Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.stsci.edu/hst/wfpc2/documents/isr/wfpc2_isr0902.pdf Дата изменения: Unknown Дата индексирования: Tue Feb 5 16:13:21 2013 Кодировка: Поисковые слова: comet

Instrument Science Report WFPC2 2009-002

WFPC2 F343N Filter Throughput Decline
S. Gonzaga and J. Biretta April 01, 2009

ABS

TRACT

Observations of the standard star GRW+70D5824 indicate a ~50% decrease in the throughput of the F343N filter during the WFPC2 mission. The decline appears to be linear with time and is approximately uniform across the field of view. Essentially identical results are obtained for the PC1 and WF3 detectors, and no features corresponding to the ~50% throughput decline are seen in external flats. Post-flight laboratory tests on the filter are needed to determine its current spectral properties.

Introduction
The WFPC2 narrow-band filters typically contain sandwiches of colored glass and pairs of multi-layered thin-film filters. These multi-layer thin-film filters are not necessarily stable over long periods of time; the literature contains various reports of wavelength change, etc., due to absorption / desorption processes and long-term annealing of the multi-layer stack. It is diffcult to assess the stability a priori, since the construction details and manufacturing processes are typically proprietary. Evidence for a throughput decrease in the F343N filter was first brought to our attention in July 2008 by a planetary observer. Herein we present analyses of standardstar data for this filter from April 1994 to the present epoch, as well as results from Earth flats (observations of the bright Earth). Our results may be used to update the Synphot throughput table at some point in the future.

Operated by the A ssociation of Univ ersities for Research in A stron omy, Inc., for the National A eron autics and Space Administration

1


Data Analysis
We looked at the photometric trend of standard star GRW+70D5824 in F343N, from April 1994 (when the CCD operating temperature was reduced from -76 C to -88 C) to recent data taken in September 2008. Analysis was restricted to PC1 and WF3 because most data were taken using those chips. We selected images that were taken within 10 days after a decontamination to minimize the effects of UV contamination on the photometry. The images were processed through the standard STScI Archive "on-the-fly" calibration, corrected for geometric distortion (by multiplying the images with the correction image, f1k1552bu.r9h), and the 34th row defect as described by Anderson & King (1999). Aperture photometry was performed using the IRAF phot package. The instrumental magnitudes reported in Tables 1 & 2 were obtained using an aperture radius of 0.5", which corresponds to r=11 pixels in PC1 and r=5 pixels in the WF chips. We also obtained measurements using r=3 pixels for PC1, and r=2 pixels for WF chips to calculate CTE loss as described by A. Dolphin (2008). Figure 1, a plot of CTE-corrected magnitude vs. Modified Julian Date, shows the throughput decline of ~50% over 15 years. A linear fit was applied to the PC1 and WF3 data using the STSDAS task tlinear. The linear fits for both chips, as shown in Table 3, are quite consistent, suggesting that the throughput decline is not localized to one section of the filter. To futher study the uniformity of the decline, we analyzed Earth flats taken with the F343N filter from two widely-spaced epochs: 1994-1995 vs. 2007-2008. We selected Earth flats taken in roughly the same post-decon interval (for these samples, more than 20 days after a decon), which had good signal-to-noise and no saturation. The rapid motion of HST creates streaks across the Earth flat images that can be removed by combining several good-quality Earth flats. This was done using the streakflat task in stsdas.hst_calib.wfpc.w_calib. The combined Earth flat from each epoch was then normalized using the normclip task in the same package, which also computes the reciprocal of the input image. Figures 2a and 2b show the 2007-2008 combined Earth flat effectively divided by the 1994-1995 combined Earth flat. This ratio image was median-filtered using a 3x3 box to remove the effects of hot pixels. Dark areas in the figure correspond to areas of low filter throughput in the 2007-2008 image. Since the F343N filter is located about 20 cm from the HST focal plane, any features on the filter should contribute only large-scale features at the detector. A single bright or dark point on this filter would map into a ~50 arcsecond diameter donut at the detector. The flats do not show any large-scale features corresponding to the 50% photometric loss. At most, a broad 10% to 15% gradient

2


might be visible across Figure 2b, running from the lower right to the upper left. Apparently the 50% throughput loss is not localized to one area of the filter, but instead must involve the entire filter and be approximately uniform on the relevant spatial scales. We cannot, of course, rule out that numerous small spots are present (few arcsecond or smaller size) which are spread across the filter in some quasi-uniform fashion. Most of the features visible in Figures 2a and 2b have known causes. The contamination "worms" which are most apparent in WF2 are from the 1994-1995 flats. These were caused by some thin contaminant on the CCD windows that repelled darker contaminants (i.e. worms are bright in the 1994-1995 image). They appeared in all UV images taken early in the mission, and slowly disappeared over the years, presumably due to the contaminant evaporating. Faint streaks are visible in the flat ratio due to Earth features. Very small (but strong) features are due to dust spots on or near the CCDs. Weaker spots or granularity are due to hot pixels, primarily from the 2007-2008 images. While the throughput loss is clearly present, we do not have any spectral information to constrain the bandpass shape, nor for that matter, the central wavelength. The filter is composed of a sandwich of two multi-layer thin-film filters. A wide range of scenarios are possible ­ one of the thin-film filter components may not have been properly bakedout during manufacture, and hence may have shifted in wavelength relative to the other component as water or gasses slowly evaporated. Under this scenario, the overall bandpass would become narrowed and shift slightly to larger or smaller wavelengths. But a wide range of other possibilities exist also. Post-flight laboratory tests are be needed to determine exactly the current spectral properties of the filter.

Table 1 . PC1 F343N observations of GRW+70D5824, taken w ithin 10 days aft er a decon. The last three co lumns indicat e the resulting instrumental magnitude, magnitude uncertainty, and C TEcorrect ed inst rumental magnitude, respectiv ely.

Image u2tx020ct u3ik030cr u4s60406r u5ls680ir u69q380ir u6hpjg0gm u8gxcg0br u8u5bg07m u9jacg06m ub0pmo01m ub0pmo02m ub0pmo03m ub0pmo04m

Date of Observation 1995-08-27 1997-08-25 1998-10-17 2000-04-19 2001-05-17 2002-06-17 2002-12-05 2003-11-15 2006-03-15 2008-09-26 2008-09-26 2008-09-26 2008-09-27

Days Since Decon 0.3807 5.2543 3.1842 1.0619 1.7126 5.1540 0.2272 1.5605 0.2753 3.8932 3.8960 3.9008 3.9036

m (0.5") 24. 24. 24. 24. 24. 24. 24. 25. 25. 25. 25. 25. 25. 344 438 455 569 636 881 776 031 005 373 383 156 139

m (0.5") error 0.02 0.022 0.022 0.024 0.025 0.029 0.026 0.051 0.034 0.045 0.045 0.036 0.036

m (0.5"), CTE-corrected 24.318 24.391 24.395 24.484 24.538 24.765 24.668 24.878 24.831 25.192 25.201 24.982 24.960

3


Table 2 . Same as Table 1, but for the WF3 detector.

Image u2tx0215t u3ik0315r u4s6040hr u5ls680sr u69q380sr u6hpjg0pm u8gxcg0nm u8u5hg07m u948hg07m u9jahg07m ub0pmp01m ub0pmp02m ub0pmp03m ub0pmp04m

Date of Observation 1995-08-27 1997-08-25 1998-10-17 2000-04-19 2001-05-17 2002-06-17 2002-12-05 2004-09-05 2005-09-22 2006-12-17 2008-09-25 2008-09-25 2008-09-25 2008-09-25

Days Since Decon 0.5592 5.4202 3.2474 1.1279 1.7751 5.2699 0.2974 2.0842 2.1356 4.5384 2.5001 2.5029 2.5078 2.5105

m (0.5") 24. 24. 24. 24. 24. 24. 24. 24. 25. 25. 25. 25. 25. 25. 251 39 49 389 649 774 834 903 056 122 187 218 226 229

m (0.5") error 0.012 0.014 0.014 0.013 0.016 0.017 0.018 0.019 0.021 0.021 0.023 0.024 0.024 0.024

m (0.5"), CTE-corrected 24.223 24.335 24.422 24.311 24.550 24.669 24.722 24.758 24.897 24.955 24.993 25.025 25.056 25.058

Figure 1 . F343N GR W+70D5824 CTE-corrected magnitude vs. Modified Julian Date.

4


Table 3 . Linear fit parameters fo r F343N CTE-corrected Magnitudes vs. Modified Julian Date: coefficients of t he f it (a, b), their standard dev iations (sig ma_a, sigma_b) , and chi squared (chi2).

Chip # of pts a PC1 WF3 13 14 15.68 15.25

sigma_a 0.85 0.56

b (slope, mag/day) 1.72E-4 1.79E-4

sigma_b 1.61E-5 1.06E-5

chi2 0.095 0.047

Figure 2a. PC1 ratio image. This effectively shows the 20 07-2008 combined Ea rth f lat divided by the 1994-1995 combined Eart h flat . This image was 3x3 pixel median-filtered to remov e the eff ects of hot pixels f rom the latter epoch. The DS9 display rang e is f ro m z1 =0.9 to z2 =1.1. Variations due to UV contamination "wo rms," scattered light, and residua l Ea rth flat features a re visible.

5


Figure 2b. The sa me rat io image as Figure 2a , st retched to show the WF chips. The "worm" features in WF2 (lower left) a re UV contaminant f eatures fro m the 1994-1995 epoch; they a re not present in the later epo ch. The DS9 display range is z1= 0.9 to z2 =1.1. Brightness variations on the spatial scales expected for f ilt er anoma lies ( ~50 arcseonds) are not stronger than about 10% to 15%.

6


Conclusions
The WFPC2 F343N filter has undergone a ~50% throughput loss during the 15 year WFPC2 mission. The loss is roughly linear with time and has an average rate of 1.75E-4 mag/day. The loss appears to be more-or-less uniform across the WFPC2 field-of-view ­ loss rates are similar in PC1 and WF3 CCDs, and no features of comparable strength appear in Earth flats taken in 2007-2008. Post-flight inspections and testing of the F343N filter are needed to determine nature of the problem and the current spectral properties of the filter. Appropriate corrections to the WFPC2 throughput tables will be made at a later date.

Acknowledgements
Our thanks to Dr. Amy Simon-Miller at NASA/GFSC for bringing this anomaly to our attention. We are also grateful to Dr. Van Dixon, Dr. Ronald Gilliland, Ms. Pey Lian Lim, and Dr. Linda Smith for their careful review of this ISR.

References
Astrometric and Photometric Corrections for the 34th Row Error in HST's WFPC2 Camera. Jay Anderson and Ivan R. King http://www.journals.uchicago.edu/doi/abs/10.1086/316432 Updated WFPC2 CTE corrections and calibration Andrew Dolphin, 2008. http://purcell.as.arizona.edu/wfpc2_calib/ WFPC2 Instrument Handbook for Cycle 17 Matt McMaster and John Biretta http://www.stsci.edu/hst/wfpc2/documents/handbook/cycle17/wfpc2_ihb.pdf

7