Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.atnf.csiro.au/computing/software/miriad/userguide/node215.html
Дата изменения: Unknown
Дата индексирования: Mon Apr 11 02:22:10 2016
Кодировка:
Поисковые слова: dark nebula

Flux density bootstrapping

The best task to correct the flux scale of an observation is mfboot - mfboot performs an analagous function to gpboot. Task mfboot also supercedes the now obsolete task plboot. It works by scaling the antenna gain table so that the flux density scale of the observed data agrees with some model.

Task mfboot models a planet as a simple disk with constant brightness temperature. It has built-in ephemerides of the planets, and so knows their apparent size and orientation. It also has a model of the brightness temperature variation with frequency. Hence mfboot can generate a nominal visibility function of a planet. This model for Mars is not sophisticated, but should be accurate to better than 10% if appropriately used.

Task mfboot can also be used when bootstrapping from a point source. Although mfboot contains models for sources such as 1934-638, these models are not relevant for 3mm observers. If using a point source, generally the user would need to give mfboot the flux density of the source.

It is implicitly assumed in the bootstrapping approach described here that the flux density calibrator is observed near simultaneously and at the same elevation as the secondary calibrator. This is required because the antenna gain changes with both elevation and with thermal changes. The most reliable approach to tying the secondary and the flux density calibrators flux scales together is through observations with the same elevation and thermal environment.

Task mfboot will apply the bandpass and antenna gain solutions to the observed data before it does a comparison with the models. While it is not generally important that the flux density calibrator data is phase calibrated, it is important that it has the antenna amplitude calibration, derived from the secondary, is applied. This is critical in tying the flux density scales of secondary and flux density calibrator.

We will discuss the various input parameters for mfboot:

• vis: These give the input visibility dataset.
• select and line: This selects the data to be used as the flux density calibrator.
• flux: If using a point source as a flux density calibrator, use this parameter to set the flux density of the source (in Jansky). If using a planet, you may set this parameter (in Kelvin) to override the brightness temperature value that mfboot would use by default.
• mode: The comparison between the model and observed data can be done using either the scalar average (effectively the amplitude of the data), the vector average (effectively the real part of the data), or the triple product of the data. The amplitude and triple product are insensitive to phase errors, and so are normally the most appropriate. The default is to use the triple product for point sources and the amplitude for planets. For planets that are not worse than mildly resolved, you might consider using the triple product for planets.

  When using scalar mode with significantly resolved planets, where there is little flux on long spacings, you might consider using the clip parameter so as to exclude data that would simply contribute noise bias.

• clip: If Mars is used as a flux density calibrator, you will want to ignore those visibilities where the deviation of the visibility function from a simple disk model is significant. In practise, it is best to ignore the visibility data outside the main central lobe of the visibility function. The way to do this is by using the clip parameter. For Mars it is recommended that you set clip=0.4, which will ignore data where the visibility function model is less than 40% of the model zero spacing value.

• device: mfboot can produce a plot of the the data (after bootstrapping) and model visibilities to give you a quantitative feel for the quality of the fit.
• options: The only available option is noapply, which causes mfboot to perform all steps except applying the bootstrap factor.

In the general case, to use mfboot with the calibration scheme recommended above is straightforward enough: you simply run mfboot on the dataset, having selected the flux density calibrator. For example:

MFBOOT
vis=vela.fixed.uv	Input multi-source, dual-band dataset.
select=source(uranus)	Select flux density calibrator.