Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.sao.ru/hq/marat/publ/articles/english.pdf Äàòà èçìåíåíèÿ: Wed Feb 24 17:32:53 2010 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 08:25:43 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: active galaxy

Astronomy Reports, Vol. 47, No. 11, 2003, pp. 903­915. Translated from Astronomicheski Zhurnal, Vol. 80, No. 11, 2003, pp. 978­991. i Original Russian Text Copyright c 2003 by Gorshkov, Konnikova, Mingaliev.

Spectra, Optical Identifications, and Statistics of a Complete Sample of Radio Sources at Declinations 10 ­12 30
A. G. Gorshkov1 , V. K. Konnikova1 , and M. G. Mingaliev
1

2

Sternberg Astronomical Institute, Universitetski i pr. 13, Moscow, 119992 Russia 2 Special Astrophysical Observatory, Nizhni i Arkhyz, Russia
Received April 7, 2003; in final form, May 8, 2003

Abstract--The results of 0.97, 2.3, 3.9, 7.7, 11.1, and 21.7 GHz observations of a complete sample of radio sources obtained on the RATAN-600 radio telescope are presented. The sample is comprised of sources from the 4.85-GHz MGB survey, and contains all sources at declinations 10 ­12 30 (J2000) with Galactic latitudes |b| > 15 and flux densities S4.85 > 200 mJy. Optical identifications have been obtained for about 86% of the radio sources with flat spectra and 59% of those with steep spectra. The spectra of the flat-spectrum sources have been decomposed into extended and compact components. c 2003 MAIK "Nauka/Interperiodica".

1. We are currently studying two complete samples of radio sources. The first sample, which was derived from the Zelenchuk Survey at 3.9 GHz [1], contains all sources in this survey with flux densities S3.9 > 200 mJy at declinations 3 30 ­6 (B1950); the right ascensions are in the range 0­24h , and the Galactic latitudes are |b| > 10 . This sample contains 160 objects, and we have been studying it since 1984. The second sample, derived from the GB6 catalog at 4.85 GHz [2], contains all sources from the catalog with flux densities S4.85 > 200 mJy at declinations 10 ­12 30 (J2000); the right ascensions are in the range 0­24h , and the Galactic latitudes are |b| > 15 . In this sample of 153 objects, 83 have flat spectra with (3.9-7.7) > 0.5 (S ) and 70 have steep spectra with (3.9-7.7) < -0.5. The main goals of our investigations of these samples are the following: (1) studying the variability of the sample sources on time scales from several days to several years (observations over the wide frequency range 0.97­ 21.7 GHz can be used to derive the main characteristics of the variability--its time scale and amplitude, as well as the spectrum of the variable component and the time dependence of its amplitude­frequency characteristics); (2) deriving the statistical properties of the radiosource spectra; (3) searching for interesting objects with unusual characteristics in both the radio and optical ranges;

(4) investigating the cosmological evolution of quasars (which requires that redshifts be found for the majority of the optical objects identified with the radio sources). Daily multi-frequency observations on the RATAN-600 telescope in 1998­1999 confirmed the presence of flux-density variability on time scales of about four days detected earlier [3­5]--so-called intraday variability (IDV). Our studies have shown that such variability is inherent to virtually all flatspectrum sources, and that IDV is characterized by a flat frequency spectrum with a mean modulation index of about 2% at 2.3­21.7 GHz. Roughly 20% of the objects display significant IDV, as is clearly visible in their structure functions. Several sources show cyclic variability on characteristic time scales of 4­25 days. The statistical properties of the longterm variability of sources in the complete sample and the properties of individual flares for the most active sources have been derived [1, 6]. Various statistical characteristics of the spectra of the sample sources have also been obtained [7, 8]. The unique radio source 0527 + 0331, found to have the most prominent long-term variability, was discovered and studied (all source names are comprised of the first four digits of their right ascension and declination for epoch J2000) [4, 9, 10]. Optical observations aimed at obtaining spectra and redshifts of objects identified with the sample sources to 21m are ongoing. Studies of the second sample were begun to enable confirmation of the results obtained for the first sampleonthe basisof a different statistically independent ensemble of radio sources. Up through 2002, three

1063-7729/03/4711-0903$24.00 c 2003 MAIK "Nauka/Interperiodica"


904

GORSHKOV et al.

series of daily multi-frequency observations were carried out for most of the sample sources over 80 days in 2000, 104 days in 2001, and 98 days in 2002. We present here the first results obtained for the second sample of radio sources. 2. OBSERVATIONS Meridian observations were carried out on the Northern sector of the RATAN-600 radio telescope at 0.97, 2.3, 3.9, 7.7, 11.1, and 21.7 GHz in June­ November 2001. The parameters of the receivers used are presented in [11]. The observations were carried out in a fixed-focus regime, as described in [12]. Readjustment of the main mirror was possible within ±1.25 of the center of the declination zone. An equal number of panels of the main mirror were mounted at all altitudes, in order to reduce the influence of variations in the radiation of the outer panels as the curvature of the circular reflector changed. The effective area remained constant over the entire frequency range. The calibrator for the observations was 1347 + 1217, whose angular size is much smaller than the horizontal cross section of the antenna beam right up to 21.7 GHz. We took the flux density of 1347 + 1217 to be 6.15, 4.12, 3.23, 2.36, 1.99, and 1.46 Jy at 0.97, 2.3, 3.9, 7.7, 11.1, and 21.7 GHz, respectively. The observations were reduced using a program package that yielded both the flux densities for individual observations and the mean flux density over an entire observing series. The reduction was based on optimal filtration of the input data using the method described in detail in [13]. Before this optimal filtration, non-linear filters were used to clean the input data of impulsive noise, jumps, and trends with time scales longer than the scale of the antenna beam in right ascension. When deriving the mean flux density over an entire observing interval, we used only those recordings for which the noise dispersion at the location of the source belonged to a single general population; the method used to reduce such recordings is described in [14]. The mean flux densities were determined via optimal filtration of the mean recording, whose ith point was the median value of all the ith points of the cleaned input recordings. As a check, we also derived the mean flux density from the relation
n

only recordings belonging to a single general population were summed. It is clear that the flux densities obtained in these two ways should be similar, and appreciable differences between them suggest the presence of bad recordings that were not removed during the initial filtration. Our experience demonstrated that such differences arose fairly rarely, testifying to the correctness of the filtration algorithm applied. Bad recordings that were not excluded were primarily those corresponding to observations made with an incorrect antenna setup. If a significant difference was observed, all the recordings were inspected visually and any considered suspicious were excluded, after which the entire reduction procedure was repeated. The errors in the measurements were determined in two ways:
1 2

=
2



2 i

A2 i

,

where is the dispersion of the residual noise in a mean recording after excluding the detected source and Ai is the tabulated antenna beam, and
n
1 2

s =
i

¯ (Si - S )2

n(n - 1)

,

¯ where S is the mean flux density given by (1). These two estimates should be similar. If they belonged to different general populations (according to the Fisher criterion), we likewise searched for bad recordings. In any case, the larger of the two estimates was adopted as the error in the measured flux density. In this approach to estimating the errors, the errors also include the rms variation of the flux density associated with intrinsic variability of the source during the series of observations. For several steep-spectrum sources for which significant linear polarization was found in [15], we constructed the flux density using the formula [16] S = 0.5S0 [1 + p cos 2(q - - )], where S0 is the total flux density of the source, p and q are the degree of linear polarization and the parallactic angle, respectively, is the angle between the plane of linear polarization of the receiver and the vertical, and is the position angle of polarization. 3. OPTICAL IDENTIFICATIONS We used accurate radio coordinates taken primarily from the JVAS (Jodrell Bank­VLA Astrometric Survey) catalog at 8.4 GHz [17] (an rms coordinate error of 0.014 ) and the NVSS (NRAO VLA
ASTRONOMY REPORTS Vol. 47 No. 11 2003

¯ S=
i

S

i

n,

(1)

where Si is the flux density of the ith observation and n is the number of observations. Introducing a weighting function is superfluous in this case, since


COMPLETE SAMPLE OF RADIO SOURCES

905

Sky Survey) catalog at 1.4 GHz [18] (mean rms coordinate errors of about 0.11 and 0.56 for right ascension and declination, respectively). The optical coordinates and B magnitudes were taken from the USNO astrometric survey [19] or APM [20]. A radio source was considered to be identified with an optical object if the difference between the radio and optical coordinates was less than 3 of the error in the radio range coordinates. A significant number of objects identified with sources in the sample had been classified earlier. All redshifts and object classifications presented here were obtained from the catalog of quasars and active galactic nuclei [21] published in 2001, the NED extragalactic database [22], and our earlier work [23­ 25]. Thirteen objects identified with sources in the sample have not yet been classified. 4. RESULTS

N 20 18 16 14 12 10 8 6 4 2 0

(a)

(b)

0.4 0.6 0.8 1.0 1.2 1.4 0.4 0.6 0.8 1.0 1.2 1.4 ­ ­
Fig. 1. Distribution of spectral indices for (a) all sources with steep spectra at 3.9 GHz and (b) sources with S spectra.

4.1. Radio Sources with Steep Spectra and the Flux-density Scale
The sample includes 70 sources with steep spectra. We did not include double-lobed sources whose total flux density at 4.85 GHz was less than 400 mJy in the sample. We did not observe ten sources because either they were larger than the size of the antenna beam at most of the frequencies observed or they were double sources that were unresolved in our observations. Table 1 presents for these sources (1)­(2) their coordinates, (3) classification (as quasars, Q, or galaxies, G), (4) redshifts, (5) references for the redshifts, and (6) B magnitudes. The spectrum of the calibrator 1347 + 1217 at 0.97­21.7 GHz was approximated by a power law, log S = A + B log . The spectra of 68% of the sources could be approximated well with such linear relations (S spectra), while the spectra of the remaining sources were approximated using parabolas of the form log S = A + B log + C log2 . The spectra of three sources flatten at high frequencies (type C+ ), probably due to the presence of a compact component, and the spectra of 15 sources flatten at low frequencies due to synchrotron selfabsorption (type C- ); this flattening is most evident at frequencies below 2.3 GHz, while the spectra remain close to linear power laws at higher frequencies. The flux densities for 36 of the steep-spectrum sources were below our detection limit at 21.7 GHz. The measured flux densities for some of the sources are underestimated, mainly at frequencies higher than 7.7 GHz, because the size of the source was comparable to the size of the antenna beam in right ascension. This underestimation is consistent with the angular sizes given in the Texas 0.365 GHz survey [26].
ASTRONOMY REPORTS Vol. 47 No. 11 2003

For these sources, the spectra were approximated using only the data at frequencies where the source angular size was smaller than the antenna beam. Table 2 presents information about the 60 sources with steep spectra. The columns give the (1)­(2) J2000 coordinates of the objects, (3) optical identifications (quasar, Q, or galaxy, G; a "+" indicates that the source has been identified but a spectrum is not yet available), (4) redshifts, (5) references for the redshifts, (6) B magnitudes, and (7)­(9) coefficients A, B , and C for the approximations to the measured spectra. A star in the last column indicates sources whose angular sizes are comparable to the size of
Table 1. Coordinates and optical identifications of extended radio sources in the sample Radio coordinates (J2000.0) R.A. 02h 17m 07.62s 03 58 57.71 04 13 42.14 09 14 19.36 1229 51.84 12 30 48.27 14 16 53.51 15 14 49.50 22 49 54.59 23 15 34.40 DEC. 11 04 10.59 10 27 17.76 11 11 44.82 10 06 38.28 11 40 24.09 12 23 33.07 10 48 40.06 10 17 00.74 11 36 30.84 10 27 18.40 Q G G G G G G EF G G 0.026 0.255 [22] [25] 9.4 17.6 0.408 0.031 0.306 0.311 0.083 0.004 0.024 [21] [21] [21] [25] [22] [21] [22] 15.9 15.4 19.5 19.9 16.0 12.9 13.0 Id z Ref. V


906

GORSHKOV et al.

Table 2. Optical identifications and approximations of the sample constants Radio coordinates (J2000.0) R.A. DEC. 00h 29m 08.953s +11 36 28.20 00 34 56.157 +10 27 52.03 +10 03 23.14 00 40 50.354 00 44 34.653 +12 11 19.32 +11 21 42.40 02 20 47.458 02 38 30.865 +10 10 07.66 02 45 14.608 +10 47 01.57 03 15 21.039 +10 12 43.12 03 27 23.109 +12 08 35.71 04 40 12.413 +11 34 03.83 07 28 32.882 +12 10 10.39 07 45 28.211 +12 09 28.85 08 04 47.974 +10 15 22.73 09 06 04.192 +11 03 27.61 +10 51 06.16 10 14 16.028 10 34 05.090 +11 12 31.95 11 00 47.733 +10 46 12.98 +11 03 23.90 11 04 34.802 11 09 46.038 +10 43 43.21 11 26 27.190 +12 20 33.10 11 30 19.246 +10 15 26.30 11 40 27.693 +12 03 07.44 11 53 03.107 +11 07 20.29 11 59 29.114 +10 46 01.00 12 04 26.711 +11 29 09.68 12 23 08.875 +10 29 01.05 12 28 36.804 +10 18 41.69 12 31 19.866 +11 22 45.03 13 06 19.248 +11 13 39.79 13 09 05.161 +10 29 39.91 13 21 18.844 +11 06 49.25 13 41 04.302 +10 32 05.96 13 47 33.425 +12 17 23.94 13 52 56.363 +11 07 07.57 14 23 30.103 +11 59 51.24 15 11 29.436 +10 01 43.73 15 22 12.151 +10 41 30.35 +11 30 23.76 15 23 27.563 15 23 56.936 +10 55 44.02 15 59 06.913 +12 10 26.95 15 59 16.840 +11 15 46.11 16 04 05.732 +11 27 59.88 16 21 10.388 +10 46 13.88 Parameters Approximation B C -0.792 -0.086 -0.636 -0.984 -0.832 -0.846 -0.108 -0.938 -0.796 -0.925 0.240 -0.815 0.081 -0.861 -0.087 -0.891 -1.067 -1.030 -0.874 -0.673 -0.224 -0.908 -0.936 -0.792 -0.138 -0.562 -0.377 -0.910 -0.505 -0.913 -1.043 -0.116 -0.920 -0.766 -0.643 -0.132 -0.885 -0.709 -0.104 -0.605 -0.786 -0.778 -0.052 -0.988 -0.463 -0.739 -0.353 -0.635 -0.874 -0.823 -0.555 -0.786 -0.115 -0.814 -0.746 -0.892 -0.743 +0.118
Vol. 47 No. 11

Id EF G G G EF G EF G EF EF EF EF Q G + EF Q EF EF G EF G EF EF EF EF Q EF G EF Q EF G EF Q EF G + + EF EF EF Q

z

Ref.

B 2. 3. 3. 2. 3. 3. 2. 3. 3. 2. 3. 3. 3. 3. 3. 3. 2. 3. 3. 3. 2. 3. 3. 2. 2. 2. 2. 2. 2. 2. 3. 2. 3. 3. 3. 2. 2. 2. 3. 2. 2. 2. 3. A 984 021 872 934 278 171 740 331 191 998 152 007 415 011 048 175 954 006 244 123 694 301 079 953 893 843 941 979 684 826 475 902 783 295 077 935 783 697 137 839 926 871 107

Comments (see text) * * *

0.057 0.188 0.228

[22] [21] [21]

15.0 18.7 18.9 19.6

0.222

[25]

19.6

1.956

[21] [25]

17.6 19.2 18.7 17.9

*

0.420

[21]

*

19.2 15.0

2.305 0.084 2.175 0.121 1.611 0.204

[21] [24] [21] [21] [21] [25]

19.0 15.4 18.4 14.9 17.5 21.0 19.1 21.1

*

1.305

[25]

21.8

ASTRONOMY REPORTS

2003


COMPLETE SAMPLE OF RADIO SOURCES Table 2. (Contd.) Radio coordinates (J2000.0) R.A. 16h 31m 45.247s 16 38 22.118 16 40 47.989 17 09 35.035 17 13 43.382 17 27 53.762 20 22 08.547 2138 26.222 215104.053 22 0116.687 22 03 45.543 22 29 57.456 22 4134.421 22 54 10.450 23 11 17.868 23 12 10.467 23 29 41.090 DEC. +11 56 02.99 +10 35 07.74 +12 20 02.08 +11 40 27.72 +10 37 24.86 +10 42 56.50 +10 01 10.81 +11 58 04.15 +12 19 50.45 +10 23 47.59 +121715.63 +11 27 37.73 +11 45 44.84 +11 36 37.90 +100815.35 +12 24 03.46 +11 17 28.60 Q EF EF + + Q Q EF EF Q + G EF Q Q Q G 0.325 0.432 1.285 0.119 [21] [21] [25] [22] 15.6 16.2 19.1 18.4 0.239 [23] 1.729 [21] 18.1 18.9 21.2 0.833 0.469 [21] [21] 18.8 19.3 20.2 19.4 1.792 [21] 18.2 Id z Ref. B A 3.306 3.144 3.382 2.865 2.869 2.784 3.651 2.979 3.114 2.832 2.895 2.794 3.062 3.283 2.950 2.842 2.758 Parameters Approximation B ­0.582 ­0.614 ­0.414 ­0.780 ­0.749 ­0.891 ­0.941 ­0.940 ­0.958 ­0.848 ­0.844 ­0.783 ­0.841 ­0.739 ­0.897 ­0.745 ­0.507 * * ­0.070 * * * ­0.084 ­0.064 * C Comments

907

(see text)

the antenna beam in right ascension at frequencies higher than 7.7 GHz. The flux density of 1140 + 1203 at 3.9 GHz is appreciably lower than the values approximated using the remaining data. Figure 1 shows the distribution of spectral indices for (a) all the sources and (b) the sources with type S spectra. The mean spectral index for the sources with S spectra is = -0.808 with an rms deviation ¯ of = 0.15, while the mean spectral index at 3.9 GHz for all the sources is = -0.815 with an rms devi¯ ation of = 0.17. In [7], we obtained for the steepspectrum sources in a sample at declinations 3 30 - ¯ 6 (B1950) the mean spectral index = -0.857 with an rms deviation of = 0.14. After taking into account sources that were obviously resolved by the antenna beam, this value decreased to = -0.81, ¯ which is virtually identical to the mean spectral index obtained for our second sample in a different range of declinations. The mean ratio of our 1.4-GHz flux densities and the flux densities from the NVSS survey [18] for
ASTRONOMY REPORTS Vol. 47 No. 11 2003

all the steep-spectrum sources with sizes much less than the size of the antenna beam below 11.1 GHz is Sapp1.4 /SNVSS = 1.019 ± 0.004. The mean ratio of our 4.85-GHz flux densities and the flux densities of the GB6 survey is Sapp4.85 /SGB6 = 1.09 ± 0.01. On average, our flux-density estimates for 0.365 GHz are in agreement with those in the Texas survey, although the accuracy of our approximations outside the range of our observed frequencies is lower than the accuracy within this range. Of the 70 steep-spectrum radio sources, 21 are identified with galaxies (17 of which have measured redshifts), 14 are identified with quasars, and 29 are in empty fields to 21m . The spectra of six objects have not yet been obtained. Figure 2 shows the redshift distributions of the galaxies and quasars. The mean redshift of the galaxies is z (G) = 0.15 ( = ¯ 0.1), while that of the quasars is z (Q) = 1.22 ( = ¯ 0.7).


908

GORSHKOV et al.

N 5 4 3 2 1 G QSO

N 18 16 14 12 10 8 6 4 2 QSO G L

0

0.1

0.2

0.3

0.4 0 0.5 1.0 1.5 2.0 2.5 3.0 z z

0

1

2

3

4 z

0.2

0.6

1.0

1.4 z

Fig. 2. Redshift distribution of galaxies and quasars identified with steep-spectrum sources.

Fig. 3. Redshift distribution for the quasars, galaxies, and BL Lac objects identified with flat-spectrum sources in the sample.

4.2. Flat-spectrum Radio Sources ( > ­0.5)
Table 3 presents for the flat-spectrum sources their (1)­(2) J2000 radio coordinates, (3) optical identifications, (4) redshifts, (5) references for the redshifts, (6) B magnitudes, and (7)­(18) flux densities and rms measurement errors. Flux densities were obtained at six frequencies for all these sources except 2203 + 1007, whose flux density at 0.97 GHz does not exceed 70 mJy, and 0448 + 1127, whose 7.7 GHz flux density we were not able to estimate because the neighboring strong source 0449 + 1121 fell in the second feed horn. The flux densities of 0449 + 1121 and 1728 + 1215 are significantly and rapidly variable, and the table presents the spectra obtained on June 5, 2001and August 30, 2001, respectively. Of the 83 flat-spectrum objects in the complete sample, 72 (86%) have optical identifications, and 11 are in empty fields to 21m . The spectrum of the optical object identified with 1603 + 1105 is stellar, probably because the real optical object is blended with a star. Forty-nine of the sources are classified as quasars and eight as galaxies (seven of which have measured redshifts). Seven sources are BL Lac objects (three of which have measured redshifts), among them 0409 + 1217, 0757 + 0956,and 1309 + 1154, which have high degrees of polarization in the radio. Six objects have not yet been classified. Figure 3 presents the redshift distribution of the quasars, galaxies, and BL Lac objects. The mean redshift of the quasars is z (Q) = 1.40 (rms deviation ¯ (Q) = 0.74), of the BL Lac objects is z (L) = 0.55 ¯ ( (L) = 0.40), and of the galaxies is z (G) = 0.26 ¯ ( (G) = 0.19).

Figure 4 shows the redshift dependence of the absolute spectral luminosity at 11.1 GHz for the flatspectrum sources in the sample calculated for a uniform, isotropic cosmological model with a zero cosmological constant, deceleration parameter q = 0.5, and H = 50 km s-1 Mpc-1 . The solid line represents the minimum luminosity that can be detected for the given flux-limited sample. Most of the flat-spectrum sources are composed of at least two components--an extended one and a compact one. To study the characteristics of the spectra of compact components, we must isolate these spectra by excluding the contribution of any extended components. Here, when we refer to the "compact component," we mean the total radiation of the jet, which is made up of clouds of relativistic electrons that have formed as a result of propagating shock waves and are located in various stages of their evolution. The formation and evolution of these clouds under the action of shocks in the relativistic jet plasma is thought to give rise to the observed variability of extragalactic radio sources [27­29]. Valtaoya et al. [30] present a semi-quantitative analysis of a "generalized model" that can be applied to trace the evolution of the spectral and temporal characteristics of flares occurring at various stages of the interaction of a shock front and cloud of relativistic electrons. Spectra covering more than six decades in frequency are constructed in [31] based on nearly coordinated, simultaneous observations carried out from ° 20 cm to 1400 A. The spectra of active sources are smooth, and are fit well by a parabolic dependence on a logarithmic scale. At the same time, the spectra of sources with low activity are less smooth and poorly
ASTRONOMY REPORTS Vol. 47 No. 11 2003


Table 3. Identifications and flux densities of flat-spectrum sources

Radio coordinates (J2000.0) R.A. DEC. 00 07 55.710 00 10 31.007 00 36 23.767 00 37 26.042 00 38 18.017 00 42 44.371 0121 29.001 0121 41.595 0143 31.090 02 03 46.657 02 11 13.177 02 25 41.910 02 42 29.171 03 02 30.548 03 09 03.625 03 2153.104 03 45 01.317 03 55 45.553 04 09 22.009 04 44 12.467 04 48 50.413 04 49 07.672 05 09 27.457 05 16 46.646 07 45 33.060 07 49 27.385 07 50 52.047 07 57 06.640 07 58 07.658 08 27 06.513
h m s

Id EF G Q EF Q EF Q Q Q Q L Q Q EF Q Q Q Q L Q Q Q L Q EF G Q L G Q

z

Ref.

B

0.97 GHz 310 90 822 235 1196 257 88 1532 348 682 160 380 1690 514 459 2000 430 960 1003 820 92 689 663 1330 2503 219 1401 1089 557 185 25 15 25 20 26 20 15 20 15 24 15 22 25 29 24 36 15 21 26 32 15 24 20 20 35 43 34 41 30 20

Flux density and its error, mJy 2.3 GHz 3.9 GHz 7.7 GHz 11.1 GHz 300 148 553 235 781 252 150 1704 220 882 297 361 1341 536 544 1844 296 595 1076 596 178 1279 552 940 3930 180 1385 1121 393 144 11 10 10 14 10 18 15 17 10 11 9 8 15 18 12 19 10 10 15 18 15 15 12 12 30 15 39 12 20 12 285 182 501 230 683 243 186 1779 173 902 317 333 1166 501 651 1748 205 527 938 500 230 1732 464 818 3726 182 1516 1130 346 141 4 2 7 9 6 6 6 18 3 5 5 4 7 5 8 4 3 5 9 13 11 21 5 5 23 4 31 9 5 5 262 257 483 215 564 215 210 1948 136 984 333 280 1156 415 835 1442 158 443 814 377 ­ 2573 467 655 2956 169 1875 1320 298 138 6 3 10 4 6 6 6 13 4 7 5 5 9 7 9 10 3 5 7 5 ­ 29 10 6 22 9 32 11 7 5 255 354 490 223 539 230 181 2030 128 1039 364 282 1266 361 1018 1471 149 398 775 340 280 3235 490 583 2502 165 2370 1433 294 127 8 4 10 4 6 7 11 14 2 8 5 4 9 10 10 10 3 5 7 8 11 30 8 6 20 11 21 17 7 7

ASTRONOMY REPORTS Vol. 47 No. 11 2003

21.7 GHz 225 650 500 226 440 210 150 1996 110 1060 420 303 1357 232 1500 1487 147 333 723 250 260 4377 589 470 1831 126 3252 1536 250 150 10 20 26 10 22 20 20 45 10 32 20 15 43 20 30 40 14 17 20 30 20 115 21 20 40 15 70 56 20 25

+10 +10 +10 +11 +12 +10 +11 +11 +12 +11 +10 +11 +11 +12 +10 +12 +12 +12 +12 +10 +11 +11 +10 +10 +10 +10 +12 +09 +11 +10



27 58 07 09 27 09 27 49 15 34 51 34 01 18 29 21 18 31 17 42 27 21 11 57 11 57 31 56 36 52

43.89 29.51 57.43 50.91 31.25 49.19 00.53 50.42 42.95 45.39 34.79 25.47 00.72 56.77 16.34 13.95 48.77 46.14 39.85 47.29 54.39 28.63 44.59 54.77 12.69 33.12 04.83 34.80 46.05 24.15

0.089 1.909 1.395 2. 0. 1. 3. 487 570 178 610

[21] [21] [21] [23] [21] [25] [21] [23] [21] [25] [21] [21] [23]s [21] [21] [25] [25] [21] [25] [25] [25] [21] [21] [25] [21]

14.1 17.2 16.5 19. 19. 19. 19. 15. 17. 19. 18. 19. 19. 18. 19. 19. 19. 19. 19. 19. 19. 17. 14. 16. 17. 7 2 7 4 1 9 3 4 4 5 2 2 0 4 9 2 1 1 7 7 0 8

COMPLETE SAMPLE OF RADIO SOURCES

0.924 2.694 0.863 2.670 0.901 1.616 1.020 2.400 1.375 1.207* 1.580 0. 0. 0. 0. 2. 214 889 280 573 295

909


Table 3. (Contd.)

910

Radio coordinates (J2000.0) R.A. DEC. 08 32 08 33 09 45 09 46 09 47 10 01 10 02 10 15 10 42 11 03 11 18 11 32 12 07 12 18 12 54 13 09 13 15 13 27 14 30 14 44 14 53 14 55 15 04 15 07 15 25 15 50 15 55 16 03 16 08 16 27 16 40
h m

Id EF Q + Q Q Q + L Q Q Q Q Q Q Q L G Q Q Q Q EF Q G Q Q L Q Q G

z

Ref.

B

0.97 GHz 3 3 7 1 7 6 2 5 9 9 9 0 0 1 5 7 5 8 9 5 8 5 4 3 5 3 8 2 6 193 724 553 358 278 352 207 185 4191 305 2410 1154 112 247 730 850 298 560 220 367 383 450 1687 340 450 1042 314 185 1671 348 481 24 30 28 20 25 20 20 31 33 20 25 25 20 33 24 23 20 30 30 20 15 30 29 30 15 20 15 25 34 30 36

Flux density and its error, mJy 2.3 GHz 3.9 GHz 7.7 GHz 254 351 358 352 230 321 250 232 2435 293 1836 617 224 222 706 990 245 420 624 211 361 327 1637 295 361 544 300 194 2217 294 315 15 7 17 10 15 11 11 11 20 7 17 10 10 8 12 12 18 15 09 10 10 10 14 15 07 13 08 12 20 13 16 210 351 308 338 215 299 302 277 1750 320 1690 443 260 217 682 1062 219 411 889 160 295 222 1745 252 223 390 274 215 2711 275 276 8 3 4 4 5 3 6 3 12 4 15 4 9 2 10 6 8 9 5 4 4 6 10 4 2 4 3 7 17 7 9 191 290 246 280 187 267 360 367 1230 345 1534 364 265 283 746 1170 205 405 916 120 195 172 2136 176 284 279 328 237 3286 259 231 5 3 6 3 6 3 8 4 10 6 15 4 9 3 6 8 10 10 7 7 4 8 16 6 3 5 3 4 27 11 6

11.1 GHz 183 274 250 250 168 266 401 446 1049 388 1591 375 281 352 888 1238 210 409 911 110 161 146 2388 144 283 252 355 254 3714 252 219 10 3 8 3 10 4 13 5 9 5 14 4 5 3 7 10 7 8 8 7 2 6 18 7 3 8 4 6 37 7 14

21.7 GHz 230 253 262 181 134 226 405 532 766 474 1543 364 318 335 924 1248 215 415 778 98 102 115 2540 98 282 220 419 257 4258 220 230 20 10 24 16 15 11 20 27 11 20 38 19 18 18 30 30 20 20 29 10 14 18 76 15 18 20 17 30 133 20 36

38.478 14.368 49.860 35.069 45.857 57.735 52.846 44.024 44.530 03.530 57.302 59.491 12.625 26.094 38.256 33.933 01.853 54.465 09.739 50.736 44.241 55.418 24.980 21.882 02.936 43.595 43.044 41.930 46.203 37.032 58.892

s

+10 +11 +12 +10 +11 +10 +12 +12 +12 +11 +12 +10 +12 +11 +11 +11 +12 +12 +10 +11 +10 +11 +10 +10 +11 +11 +11 +11 +10 +12 +11



40 23 05 17 13 15 16 27 03 58 34 23 11 05 41 54 20 23 43 31 25 51 29 18 07 20 11 05 29 16 44

19.68 36.25 31.32 06.13 53.99 49.70 14.59 07.07 31.73 16.61 41.72 42.63 45.88 05.27 05.89 24.56 52.63 11.16 26.86 56.40 57.57 45.86 39.20 44.99 44.09 47.45 24.38 48.68 07.78 07.11 04.23

2.979 1.007 1.760 1.530

[21] [24] [21] [21] [24] [21] [24] [21] [21] [24] [21] [21] [21] [24] [25] [21] [21] [25] [21] [21] [21] [21] [21] [25] [22]

1. 0. 2. 0. 0. 1. 0.

028 917 118 540 896 403 870

0.261 0.950 1.710 0.851 1.770 1.833 0.331 0.436 0.360 1.226 1.216 0.078

18. 19. 18. 18. 17. 19. 19. 17. 18. 17. 16. 19. 19. 16. 18. 18. 19. 17. 17. 20. 18. 13. 17. 16. 14. 18. 17. 18. 15.

GORSHKOV et al.

ASTRONOMY REPORTS Vol. 47 No. 11 2003


Table 3. (Contd.) ASTRONOMY REPORTS Vol. 47 No. 11 2003

Radio coordinates (J2000.0) R.A. 16h 45m 54.675s 17 06 20.498 17 22 44.582 17 28 07.051 17 46 56.965 20 3154.995 20 35 22.334 20 49 45.865 2123 13.359 2145 18.776 2157 12.862 22 00 07.933 22 03 30.953 22 22 52.991 22 32 36.409 22 33 58.450 23 00 18.317 23 10 28.517 23 30 09.952 23 30 40.853 23 47 36.406 23 50 02.031 DEC. +11 13 52.64 +12 08 59.81 +10 13 35.77 +12 15 39.48 +11 27 17.35 +12 19 41.34 +10 56 06.78 +10 03 14.40 +10 07 54.96 +11 15 27.30 +10 14 24.80 +10 30 07.90 +10 07 42.58 +12 13 49.82 +11 43 50.89 +10 08 52.10 +10 37 54.08 +10 55 30.68 +12 28 28.60 +11 00 18.71 +11 35 17.89 +11 06 36.71

Id EF + Q Q EF Q Q EF Q Q Q + EF + Q Q Q G G Q + +

z

Ref.

B 0.97 GHz 583 19.0 115 380 256 555 875 1156 185 930 364 215 300 290 7822 395 79 303 510 1180 468 275 20 17 30 20 20 17 25 20 30 18 20 20 15 47 25 10 18 26 20 40 15 2.3 GHz 365 159 378 405 345 876 780 434 646 368 272 269 220 20.9 260 6297 320 130 274 343 1255 267 256 10 07 18 08 10 13 20 08 10 07 12 12 13 15 26 24 10 8 19 20 20 6

Flux density and its error, mJy 3.9 GHz 301 187 374 524 290 902 785 500 513 413 267 296 325 252 5316 360 168 447 269 1183 220 257 4 2 7 6 4 6 11 3 3 4 3 8 3 2 28 13 4 5 6 10 5 2 7.7 GHz 246 199 370 611 260 850 842 480 403 460 269 319 276 230 4266 350 159 357 206 1053 183 269 6 3 6 8 3 8 10 4 4 4 4 5 3 3 36 10 9 4 5 10 5 3 11.1 GHz 239 197 369 635 249 914 942 492 368 499 318 344 215 228 4264 315 161 395 190 1026 186 279 8 3 12 10 4 8 10 4 3 5 2 10 3 3 40 18 7 4 8 10 6 3 21.7 GHz 250 173 351 680 195 1034 1040 511 336 571 428 358 102 200 4686 275 148 427 153 923 153 280 30 11 34 29 10 20 47 22 18 16 16 26 10 10 70 20 15 22 20 14 20 17 COMPLETE SAMPLE OF RADIO SOURCES

0.732 0.588 1.215 0.601 0.932 0.550 0.761

[25] [25] [21] [21] [21] [21] [21]

21.6 21.7 18.2 16.6 18.1 18.0 18.4 21.7

1.037 1.854 2.816 0.494 0.144 1.489

[21] [23] [23] [25] [23] [21]

17.1 17.6 19.3 18.9 17.5 17.8 18.3 19.7

The optical spectrum of the object identified with the source 0449+1121 was obtained on the 6-m telescope of the Special Astrophysical Observatory of the Russian Academy of Sciences. The spectrum shows only a continuum, with no lines, and we have classified the object as a BL Lac object [25]. The references to the redshift presented in the 2001catalog of quasars and active galaxies (z = 1.207) do not include observational data for this object.

911


912

GORSHKOV et al.
33

L, 10 1000 100 10 1 0.1

erg/Hz

Flux density, mJy 1000

100

0.01 0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0 z

1

Frequency, GHz

10

Fig. 4. Redshift dependence of the absolute spectral luminosity at 11.1 GHz for flat-spectrum sources from the sample. The stars denote BL Lac objects and galaxies, while the circles denote quasars. See text for an explanation of the solid curve.

Fig. 5. Decomposition of the spectrum of 1327 + 1223 in June 2001. The solid curve shows the observed spectrum, the dotted curve the spectrum of the compact component, and the dot­dash curve the spectrum of the extended component.

described by parabolic fits. The spectra of variable sources at various stages of development of the variability are studied in [32] based on data obtained in 1965­1973 at 6.6 and 10.7 GHz [33], in 1965­1984 at 4.8, 8.0, and 14.5 GHz [34], and from 1978 to the present at millimeter wavelengths [35]. These spectra are also approximated well by logarithmic parabolas. In our subsequent analysis, we assume that the spectra of the radio sources are comprised of two components: a power-law component of the form log S = S0 + log and a compact component that can be represented by a quadratic function of the form log S = C +0.5(log - B )2 /A, where C is the logarithm of the peak flux density, B is the logarithm of the peak frequency, and A is the logarithmic interval from the peak frequency to the frequency at which d log S/d log = 1 (the curvature parameter of the spectrum) [31]. We decomposed the spectra into components by finding the solution with the minimum residual
s c [Si - (Si + Si )]2 ,

where Si is the measured flux density at the given s frequency, Si the flux density of the power-law comc ponent, and Si the flux density of the compact component. We went through this procedure for nearly all the sources using the data at 0.365 GHz [26].

We believe that the spectral characteristics of the extended components of flat-spectrum and steepspectrum sources are similar. Accordingly, we considered the decomposition to be successful if the spectral index for the extended component lay in the range = -0.5 ... -1.1. The spectra of the extended components were assumed to remain constant over all observing epochs. We present the characteristics of the spectra at epochs June­November 2001 below. We considered variations of these characteristics within only a narrow time interval from this epoch (one year). The spectra of the sample sources can be divided into four groups. (1) The spectra of 28 sources can be described with two components: an extended component with a power-law spectrum and a compact component that is approximated well by a logarithmic parabola with its maximum at a frequency no higher than 25 GHz. The spectrum of the extended component can be determined more accurately. The mean spectral index ¯ for the extended components is = -0.79, which is close to the mean spectral index obtained for the steep-spectrum sources in the sample, confirming the correctness of the procedure used for the spectral decomposition. Figure 5 shows the spectral decomposition for 1327 + 1223 as an example, while Fig. 6 show the decompositions for 2035 + 1056 at epochs (a) June 2002 and (b) June 2001. For 23 objects, the extended component comprises from 18% (1706 + 1208)to 100% (1042 + 1203) of the total flux density at 0.97 GHz. The extended
ASTRONOMY REPORTS Vol. 47 No. 11 2003


COMPLETE SAMPLE OF RADIO SOURCES

913

Flux density, mJy () 1000 (b)

N 10 8 6 4

()

(b)

1000

2 0 1 10 Frequency, GHz 10 0 5 10 15 20 25 30 0 10 20 30 40 50 Peak frequency, GHz
Fig. 7. Distribution of the peak frequencies of the compact components (a) in the rest frame of the observer and (b) in the rest frame of the source. The shaded region in the left diagram corresponds to sources with measured redshifts.

Fig. 6. Same as Fig. 5 for 2035 + 1056 at epochs (a) June 2002 and (b) June 2001.

components of five of the objects, such as 0121 + 1127 and 2203 + 1007, are small. The mean redshift of the quasars of this group is z = 1.58, with an rms deviation of = 0.61. Figure 7 ¯ shows the distribution of the peak frequencies of the compact components in the rest frames of the observer and source. The observed quasar peak frequencies show a dependence on redshift: max = 8.5 GHz ¯ for quasars with redshifts z > 1.3 and max =14 GHz ¯ for those with z < 1.3. In the rest frame of the source, this dependence disappears, and the mean peak frequencies for these redshifts become 25 and 26 GHz, respectively. We did not find any statistically significant correlations between the approximation coefficients for the compact components in the source rest frame and the dependences of these coefficients on the absolute spectral radio luminosity of the source. This same result was obtained for the compact components of sources with declinations 3 30 ­6 [8]. The flux densities of most of the sources in this group vary slowly, and their variability indices V = (Smax - Smin )/(Smax + Smin ) do not exceed 0.05. An exception is 1722 + 1013, whose flux density displays appreciable variability, with the variability index at 21.7 GHz being V = 0.32 over six months. In November 2001, the observed spectrum of this source rose with frequency, and the spectrum maximum was located beyond our observing interval. The time variations in the spectra of most sources are characterized by a shift of the maximum toward lower frequencies during the evolution of an isolated flare. (2) The spectra of 22 sources can be decomposed into a power-law component and a parabolic component whose peak frequency is appreciably higher than our highest frequency. The mean spectral index
ASTRONOMY REPORTS Vol. 47 No. 11 2003

and rms deviation for the extended components of this group do not differ from those for the previous group. The mean redshift for the quasars and BL Lac objects of this group is z = 0.99, with an ¯ rms deviation of = 0.57. The accuracy of the approximation is not sufficiently high to permit quantitative estimates. This group includes more active sources, a large fraction of which have their spectral maxima at frequencies above 30 GHz. The time variations of the spectra of such sources are characterized by variations in the spectral index corresponding to the section of the parabola that grows with frequency. For example, the spectral index of the compact component of 1015 + 1227 between 7.7 and 11.1 GHz varied from +0.38 to +0.55 over three years. The peak frequencies of the compact components of the three sources of this group with the highest redshifts [1118 + 1234 (z = 2.118), 1504 + 1029 (z = 1.883), and 1608 + 1029 (z = 1.226)] are at frequencies lower than 25 GHz at other observing epochs. The spectrum of the compact component of the galaxy 0010 + 1058 in September 2000 can also be approximated by a parabola with its maximum within our frequency range (10 GHz). Seven sources in this group have prominent extended components in our frequency range, and we can detect the compact component only at our highest frequencies. There is no fundamental difference between the first and second group of sources, and the peak frequencies are distributed fairly uniformly when observations over a broader range of frequencies are considered. The higher activity of sources in the second group is associated with their higher peak frequencies [30].


914

GORSHKOV et al.

(3) The spectra of 17 sources cannot be decomposed using two-component models. We believe that their spectra include contributions from several compact components. The measured points in the spectra are insufficient to enable an unambiguous decomposition into components. Only for two sources whose extended components were weak and for which there were observations at 0.365 GHz were we able to separate their spectra into two components, each of which has a parabolic spectrum. For example, the spectrum of 1453 + 1025 is the superposition of two parabolas with their maxima at 1.3 and 20 GHz. Most sources in this group remained complex at all observing epochs. The mean redshift of the quasars is z = 1.76, with rms deviation = 0.87. Many sources ¯ display appreciable flux-density variability at high frequencies, V21.7 > 0.13 over a year. Several sources also display appreciable variability at low frequencies, comparable in amplitude to the variability at high frequencies. For example, the variability index of 0409 + 1217 at 2.3 GHz is V = 0.33, while the variability index at 21.7 GHz is V = 0.29. More than halfofthe BL Lac objects belong to this group. (4) The 0.365­21.7 GHz spectra of 11 sources either can be approximated by a power law with an index from = -0.04 (0037 + 1109) to = -0.5 (1455 + 1151, 1507 + 1018) or flatten at frequencies above 3.9 GHz. This flattening is probably due to a compact component that is present at higher frequencies, although it is not possible to separate out the spectral components of these sources using the model considered. Sixty per cent of the identified sources in this group are galaxies, and sources that were observed more than once do not display significant variability. We were not able to interpret the spectra of 0833 + 1123 and 2310 + 1055. There are no flux-density measurements for the former source at frequencies below 0.97 GHz, while our observations show a flux density at 0.97 GHz of 724 mJy. In the case of the latter source, a nearby source falls into the antenna beam at several frequencies. 5. CONCLUSION We have derived flux densities in the range 0.97­ 11.1 GHz for all the sample sources with steep spectra and up to 21.7 GHz for one third of the spectra. The spectra of 68% of the sources can be approximated with power laws over this entire frequency range, and the spectra of 28% of the sources display self-absorption at low frequencies. The spectra of three sources flatten toward high frequencies, probably due to the presence of compact components whose peak frequencies are above our studied frequency range.

We have also derived 0.97­21.7 GHz flux densities for all the sample sources with flat spectra, and divided their spectra into extended and compact components. The extended components can be approximated by logarithmic power laws with spectral indices = -0.5 ... -1.1, while the compact components were described using logarithmic parabolas. We were able to obtain a satisfactory spectral decomposition for 50 sources, consistent with available flux-density measurements at low frequencies. The contribution of the extended component at 0.97 GHz can range from 0 to 100% for various sources. The spectra of 17 sources could not be decomposed using such two-component models, probably because they include the contributions of several compact components. We found no statistically significant correlations between the spectral parameters and the absolute spectral radio luminosities of compact components whose peaks are below 25 GHz in the source rest frame. All the sources have been optically identified. A significant number of the optical objects associated with the sample sources were classified earlier; work on obtaining optical spectra of the remaining optical objects is ongoing. Fifty-nine per cent of the steep-spectrum sources are identified to 21m ; among these, 52% are galaxies with a mean redshift of z = 0.15, 34% are quasars ¯ with z = 1.22, and the spectra of six objects have not ¯ yet been obtained. Eighty-six per cent of the flat-spectrum sources have optical identifications; among these, 68% are quasars with a mean redshift of z = 1.40, 10% are ¯ BL Lac objects with z = 0.55, 11% are galaxies with ¯ z = 0.26, and the remaining objects have not yet been ¯ classified. 6. ACKNOWLEDGMENTS This work was supported by the Russian Foundation for Basic Research (project no. 01-02-16331), a grant of the "Universities of Russia" program (project no. .02.03.005), and a grant from the State Science and Technology Program "Astronomy." REFERENCES
1. A. G. Gorshkov and V. K. Konnikova, Astron. Zh. 72, 291(1995) [Astron. Rep. 39, 257 (1995)]. 2. P. C. Gregory, W. K. Scott, K. Douglas, and J. J. Condon, Astrophys. J., Suppl. Ser. 103, 427 (1996). 3. D. S. Heeschen, Astron. J. 89, 1111 (1984). 4. A. G. Gorshkov, V. K. Konnikova, and M. G. Mingaliev, Astrophys. Space Sci. 278, 93 (2001). 5. A. G. Gorshkov, V. K. Konnikova, M. G. Mingaliev, et al., Astron. Zh. (in preparation).
ASTRONOMY REPORTS Vol. 47 No. 11 2003


COMPLETE SAMPLE OF RADIO SOURCES 6. A. G. Gorshkov, V. K. Konnikova, and M. G. Mingaliev, Astron. Zh. (in preparation). 7. A. M. Botashev, A. G. Gorshkov, V. K. Konnikova, and M. G. Mingaliev, Astron. Zh. 76, 723 (1999) [Astron. Rep. 43, 631(1999)]. 8. A. G. Gorshkov, V. K. Konnikova, and M. G. Mingaliev, Astron. Zh. 77, 407 (2000) [Astron. Rep. 44, 353 (2000)]. 9. A. G. Gorshkov and V. K. Konnikova, Astron. Zh. 74, 374 (1997) [Astron. Rep. 41, 325 (1997)]. 10. A. G. Gorshkov, V. K. Konnikova, and M. G. Mingaliev, Astron. Zh. 77, 188 (2000) [Astron. Rep. 44, 161 (2000)]. 11. A. B. Berlin, A. A. Maksyasheva, N. A. Nizhel'skii i, et al., in Abstracts of the XXVII Radio Astronomy Conference, St. Petersburg (1997), Vol. 3, p. 115. 12. N. S. Soboleva, A.V.Temirova, andT. V.Pyatununa, Preprint Spets. Astrofiz. Obs. (1986). 13. A. G. Gorshkov and O. I. Khromov, Astrofiz. Issled. (Izv. SAO) 14, 15 (1981). 14. A. G. Gorshkov and V. K. Konnikova, Astron. Zh. 73, 351(1996) [Astron. Rep. 40, 314 (1996)]. 15. H. Tabara and M. Inoue, Astron. Astrophys., Suppl. Ser. 39, 379 (1980). 16. A. D. Kuz'min and A. E. Solomonovich, Radio Astronomical Methods for the Measurement of Antenna Parameters [in Russian] (Sov. Radio, Moscow, 1964). 17. I. W. A. Browne, Mon. Not. R. Astron. Soc. 293, 257 (1998). 18. J. J. Condon, W. D. Cotton, E. W. Greisen, et al., Astron. J. 115, 1693 (1998). 19. D. Monet, A. Bird, B. Canzian, et al., USNO-SA1.0 (U.S. Naval Obs., Washington, 1996). 20. R. L. Pennington, R. M. Humphreys, S. C. Odewahn, et al., Publ. Astron. Soc. Pac. 105, 103 (1993).

915

21. M. P. Veron-Cetty and P. Veron, Astron. Astrophys. 374, 92 (2001). 22. NASA/IPAC Extragalactic Database. http://nedwww.ipac.caltech.edu. 23. V. Chavushyan, R. Mujica, J. R. Valdes, et al., Astron. Zh. 79, 771 (2002) [Astron. Rep. 46, 697 (2002)]. 24. V. L. Afanas'ev, S. N. Dodonov, A. V. Moiseev, et al., Astron. Zh. 47, 6 (2003) [Astron. Rep. 47, 458 (2003)]. 25. V. L. Afanas'ev, S. N. Dodonov, A. V. Moiseev, et al., Astron. Zh. (in preparation). 26. J. N. Douglas, Bull. Am. Astron. Soc. 19, 1048 (1987). ¨ 27. R. D. Blandford and A. Konigl, Astrophys. J. 232, 34 (1979). 28. A. H. Marscher and W. K. Gear, Astrophys. J. 298, 114 (1985). 29. P. A. Hughes, H. D. Aller, and M. F. Aller, Astrophys. J. 374, 57 (1991). 30. E. Valtaoya, H. Terasranta, S. Urpo, et al., Astron. Astrophys. 254, 71(1992). 31. R. Landau, B. Golisch, T. J. Jones, et al., Astrophys. J. 308, 78 (1986). 32. A. G. Gorshkov, in Abstracts of the XXVII Radio Astronomy Conference, St. Petersburg (1997), Vol. 1, p. 176. 33. B. H. Andrew, J. M. Macleod, G. A. Harvey, and W. J. Medd, Astron. J. 83, 863 (1978). 34. H. D. Aller, M. F. Aller, G. Latimer, and P. E. Hodge, Astrophys. J., Suppl. Ser. 59, 513 (1985). 35. H. Terasranta, M. Tornikoski, E. Valtaoja, et al., Astron. Astrophys., Suppl. Ser. 94, 121 (1992).

Translated byD. Gabuzda

ASTRONOMY REPORTS

Vol. 47 No. 11 2003