[daip] integrated flux error equation

[Eric Greisen]

Casey Law writes:
 > Hello daip,
 >   I am trying to reproduce the error analysis of source detection algorithms
 > in AIPS (JMFIT and SAD).  I can't find an explicit description of how the
 > integrated flux error is calculated, but the funny thing is that I do see a
 > trend between the peak flux error and the integrated flux error like this:
 > 
 > err_Si = err_Sp * (A_src/A_beam + 0.75),
 > 
 > where A_src/A_beam is the ratio of the area of the source to the area of the
 > beam.
 >   When I plot up the fluxes normalized by the beam like this, I see a tight
 > correlation (particularly with the 0.75).  Is this the actual error equation
 > used in AIPS source detection?  Or do you have any ideas why this
 > correlation would exist?
 > 
Looking at the code it is

err_Si = err_Sp * A_src/A_beam * sqrt (1 + 2 A_beam/A_src)

and is based on work by Condon.  The pure math errors from fitting do
not convey correct errors at all so we depend on Condon.  The err_Sp
is more or less the image rms as well.

The code is provided to you and you can look yourself.

Eric Greisen