leak suppression comparable to that of the WFPC2 while using much higher transmission filters. In Cycle 14 we obtained new calibration data to check the impact of red leaks on observations. The results are described in ACS ISR 2007-003. In Table 5.6 we show the ratio of in-band versus total flux for a few UV and blue HRC filters, where the cutoff point between in-band and out-of-band flux is defined as the filter’s 1% transmission point. The same ratio is also listed for the equivalent filters in WFPC2. Correction factors for different stellar spectral types and non-stellar spectra can be found in the ISR. Clearly, red leaks are not a problem for F330W, F435W, and F475W. Red leaks are more important for F250W and F220W. In particular, accurate UV photometry of objects with the spectrum of an M star will require correction for the redleak in F250W and will be essentially impossible in F220W. For the latter filter a red leak correction will also be necessary for K and G stars.
Table 5.6: In-band flux as a percentage of the total flux.5.5 Ultraviolet Imaging with the SBC
The Solar Blind Channel is the ACS camera optimized for UV imaging. The SBC uses the same optical train as the HRC and is comparable in performance to the FUV MAMA of STIS. The use of the repeller wire increases the quantum efficiency of the detector by ~30%, but adds a halo to the PSF. Bright object limits are discussed in detail in Section 7.2.
5.5.1 Filter Set
Several filters are available for use with the SBC, including a Lyman a narrow band filter (F122M), a long pass quartz filter (F150LP), MgF2 filter (F115LP), and a CaF2 filter (F125LP). The SBC also includes two additional long pass filters (F140LP and F165LP) as well as prisms (discussed in Chapter 6). A list of filters is given in Table 5.3
5.5.2 Red Leaks
The visible light rejection of the SBC is excellent, but users should be aware that stars of solar type or later will have a significant fraction of the detected flux coming from outside the nominal wavelength range of the detector. This is discussed in greater detail in Section 4.4.2.
Table 5.7: Visible-light rejection of the SBC F115LP imaging mode.
which have l < 1800 Å
which have l < 3000 Å5.5.3 SBC Imaging Filter Shifts
The SBC focal surface, like that of the HRC, is tilted significantly with respect to the chief ray. Because the MAMA detector is a STIS spare, its window is approximately parallel to the MCP surface and the whole detector must tilt to achieve good focus over the field. Because the window is therefore tilted, "lateral color" is introduced, which would result in dispersion-induced degradation of the PSF, so the filters are canted in the opposite direction to that of the window to ameliorate the color. The filter thickness is matched to the mean index of refraction over its bandpass to maintain focus. These result in unavoidable image location offsets between filters. In contrast, the WFC and HRC filters and windows are normal to the chief ray and the detector surfaces are tilted within their housings to match the focal surface tilt. In Table 5.8, we list the shifts for each SBC imaging filter with respect to the F115LP filter. No pointing compensations are made for these offsets.
Table 5.8: Shifts between SBC imaging filters.
in the x,y directions5.6 ACS Point Spread Functions
The ACS point spread function has been studied in ground test measurements, and by using on-orbit data and models generated by the
Tiny TIMsoftware (http://www.stsci.edu/software/tinytim/tinytim.html) developed by J. Krist and R. Hook. As with other HST instruments, the ACS point spread function is affected by both optical aberrations and geometric distortions. Point sources imaged with WFC and HRC experience blurring due to charge diffusion into adjacent pixels because of CCD subpixel variations, which reduces the limiting magnitudes that can be reached by WFC/HRC. The SBC PSF and the long-wavelength HRC PSF are also affected by a halo produced by the detectors themselves.5.6.1 CCD Pixel Response Function
The sharpness of the CCD PSF is somewhat degraded by charge diffusion into adjacent pixels. The effect is usually described in terms of the pixel response function (PRF), which gives the distribution of flux from within the pixel into adjacent pixels. Charge diffusion results in ~0.5 magnitude loss in the WFC limiting magnitude at short wavelengths (the worst case). At longer wavelengths and at all wavelengths for the HRC the reduction in the limiting magnitude is ~0.2 magnitudes or less. Due to variations in the CCD thickness, charge diffusion is not constant over the field of view. At different wavelengths, the CCD pixel response functions can be represented by the following kernels (for the center of the field):
Figure 5.8:Kernels representing CCD pixel functions for HRC and WFC.at l = 4000 Å,
at l = 5500 Å, and
at l = 8000 Å.
More details on ACS CCD charge diffusion are given in ACS ISR 2006-01. For details on CTE-induced photometric losses for ACS/WFC and techniques to correct for them, see ACS ISR 2006-01.
5.6.2 Model PSFs
Table 5.9 and Table 5.10 give ACS model PSFs in the central 5 ¥ 5 pixel region in two wavelength bands (filters). Numbers listed are the fraction of the total energy received in each pixel. The models have been generated using Tiny TIM, taking into account the HST optical aberrations and obscurations as well as the CCD pixel response function. Field dependent geometrical distortions are included. The real PSF will also differ from the model because of the jitter in the HST pointing, HST focus variation (focus breathing), and other instrumental effects, some of which are briefly discussed below. For further details on the PSF variations and an effective procedure to model them, see ACS ISR 2006-01. The SBC PSF is shown Figure 5.10.
Table 5.9: Model ACS CCD PSFs.Table 5.10: Model ACS SBC PSFs.5.6.3 Encircled Energy
In general, the ACS channels encircled energy distribution has been found to be within the original instrument specifications. Figure 5.9 and Figure 5.10 show the ACS encircled energy curves derived from on-orbit images. Tabulated values of the encircled energy for most filters are available in Sirianni et al. (2005PASP 117.1049).
5.6.4 Geometric Distortions
Geometric distortions produce a significant impact on the shape of the PSF in all three of the ACS channels, as can readily be seen in Figure 5.11 and Figure 5.12, which display WFC and HRC PSF images. The log stretch enhances the spider diffraction patterns, which the distortion renders non-perpendicular, and the outer Airy rings, which appear elliptical. The distortion owes primarily to the tilt of the focal surface to the chief ray at the large OTA field angles of the ACS apertures. The linear, field-independent approximation for the WFC produces a difference in plate scale of about 8% between the two diagonals of the field and, in the HRC and SBC, about a 16.5% difference in scale between orthogonal directions rotated about 20× from the aperture edges. Field-dependent distortions, measured as actual vs. predicted distances from field center, amount to about 2% peak in the WFC and about 1% in the HRC and SBC.
The distortions render the pixels, as projected on the sky, trapezoidal in shape and their area varies over the field by about 19% and 3.5% in the WFC and HRC/SBC, respectively. These variations have significant ramifications concerning appropriate techniques for flat-fielding and photometric calibration, especially when complicated by resampling in order to combine dithered image sets. Related issues are the manner in which the halation effects of the HRC and SBC detectors are removed and the treatment of spectra from the prisms and grism, which are not subject to the same distortion effects.
More details concerning geometric distortions in ACS can be found in ACS ISR 2002-02 and ACS ISR 2004-15. An introduction to calacs, and to multidrizzle which applies corrections for geometric distortion, is available on-line at:
http://www.stsci.edu/hst/acs/analysis/multidrizzle.
Figure 5.9: Encircled energy for the CCD channels.Figure 5.10: Encircled energy for the SBC.Figure 5.11: ACS WFC PSF - F625W.Figure 5.12: ACS HRC PSF - F625W.5.6.5 PSFs at Red Wavelengths and the UV
The CCDs used in the HRC and WFC suffer from a halo that is caused by very red photons passing through the device and being scattered back into the detector by the mounting substrate. This creates a large halo in HRC images beyond 7000 Å and WFC images past 9000 Å. At 8000 Å in the HRC, the halo contains about 10% of the light. At 10,000 Å, it contains about 30% and increases the surface brightness of the PSF wings by over an order of magnitude, overwhelming the PSF diffraction rings and spikes. A discussion of this may be found in Gilliland & Riess, (2002) HST Calibration Workshop at:
In the F850LP filter, in particular, extremely red stars show a progressive loss of flux in small to moderate sized apertures as a function of color. A paper by Sirianni et al. (2005, astroph/0507614) is available at:
http://adcam.pha.jhu.edu/instrument/photometry/sirianni.pdf
See also Bohlin ACS ISR 2007-06. These papers provide a detailed recipe to correct for this effect. This halo effect is only partially treated by the Exposure Time Calculator. Observers can use synphot (see Section 9.3.2) to accurately calculate the photometry of red sources in the SDSS z-filter.
Long wavelength photons that pass through the CCD can also be scattered by the electrode structure on the back side of the device and will create two spikes that extend roughly parallel to the x-axis. These spikes are seen at wavelengths longer than 9500 Å in both the HRC and WFC (see Figure 5.13 and Figure 5.14).
Figure 5.13: ACS WFC PSFs (10" x10"). FR914M images are saturated.Figure 5.14: ACS HRC PSFs (3.25" x3.25").In the UV the core of the PSF becomes rather asymmetrical due to midfrequency optical surface errors. In the SBC, a halo is created by charge migration at the microchannel plate surface. This effect, seen previously in STIS MAMA images, broadens the PSF core and redistributes a small portion of flux into a broad halo that can be approximated by a Gaussian with FWHM ~20 pixels. The peak flux for a point source centered on a pixel is reduced by 30% to 40% depending on wavelength.
The encircled energy curves presented in this handbook and incorporated into the ETC include all of the scattering effects discussed here.
5.6.6 Residual Aberrations
Residual aberration levels at the center of the field in each camera are 1/30 wave (HRC) and 1/20 wave (WFC) rms at 5500Å (excluding defocus). Coma and astigmatism are minimized at the field center of each camera. The ACS PSF varies far less over the field of view than those of WFPC2 and STIS. WFPC2 especially suffers from a variable obscuration pattern that significantly alters the PSF structure depending on field position. Lacking the additional obscurations present in WFPC2, ACS PSF variations are instead due to changes in aberrations and charge diffusion.
At the extreme corners of the WFC field, increased astigmatism slightly elongates the PSF core. The axis of elongation rotates by 90× if the system passes through focus due to breathing. This may affect ellipticity measurements of small galaxies with bright cores at the field edges. Focus variations in the WFC, which alter the amount of light in the peak, are largely due to detector surface height irregularities and amount to the equivalent of 5 microns of breathing (1/18 wave rms). The largest focus offset is along the gap between the two CCDs. Variations in the width of the PSF core are dominated by changes in CCD charge diffusion, which is dependent on the thickness of the detector (12 to 17 microns for the WFC). The PSF FWHM in F550M, for example, can vary by 20% (0.10 to 0.13 arcseconds) over the field.
The PSFs in the HRC and SBC are reasonably constant over their fields. The HRC FWHM is 0.060 to 0.073 arcseconds in F550M. More details on ACS PSF field variations are provided in ACS ISR 2003-06. The Tiny Tim PSF simulator includes field dependent aberrations and charge diffusion and may be used to estimate the impact of these variations.
|
Your Company Your Company's Web Site Voice: (555) 555-1234 Fax: (555) 555-5678 Your Company's E-mail Address |
