[daip] GBT data in aips

[Eric Greisen]

Anish Roshi wrote:
> 
> Hi Eric,
> 
> I have a GBT spectral line data set at 4.8 GHz which I want to combine 
> with VLA data taken at the same freq. I have created a uv fits file 
> (random position single-dish data) with the GBT spectra calibrated in 
> 'Jy'. In aips I checked the BUNIT in the header of the uvfits and it is 
> in 'Jy'.
> 
> I created an image using sdgrd task after setting the convolving beam 
> parameter in XPARM and YPARM. I used REWEIGHT (1) as 0 and REWEIGHT (2) 
> as 0.05. The image has units of 'K' (ie BUNIT is set to 'K' by sdgrd).
> 
> Question:
> 
> 1. How do I get the units in the image file in 'Jy/beam' ? I believe 
> that is the units I need if I want to combine the GBT data with the vla 
> data.
> 
> 2. How is the 'K' unit calculated ? Let's say S is the flux density of 
> the spectral line and the convolving beam is Omega, then T (K) = 
> S*lambda^2/Omega/2/k, where  lambda is the wavelength and k is the 
> Boltzmann constant. Is that right ?

Good question - it appears that SDGRD assumes that the input units are K 
and forces the output header to say K.  I see no place where any scaling 
of the input numbers is done (it is OOP code and that means that it is 
hard to find where anything is done).  Note that the input units of Jy 
requires you to have converted a change in system temperature into a 
measured total flux correcting for beam efficiencies etc.  The value at 
each pixel is the integral flux over the beam area centered on that 
pixel.  You need to divide the image values by the beam area in pixels 
and then force the units to 'JY/BEAM' - or at least that sounds like a 
reasonable thing to do.  It has been ~15 years since I worked on this 
part of the code so I am more than a little rusty...

Eric Greisen