Hydrogen line driftscan mapping

[Eduard Mol]

This week I assembled my 3 metre dish again and started doing driftscans at 1420 MHz. Everyday I point the dish 2.5 degrees higher. It’s going to take a few weeks to map the entire northern sky, but it has already been a lot of fun to see the first driftscan results come in. The plots below are from the first driftscan at 0 degrees declination. 

It’s been 8 years since my first HI detection, and seeing the “S-curve” on a driftscan plot still makes me excited!

[Alex P]

Excellent Quality&Resolution

[b alex pettit jr]

Eduard,

FFT size ?

Integration time per spectra ?

What software did you use for the blue 3D plot ?

>>  Your Data and this Plot is the best H1 presentation I've seen .

NICE !

Thanks,

Alex

[Eduard Mol]

Hi Alex, here are some extra details:

The airspy mini has a bandwidth of 6 MHz. I am recording spectra with 1024 FFT points, so 5.86 KHz/bin.

Integration time is 146 seconds/ spectrum. 

The blue plot is from my own custom python script, I am using Matplotlib for all the plotting. In this case, I added a small constant multiplied by the spectrum number to each spectrum to create the 3d- like effect. 

The waterfall plot is from the pyplot.contourf() function.

If you are interested I can send you the jupyter notebook (however, the code is still "under construction" at the moment and missing things like plotting against RA instead of time and Vlsr conversion). 

Best regards,

Eduard

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/sara-list/322410290.372537.1776522610183%40mail.yahoo.com.

[b alex pettit jr]

Hi Eduard,

Yes, it obvious you are using an FFT above 512 ..

( and, if you are not going beyond the MW, 3MHz would be fine .. I use 1.2 )

Here's an idea

Your beamwidth is 4 degrees .. you 'should' be able to resolve amplitude changes in periods of 1/10th = 0.4 degrees.

At 240 secs per degree, you might try 30 or 60 seconds seconds ..

Your plots have a very low, flat noise floor ; increasing the angular resolution would be an interesting test.

Thanks for the offer on the *.py code, but I'll stay with Rinearn or Matlab.

Regards,

Alex

=============================================================

[Stephen Arbogast]

Hi Eduard,

Very nice  plots!  I like  Jupyter Notebooks, python ...   Please send your notebook  to my  email address.

Thanks,

Stephen

[Stephen Arbogast]

On  second  thought .....   I  think  others  would  like to see your  code.  If  you don't mind   please  share  here   with all of us.   I have   done   Computational  Physics  using  Jupyter Notebooks.  They  are  very  good for  collaboration.

Stephen

[Stephen Arbogast]

Here   are  two  examples. Computational  Physics  and  Digital  Signal  Processing...

Google  Computational  Physics  by  Stephen  Koonin.  It  is  a  classic  from  1985  in  Fortran   but  easy  to  translate into  Python.

Digital  Signal  Processing...    https://github.com/spatialaudio/digital-signal-processing-lecture

Stephen

[Eduard Mol]

Hi Alex, 

With my ~5 degree half power beam width, I am already oversampling the beam at 0.63 degrees/ spectrum. For mapping I don't see much benefit in increasing the resolution even further. Yes there are small changes in the appearance of the spectrum even with a 30 sec integration time, but translating this into a brightness distribution on the sky at a much higher resolution than the beam width would maybe not be possible. There's a reason why professional radio astronomers use aperture synthesis after all...

I know planetary imagers use techniques like wavelet sharpening and drizzle to go a bit beyond the resolution that the seeing and their equipment can offer in real-time, however I don't know much about the details behind these image processing techniques and how far one can stretch this. Has anyone here tried any of these techniques from optical astronomy on radio astronomy data?

As for the 6 MHz bandwidth: I can record at 1/2 the bandwidth, but IMO 5.8 KHz is a good enough resolution for HI work. Furthermore, I am expecting to get some data on HVCs and extragalactic sources as well, and the extra bandwidth will help with that. It's also nice to have some extra bandwidth for the continuum outside the hydrogen line, so far I have been able to detect continuum radiation from the Galactic plane near Scutum and Aquila for example. 

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/sara-list/510156012.427965.1776545998841%40mail.yahoo.com.

[Eduard Mol]

Hi Stephen, 

I do not want to share the notebook publicly (at least not yet), because it is still unfinished and missing lots of features as mentioned in an earlier post. 

Also my notebooks are often very specific to one type of observation or project, so maybe not very useful for others. 

(especially now that anyone can generate code that's maybe even better than mine in under a minute using LLM coding models...)

I don't have your email address btw. if you have mine, can you please send me an email so I can send you the notebook?

Best regards, 

Eduard

To view this discussion visit https://groups.google.com/d/msgid/sara-list/44b57134-35b6-427e-97b5-eae2285abab0n%40googlegroups.com.

[b alex pettit jr]

Hello Eduard,

Okay. Looks like you have gone in detail through all the math :)

It appeared on your plots that the brighter of the plots had high freq/vel resolution but that the

angular /time axis showed the regions being compressed... I am not too familiar with aperture synthesis

as applied to radio astronomy, but yes, I can understand how that technique might be done.

It would be interesting to try an "optical unsharp mask" procedure on an RA data set !

Thanks,

Alex

Res with my 1.2m  @ 1.25deg/frame  =  1/10 Beam-Width

============================================================

On Sunday, April 19, 2026 at 04:52:37 AM EDT, Eduard Mol <eddiem...@gmail.com> wrote:

Hi Alex, 

With my ~5 degree half power beam width, I am already oversampling the beam at 0.63 degrees/ spectrum. For mapping I don't see much benefit in increasing the resolution even further. Yes there are small changes in the appearance of the spectrum even with a 30 sec integration time, but translating this into a brightness distribution on the sky at a much higher resolution than the beam width would maybe not be possible. There's a reason why professional radio astronomers use aperture synthesis after all...

I know planetary imagers use techniques like wavelet sharpening and drizzle to go a bit beyond the resolution that the seeing and their equipment can offer in real-time, however I don't know much about the details behind these image processing techniques and how far one can stretch this. Has anyone here tried any of these techniques from optical astronomy on radio astronomy data?

As for the 6 MHz bandwidth: I can record at 1/2 the bandwidth, but IMO 5.8 KHz is a good enough resolution for HI work. Furthermore, I am expecting to get some data on HVCs and extragalactic sources as well, and the extra bandwidth will help with that. It's also nice to have some extra bandwidth for the continuum outside the hydrogen line, so far I have been able to detect continuum radiation from the Galactic plane near Scutum and Aquila for example. 

om.