If you drive from Cambridge (UK) to Wimpole, you'll see some impressively large radio telescopes that belong to the Mullard Radio Astronomy Observatory (MRAO).
However, there's much more that's not visible from the road. Hidden behind the trees, MRAO has a prototype SKA-Low array (from before the full installation in Australia), and three dishes from a HERA prototype.
The MRAO itself has a fascinating history, notably including the discovery of the first pulsar by Jocelyn Bell using the wonderfully named Interplanetary Scintillation Array, which consisted of over four thousand dipole antennas spread across nine acres. In WWI the site was a mustard gas factory, with train station and sidings. The train tracks have long since gone, but the station building remains. Inside hangs a large, coloured but faded image titled "GALACTIC RADIO EMISSION AT 38 Mc/s". This appears to be a coloured visualisation based upon the black & white figure in pages 654-655 of a 1957 paper [0].
The above 1957 paper illustrates a survey of half the celestial sphere at 38 MHz. In comparison, this specific MeerKAT image from the article [1] appears to be a 1.28 GHz measurement focusing on the galactic center (6.5 square degrees) [2]. So it's not a 100% like-for-like comparison, but interesting nonetheless to see how much the detail has improved in the past ~70 years!
The MRAO is a fascinating place, with things left as they were the last time an instrument was used. The floor of the hut where the array cables were aggregated for connection to the cable back to the Cavendish is covered in little plastic caps from the connectors, discarded as the instrument was being set up.
The article talks about HERA; MRAO hosts the prototype for that. IIRC, they experimented with methods to build the dishes with off-the-shelf parts - such as drainpipes to build the ring.
there is an antena farm on the way from the city to my place that I use as a reference land mark for new visitors, which I call "area 52", which also serves as a kind of personality test, where most will laugh and say they know where it is, but some few who are uncomfortable as it's a long wave sigint base, marked on all the flight maps, that they are clearly wishing not to have seen or looked at, be in a conversation referencing, and now marked for life grimly waiting for a knock on the door.
Do these interferometer type of arrays attempt to have the dishes at the same level essentially a flat plane? Or can they adjust for that with the timing used to "sync" them together so one dish might be in a dip while another is higher on a mound. TFA mentions 150km diameter array. Does the earth's curvature come into play as well?
> The image (above) shows long radio-emitting filaments up to 150 light–years long unspooling from the heart of the galaxy. These structures, whose origin remains unknown, were first observed in 1984, but the new image revealed 10 times more than had ever been seen before.
I don't understand how this isn't the biggest news in astronomy. Gigantic filaments of energy passing through the milky way
Well.., I've been more busy with writing code lately so that the first question coming to mind was, how many bytes is an array of one square kilometer? And I assume it's a two-dimensional array.
I'm thinking that you'd need to print out the array to be able to properly measure this. So we'd need to decide on the print itself. Do we use A4 paper, old school green bar? What size font is used? We'd also need to decide on the contents of each element in the array. Let's say they are floats with a set limit of precision. So an A4 is 210mm x 297mm. 1km / 210mm = 4761.9 1km/297mm = 3367.0. If we go with 4bytes per float, that's 4761.9 * 3367.0 * 4 = 64,133,269.2 bytes => 61.16 MB
It's also before coffee, so my logic might not be right for basic math yet
Every sensor in the array is sampling at frequency, so - first order - you can use that sampling frequency and the sample size, you get an idea of the input bandwidth in bytes/second. There are of course bandwidth reduction steps (filtering, downsampling, beamforming)...
Sorry, not sure I follow from what I said (explaining how much data sensors produce) to 'increasing the sampling frequency' ? You're usually sampling at larger width to then put specifically taylored pass-band filter and removing aliasing effects and then downsampling. This is a classic signal acquisition pattern : https://dsp.stackexchange.com/questions/63359/obtain-i-q-com...
None of this changes the actual real amount of data you have at the end of the day though after all is said and done, that's what I mean, so long as you don't botch it and capture too little. In computing terms, the amount of real data in a compressed archive and the uncompressed original is the same, even if the file size is larger for the latter.
On SKA from what I understand they're sampling broadband but quickly beamform and downsample as the datarates would be unsustainable to store over the whole array.
Right, that makes sense, you'd be looking at an insane amount of data across the ranges that these sensors can look at. But they would still need to preserve phase information if they want to use the array for what it is best at and that alone is a massive amount of data.
I think they preserve timestamped I,Q data. Know some people looking at down-sampling, preselecting those signals for longer term storage and deeper reprocessing and they seem to have a 24h window to 'analyze and keep what you need'.
We're still in technological phase where ADCs are far more advanced than storage and online processing systems, which means throwing away a lot. But I have high hopes for a system where you upgrade computing, network, storage (and maybe ADCs...) and you get an improved sensor. Throw man-hours at some GPU kernel developers and you get new science. The limit seems more now about enough people and compute to fully exploit the data than technological...
Too late to edit: any idea of the resolution that the I,Q data is sampled at (bandwidth, bit depth)? I've been in one of these installations a while ago and the tourguide had really no clue about any of the details (I think he was the son of one of the scientists)?
If you drive from Cambridge (UK) to Wimpole, you'll see some impressively large radio telescopes that belong to the Mullard Radio Astronomy Observatory (MRAO).
However, there's much more that's not visible from the road. Hidden behind the trees, MRAO has a prototype SKA-Low array (from before the full installation in Australia), and three dishes from a HERA prototype.
The MRAO itself has a fascinating history, notably including the discovery of the first pulsar by Jocelyn Bell using the wonderfully named Interplanetary Scintillation Array, which consisted of over four thousand dipole antennas spread across nine acres. In WWI the site was a mustard gas factory, with train station and sidings. The train tracks have long since gone, but the station building remains. Inside hangs a large, coloured but faded image titled "GALACTIC RADIO EMISSION AT 38 Mc/s". This appears to be a coloured visualisation based upon the black & white figure in pages 654-655 of a 1957 paper [0].
The above 1957 paper illustrates a survey of half the celestial sphere at 38 MHz. In comparison, this specific MeerKAT image from the article [1] appears to be a 1.28 GHz measurement focusing on the galactic center (6.5 square degrees) [2]. So it's not a 100% like-for-like comparison, but interesting nonetheless to see how much the detail has improved in the past ~70 years!
[0] https://adsabs.harvard.edu/pdf/1957MNRAS.117..652B ("RESULTS OF A SURVEY OF GALACTIC RADIATION AT 38 Mc/s")
[1] https://physicsworld.com/wp-content/uploads/2025/03/2025-02-...
[2] https://arxiv.org/pdf/2201.10541 ("The 1.28 GHz MeerKAT Galactic Center Mosaic")
The MRAO is a fascinating place, with things left as they were the last time an instrument was used. The floor of the hut where the array cables were aggregated for connection to the cable back to the Cavendish is covered in little plastic caps from the connectors, discarded as the instrument was being set up.
The article talks about HERA; MRAO hosts the prototype for that. IIRC, they experimented with methods to build the dishes with off-the-shelf parts - such as drainpipes to build the ring.
there is an antena farm on the way from the city to my place that I use as a reference land mark for new visitors, which I call "area 52", which also serves as a kind of personality test, where most will laugh and say they know where it is, but some few who are uncomfortable as it's a long wave sigint base, marked on all the flight maps, that they are clearly wishing not to have seen or looked at, be in a conversation referencing, and now marked for life grimly waiting for a knock on the door.
Do these interferometer type of arrays attempt to have the dishes at the same level essentially a flat plane? Or can they adjust for that with the timing used to "sync" them together so one dish might be in a dip while another is higher on a mound. TFA mentions 150km diameter array. Does the earth's curvature come into play as well?
> The image (above) shows long radio-emitting filaments up to 150 light–years long unspooling from the heart of the galaxy. These structures, whose origin remains unknown, were first observed in 1984, but the new image revealed 10 times more than had ever been seen before.
I don't understand how this isn't the biggest news in astronomy. Gigantic filaments of energy passing through the milky way
FROM TFA: "The elongated radio filaments visible emanating from the heart of the galaxy are 10 times more numerous than in any previous image"
I love physics names. A array of telescopes with collecting are of 1 square kilomter. Square kilometer array. Makes it so clear what it is.
"Makes it so clear what it is."
Well.., I've been more busy with writing code lately so that the first question coming to mind was, how many bytes is an array of one square kilometer? And I assume it's a two-dimensional array.
I'm thinking that you'd need to print out the array to be able to properly measure this. So we'd need to decide on the print itself. Do we use A4 paper, old school green bar? What size font is used? We'd also need to decide on the contents of each element in the array. Let's say they are floats with a set limit of precision. So an A4 is 210mm x 297mm. 1km / 210mm = 4761.9 1km/297mm = 3367.0. If we go with 4bytes per float, that's 4761.9 * 3367.0 * 4 = 64,133,269.2 bytes => 61.16 MB
It's also before coffee, so my logic might not be right for basic math yet
It's a radio telescope, how would you imagine translating that to bytes?
Here's an article mentioning the data transmission rates in SKA, up to 20 terabits per second:
https://www.skao.int/en/explore/big-data
Every sensor in the array is sampling at frequency, so - first order - you can use that sampling frequency and the sample size, you get an idea of the input bandwidth in bytes/second. There are of course bandwidth reduction steps (filtering, downsampling, beamforming)...
This makes no sense though? Given the Nyquist theorem, simply increasing sampling frequency past a certain step doesn't change the outcome.
Sorry, not sure I follow from what I said (explaining how much data sensors produce) to 'increasing the sampling frequency' ? You're usually sampling at larger width to then put specifically taylored pass-band filter and removing aliasing effects and then downsampling. This is a classic signal acquisition pattern : https://dsp.stackexchange.com/questions/63359/obtain-i-q-com...
Actually, it does. You can decimate the higher sample rate to increase dynamic range and S/N ratio.
Also, for direct down conversion, you can get better mirror frequency rejection by oversampling and filtering in software.
None of this changes the actual real amount of data you have at the end of the day though after all is said and done, that's what I mean, so long as you don't botch it and capture too little. In computing terms, the amount of real data in a compressed archive and the uncompressed original is the same, even if the file size is larger for the latter.
Aren't they sampling broadband for later processing?
On SKA from what I understand they're sampling broadband but quickly beamform and downsample as the datarates would be unsustainable to store over the whole array.
Right, that makes sense, you'd be looking at an insane amount of data across the ranges that these sensors can look at. But they would still need to preserve phase information if they want to use the array for what it is best at and that alone is a massive amount of data.
I think they preserve timestamped I,Q data. Know some people looking at down-sampling, preselecting those signals for longer term storage and deeper reprocessing and they seem to have a 24h window to 'analyze and keep what you need'.
We're still in technological phase where ADCs are far more advanced than storage and online processing systems, which means throwing away a lot. But I have high hopes for a system where you upgrade computing, network, storage (and maybe ADCs...) and you get an improved sensor. Throw man-hours at some GPU kernel developers and you get new science. The limit seems more now about enough people and compute to fully exploit the data than technological...
Too late to edit: any idea of the resolution that the I,Q data is sampled at (bandwidth, bit depth)? I've been in one of these installations a while ago and the tourguide had really no clue about any of the details (I think he was the son of one of the scientists)?
Remembering now that SKA is many, many things and I was talking of the higher-bands array still being installed, but many infos on the -low and -mid arrays are available, for example https://www.skao.int/sites/default/files/documents/Year_In_T...
That's an incredible resource, thank you again. I've been reading for the last hour and it is only getting better.
This is also an interesting development:
https://en.wikipedia.org/wiki/Low-Frequency_Array
Fascinating stuff, thank you for the details and the view of a possible path forward.
Are you deliberately obtuse to the play on words of an array being used from a programmer's use of the word in contrast to an array of antennas?
If you have to ask, you know the answer. :-p