Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.stsci.edu/hst/acs/documents/isrs/isr9907.pdf Дата изменения: Mon Oct 25 22:48:17 2004 Дата индексирования: Sat Dec 22 23:29:02 2007 Кодировка: Поисковые слова: m 2-9

·

ACS Instrument Science Report 99-07

The Solar Blind Channel Bright Object Limits for Astronomical Objects
F.R. Boffi and R. C. Bohlin August 31, 1999

ABSTRACT Tables of limiting magnitudes for all observation modes of the ACS Solar Blind Channel are calculated for a range of stellar energy distributions, using Kurucz model fluxes and observed standard stars. For K and M stars, the observed Bruzual flux distributions extend down to only 2000 A. New prototypes for K and M stars are extracted from the IUE archive for determining the bright object limits for these late Morgan-Keenan spectral classes.

1. Introduction
The Solar Blind Channel (SBC) for the ACS uses the STIS FUV flight spare MAMA detector (STF7). Therefore, a similar bright object protection problem arises. Differences between the STIS and the ACS Bright Object Protection procedures are discussed in ISR ACS 98-03 by Cox et al. (1998). The ACS SBC detector is subject to damage at high local and global illumination rates. The limits are: · · The global input count rate must not exceed 2.0x105 counts/sec over the entire MAMA The local input count rate must not exceed 50 counts/sec per pixel

Limiting magnitudes for all SBC observation modes of single point sources are always set by the local rate limit of 50. If the local rate check detects a violation of the local limit, the SBC is shuttered and the observation is lost (Cox et al. 1998). Our tabulations of the limits vs. spectral-type will help users of the Solar Blind Channel. The maximum allowed count rates are translated into limiting V magnitudes for convenience.

1


2. Samples of stellar energy distributions
In order to calculate the maximum allowed count rates for each observation mode of the ACS SBC, a range of spectral energy distributions, i.e. spectral types, are selected. The selection procedure is analogous to Leitherer et al. (1996). Input flux distributions are Kurucz (1992) line-blanketed LTE model atmospheres for a grid of 16 models covering O to K stars for the dwarf luminosity class V. The corresponding temperatures for each spectral type are from Schmidt-Kaler's (1982) spectral-type vs. Teff calibration. The Kurucz models can be found on-line in the synphot libraries (dir crgrid$k93models/) and correspond to the log g values in Table1. The Kurucz continuum models in Table 1 are the same as in Leitherer et al. (1996). However, for the later spectral types, the chromospheric emission is stronger than the continuum in the far-UV (Leitherer et al. 1996). Therefore, spectra of actual, observed stars must be used to calculate the visual brightness limits corresponding to the limiting count rates in the UV. Some input spectra for actual stars are from the undocumented Bruzual-Persson-Gunn-Stryker (BPGS) spectrophometry atlas, which is on-line in the synphot libraries (dir crgridbpgs$). These stars are listed in Table 2, where they are given the name (bpgs_number), their spectral type, the logarithm of the effective temperature (log Teff) and the V magnitude (they are normalized to 0). However, the BPGS fllux distributions for K and M stars lack information below 2000A. For these coolest spectral types, the IUE database has been mined for the nine new standard stars also listed in Table 2. In addition, five IUE spectra of type G0V and the solar spectrum from Colina, Bohlin, & Castelli (1996) are included to provide an overlap with the BPGS data. Table 2 also lists temperatures from Schmidt-Kaler and V magnitudes for the IUE stars. These new IUE spectra can be found at http://www.stsci.edu/instruments/observatory/calobs.html, while the flux distributions of the K-M stars are illustrated in Figure 1. At wavelengths shorter than ~1900 A, there may be some contribution to the IUE continuum from scattered red light. Table 1: Kurucz models
Spectral Type O5V B1V B3V B5V B8V A1V A3V A5V log Teff 4.648 4.405 4.271 4.188 4.077 3.965 3.940 3.914 5.0 3.9 3.9 3.9 4.0 4.1 4.2 4.3 log g

2


Table 1: Kurucz models
Spectral Type F0V F2V F5V F8V G2V G5V G8V K0V log Teff 3.857 3.838 3.809 3.792 3.768 3.760 3.746 3.720 4.3 4.3 4.3 4.4 4.4 4.5 4.5 4.5 log g

Table 2: Bruzual and IUE Stars
Name bpgs_1 bpgs_5 bpgs_14 bpgs_31 bpgs_33 bpgs_38 bpgs_36 bpgs_45 bpgs_46 bpgs_44 bpgs_49 CVn CrB 1 Ori HD26337 HD206860 Sun Cep Ser Spectral type O5 B1V A0V F4V F6V F8V G0V G2V G2V G5V G8V G0V G0V G0V G0V G0V G2V K1IV K0III/IV log Teff 4.648 4.405 3.978 3.818 3.798 3.792 3.780 3.768 3.768 3.760 3.746 3.780 3.780 3.780 3.780 3.780 3.768 3.706 3.676 ... ... ... ... ... ... ... ... ... ... ... 4.26 5.22 4.41 7.10 5.94 -26.75 3.21 3.26 V

3


Table 2: Bruzual and IUE Stars
Name Boo Aql HD146051 Cet µ Peg Gru HD120323 Spectral type K2III K3II M1III M2III M2III M5III M5III log Teff 3.646 3.617 3.570 3.559 3.559 3.522 3.522 -0.04 2.72 2.74 2.53 3.48 2.11 4.19 V

3. Calculation of the limiting magnitudes
Limiting magnitudes for all imaging modes are computed with the ACS CGI Exposure Time Calculator (ETC) developed by R. Shaw. For the Kurucz models, the ETC calculates the source count rate (C) in the central pixel for a V=0 star, assuming the stellar source is centered on that pixel. Then, the limiting magnitude is calculated by comparing C to the local limit of 50 counts/sec. Since the spectroscopic ETC is not yet available, an IDL program calculates the source count rates for the two prisms. For the undocumented BPGS models and the IUE spectra, computations of the limiting magnitudes follow a similar procedure. These limiting magnitudes for each SBC mode are plotted in Figures 2 through 9. In the figures, the dashed lines are the Kurucz models, while the triangles represent the real stars (both the BPGS sample and the IUE). The earlier types agree with the Kurucz models, while the real stars corresponding to spectral types cooler than log Teff=3.75 deviate from the Kurucz models of the continuum, because the dominant source of UV flux is chromospheric emission (see section 2). Our set of SBC bright-object limits are based on the Kurucz model atmospheres and on the observed UV fluxes of stars, assuming zero reddening. Since there is some spread in the fluxes of a star of a given spectral class, the new limits are obtained by defining an envelope to the theoretical and observational limits as follows: Above log Teff=3.9, the limits from the observations are all within one mag fainter than the models, so that a safe faint limit is defined by the model plus one mag safety factor. From log Teff= 3.75 to 3.9, the stars scatter by ~2 mag fainter than the models, so that the safe limit is set to the locus of Kurucz models plus two mag. Below 3.75, the models are not applicable, so that the limit is set by the average of the results from the new IUE standards plus 2 magnitudes. This envelope is the bold solid line in Figs. 2 through 9. These limiting magnitudes are provided in Table 3 as a function of Morgan-Keenan spectral class and are typically more conservative than the values in Cox et al. (1998), even for hot stars, because of the different throughput tables for the SBC components.

4


Two minor concerns about the above analysis remain: The IUE does not measure the chromospheric emission at Lyman-alpha, because of contamination from the Earth's geocorona. Therefore, the data at log Teff < 3.75 in Figure 2 for F122M are only lower limits. A cool star fainter than the 2.4 mag limit shown at the left of Figure 2 with a very bright Lyman-alpha coronal emission line could cause a bright object violation. No value is tabulated for the K0V star and the F122M filter in Table 3. Also, some concern is raised by the BPGS data point at G0V log Teff=3.780, which lies above our heavy solid line in Figures 2-9. This star has anomalously large UV fluxes compared to neighboring F and G stars in the BPGS list. The five IUE G0V stars and the Sun all lie well below the BPGS G0V star, suggesting that BPGS data point is invalid. An even more serious local rate violation could arise from the inadvertent observation of a cool star with an undetected hot companion. Following Leitherer (1996), such a douTable 3: Bright Object Limits for the SBC modes
Spectral type O5V B1V B3V B5V B8V A1V A3V A5V F0V F2V F5V F8V G2V G5V G8V K0V Doublea AG Pegb log Teff 4.648 4.405 4.271 4.188 4.077 3.965 3.940 3.914 3.857 3.838 3.809 3.792 3.768 3.760 3.746 3.720 ----f122m 17.3 16.4 15.5 14.8 13.4 10.8 9.3 8.4 6.9 6.2 4.5 3.4 2.4 2.4 2.4 ---15.6 14.2 f115lp 20.7 19.9 19.1 18.5 17.4 15.4 14.3 13.6 12.3 11.6 10.1 9.0 7.3 6.7 6.6 6.6 19.0 17.5 f125lp 20.6 19.8 19.0 18.4 17.4 15.4 14.3 13.6 12.3 11.6 10.1 9.0 7.3 6.7 6.6 6.6 18.9 17.4 f140lp 20.0 19.2 18.5 18.0 17.0 15.3 14.3 13.6 12.4 11.6 10.1 9.0 7.3 6.7 6.2 6.2 18.3 16.9 f150lp 19.5 18.6 18.0 17.4 16.5 15.1 14.2 13.6 12.4 11.6 10.1 9.0 7.3 6.7 5.9 5.9 17.8 16.5 f165lp 18.2 17.5 16.8 16.3 15.4 14.2 13.7 13.4 12.3 11.6 10.1 9.0 7.3 6.7 5.3 5.3 16.5 15.0 pr110l 16.7 15.9 15.2 14.6 13.7 12.5 11.9 11.7 11.0 10.4 8.9 7.7 5.9 5.3 4.1 4.1 15.0 15.3 pr130l 16.5 15.7 15.0 14.5 13.5 12.2 11.7 11.5 10.6 9.9 8.3 7.2 5.5 4.9 3.9 3.9 14.8 15.4

5


a. System made of total light in the iting magnitude O5V case. b. Star with a flux

a main sequence late-type star with an O5V star contributing 20% to the V band. In the UV, the O5 component dominates and sets the same limfor types A-M. A one magnitude safety factor has been added, as for the distribution like AG Peg.

ble star could have an undetected O5 like companion with 20% (1.7 mag fainter) of the cool star flux in the V band. Thus, the limits for the entry "Double" in Table 3 are 1.7 mag less than the O5 entries. An actual symbiotic star AG Peg with V=8.4 (Keyes, private communication) produces similar limits to the "Double" simulation. Since AG Peg is an extreme case of the ratio of UV to visual flux for such symbiotic stars, no safety factor is added. Hopfully, these unknown cool stars with hot companions are rare enough to cause few cases of shuttering due to local rate limits. The FUV-MAMA on STIS has a sensitivity similar to the ACS SBC channel and has not yet encountered this shuttering situation. If the problem becomes a common occurrence for either STIS or ACS, observations with the broader filters in the UV will have to be restricted, for example within some degrees of the galactic plane.

4. Practical Considerations
In performing the screening for bright object violations, the STScI must perform a labor intensive special plate analysis to get colors for the faint stars in the field of the SBC. Therefore, the worst case of O stars is assumed using the V mag limits from the first row of Table 3 for the first cut analysis. However, since a small fraction of the general field stars are so hot, the actual damage limit for the SBC MAMA is used, unless there are astrophysical reasons to suspect the presence of hot, unreddened stars. Damage to the SBC in the form of a loss of sensitivity in the pixels under the source occurs at about 10,000 input counts/sec. The counting electronics saturate, and the output counts/sec can never exceed the global limit for one point source. However, damage begins at ~200x (6mag) brighter than the limits in Table 3. For the general field, a GO observation will automatically pass screening, if there are no stars within 6 mag of the limits in the first row of Table 3, eg. V=14.7 for the worst case of F115LP. If brighter stars are present, then a GO may be asked to change targets, select another filter, or provide UV flux information for the offending stars. If these remedies are not available, then the STScI may measure the colors of the offending stars. If the colors are consistent with spectral types later than A, then the remaining worry is the presence of hot companion to a cool star, i.e. the "Double" limits of Table 3. Again, damage may occur for double stars that are more than 6 mag brighter in V than these Double Star limits, eg. V=13 in the worst case of the F115LP filter. If there are still offending stars, then the GO must again be requested to change targets, select another filter, or provide UV flux information. For such bright stars, IUE spectra are often available. Lacking these remedies, the observation may be disallowed.

6


5. References
Colina, L., Bohlin, R. C., & Castelli, F. 1996, AJ, 112, 307. Cox, C., Boffi, F.R., & Clampin, M. 1998, ISR ACS 98-03, (Baltimore:STScI). Kurucz, R. L. 1992, in IAU Symp.149, The Stellar Populations of Galaxies, ed. B. Barbuy & A. Renzini (Dordrecht: Kluwer), 225. Leitherer, C., Kinney, E., Baum, S., &Clampin, M. 1996, ISR STIS 96-024, (Baltimore:STScI). Leitherer, C. 1996, TIR STIS 96-005, (Baltimore:STScI). Schmidt-Kaler 1982: in Landolt-Boernstein, New Series, Group VI, Vol. 2b, ed. K. Schaifers and H.H. Voigt (Berlin:Springer).

6. Acknowledgements
We would like to thank Cheryl Pavlovsky and Steve Hulbert for their kind and essential help with Framemaker. Thanks to Tony Keyes for the flux distribution of AG Peg. Useful discussions with Claus Leitherer are gratefully acknowledged.

7. Figure Captions
Fig. 1: Flux distributions for the nine new IUE cool standards stars. Absolute scaling is arbitrary. Many cool stars have strong chromospheric emission lines, such as OI 1302A and SiII 1814A in the region of SBC sensitivity. Straight dotted lines connect the IUE SWP short wavelength region across a sea of noise to the good part of the LW data. Other dotted line regions flag saturation, i.e. unreliable fluxes. The star labels lie below the flux plot for each star. Fig.2: The V limiting magnitude (Vlim) vs. the logarithm of the effective temperature Teff for the observing mode with the F122M filter: the dashed line with asterisks at the points from Table 1 represents the Kurucz star models, while the triangles represent the real stars, both the BPGS sample and the IUE. The curved bold solid line is an envelope to the theoretical and observational limits and is constructed by applying a 2 magnitude safety factor to the Kurucz models at Teff=3.75-3.9 and a one magnitude safety factor to the Kurucz limits at Teff>3.9 and then fitting a spline function. The straight flat line segment below ~3.75 is the average of the cool star results plus 2 mag. In the case of this F122M filter, the plots are only lower limits below ~3.75, because the Lyman-alpha coronal line is not measured by IUE. Fig. 3: The V limiting magnitude (Vlim) vs. the logarithm of the effective temperature Teff for the observing mode with the F115LP filter. The symbols and lines have the same meaning of Fig. 1.

7


Fig.4: The V limiting magnitude (Vlim) vs. the logarithm of the effective temperature Teff for the obesrving mode with the F125LP filter. The symbols and line have the same meaning of Fig.1. Fig.5 The V limiting magnitude (Vlim) vs. the logarithm of the effective temperature Teff for the obesrving mode with the F140LP filter. The symbols and line have the same meaning of Fig.1. Fig.6 The V limiting magnitude (Vlim) vs. the logarithm of the effective temperature Teff for the obesrving mode with the F150LP filter. The symbols and line have the same meaning of Fig.1. Fig.7 The V limiting magnitude (Vlim) vs. the logarithm of the effective temperature Teff for the obesrving mode with the F165LP filter. The symbols and line have the same meaning of Fig.1. Fig.8 The V limiting magnitude (Vlim) vs. the logarithm of the effective temperature Teff for the obesrving mode with the PR110L filter. The symbols and line have the same meaning of Fig.1. Fig.9 The V limiting magnitude (Vlim) vs. the logarithm of the effective temperature Teff for the obesrving mode with the PR130L filter. The symbols and line have the same meaning of Fig.1.

8


Fig. 1

9


Fig. 2

Fig. 3

10


Fig. 4

Fig. 5

11


Fig. 6

Fig. 7

12


Fig. 8

Fig. 9

13