On average the 
(the rms of background fluctuation of a map) is
about 50 mJy, but for fields near Cyg A and Cas A the 
may be as
large as 200 mJy, while the lowest 
is about 30 mJy.
Information of all 152 fields of view including field-centers and RMSs of
background fluctuation are shown in Table 3
.
In each field of view sources with 
are
searched and only those with
are presented in this catalog. Sources
with 
are presented only when they have counterparts in other
catalogs. As the source searching procedure is
carried out in a CLEAN map, the distorted area at the foot of the strong
sources is rather limited.
A source searching program was developed by
Zhang (1995) and Cao (1995). The program
checks the surroundings of a maximum first. If a feature like a
straight line or
an arc of a circle is found, the program will mark the small area as not a real
source and then search for the next maximum. Only for fields of view
around Cyg A and Cas A, this situation was encountered rather often.
If none of such features
is found, the program will go on to analyze the
surroundings further. Next, the program will measure the position and the
intensity by fitting a small area, usually 
pixels, with a
Gaussian
function which has the same width as the synthesis beam. Sometimes the program
also measures positions and intensities of subpeaks within the small area.
At this stage the source searching program is also used to integrate the flux
densities of the
sources contained in the area. The integral flux is not given for sources
whose integral fluxes are less or equal to their peak fluxes.
The boundary is determined automatically by the program.
The limitation of boundary is reached that when the intensity is either
smaller than 
(for sources with 
) or down to 
(for
sources with S/N<10). Because the integration is carried out for values
on the pixels, for sources with small S/N the integral flux may less than
the peak flux.
The zero level and noise at the position of each source is estimated by taking the average of its surroundings (A belt with 7 pixels in R.A. direction and 15 pixels in DEC. direction around the beam). In fact, zero levels of different sources are almost equal to zero (only a few mJy from zero), and the noise of different sources are found to be nearly the same within each field of view before the primary beam correction. But there are somewhat bigger differences of noise between some different fields of view, for example, some fields near very strong sources have larger noise. It may be caused by the imperfect phase and gain calibration, and the primary beam correction will cause an increase of the noise at the edge of the field.
No attempt was made to measure source-angular-size, because most of them are unresolved by our telescope and some may not be a single source.
Research aimed at establishing flux density standards covering a wide range of wavelengths have been done by many authors. The absolute flux density systems of Baars et al. (1977 =BGPW) and RBC (Roger et al. 1973) are among the most widely used. Laing et al. (1980) gave the calibrated flux density spectra of 165 3CR sources at frequency ranges 10-178 MHz and 750 MHz-15 GHz. Riley (1988) presented the flux densities at 408 MHz of a number of sources. Flux density spectra of sources at different frequencies and in different sky regions have been collected by Kühr et al. (1981), Veron et al. (1974), Gregorini et al. (1984), Long et al. (1966), Williams et al. (1967), and Kellermann et al. (1969).
We rely on the reference sources to determine the flux density scale in terms of the recorded values on map-plane. The BGPW absolute flux scale is used as the flux scale of the Miyun Survey. The catalogs of 6C (6C1--6C6) and the 87GB (Condon et al. 1991) are used to make a reference source list. Spectra of selected sources from 6C and 87GB catalogs were assumed to be a straight power law spectra. As the frequency ratio of the two catalogs is about 32, the spectral index error would not be too large. For some sources, when other frequency data are available, quadratic curve fitting was adopted.
About 7200 sources were detected in two or more fields of view. If a source appeared in more than one field of view the flux density is taken from the field in which the source has the largest ratio of signal to noise.
The primary antenna pattern is corrected before proceeding to the flux
density calibration.
The width of primary beam used for the demodulation
is 
measured by Kang et al. (1985).
As the calibration of the flux density is done in the
map-plane, the AGC system has no effect on the calibration.
After primary beam correction and flux scaling of the 152 maps, 70 sources
which have straight-line spectra were selected from the preprint of
MPI (Kühr et al. 1981) to check our flux scale.
The factor between the Miyun
survey and BGPW system is 
. Fig. 2 (click here) shows the result.
Figure 2: The comparisons of flux densities of 70 straight-line spectra
sources selected from the preprint of Kühr et al. (1981).
Ordinate: values of
observed fluxes devided by calculated fluxes. Abscissa: the observed
fluxes in Jy/Beam
Coordinates conversion from map-plane grid to celestial equator
system 
, 
on epoch 1950.0 were done by
using synthesis formulae of NCP coordinates system. The formulae we used
are:
is the phase tracking point which is taken
as the field centre, i.e. X=0,Y=0 .
Systematic shifts in the apparent positions of sources
can be caused by large scale gradients in the
ionosphere.
Self-calibration can also cause a constant position shift.
On the average shifts of tens of arcsec for most fields
in this survey were found.
To correct the systematic errors, some sources taken from 6C and
Texas catalogs (Douglas et al. 1973) are used as position references.
Monte Carlo method is not employed as it is
sensitive only to the background fluctuation and not to systematic errors.
Factors which cause position uncertainties mainly come from background
fluctuation and wide synthesis beam. The relation between the
rms uncertainties 
(or
) of a source and the peak flux density Sp can be
expressed by a quadratic sums of two terms (Ball 1975). The first terms
are intensity-independent errors 
, 
and the second terms are inversely proportional to Sp. The equations
used to analyze the errors 
and 
were
taken from the preprint of the NVSS (Condon et al. 1995). We have:
where 
and 
are FWHP of restoring beam in
and 
diretions respectively.
To determine the 
, 
, a test
area of 
to 
, 
to 
was selected.
By comparing positions of strong point sources in this area with that in
6C catalog, the rms offsets of
, 
were obtained. They are 
in right ascension and 
in declination. In this survey
and 
vary with different field of view. On the average
is 50 mJy except for that in fields near the Galactic plane,
and the synthesis beam is about 
.
Figures 3 (click here) and 4 (click here) show the results of the comparisons between this survey and 6C survey.
Figure 3: The RMSs in right ascension direction between the Miyun
and the 6C catalog. Sources used in this measurement are within the range:
to 
, 
to 
.
Abscissa: flux density in Jy/Beam
Figure 4: The RMSs in declination direction between the Miyun
and 6C catalogs measured with sources in the range of
to 
, 
to 
.
Abscissa: flux density in Jy/Beam
The random uncertainties of a source, 
depend on background fluctuations, 
, in each field.
On the average,
about 60% of the sources have
apparent flux densities within 
mJy of their ``true" flux density.
A possible systematic effect on flux densities is that the
self-calibration may intensify the flux densities
as it may enhance the weighting toward strong sources after
several runs of the self-calibrating procedure.
The flux scale is slightly going up from low flux density to high
flux density.