General primary flux calibrator question...

[cat001]

Hello,

I was told the absolute flux uncertainties for CABB are:
at 3 mm: 30 to 50%
at 7 mm: 10 to 20%
at 15 mm: 5 to 10%
at 3, 6cm: 5%?

Have these values changed? and what were the absolute flux uncertainties before CABB?

Thanks
Catarina

[Mark.Wieringa]

Hi Catarina,

the values you quote sound reasonable to me. At high frequency the weather and phase stability during the observations would play an important role.
The values would not have changed much since CABB arrived, except for data affected by RFI bad enough to mess up the tsys values.
Jamie may have some actual data on this from comparisons with VLA flux calibrators.

Cheers,

Mark

[ste616]

Hi Catarina,

This is a persistent question, and there is a lot of uncertainty on what to expect for flux calibration accuracy.

Certainly, the numbers you have quoted are commonly quoted, but not very accurate. Based on some results comparing our fluxes to those measured with the EVLA at the end of 2010, we can certainly say that our flux calibration accuracy is better than 10% in the 15mm and 7mm bands. We have repeated this experiment with even more sources and observing time earlier this year, and results from these observations should be available in the middle of this year. This should go some way to answering this question for all the ATCA bands except 3mm.

For 3mm observing, while doing some observations of stars expected to follow the Rayleigh-Jeans brightness profile, we were able to measure an 8 mJy calibrator to a precision of 10% (RMS limited) when cross-calibrating with Uranus. So there's really no reason to expect terrible calibration in 3mm either. Of course this is all weather and observing-technique limited, but with the recommended procedures documented in the Users Guide, I would suggest that a 10% flux calibration accuracy is not unreasonable over the entire frequency range of the ATCA.

[cat001]

Thanks Jamie and Mark.
It is good to hear that we can achieve 10% uncertainty at 3mm.

On the same topic, is there a way to determine what the % flux uncertainty is after the observations?


Thanks
Catarina

[ste616]

Hi Catarina,

From a single observation of a single source, it isn't really easy to compute an absolute uncertainty on its flux density. This is because any number of effects can come into the calculation (eg. opacity, decorrelation), and each of these effects may change the measured flux density in a difficult-to-determine fashion.

If you have many observations of the same source, and you don't think it should be varying over time, you can estimate how well the flux scale is applied by looking at how the measurements vary.

We are trying to estimate the calibratability of the ATCA by comparing the flux densities of a large number of sources against those found with the EVLA and Planck. This allows us to see systematic effects and to hopefully correct them.

If you're asking for general advice on this, I would recommend checking the flux density of your phase calibrator(s) after they have been tied to your flux calibrator. If their fluxes are close to those in the calibrator database, you can assume then that your flux calibration is correct.

After all that, I'm sorry to say I haven't got a good answer for you. If you let me know your precise situation, I may be able to give you better advice. And until we get a better understanding of our absolute flux scale through our continuing comparison with Planck and the EVLA, I won't be able to give you exact figures either.