c Pleiades Publishing, Ltd., 2011. ISSN 1063-7729, Astronomy Reports, 2011, Vol. 55, No. 1, pp. 13–18. c A.S. Shugarov, N.V. Chupina, A.E. Piskunov, N.V. Kharchenko, 2011, published in Astronomicheskii Zhurnal, 2011, Vol. 88, No. 1, pp. 16–21. Original Russian Text
An Experimental Test of the Photometric Calibration of the Guide Star Catalog for the Spectrum-UV (WSO/UV) Project A. S. Shugarov1 , N. V. Chupina1 , A. E. Piskunov1 , and N. V. Kharchenko2 1
Institute of Astronomy, Russian Academy of Sciences, ul. Pyatnitskaya 48, Moscow, 119017 Russia 2 Main Astronomical Observatory, National Academy of Sciences of Ukraine, ul. Akademika Zabolotnogo 27, Kiev, 03680 Ukraine Received 2010; in final form, July 1, 2010
Abstract—A test of the photometric calibration accuracy for the guide-star catalog (Master Catalog) of the Spectrum-UV project (World Space Observatory—Ultraviolet) has been performed using CCD observations in a spectral band close to that of the guide sensor system of the T-170M space telescope. The mean photometric uncertainty of the Master Catalog at 14m −17m is 0.23m ; variations in this uncertainty over the sky are within a factor of two. About 2% of the stars in the Master Catalog were found to have photometric errors in excess of 2m . We analyze the correspondence of large photometric errors to flags of the Master Catalog. DOI: 10.1134/S1063772911010082
star catalog, and sends information on the coordinates of the current telescope pointing to the onboard control computer. As soon as the telescope is successfully pointed toward the object of interest with an accuracy of 0.1 , the information from the GSS is used to stabilize the spacecraft with an rms accuracy of 0.03 . The guide-star catalog to be used for the GSS is a special MC developed at the Institute of Astronomy [6]. The MC satisfies the following requirements: a high-density grid of guide stars everywhere on the sky, including positions near bright stars and in the fields of star clusters; high accuracy stellar coordinates; an observing epoch close to the time of the observatory’s launch; completeness to 17m in a photometric band compatible with the GSS. The MC was developed based on the 2MASS catalog [1], by reducing its infrared J magnitudes to optical RJ magnitudes [2]. The reference photometric system selected for this reduction was that of the optical RU magnitudes from the UCAC2 catalog [3]. All stars in the 2MASS catalog have photometric data-quality flags, comprising a fairly cumbersome system of symbols. We simplified this system and adjusted it to our particular goal: marking potentially “bad” stars in the MC. Based on the data from the 2MASS catalog, we assigned flags to each MC star, designated E, P , N , each of which can take values of 0 or 1. The photometric-accuracy flag E = 1 if the error e,
1. INTRODUCTION Our earlier paper [1] described the photometric calibration of the 2MASS catalog [2] performed to benefit the Spectrum-UV space project [3] and compile a Master Catalog (MC) of guide stars for the project. A comparison of our results with the approximate photometry in the UCAC2 catalog [4] indicated that they enable adequate reduction of infrared to optical photometry, satisfying the requirements of the Spectrum-UV pointing system. The current paper describes the results of a test of this transformation and our analysis of the correspondence between the photometric errors and MC flags via a comparison of the transformed magnitudes and actual optical observations of selected fields with the Zeiss-600 telescope at Mt. Terskol. 2. GUIDE-SENSOR SYSTEM AND GUIDE-STAR CATALOG OF THE SPECTRUM-UV PROJECT Fine pointing of the T-170M telescope of the Spectrum-UV project and its stabilization will be carried out using a guide-sensor system (GSS) [5] possessing three guide sensors at the edges of the telescope’s field of view. Each sensor contains a cooled “Elar” 1024 × 1024-pixel CCD detector as well as electronic boards. When pointing the telescope, the GSS obtains images of three fields on the celestial sphere, automatically identifies stars in these fields with the guide13
14
SHUGAROV et al.
Table 1. Photometric uncertainty of the MC for various fields Field
α
No. of deviations
σ
δ 15 −17 m
m
14 −16 m
m
>1
m
>2
m
>3
m
No. of stars
ASCC 101
19.227h
36.33◦
0.19m
0.13m
3.6%
5
4
2
140
ASCC 103
19.603
35.67
0.19
0.14
2.7%
8
5
5
291
ASCC 113
21.200
38.60
0.20
0.14
1.7%
4
3
1
233
IC 4996
20.275
37.63
0.29
0.26
4.6%
13
7
3
281
NGC 118
0.783
85.25
0.24
0.17
3.7%
13
0
0
345
NGC 6633
18.454
6.51
0.19
0.14
1.5%
2
2
2
135
NGC 6694
18.755
−9.38
0.23
0.20
3.6%
8
5
4
219
NGC 6705
18.851
−6.27
0.20
0.22
12.5%
23
19
15
184
NGC 6709
18.855
10.32
0.13
0.12
3.9%
5
3
3
128
NGC 6728
18.979
−8.97
0.24
0.25
1.9%
3
3
3
154
NGC 6811
19.621
46.39
0.25
0.15
5.0%
7
4
4
140
NGC 6830
19.850
23.10
0.25
0.20
9.1%
13
12
6
142
NGC 6866
20.065
44.16
0.26
0.23
1.1%
2
1
0
187
NGC 7063
21.406
36.49
0.22
0.18
0.5%
1
1
1
220
NGC 7209
22.088
46.48
0.23
0.17
2.5%
6
0
0
244
NGC 7243
22.252
49.90
0.24
0.16
3.2%
5
1
1
156
R1
23.150
10.47
0.32
0.32
5.4%
4
0
0
74
R2
1.316
7.65
0.15
0.21
11.5%
7
1
0
61
SA 113
21.683
0.42
0.28
0.23
4.6%
7
0
0
152
SA 115
23.733
1.07
0.32
0.32
2.9%
2
0
0
68
defined as the rms from the eJ and eH magnitude uncertainties in the 2MASS catalog, exceeds 0.5m . The photometric-distortion flag P = 1 if either the photometry of the source is uncertain in one of the Table 2. Mean photometric uncertainties of the MC from the 20 fields Range 12m –14m
Δ
σ
0.039m
0.23m
14–15
0.038
0.19
15–16
0.003
0.22
16–17
–0.009
0.25
17–18
0.130
0.31
18–19
0.420
0.35
2MASS bands or the photometric measurement is distorted and/or contaminated. The close-neighbor flag N = 1 if 2MASS contains indications that the source is multiple. These flags should be taken into account when using the MC. If [E, P, N ] = 1, the object should not be used to establish the photometric system and/or carry out testing. Distorted coordinates are possible when N = 1, so that using the object for coordinate and identification problems requires caution. At least one of the three flags is equal to 1 for 43% of all the MC stars. Among these, 41% have P = 1, 14% have N = 1, and 0.04% have E = 1. 3. OPERATING CONDITIONS OF THE GSS The following limitations and conditions should be kept in mind when using the GSS: ASTRONOMY REPORTS Vol. 55
No. 1
2011
SPECTRUM-UV (WSO/UV) PROJECT
15
1.0m
Δ
0.5
0
–0.5
–1.0 13
14
15
16
17
18m
Rj Fig. 1. Test for the presence of a brightness equation in the MC from 20 fields: relation between the magnitude differences Δ = RJ and RJ magnitudes.
Δ
0.5m
0
–0.5 –0.5
0 V–R
0.5m
Fig. 2. Test for the presence of a color equation in the region of NGC 7209: relation between the magnitude difference Δ = RJ − m and the V −R color index.
—the MC is complete to RJ = 17m ; —the rms uncertainty of the MC magnitude reductions to the RJ band is 0.1m −0.7m ; —the MC RJ photometric band is based on observations at 580–640 nm [4]; the GSS band is broader (200–1000 nm), with a maximum in the red (a CCD detector without a filter); ASTRONOMY REPORTS
Vol. 55 No. 1 2011
—the relative GSS sensitivity limit for 1 s exposure times is 17m , with the signal-to-noise (S/N) ratio being 10–20; – the GSS field of view will predominantly contain faint stars (16m −17m ); —for the conditions described above, one or two stars will be observed in the GSS field of view at high Galactic latitudes.
16
SHUGAROV et al.
Table 3. Statistics of the observations in terms of flags and deviations No. of stars
Flags∗ N P
>1m
>2m
>3m
3543
0.65m
138
70
49
2887
0.68
114
67
48
402
1.61
94
65
46
1X (for stars <17 )
275
1.79
83
63
46
X1
117
1.82
48
38
21
X1 (for stars <17m )
105
1.84
46
38
21
01
7
0.28
0
0
0
01 (for stars <17m )
7
0.28
0
0
0
110
1.85
48
38
21
98
1.86
46
38
21
XX m
XX (for stars <17 ) 1X m
11 11 (for stars <17m ) ∗
No. of deviations
σ
X means any state of the flag.
These factors indicate that, at high Galactic latitudes, the GSS will work at close to the limits of its sensitivity and the completeness of the MC. As a consequence, situations will occur when some of the MC stars will be considerably fainter than the GSS sensitivity limit or, on the contrary, some stars fainter than 17m in the catalog will appear in the field of view because they are actually brighter, hampering the work of the automatic identification algorithms. Because of this, the brightness of the observed stars becomes one of the most important identification factors. The photometric accuracy of the MC is subject to certain limitations due to the observed S/N, the uncertainties of the initial 2MASS catalog, errors in the reduction of its infrared magnitudes to optical magnitudes, and incomplete coincidence of the GSS spectral range with the RJ band. To test the GSS’ ability to work successfully and determine the actual photometric accuracy of the MC, we obtained ground-based observations in a spectral band matching that of the space telescope’s GSS as closely as possible. 4. OBSERVATIONS AND REDUCTION We used the S1C CCD camera mounted on the Zeiss-600 telescope of the Mt. Terskol Observatory between August 13 and 21, 2009 to observe 20 fields, mainly those containing star clusters (observer A.S. Shugarov). We chose these fields because we wanted to have as many stars as possible, observed under the same conditions. The coordinates of our
fields (equinox J2000.0) were taken from [7, 8] and are presented in Table 1. We performed our observations both without a filter and with filters realizing the V and R bands, with the exposure time of 120 s providing an S/N of about 50 when observing 17m stars without a filter. Our selected fields were distributed on the sky so that we were able to check the accuracy of the MC both in the Galactic plane, where the reduction uncertainties may be largest due to the high interstellar extinction, and far from the plane. We used the MaximDL code to perform the photometric reduction of our CCD images for 3543 stars in these fields. We determined the photometric zero point in each of the fields using the criterion of a zero mean difference between the observed magnitudes and those from the MC in the range 15m −17m , since we are interested in the relative accuracy of the MC photometry. Thus, we determined the magnitudes m of all the stars in a system close to the instrumental system of the GSS. 5. ANALYSIS OF THE MC PHOTOMETRIC CHARACTERISTICS We assumed that the spectral sensitivity of the CCD camera mounted on the ground-based telescope used approximately coincided with that of the GSS of the T-170M space telescope, and that the magnitudes m derived from our CCD observations were considerably more accurate than the RJ magnitudes from the MC. In this case, the differences ASTRONOMY REPORTS Vol. 55
No. 1
2011
SPECTRUM-UV (WSO/UV) PROJECT
17
Fig. 3. Example of the most frequent type of large deviations in the MC: one of the stars in a close double (circled) is much too bright. Left: in the MC, the stars have RJ = 11.03m and 14.97m ; Right: in our CCD observations, the stars have m = 15m and 17m .
Δ = RJ − m between the MC and CCD magnitudes and their rms scatter σ will indicate the photometric quality of the MC in the systematic and random sense. Table 1 presents the values of σ for each of the fields in the ranges 15m −17m and 14m −16m . We excluded stars with Δ > 1m from these calculations. The Δ and σ values for different fields indicate that: —the mean accuracy of the MC differs by up to a factor of two for different fields; —the mean accuracy in a field is not correlated with the Galactic latitude or the stellar density in the field; —the number of large deviations from the MC magnitudes is correlated with the stellar density in the field (not presented in Table 1). The photometric uncertainty of the MC, σ, as a function of magnitude is presented in Table 2. In the working range of the GSS, 14m −17m , σ is 0.23m , which is satisfactory for automatic identification. Our measurements in the range 17m −19m are not representative, as they were performed for only some of the fields. To test for the presence of systematic errors such as the so-called brightness and color equations, we plotted Δ versus magnitude for all the fields (Fig. 1) and Δ versus the V −R color index for several fields (Fig. 2), in the region of sky containing the cluster ASTRONOMY REPORTS
Vol. 55 No. 1 2011
NGC 7209. Within the errors, we find no brightness and color equations in the working range of the GSS, 14m −17m , testifying to the good quality of the reduction of the infrared magnitudes from the initial 2MASS catalog to the optical RJ magnitudes. We performed a statistical analysis of the deviations and their relation to the flags. The statistics for the P, N flags are collected in Table 3; no cases of the E = 1 flag were encountered among the measured stars. In this case, the σ values were obtained without rejecting large deviations. In the MC, 43% of stars have at least one non-zero flag, but there are only 11% stars with non-zero flags among the measured stars. This considerable difference is due to a selection effect: the reduction was performed for isolated stars, suitable for accurate photometry. On average, such stars have better photometry in the MC. When processing our fields, we looked for large deviations (Δ > 2m ), also in high-density stellar fields not suitable for photometry. We recorded the detected large deviations even when accurate photometry was impossible. Thus, qualitatively speaking, the indicated selection effect was much less important for large than for small deviations. The statistics of the deviations for stars brighter than 17m show that: —the number of deviations in excess of 2m that are able to hinder the operation of the GSS star-
18
SHUGAROV et al.
identification algorithms is about 2%, while the number in excess of 1m is about 4%; —94% of all stars with deviations in excess of 2m have the flag P = 1, although 70% of all stars possessing this flag have normal photometry; —in 35% of cases, the flag P = 1 goes together with the flag N = 1; this combination gives 60% of the total number of deviations exceeding 2m that have the flag P = 1, i.e., the simultaneous presence of these two flags increases the probability of a large deviation; —the flag N = 1 with P = 0 is rare; all such stars have good quality photometry. The detected correlation of the P and N flags with large deviations can be used for the development of star-identification algorithms. Our analysis shows that the appearance of considerable photometric deviations is related to problems of the 2MASS photometry in the H band. In cases of unreliably recorded fluxes (e.g. for faint images) and/or distortions due to a close companion, the 2MASS catalog presents either an upper limit for the expected flux or the integrated flux from the unresolved sources. The resulting J − H color index then becomes extremely negative, and the calibration formula from [1] causes a large deviation. More than half of all the large deviations in the MC are found for close double stars (Fig. 3), with one star in the pair having sufficiently accurate photometry and the other being too bright. 6. CONCLUSIONS The following conclusions follow from this study. (1) The average uncertainty of the RJ magnitudes in the MC for the measured fields is 0.23m in the range 14m −17m ; variations of the MC photometric uncertainty over the sky are within a factor of two. (2) There are no brightness or color equations for the MC magnitudes. (3) About 2% of all MC stars have photometric errors in excess of 2m , about 94% of which are flagged. (4) The vast majority of all flagged stars have good photometry.
ACKNOWLEDGMENTS The authors wish to thank M.V. Andreev (Mt. Terskol Branch of the Institute of Astronomy) and V.V. Sasyuk (Engelhardt Astronomical Observatory) for their assistance during the observations. This study was supported by the Program of State Support for Leading Scientific Schools of the Russian Federation (grant NSh-4354.2008.2). REFERENCES 1. A. E. Piskunov, N. V. Kharchenko, and N. V. Chupina, Pis’ma Astron. Zh. 34, 285 (2008) [Astron. Lett. 34, 256 (2008)]. 2. R. M. Cutri, M. F. Skrutskie, S. Van Dyk, et al., The 2MASS All-Sky Catalog of Point Sources (Univ. of Massachusetts, MA, 2003). 3. B. M. Shustov, A. A. Boyarchuk, A. A. Moisheev, et al., in Ultraviolet Universe II, Proceedings of an All Russian Conference, Ed. by B. M. Shustov, M. E. Sachkov, and E. Yu. Kil’pio (Yanus-K, Moscow, 2008), p. 7 [in Russian]. 4. N. Zacharias, S. E. Urban, M. I. Zacharias, et al., Astron. J. 127, 3043 (2004). 5. A. S. Shugarov, in Ultraviolet Universe II, Proceedings of an All Russian Conference, Ed. by B. M. Shustov, M. E. Sachkov, and E. Yu. Kil’pio (Yanus-K, Moscow, 2008), p. 78 [in Russian]. 6. N. V. Chupina, A. E. Piskunov, and N. V. Kharchenko, in Ultraviolet Universe II, Proceedings of an All Russian Conference, Ed. by B. M. Shustov, M. E. Sachkov, and E. Yu. Kil’pio (Yanus-K, Moscow, 2008), p. 60 [in Russian]. ¨ 7. N. V. Kharchenko, A. E. Piskunov, S. Roser, et al., Astron. Astrophys. 438, 1163 (2005). ¨ 8. N. V. Kharchenko, A. E. Piskunov, S. Roser, et al., Astron. Astrophys. 440, 403 (2005).
Translated by N. Samus’
ASTRONOMY REPORTS Vol. 55
No. 1
2011