Hacker News Reader

Comment by woeirua | Hacker News Reader

Is the loss of Arecibo a big blow to the scientific community? I was of the impression that Arecibo was long past it's heyday and in general that large arrays of smaller receivers were now considered superior.

View on HN

Allow me to gently contradict the previous responders. The community of radio astronomers is feeling the loss of Arecibo very keenly.

I've mentioned this in a previous comment [1] so I won't rehash it again, but here [2] is a statement from the NANOGrav collaboration - we have been using the Arecibo telescope on a weekly basis (along with the Green Bank telescope) to observe a large number of radio pulsars for over 15 years now, and when we announce the first detection of very low frequency gravitational waves, Arecibo data will play a critical role.

This loss really hurts. The US community is scrambling, and we are going to have to rely on the generosity of our international partners going forward - there's just not going to be enough telescope time to do all the science that we'd judged as very important through competitive review processes.

(And that's before we get into the unique radar capability, etc.)

[1] https://news.ycombinator.com/item?id=25154898 [2] http://nanograv.org/announcement,/press/2020/12/02/Arecibo.h...

I mentioned NANOGrav earlier[1] as well. After 15 years they were about to really drive down the noise for each additional year of data.

They were worried about a 6 month delay, as it would seriously hurt their dataset. I assume they're devastated by the complete loss.

[1]: https://news.ycombinator.com/item?id=25046378

Very suspicious in my mind this happened soon after the Chinese 500-meter FAST telescope came online. Is this what you mean by relying on international partners? Maybe it's time for the US to build a new telescope.

I would also encourage everyone to read what this loss means to the local communities in Puerto Rico [0], and most particularly, to the young kids who fell inspired by it.

"Kevin Ortiz Ceballos, a physics student and aspiring astronomer at the University of Puerto Rico, said in an email that it had always been his dream to work at Arecibo.

Now he said, “It feels like the rug has been pulled out from under us and our dreams to work in astronomy in Puerto Rico has vanished.”"

[0] https://www.nytimes.com/2020/11/19/science/arecibo-observato...

I work at UW astronomy. It’s mostly symbolic. The main capability loss is that Arecibo was able to transmit, which is unique at its size. As an example, that was useful to rescue the ISEE-3 explorer when it passed through a comet trail in the 1980s. Only Arecibo was powerful enough to blast a radio message through the comet trail. (PDF reference: https://ipnpr.jpl.nasa.gov/progress_report/42-84/84T.PDF)

Arecibo was certainly still used actively, but it’s not the only observatory in its class - its not like the LHC or Hubble blowing up or something.

Asking as a non-astronomer: my understanding is that other telescopes in this class are synthetic/multiple aperture, and not single-aperture like Arecibo (with a small number of exceptions, like FAST).

Is there any science you can do on extremely-large single-aperture telescopes that you can't do on comparable synthetic/multiple aperture telescopes?

Arecibo was initially built to excite plasma balls in the atmosphere to study them, and was also used for other transmitter uses that required a super powerful transmitter, initially designed for that plasma use.

Built to “excite plasma balls” sounds like it would be creating plasma balls if it’s own. It was used as a scatter radar to study plasma in the upper atmosphere.

For those interested: FAST has a diameter of 500m, compared to 300m for Arecibo (https://en.wikipedia.org/wiki/Five-hundred-meter_Aperture_Sp...)

There’s also https://en.wikipedia.org/wiki/RATAN-600, with a diameter of 575m, but that isn’t a fully filled dish.

Though, as the linked article explains, you can’t use the entire 500m as a single aperture. You can only use 300m of it at a time, but you can “point” it in different directions by using different 300m windows.

While maximum resolution is limited by the diameter or the largest baseline in an array, there is one advantage to single dishes: There is no lack of small baselines. Which implies that you can only really get images of large, dim structures with a single dish.

Also sensitivity is given by total collecting area, which is very competitive for a single huge dish compared to a limited number of small dishes.

Answering as a non-astronomer: Arecibo had two major advantages over synthetic aperture arrays. The first was the "seeing" power or total sensitivity. Because it's a large single dish it collects far more photons than multiple smaller dishes.

The receivers were also cooled with liquid helium IIRC so the noise floor was ridiculously low. Those features combined made for not just high resolution but high sensitivity observations. Most telescope arrays (to my understanding) have a higher noise floor.

The second feature was the ability to transmit. Arecibo has been used for decades as a radar to image asteroids and such. Many asteroids are too small and dark to effectively image with optical telescopes where Arecibo's radar is no problem. It also gets extremely accurate distance and rotation measurements.

I studied pulsars in University and worked with Arecibo, Green Bank, did some data analysis with VLA, and worked on some undergrad projects using low frequency arrays.

The details get complex, but the gist of it is that arrays let you see bright small scale things, while single aperture telescopes let you see large scale that are dim.

Additionally, an array suffers with sensitivity and noise issues if they don't have enough antennas covering discrete distance pairings. This usually occurs at short baselines since array elements often have a minimum spacing. Single aperture telescopes have a large instantaneous gain factor due to their large collecting area, which is really hard to make up without an exceedingly large number of array elements. For example, Arecibo has the same collecting area as ~148 VLA dishes.

Data handling for an array is complex as well, as you need to mix and process the data from each telescope to sync them together (correlation), while a single aperture telescope doesn't need this step.

On the other hand, an array is generally much less sensitive to instrument noise as the cumulative signal means noise from a single array element has a much lower impact overall. Also, it generally dodges the single-point failure modes by being essentially redundant. You can still operate with one or more elements missing/out of action.

Each telescope type has it's pros and cons, and there are equally valid use cases for each one.

One thing that I imagine is quite difficult to perform on an array is the type of planetary radar Arecibo specialized in by nature of it's high-power S-Band radar (2.38GHz). With AO's S-Band radar, it was possible to get highly detailed imaging and orbital parameters of near-Earth Asteroids with short observations, while determining orbital parameters with optical observations can take numerous observations (and doesn't generally give you the same detailed 3d models). NASA's Goldstone Observatory is the next best telescope for this, but is nearly 20x less sensitive due to it's size and different operating frequency.

There was still way more requests for observing time and the total amount of time available in the last cycles. Also there is a lot of value to have long-term datasets that are recorded with the same instruments as it removes a lot of systematics. Nanograph is being badly bitten by that, as they use pulsar timing data over 10 or 12 years. (And were likely only 1 or 2 years away from results, which may now be delayed for several years if they have to switch to timing with another instrument.)

Doesn't the instrument itself change over those long time periods?

Yes, there is some changes. Clocks drift, analog receivers age, and so on.

Are requests a valid measurement of value? Why didn't the astronomer community band together to save it?

They've funded a good number of very expensive telescopes the last 20 years, the fact they didn't fund refurbishment here sees to tell a great deal about its actual value over building new radio telescopes.

The astronomer community couldn't save it the same reason they can't build a new one - these are national projects, funded from national funds. The scale of academic science funding is simply on the wrong scale for major projects like Arecibo, or CERN, or the NIF. No grant of a few thousand dollars is going to crowdsource a radiotelescope. In any event, the NSF would be the one giving the astronomers the funds in the first place, so if it was important, they would simply have repaired or replaced the telescope, if this was the intent.

What’s a good budget to get it done?

I heard estimates that fixing it could have cost up to $150M.Yet just in the last few years, Astronomers were able to find funding for:

MeerKAT: $330M

ELT: > $1B

GMT: $1B

Thirty Meter: $2.4B

Vera Rubin: $473M

James Webb: $10B++

WFirst: $4B

FAST (China): $180M

SKA: $1B

NOEMA: $70M

The last three are radio telescopes, which show both the range of funding available for new radio telescopes, and also serve as evidence we've gone beyond Arecibo.

I am no astro-rocket-surgeon, but I've been led to believe the problem with Arecibo is that it's giant dish was only an advantage for sending signals, while astronomy's overwhelming needs are to receive fainter signals which networks of dishes like SKA can do a much more amazing job at.

Note that FAST was built by China alone, SKA is an international collaboration but without US involvement and NOEMA does have a partnership agreement for data access with the university of Michigan, but is operated and payed for by IRAM, a collaboration between France, Germany and Spain. Meerkat (at the top of the list) is also a radio telescope and while it is not build solely by South Africa and Germany, I have not heard of any US involvement that goes beyond US citizens hiring on at Hartrao or Cape Town. And in the last few years NSF has not just stopped funding Arecibo Observatory, but also Greenbank Observatory. (But they were more successful in finding their on funding and staying operational.)

And while the planetary radar at Arecibo was one of a kind, and something that simply no other facility has right now, it is not the only food thing about AO. The huge collecting area and the low noise receivers were actually rather good at finding faint signals. Networks of dishes shine when it comes to resolving power, but they have their own drawbacks such as poor coverage of the u-v plane or the inability to image extended faint objects.

This statement ignores the fact that the radar capability put Arecibo in it's on class when it comes to Planetary Astronomy.

The capability to study near-Earth Asteroids and perform radar observations of the Moon and other large objects in our Solar System has definitely been diminished with this loss. Even Goldstone, the next most powerful radar telescope, is 20x less sensitive.

Not to mention the pairing of Arecibo's radars with the LIDAR system and it's HF Ionospheric heater system. I don't think saying Arecibo was not the only observatory in it's class is being rather disingenuous.

> arrays ...

e.g. Notionally 16km 'wide' https://en.wikipedia.org/wiki/Atacama_Large_Millimeter_Array

Or 6km https://en.wikipedia.org/wiki/Australian_Square_Kilometre_Ar...

and future the ~3000km (and total collection area of 1km^2) https://en.wikipedia.org/wiki/Square_Kilometre_Array

Correct, smaller arrays are superior, cheaper and less susceptible to single-point failure modes.

Reply on news.ycombinator.com