Update "flux.ctl" to include additional Southern Sources

[stuartdweston]

The current "flux.ctl" is very sparse with calibrators for the Southern Hemisphere.

ATNF with the Compact Array have a program of monitoring Southern Sources, and provide a ATCA Calibrator Database (https://www.narrabri.atnf.csiro.au/calibrators/). These are too weak for small (12m) dishes to see, but larger dishes will. Looking at the FS flux.ctl the same spectral fit from Ott et al 1994 is used :

log S [Jy] = a + b * log v [MHz] + c * log^2 v [MHz]

See https://www.narrabri.atnf.csiro.au/calibrators/calibrator_database_documentation.html#interpreting-flux-densities

So we need to include some of these calibrators. Some are variable so is there a way for the FS to update "flux.ctl" from the ATNF calibrator database automatically ?

[varenius]

A "simple" way could be to write a script which could be run via e.g. cron, which just reads your ATCA database and overwrites the flux.ctl file regularly.

[stuartdweston]

Request to reopen by Ed

[stuartdweston]

Here are the callibrators I have added locally for the Southern Hemisphere, they are monitored by ATCA and are > 10 Jy, so our 30m does see them. I doubt our 12m will see them though. In my comment I have put the date that the model coeff's were updated by ATCA.

ATCA Callibrators
Source of data of monitored sources is obtained from the ATCA Calibrator Database
https://www.narrabri.atnf.csiro.au/calibrators/calibrator_database.html

Source Date Model Coeff Flux Scatter Freq
1226+023 2021-01-30 1.930E+00,-9.798E-01,3.611E-01 23.418 0.246 @ 6600 MHz
2251+158 2010-12-18 7.150E-01,6.562E-01 17.897 0.191 @ 6600 MHz
1253-055 2021-01-30 8.203E-01,5.343E-01,-9.819E-02 15.568 0.089 @ 6600 MHz
1921-293 2020-09-20 3.719E-01,1.258E+00,-6.501E-01 9.252 0.025 @ 6600 MHz
1934-638 2021-02-21 1.528E+00,-1.044E+00,-1.104E-01 3.965 0.006 @ 6600 MHz
2134+004 2020-09-20 6.702E-01,1.131E+00,-8.400E-01 10.787 0.032 @ 6600 MHz
0723-008 2021-02-08 1.198E+00,-4.557E-02,-3.223E-01 8.794 0.135 @ 6600 MHz
1830-210 2021-01-08 1.005E+00,5.835E-01,-3.969E-01 16.467 0.069 @ 6600 MHz
          freq MHz  ---- flux 10** ----   "     FS
source type min max log log(f) 2log(f) size model
1226p023 p 6400 6800 1.93 -0.9798 0.3611 1 gauss 100 1s
1934m638 p 6400 6800 1.528 -1.044 -0.1104 1 gauss 100 1s
2251p158 p 6400 6800 0.715 0.6562 0.0 1 gauss 100 1s
1253m055 p 6400 6800 8.203 0.5343 -0.09819 1 gauss 100 1s
2134p004 p 6400 6800 0.6702 1.131 -0.84 1 gauss 100 1s
0723m008 p 6400 6800 1.198 -0.04557 -0.3223 1 gauss 100 1s
1830m210 p 6400 6800 1.005 0.5835 -0.3969 1 gauss 100 1s

So as per your comments above I need to change "type" and "model". They are point like sources for the 30m so good for pointing as well. So for "c" or "p" can we have a "b" (both) ?

The above models also work for X-Band, so perhaps the min/max freq can be adjusted. I am currently using them for C-Band to point on and callibrate.

[stuartdweston]

There are two additional calibrators used at Park's for Pulsar Work that other large dishes should also see, Hydra A and PKS1934-63 (which I think we have from ATCA 1934m638).

Hi Stuart,

Appended below is the absolute flux calibration information from fluxcal.cfg (a file maintained by the pulsar community). It has some absolute flux calibrators that have been used for single dishes in the Southern hemisphere; these are valid up to 8.6 GHz, including 3C218 (Hydra A, RA = 09:18) and PKS1934-63.

Cheers,
Willem

##########################################################################
#
# Format 2:
#
# Flux in Jy for a frequency in GHz is:
# log10(S) = a_0 + a_1*log10(f) + a_2*(log10(f))^2 + ...
#
# See e.g. Baars et al. 1977 for format 2 references
#
# &NAME  RA       DEC      a_0  a_1  a_2  ...
#        (J2000)  (J2000)
#
# NOTE:
# Fluxes from Perley, R. A. and Butler, B. J. 2017ApJS..230....7P
# (http://adsabs.harvard.edu/abs/2017ApJS..230....7P
#	
# An Accurate Flux Density Scale from 50 MHz to 50 GHz
# The flux-density scale of Perley & Butler is extended downward to ∼50 MHz by utilizing recent observations with the Karl G. Jansky Very Large Array (VLA) of 20 sources between 220 MHz and 48.1 GHz, and legacy VLA observations at 73.8 MHz. The derived spectral flux densities are placed on an absolute scale by utilizing the Baars et al. values for Cygnus A (3C405) for frequencies below 2 GHz, and the Mars-based polynomials for 3C286, 3C295, and 3C196 from Perley & Butler above 2 GHz. Polynomial expressions are presented for all 20 sources, with accuracy limited by the primary standards to 3%-5% over the entire frequency range. Corrections to the scales proposed by Perley & Butler, and by Scaife & Heald are given. adsabs.harvard.edu) valid
# from 50 MHz to 50 GHz. Note that 3C48, 3C138, 3C147 and 3C380
# are time variable with time scales for significant variations
# of a few years. For suitability see Table 10 in the paper.

&J0444-2809  04:44:37.708 -28:09:54.403 0.9710 -0.8938 -0.1176

&J0519-4546  05:19:49.723 -45:46:43.855 1.9380 -0.7470 -0.0739
aka PicA
aka Pictor
aka PictorA

&3C218  09:18:05.669 -12:05:43.95 1.7795 -0.9176 -0.0843 -0.0139 0.0295
aka HydA
aka Hydra
aka HydraA
aka J0918-1205

# PKS J1934-6342
# This is primary calibrator no. 1 for MeerKAT.
# Position: Simbad ICRS
# Flux model: "A revised flux scale for the AT compact array", Reynolds, 1994
# This version has coefficients quantised to 4 significant digits and remains within 0.2% of original model.
# The model is valid between 408 and 8640 MHz.
# Original model parameters given in MHz scale:
# b0 = -30.7667
# b1 = 26.4908
# b2 = -7.0977
# b3 = 0.605334
# to convert to GHz scale:
# a0 = b0 + 3. * b1 + 9 * b2 + 27 * b3 -> a0 = 1.1704
# a1 = b1 + 6 * b2 + 27 * b3           -> a1 = 0.2486
# a2 = b2 + 9 * b3                     -> a2 = -1.6497
# a3 = b3                              -> a3 = 0.605334
&PKS1934-63  19:39:25.02671 -63:42:45.6255 1.1704 0.2486 -1.6497 0.605334
aka J1939-6342
aka 1934-638

[wehimwich]

So as per your comments above I need to change "type" and "model". They are point like sources for the 30m so good for pointing as well. So for "c" or "p" can we have a "b" (both) ?

Actually c is both, p is not a flux calibrator. :) As far as I remember, this is only used in gnplt by the user as part of selecting the data to work with.