22.2GHz water maser observations with a small dish

[Ed astro]

Hi all

After the successful detection of two methanol masers at 12.2 GHz with my 1 metre dish I wondered whether it was possible to detect water masers at 22.2 GHz as well. Recently I did my first successful observations. The setup I am using is broadly similar to what I used to detect the 12.2 GHz methanol line: a 1 metre dish on a HEQ5 mount, with an LNB placed in the focal point which amplifies the incoming signal and converts it down to a lower frequency within the range of my SDR receiver. The main difference is in the LNB: I am using a norsat 9000LDF to receive at 22.2 GHz. It has an LO frequency of 20250 MHz so the water line is mixed down to 22235-20250=1985 MHz. This is unfortunately just outside the tuning range of my airspy mini SDR so I am using a NeSDR XTR instead, which has an extended tuning range up to 2300 MHz. Since the norsat LNB came without a feed I made my own rectangular horn feed from a sheet of brass foil. It was designed using the HDL-ANT program by W1GHZ; the length of the horn is 25mm, the dimensions of the opening are 24X20mm.

Last weekend I did my first successful observation of W49, one of the strongest water maser sources in the sky. The Arcetri catalog of H2O maser sources by Valdettaro et al. (2001) reports a peak flux density of 31000 jansky, making W49 one of the brightest objects in the sky at microwave frequencies! However, like many other water masers W49 is highly variable on timescales of months to years. Flux density values published years ago should therefore be taken with some caution. During the observation of W49 spectra were recorded at 5 minute intervals for about 1.5 hours. A heatmap of all the 5- minute spectra is shown below:

The dish was pointed 1 degree away from the target twice during the observation, this was done in order to check whether or not a received signal was coming from the direction of W49. The narrow signal at 22234.9 MHz is an artifact of the SDR. The strong signal at 22235.75 MHz disappeared when the dish was pointed “off-target”, so it is most likely coming from the direction of W49. There is some frequency drifting, this is because the LO frequency of the LNB is unfortunately not very stable. After averaging all the on-target spectra, some weaker peaks could be seen as well. I did not convert frequency to radial velocity because the LNB LO frequency is still rather uncertain.

Finally, the star forming region W51 was also observed. In October of last year, a strong flare of the H2O maser in this star forming region was reported (see ATel #15002: https://www.astronomerstelegram.org/?read=15002). The signal was indeed quite strong, even stronger than W49. Maybe the flare is still ongoing. However, I have no way to determine the flux density of the maser at the moment so this is just speculation.

Anyway, this is pretty much all I have to say about this project right now. This post has become way too long but I hope you all don’t mind 😉

Eduard

[Maxim Usatov]

This is very cool!

[Ed astro]

Hi all,

Last weekend I observed two more 22.2 GHz water maser sources in star forming regions.

The first one was W3. This star forming region is associated with the bright nebula NGC 896 in Cassiopeia. The protostars themselves are completely obscured by thick clouds of dust and gas, and are therefore  invisible with "regular" visible-light telescopes. With a radio telescope we can detect the maser sources associated with these young stars, because the radio waves easily pass through the dust clouds. 

Just like during the observations last week I pointed the dish 1 degree away from the target twice during the observation to check whether the signal was indeed coming from the direction of the target source.

 

The other source is Orion KL (the Kleinmann- Low nebula). This is one of the closest star forming regions, it is embedded inside the well- known Orion nebula. Orion KL was much lower in the sky than W3 at my location (about 30 degrees), and because of this there was more absorbtion and thermal noise from water vapour in the atmosphere. Despite these issues the water maser was still succesfully detected. Besides the main peak at 22229.8 MHz there are several weaker lines in the spectrum as well.

best regards,

Eduard

 

 

 

 

[StarryEyed]

Eduard. 

Outstanding work.

How well does the HEQ5 hold up with a metre dish?

Kevin. 

 

[Ed astro]

Hi Kevin,

The whole dish setup (including the counterweight, the LNB and the supports) is about 10 kg, so it is a bit on the heavy side but it should still be well within the limits of the mount. So far I have not had any problems with the 1 metre dish mounted on the HEQ5.

It does catch a lot of wind though, so it can only be set up with calm weather.
 

 

[Marvin Jenkins]

Big WOW, giant WOW. I know I am looking at something amazing, but I have not understood a single word anyone has said, except the bit about an HEQ5 holding a dish. Think of Baldrick in Blackadder and you have me.

Marv

[Ed astro]

Well Marv that is mostly my fault since I did not even really try to explain this 😉 

What you are seeing are spectra of water masers. When water molecules in a gas cloud are subjected to strong infrared radiation or collisions with other molecules, they can emit strong microwave radiation at a frequency of 22235 MHz (wavelength of about 13.5 millimeters.) These conditions are usually found inside star forming regions. When we point a radio telescope at a star forming region and plot the intensity of the received microwave radiation against frequency, we see a spike at or near 22235 MHz. Due to Doppler shift, this spike is usually a few MHz above or below the rest frequency depending on how fast the source is moving towards or away from us. In many cases we see a number of different spikes in the spectrum, these all correspond to different sources within the same star forming region, which are moving at different velocities.

I hope this makes things a bit more clear!

Best regards,

Eduard

[Victor Boesen]

This is excellent!! I've never heard of radio emissions at these frequencies from water masers. What type of software did you use to receive the emission? I suppose you must have done a lot of oversampling and smoothing to receive it judging by the y-axis interval on your frequency domain plots?

This is most definitely I would be looking forward doing at some point!

Victor

[Ed astro]

Hi Victor,

I am using the SDR# IF average plugin for averaging the spectrum over a longer time period.

The methods for making spectra of these masers are very similar to hydrogen line observing- you can use the same software and observation routine. The main difference is that the maser lines are usually 10- 100 times weaker, so you need to integrate a lot longer to get good SNR. Typically I make  tens of spectra with each an integration time of a few minutes, and average them all afterwards. I keep the resolution low at 256 bins (12 KHz for an SDR with a bandwidth of 3 MHz, just enough to resolve the lines while keeping the noise low.

Observing these water masers is challenging but also rewarding, they are bright and variable radio sources but they are rarely observed by amateurs. So if you are done with hydrogen line observing at some point this might be an interesting project.

[Victor Boesen]

On 08/03/2022 at 15:37, Ed astro said:

Hi Victor,

I am using the SDR# IF average plugin for averaging the spectrum over a longer time period.

The methods for making spectra of these masers are very similar to hydrogen line observing- you can use the same software and observation routine. The main difference is that the maser lines are usually 10- 100 times weaker, so you need to integrate a lot longer to get good SNR. Typically I make  tens of spectra with each an integration time of a few minutes, and average them all afterwards. I keep the resolution low at 256 bins (12 KHz for an SDR with a bandwidth of 3 MHz, just enough to resolve the lines while keeping the noise low.

Observing these water masers is challenging but also rewarding, they are bright and variable radio sources but they are rarely observed by amateurs. So if you are done with hydrogen line observing at some point this might be an interesting project.

Thank you Ed, very useful!

I don't think I'll be done with hydrogen line in the near future, but would also like to play with new stuff. I just started on university the summer last year and have yet to play with the 3m dish available on campus 

Should be nice for hydrogen line, although for higher frequencies in X-band, Ku, K band and etc there are also a couple of solid 1m dishes available.

Plenty to observe out there!!

Victor

[Ed astro]

Hi Victor,

On 09/03/2022 at 21:49, Victor Boesen said:

Thank you Ed, very useful!

I don't think I'll be done with hydrogen line in the near future, but would also like to play with new stuff. I just started on university the summer last year and have yet to play with the 3m dish available on campus 

Should be nice for hydrogen line, although for higher frequencies in X-band, Ku, K band and etc there are also a couple of solid 1m dishes available.

Plenty to observe out there!!

Victor

These look like some really nice dishes to play around with!

As mentioned before water masers are difficult to observe due to their high frequency (you need really accurate pointing and tracking) but there are other options at lower frequencies as well. OH (hydroxyl) has four maser lines at 1612, 1665, 1667 and 1720 MHz. These frequencies are quite close to the hydrogen line, so you could use the same SDR receiver and LNA, and probably even the same feed that you use for hydrogen line observations if it performs well enough at 1.6 GHz. OH masers are generally a lot weaker than water masers though, but with a 3 metre dish you should be able to detect a few of them.

 

[Victor Boesen]

18 minutes ago, Ed astro said:

Hi Victor,

These look like some really nice dishes to play around with!

As mentioned before water masers are difficult to observe due to their high frequency (you need really accurate pointing and tracking) but there are other options at lower frequencies as well. OH (hydroxyl) has four maser lines at 1612, 1665, 1667 and 1720 MHz. These frequencies are quite close to the hydrogen line, so you could use the same SDR receiver and LNA, and probably even the same feed that you use for hydrogen line observations if it performs well enough at 1.6 GHz. OH masers are generally a lot weaker than water masers though, but with a 3 metre dish you should be able to detect a few of them.

 

Hydroxyl is definitely on my TO-DO list because of the reason you also mention, required equipment/software is somewhat similar between the two. I have a filtered LNA with 30dB gain centered at 1688MHz and a pass band of 80MHz. This should perform nicely for at least three of the two of the frequencies you mention.

Have you tried observing OH masers?

Victor

[Ed astro]

Hi Victor,

That filtered LNA could be really handy- there is often quite a bit of RFI above 1.7 GHz which can cause some problems when trying to detect weak signals. What is the noise figure of your LNA? Low system noise is really important for this kind of work.

I have detected four OH masers with my 3 metre dish over the past few years: these were the red giant stars NML Cygni, OH26.5+0.6 and V669 Cas at 1612 MHz, and  the star forming region W3(OH) at 1665 MHz. I used a G8FEK LNA and homemade filters for 1612 and 1665 MHz. These observations were quite challenging, it required tens of minutes to several hours of data recording to get good results. 

 

[Ed astro]

Hi all,

More observations have been done at 22.2 GHz over the past two weeks. I have pointed the dish at a number of different sources: NGC 2071, NGC 7129, W75, Cepheus A, IRAS 06053-0622 and G25.8-0.18. Unfortunately, none of these sources were detected, they are probably too weak right now for detection with a small 1 metre dish. 

W49 and W51 were successfully observed again. It seems like both sources have become weaker since the first observations 5- 6 weeks ago. However, since the vertical axis is not calibrated it is hard to tell which part of the change is real, and what part is due to pointing errors or gain variations. On the other hand, the different maser features (peaks) in the spectrum of a source usually vary independently of each other, while gain variations or pointing errors should influence the measured intensity of all the peaks equally. I plotted my most recent spectrum of W49 (from March 26) over an earlier spectrum made on February 12. The strongest feature has significantly decreased in intensity, while the weaker peaks have remained almost the same. It is therefore likely that the decreasing intensity of the main peak is (mostly) due to the intrinsic variability of the source.

The same was done for W51. However, W51 lacks any sufficiently strong peaks besides the main peak to compare with. 

It would be interesting to observe these sources again in one or two months.

Edited March 29, 2022 by Ed astro

[mdstuart]

And no issues if it is cloudy! Fascinating read.

Mark

[Ed astro]

Hi all,

Another month has now passed since my previous observations, so I decided it was time to check W49 and W51 again. This morning W49 was observed for about an hour.  

I also plotted todays spectrum over my previous spectra from March and February. Overall, the intensity of the main peaks seem to have increased compared to the earlier observation in March, but it is unclear whether this is a real change of the source itself or if it is due to more favourable conditions (less clouds/ moisture in the atmosphere).  However, it is also clear that the peaks are changing in intensity relative to each other- likely due to the intrinsic variability of the source. The fact that changes can be detected in the spectrum over a period of a few weeks or months means that the source itself must be quite small (at most a light-month across). 

I will try to observe W51 again this weekend.

Eduard 

[Ed astro]

Hi all, 

This morning I observed W51. Its brightness seems to have significantly decreased again since my last observation five weeks ago. Some of this decrease may be due to less favourable observing conditions (W51 was lower in the sky and there was a thin layer of clouds). However, it is unlikely that such a large decrease in intensity (>10-fold since february) is only the result of variations in gain or moisture, given that I get more or less the same numbers each time I point my dish at the sun or W49.

[Victor Boesen]

Excellent Ed! Out of curiosity, do these masers vary periodically in intensity, or do some exist for a certain amount of time and then fade out? Surely they must rely on the star formation and molecules in the gas clouds?

Victor

[Ed astro]

Hi Victor,

Most of these masers vary irregularly. Usually there is always some maser emission detected from these strong sources, but individual peaks in the spectrum can emerge and disappear in a few months to several years. Sometimes one of these peaks increases rapidly in brightness and fades away again in only a few months or weeks. Such flares are unpredictable and are still not very well understood

There are also examples of masers which vary periodically, such as the OH- masers in the atmospheres of Mira variables and OH/IR stars.

Anyway, I think it is great to see that some of this variability is detectable with my amateur setup. After all, if you can only “see” a handful of “maser stars” with your small radio telescope then these stars better be very interesting, otherwise you will soon get bored…

Eduard

 

 

[Ed astro]

Hi all,

I realised that this thread has slowly turned into a monthly update on the variable masers in W49 and W51... This time, I have of course new observations of these sources, but I have also something new to share about this project. Recently I have tried using the 28th harmonic of a Leo Bodnar GPSDO (GPS-disciplined oscillator) as a reference signal for correcting the frequency drift and offset of the LNB. I made a small waveguide antenna out of a piece of copper tube with an inner diameter of 8mm (this antenna is essentially a miniaturized version of the "cantenna" design commonly used for 1420 MHz hydrogen line feeds). The cutoff frequency of this waveguide is about 11 GHz, so any unwanted harmonics below that frequency should be filtered out. I also added a conical horn section with an opening of 35mm to make the antenna more directional. The 28th harmonic is of course extremely weak and can only be detected at short distance, but that is actually desirable for my project (a stronger signal would disturb the observations). 

Recently I did some observations of W49 and W51. Thanks to the reference signal I was able to compensate the frequency offset and drifting of the LNB. It was also finally possible to calculate radial velocity (doppler shift). This will make it easier later on to compare my own spectra from different observations to each other and to spectra published in the scientific literature.

To be continued...

Eduard

 

 

 

 

[Ed astro]

Hi all,

For comparison I have plotted the most recent spectra of W49 and W51 together with the earlier observations. Because I did not have a reference signal during the earlier observations I just tried to align the peaks (note that this may not be entirely correct). 

The wide peak at -6 km/s in the spectrum of W49 has significantly increased in intensity over the last two months compared to the narrower peak at 5.5 km/s. W51 on the other hand does not seem to have changed very much since last month apart from a slight increase in intensity.

[Victor Boesen]

Very interesting to follow along in this thread Ed!! It has made me curious, is there anywhere we can have a look at some of the code or etc. you use for frequency calibration, LSR velocity and etc? Would love to have a look

Victor

[Ed astro]

Hi Victor,

I have not published the code yet, there is no GitHub page or anything for this project. For LSR correction I have not written any code myself but I just use the online LSR calculator from ATNF (https://www.narrabri.atnf.csiro.au/observing/obstools/velo.html)

Programming is really not my expertise though… 

Eduard

 

 

[Victor Boesen]

1 hour ago, Ed astro said:

Hi Victor,

I have not published the code yet, there is no GitHub page or anything for this project. For LSR correction I have not written any code myself but I just use the online LSR calculator from ATNF (https://www.narrabri.atnf.csiro.au/observing/obstools/velo.html)

Programming is really not my expertise though… 

Eduard

 

 

Thanks Ed, I've never seen that site before. It looks very useful!

No worries with the programming, I was just too quick to assume you had programmed a script to do the work for you 

[Ed astro]

Hi all, 

I have done new observations of W49 and W51 on July 14 and August 8. With six observations in total now, the plots where all the spectra are plotted together are starting to look rather messy and confusing. Therefore, I was experimenting with different ways to visualise the changes in the spectra over time using Matplotlib/Pyplot. First I tried to make "stacking graphs". The nice thing about this type of graph is that the shape of the peaks is still visible. On the other hand, it is difficult to avoid too much overlap between strong peaks while also showing the weaker features.

Another way to display the changes is by plotting a heatmap of all the spectra. What I like about this type of diagram is that changes in the peaks are very well displayed. However, the shape of the peaks is not shown of course. 

The peak at -6 km/s in the spectrum of W49 was in outburst around May 28, and it is now back at the same intensity as before the outburst. Furthermore, a new weak feature emerged at +11 km/s. The intensity of W51 decreased more than 10-fold in the period February- April, but it did not change much since then. 

Last night I also tried to detect another maser (OH43.8-0.1), but this attemt was not successful.