Jy/beam to K conversion using imstat

[renovatio]

Hi all,

I was wondering whether somebody could explicitly explain to me how IMSTAT calculates the conversion factor for Jy/beam to Kelvin in MIRIAD? It explains in the help file for imstat but it is not very clear for me. 

apparently it sums all the pixels within the region specified and divide it by the sum of the beam. Then divide it by [(2k/lambda^2) * omega] to convert it into Kelvin. SO, something like; 

[the sum/the sum of the beam]/[(2k/lambda^2) * omega]

So, how we actually calculate the 'omega'? k is the boltzmann constant, lambda is the lambda of the line of interest, so what about the meaning of the other parameters actually?

many thanks in advance!

cheers,

[melvyn wright]

S = 2kT/lambda^2 * omega is used to calculate S/T in JyperK units.

The beam area omega can be calculated for a Gaussian , as in the Miriad
task obsrms,
or measured by
counting pixel values.

histo, and imlist options=stat
do this explicitly.

e.g.

s> imlist in=J0607-0850_spw1.cmr options=stat
ImList: version 19-mar-2007
***** Listing for Image = J0607-0850_spw1.cmr *****

------------------------------------------------------------

Image Statistics
bunit: JY/BEAM frequency: 239.149796 K/Jy: 13.0835
Axes: RA---SIN x DEC--SIN Beam: 1.48 x 1.10 arcsecs
Bounding region is Blc=(1,1,1),Trc=(95,45,1)
Effective beam area: 0.1508E+02 pixels (Used to normalize total flux)

Axis: 3 FREQ-OBS
plane Frequency Total Flux Maximum Minimum Average
rms xbar ybar
1 239.150 1.3443E+00 1.3988E+00 -8.1006E-02 4.7431E-03
5.9226E-02 4.7322E+01 2.2052E+01
/o/wright/orims> histo in=J0607-0850_spw1.cmr
Histo: version 01-dec-2011
Finding image minimum and maximum
Object j0607-0850
Mean 4.74308E-03 Rms 5.92258E-02 Flux 1.34425E+00 Jy (4275 points)
Maximum value 1.398840E+00 at (48,23,1,1)
Minimum value -8.100640E-02 at (95,23,1,1)


Hope this helps.

Melvyn

> --
> You received this message because you are subscribed to the Google Groups
> "miriad" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to miriad+un...@googlegroups.com.
> To post to this group, send email to mir...@googlegroups.com.
> Visit this group at http://groups.google.com/group/miriad?hl=en.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>
>

[renovatio]

Dear Melvyn,

Thanks a lot for the info!

What about actually applying this conversion factor to the cube properly and creating a new cube has the unit of Kelvin? using 'maths'? 

The unit of the cube is in J/beam. I think we can divide the cube by the conversion factor using maths to perform the conversion factor to each pixel, and change the 'bunit' keyword in the header(?). What was the task for editing the header in Miriad? or maybe there is another task does all these things at once?

Cheers!


21 Şubat 2013 Perşembe 14:11:43 UTC tarihinde renovatio yazdı:

[renovatio]

Correction: not dividing, we should multiply the cube by the conversion factor using maths. Thats is I think. 

Cheers,

22 Şubat 2013 Cuma 11:38:24 UTC tarihinde renovatio yazdı:

[melvyn wright]

Hello Renovatio,

If the units of the uv data are Jy, then the image (from invert) has
units Jy per synthesized beam.
Typically, we deconvolve the image (clean, mosmem etc) to remove the
synthesized beam
sidelobes. The units are then JY/PIXEL. Then, we usually convolve
this image with a Gaussian
beam, so the units are JY/BEAM with a well defined Gaussian beam. If
you divide
by PIXELS/BEAM , you will have units JY/PIXEL, but the image has still
been convolved
by the Gaussian beam, so the interpretation of the image is not so clear.

The analysis tasks look at the units JY/BEAM or JY/PIXEL in converting
to temperature
units.

Melvyn

[Peter Teuben]

I should also advertise again that miriad is open source, if you have
miriad, you should see all program source
code under $MIRPROG, and looking in $MIRPROG/analysis/imstat.for you
can read some decent comments
on how exactly for example omega is computed. Of course some of those
source code comments don't
make it into the help section (all lines between c+ and c-).

If you do NOT see $MIRPROG/analysis/imstat.for, you have a
corrupted/unsupported miriad, and I will for example
not be able to help you debug your problem in case there is a version
mismatch.

The global version of miriad is stored in the file $MIR/VERSION, and
this is probably 4.3.4 for most of you. If it's more
than two minor versions away from this, you are probably running an
outdated version. At least you need to update
all flux tables and baseline files in $MIRCAT:

cd $MIRCAT
cvs update

peter

[renovatio]

Many thanks to both of you, dear Melvyn and Peter.

Dear Melvyn, thanks a lot for the detailed explanation about how the unit evolves through the reduction processes (invert-cleaning(mossdi)-restor). uv data has the unit of Jy, so the final unit is in Jy/beam for the fully reduced+cleaned cube. Therefore the unit of the conversion factor from Jy/beam to K is K/[Jy/beam]. Multiplying the cube by this conversion factor gives the unit of Kelvin, so each pixel will have data in Kelvin (Tmb) that the 0th moment can be created in the unit of K.km/s instead of (Jy/beam).km/s. Thanks again.

Dear Peter, the version I have is 4.3.4 as you suggested. I also checked the $MIRPROG/analysis/imstat.for and it was there. Cheers!


Best,

22 Şubat 2013 Cuma 16:23:19 UTC tarihinde teuben yazdı: