where we stand

rosten
Posts: 19
Joined: Fri Feb 14, 2014 3:40 pm

where we stand

Postby rosten » Tue Apr 15, 2014 9:14 am

Hi folks,

There's been discussion at the council level and it seems that we are converging on a strawman idea for a VLASS, which I want to get your feedback on. As a galactic working group we still have some outstanding issues:

--the time required to do a Ku band galactic plane thermal survey. The bright methanol maser would also be captured, giving some line science in addition to continuum. See discussion in some of the other topics on this forum, but it sounds like the time required would be less than 2000 hours. Is there enough momentum to keep this in the VLASS as a separate tier, or is this an effort which would best be pursued as a PI-driven proposal?

-- there was discussion amongst the extragalactic, transient, and galactic working group leads about a compromise VLASS that could serve all parties. Right now the idea is an S-band survey of ~21,000 sq. deg extragalactic to 50 uJy between -10<dec<+75, using B-config [6350 hours including 25% overhead], and ~2800 sq. deg of the Galactic Plane and Bulge to 30 uJy using A-config [2350 hours including 25% overhead]. The total time request for the VLASS would be ~8,700 hours, i.e., 20% VLA usage over 5 years. My questions to this group are: the 30 uJy rms is a thermal noise limit, is this realizable in the galactic plane? How does this compare to what you could do at L band with the GMRT? What is the science yield for going to this sensitivity? If Shami & co. could pull some numbers out of their white paper for pulsar yields I would appreciate it, and Greg Sivakoff also mentioned compact objects in our telecon -- ditto for expected source yields.

claw
Posts: 26
Joined: Mon Mar 03, 2014 7:18 pm

Re: where we stand

Postby claw » Thu Apr 17, 2014 1:22 pm

Hi Rachel,
One other point that just came up in the technical WG is identifying the science potential in extended sources. The range of possibilities in antenna configuration is too large to simulate fully, yet each one has bearing on sensitivity, spatial scale, scheduling and more. To do our simulations of the configuration/beam, we need more guidance on the science as a function of spatial scale.
So far, much of the science case has being discussed is at small scales, in which case we won't need to know anything beyond the resolution for a given config/band. That certainly is the rationale behind the latest consensus survey definition, which suggests A config in the Galactic plane.
I'm guessing that the Galactic community probably has the most interest in larger scales. If so, please chime in and help guide the process for the technical WG.

thanks,
casey

lsjouwerman
Posts: 21
Joined: Mon Jul 18, 2011 3:49 pm

Re: where we stand

Postby lsjouwerman » Thu Apr 17, 2014 6:55 pm

rosten wrote:--the time required to do a Ku band galactic plane thermal survey. The bright methanol maser would also be captured, giving some line science in addition to continuum. See discussion in some of the other topics on this forum, but it sounds like the time required would be less than 2000 hours. Is there enough momentum to keep this in the VLASS as a separate tier, or is this an effort which would best be pursued as a PI-driven proposal?


I do not understand the number you quote - 2000 versus 3000 hours. To have a homogeneous survey sensitive to (pre-dominantly high-frequency) thermal emission you must use the upper part of the band as I've explained in this forum on several topics, and to stay within any reasonable data rate limit you cannot scan faster than about 0.8 sqdeg/hr. The full 2800 sqdeg thus takes just over 3000 hours. I've done this calculation over and over again, so you are making some assumptions that I do not find back in this forum or your reasoning. Please explain your assumptions here.

rosten wrote:-- there was discussion amongst the extragalactic, transient, and galactic working group leads about a compromise VLASS that could serve all parties. Right now the idea is an S-band survey of ~21,000 sq. deg extragalactic to 50 uJy between -10<dec<+75, using B-config [6350 hours including 25% overhead], and ~2800 sq. deg of the Galactic Plane and Bulge to 30 uJy using A-config [2350 hours including 25% overhead]. The total time request for the VLASS would be ~8,700 hours, i.e., 20% VLA usage over 5 years. My questions to this group are: the 30 uJy rms is a thermal noise limit, is this realizable in the galactic plane? How does this compare to what you could do at L band with the GMRT? What is the science yield for going to this sensitivity? If Shami & co. could pull some numbers out of their white paper for pulsar yields I would appreciate it, and Greg Sivakoff also mentioned compact objects in our telecon -- ditto for expected source yields.


Again, I do not understand why you say that the request for Ku would be in A-config. This is completely out of the blue and I would stronlgly disagree to this for a thermal survey. Sure in A-config you have higher resolution which may be better for identifying pulsars (which they say they won't detect anyway at Ku band at the GUTS sensitivity), but this would hurt the goal of being sensitive to thermal and extended emission. One would expect thermal sources to possibly be a bit extended (at any resolution). In A-config/Ku the largest angular scale is a couple of arcsec, whereas it is about an arcmin in C-config. In C-config the angular resolution is sufficient to register sources in the optical/IR, do targeted follow up for pulsars and transients, and matches well the S and L band in B-config so I don't understand this change of parameter unless it is a typo. Also note that in C-config the coherence time is much longer and calibration is much easier. And you're not setting yourself up with a data reduction and analysis nightmare (A-config uses 100x more pixels in the imaging and larger data sets are much harder to handle). I also think your quote of the sensitivity in Ku band is very optimistic, but I'm fine with that. Note that a better sensitivity requires slower slew rates, thus an increase in total observing time. The Ku calculations in GUTS are for the instantaneous sensitivity at the highest slew rate (see above) so to go deeper you need at least 2-3 times the 3000 hours in the previous section. Please explain your switch to A-config.

Lorant ()

claw
Posts: 26
Joined: Mon Mar 03, 2014 7:18 pm

Re: where we stand

Postby claw » Thu Apr 17, 2014 8:43 pm

Hi Lorant,
I think Steve explained the way we calculate the survey speed most completely in the topic I started on the VLASS calculator. Take a look in the middle of the discussion:
viewtopic.php?f=62&t=143&start=10

As we've discussed, you are free to target a sensitivity at the top of your band. However, the quoted sensitivity should be the specification at the band center (which would be a lower, more sensitive value). That will allow us to compare the different surveys properly. If by targeting the sensitivity at the top of the band you get a 3000 hour survey, then that is fine. But a Ku-band, 130 microJy survey spec at band center will take less than 2000 hours to cover the Galactic plane.

casey

lsjouwerman
Posts: 21
Joined: Mon Jul 18, 2011 3:49 pm

Re: where we stand

Postby lsjouwerman » Fri Apr 18, 2014 4:36 am

claw wrote:But a Ku-band, 130 microJy survey spec at band center will take less than 2000 hours to cover the Galactic plane.


I think the point I really want to make is that regardless of the final sensitivity, if you want a homogeneous sensitivity at the top of the band - which we want as the ultimate spec - you are bound by the maximum survey speed at the top of the band at a given maximum data rate (which may only be as low as 25 MB/s whereas I used 36 MB/s which can be thought of as standard MB/s at 3-bit continuum plus some MB/s for line resources). That is what really determines the hours needed. Whatever the final sensitivity is I do not care as it will be low enough for our purposes; I calculated it to be something to show a number for comparison, but that sensitivity is not what is driving the thermal survey and thus please do not use it as the spec for your calculations. I'd be happy to see what (better sensitivity) you get for final sensitivity using 3200 hours, and please consider GUTS or the Ku band Galactic Plane survey amended for that sensitivity in your comparisons at the center of the band.

claw
Posts: 26
Joined: Mon Mar 03, 2014 7:18 pm

Re: where we stand

Postby claw » Fri Apr 18, 2014 4:59 pm

Fair enough. Using the calculator, a depth of 90 mJy at band center will require ~3000 to observe the Galactic plane defined in GUTS. (Going to 80 microJy bumps it up to 3600 hours, so I thought it better to round the time down and keep a nice round depth value.)
So, going back to Rachel's question: when comparing GUTS to other surveys, it is a 3000 hour survey, which is too large to be requested through the standard proposal process.

casey

lsjouwerman
Posts: 21
Joined: Mon Jul 18, 2011 3:49 pm

Re: where we stand

Postby lsjouwerman » Fri Apr 18, 2014 5:37 pm

Thank you Casey, I will quote the 90 microJy/bm at the center of the band if I get to update the GUTS white paper. I also agree it is too big for a PI driven project.

I still like to hear from Rachel why she quotes A-config for the Ku band part.

claw
Posts: 26
Joined: Mon Mar 03, 2014 7:18 pm

Re: where we stand

Postby claw » Fri Apr 18, 2014 5:56 pm

The reference to A config was in the context of the consensus survey, which is at S band. The consensus was drawn up last week by the science WG chairs. It is by no means final, but it is starting to sound like a plan that most people can get behind. If you think a revised GUTS white paper would help make the high-frequency argument better, then you should move quickly before the opinions become fixed.

casey

lsjouwerman
Posts: 21
Joined: Mon Jul 18, 2011 3:49 pm

Re: where we stand

Postby lsjouwerman » Fri Apr 18, 2014 6:08 pm

I can understand B-config (or A-config) for S band, but that should not determine an extended/large config for the Ku band part. It makes sense (for spex measurements) to have a similar beam so a 5x smaller range of baselines (ie about C-config) in Ku, which also allows to detect/compare larger (few tens of arcsec) structures seen in the IR. In A-config you won't detect anything larger than a couple of arcsec, and this high resolution is not needed, actually less sensitive to thermal structures. Also the caliration and data reduction benefit much from a smaller array configuration. In short, I do not see why Ku was switched to A-config - it should be completely independent of the all-sky S config (apart from spex measurements, which is an argument against the same/larger config in Ku). Doing a full Galactic plane (thermal) survey in A-config at Ku band simply does not make sense - I've not heard any argument in favor of this - save A-config for follow-up if really needed.

rosten
Posts: 19
Joined: Fri Feb 14, 2014 3:40 pm

Re: where we stand

Postby rosten » Mon Apr 21, 2014 11:46 am

Lorant,

There was no mention of array configuration for Ku band -- I was referring to the S band specification. And we were awaiting the final resolution of the Ku band thermal survey time, due to the issues Casey and Steve brought up in the discussion. It hasn't been discarded from consideration, and if the required time is indeed ~3000 hours then it should be considered within the scope of the VLASS. The leaders of the other working groups are concerned that it is too dissimilar from the main survey being proposed (or that was being discussed as of about a week ago). I personally think the Ku band survey has a lot of potential for legacy value and overlap with surveys at higher frequencies.


Return to “Galactic Working Group”

Who is online

Users browsing this forum: No registered users and 1 guest