Abstract. This paper presents results and describes the improved data processing algorithm of the low frequency sky survey of discrete sources carried out with the UTR-2 radio telescope. The measurements were conducted within the frequency range 10 to 25 MHz. Coordinates and flux densities of the sources detected were obtained. Identification with sources from the 4C survey has been done. The resulting catalogue contains parameter estimates for 483 sources on a set of frequencies within the UTR-2 range.
Keywords: Discrete source, Catalogue, Northern sky
Figure 1. Coverage of the celestial sphere by the present part of the survey.
Table I. List of sources not included in present part of the catalogue.
RA 1950 DEC 1950 Name
01h 53m 28.58d 3C 55 03h 18m 41.17d 3C 84 04h 15m 38.03d 3C 111 04h 34m 29.50d 3C 123 05h 00m 38.00d 3C 134 05h 23m 32.47d 3C 141 16h 26m 39.63d 3C 338 22h 44m 39.40d 3C 452
As before, the data processing algorithm was conditionally divided into the primary and the secondary stage. In the first stage coordinates and flux density of the source are estimated using the data of a single observation. In the second stage data obtained are subjected to corre- lation analysis over all observations, in order to determine parameters of the observed sources.
The primary processing algorithm is based on an optimal fitting to experimental records of an ideal response to a discrete source. Thus, a trial source is placed at the nodes of a two-dimensional grid (with size of the cell 0.06d in right ascension (K) and 0.05d in declination (d) ). In each node the optimum amplitude of discrete sources and its error are calculated using the data of all five radiotelescope beams according to the least squares method. (Note that in the former algorithm the ideal response was fitted to the measured data in right ascension only, while the source declination was calculated through interpolating the responses of three neighbor radiotelescope beams). The responses corresponding to local maxima of the signal amplitude are passed over for secondary processing.
It is the algorithm of secondary processing that has undergone the most serious changes. A simultaneous full correlation analysis of the results of primary processing of all the observations was previously impossible because of the low computer potential. That reduced significantly the reliability of the algorithm because most of the responses detected, as a result of primary processing happened to be false. In fact, the only information, which allows one to distinguish between true and false responses, is the rms error of an optimum fit of the ideal response to the measured data. But, as can be inferred from our numerous observations, ionospheric distortions of the reception pattern often make this information unreliable at decameter wavelengths. In order to detect true responses in the new algorithm we construct a two-dimensional map of source density distribution in the RA and DEC coordinates. Since false responses are distributed uniformly over the coordinates, their correlation is insignificant. On the other hand, coor- dinates of the true sources obtained in diffierent sessions do not differ considerably, and hence they form maxima on the density map. Then averaging within these maxima yields parameters of the true sources.
Table II. Levels of UTR-2 noises in the declination range +30d to +40d.
MHz 10 12.6 14.7 16.7 20.0 25.0 Jy 12.2 6.6 4.3 3.2 1.1 0.8 Jy 16.5 10.8 8.1 6.4 4.6 3.2 Jy 20.5 12.7 9.2 7.2 4.7 3.3
formula
where ?? is the background radiation level and ?? c the confusion level. Estimates of the random noise, ?? , calculated for operating frequencies of the UTR-2 with account of the sky brightness temperature and the radiotelescope parameters (Megn et al., 1978) are given in Table II. In order to estimate the confusion noise, ??c , some additional measurements were carried out using the method described in the paper Sokolov, 1988. According to this technique, several observations of the sky area remote from intense radio sources within the declination range +30d to +40d were conducted. In the single-beam operation mode of the UTR-2 the frequencies of all 30 receivers were set with a step greater than their frequency bandwidth. In this case all 30 channels show equal confusion noise levels but different random ones. The confusion noise can be estimated by averaging the signals of all the channels. Since the random noise of different channels is statistically independent, it decreases p 30 times as a result of averaging over the channels.
Measured values of the confusion level, ?? , at different frequencies are given in Table II. Note that the experiment was carried out at 25 MHz and confusion noise estimates at other frequencies were obtained like in paper Durdin & Terzian, 1972, with account of the average spectral index of extragalactic radio sources and solid angles of the UTR-2 pattern.
The estimates of the full noise, ??, are given in the same table. According to Table II and with S ?? 5 the minimum flux density is 37.5 Jy at the middle frequency of the UTR-2, 16.7 MHz. It should be kept in mind that the S ?? 5 criterion is mostly used for high frequency observations, where strong multiplicative interference is absent. At decameter wavelength their influence is substantial, and, generally speaking, the use of this criterion for a single-frequency observations might be inappropriate.
But the UTR-2 sky survey implements simultaneous observations at a number of low frequencies, and it has been shown (Sokolov, 1985) that acceptable for further statistical analysis values of the source cata- logue completeness and reliability are reached at the flux-densities level S_16.7 3.5, where S_16.7 is a flux density at 16.7 MHz as read from linear approximation (in log S -log_nu scale ) of spectrum within the UTR-2 frequency range.
formula
and ?? are sources's coordinates differences on right ascension and declination respectively, we obtain the threshold equal to 0.3d. For comparison, the half-width of the UTR-2 pattern is shown with the dashed line on the same figure. Results of comparing the 4C and UTR-2 source coordinates can be represented in the form of histograms. Such histograms of the source number distribution against deviations in the RA and DEC coordinates are shown in figure 3 and figure 4, respectively. The histogram for the angular distance deviation (R) is shown in figure 5.
Figure 2. Average scatter of coordinates for a single measurement versus frequency.
Figure 3. Histogram of right ascension deviations for identified UTR-2 sources.
Figure 4. Declination deviation histogram for identified UTR-2 sources.
Figure 5. Angular distance deviation histogram for identidied UTR-2 sources. The mean coordinate differences for the identified sources from the UTR-2 and 4C surveys detected in declination range +30 Æ to +40 Æ are as follows: ....
Acknowledgements
This work was performed with support of INTAS97-1964 grant.
References
Braude, S.Ya., Megn, A.V., Ryabov, B.P., Zhoock, J.N.: 1978, Astrophys. Space Sci. 54, 3.
Braude, S.Ya., Sokolov, K.P., Zakharenko, S.M.: 1994, Astrophys. Space Sci. 213, 1.
Durdin, J.M., & Terzian, Y.: 1972, Astron. J.77, 637.
Gower, J.F.R., Scott, P.F., & Wills, D.: 1967, Mem. Roy. Astron. Soc. 71, 49.
Megn, A.V., Sodin, L.G., Sharykin, N.K., Braude, S.Ya., Melyanovsky, P.A., Inyutin, G.A., Goncharov, N.U. Antennas (in Russian). Collection of papers. 26 Release./Edited by A.A. Pistolkorse. - M.:Svyaz, 1978. - 200 s., il.
Sokolov, K.P.: 1985, PhD Tesis, Space Research Institute of the USSR Academy of Sciences.
Sokolov, K.P., 1988, Astron. J. (Soviet). 65, 236.5