Galactic WG Summary so far (comments please!)

elisabethmills
Posts: 13
Joined: Tue Mar 11, 2014 3:58 pm

Re: Galactic WG Summary so far (comments please!)

Postby elisabethmills » Thu Mar 27, 2014 3:14 pm

To clarify: I mean that I would like to see the parameters (and science case) that would be ideal for a Galactic low-frequency, high-resolution survey.

What will the proposed all-sky component of the Extragalactic working group accomplish for Galactic science? Where is it an improvement over what can be done from GloStar? Is A or B configuration more optimal for low-frequency observations of the plane? Can it be made to yield useful kinematic information? (S band has no useful molecular lines, so no maser kinematics will be forthcoming there). Are there technical problems unique to the plane for imaging or calibration that need to be considered?

chomiuk
Posts: 12
Joined: Fri Feb 28, 2014 4:16 pm

Re: Galactic WG Summary so far (comments please!)

Postby chomiuk » Fri Mar 28, 2014 10:17 am

I can't make the telecon today, so a few notes.
Thanks for your comments, Betsy. I appreciate that you've taken the time to compile all the past/proposed MW surveys. And thanks for the notes on optically-thick sources.

I think the one region at S band that is of significant interest and is not covered by CORNISH, etc (not sure of GLOSTAR coverage) is the Galactic bulge. This is the portion of the sky that is most promising for identifying large samples of pulsars, low-mass X-ray binaries, novae, symbiotic stars, and cataclysmic variables. You don't have to be a person who cares about variable stars for long to know that nearly all of them are located in Sco or Sgr.

The transients people have discussed a deeper-drilling field in the Galaxy in the typical high-res S-band mode, and there has been a fair amount of support for that. That will likely fulfill many of these science goals.

If people feel strongly that another strategy should be used in the Galactic plane, I suppose I can sign off on that. But I'd just like to warn that ~3000 hours is a very large 'tier', and we should think about the value of aligning our observational plans with other groups, to make sure that the collective Galactic voice is heard, and not dismissed if we are simply asking for too much time in too distinct an observing mode.

gsivakoff
Posts: 17
Joined: Tue Mar 25, 2014 11:28 am

Re: Galactic WG Summary so far (comments please!)

Postby gsivakoff » Fri Mar 28, 2014 9:07 pm

Based on our discussion today, I've put together a potential three tier Galactic plan, which would both play well with others while creating a high frequency legacy.

  • Tier 1: S band All Sky Survey [which I want to nickname SASSY] (dec > -40)
    This was taken from the Extragalactic WG definition tier 1, but we should make the science case for this tier, because I think there is Galactic science for nearby sources and halo sources. Also, regardless of the array chosen, we want to make sure that we work with the Technical Working Group to develop ideas to do parallel processing that will sacrifice point-source sensitivity for a larger beam and thus probe extended sources.
    • Galactic Science Driver(s):
      Halo compact objects and radio stars, Nearby disc compact objects and radio stars
    • Area:
      34120 sq. deg (31100 sq. deg for dec > -30 only)
    • Cadence:
      1 epoch
    • Single Epoch RMS:
      100 muJy/beam
    • Stacked Epoch RMS:
      100 muJy/beam
    • Stacked Epoch RMS equivalent at K_u band:
      nu^-1.5: 9 muJy/beam
      nu^-1.0: 20 muJy/beam
      nu^-0.5: 45 muJy/beam
      nu^+0.0: 100 muJy/beam
      nu^+0.5: 230 muJy/beam
      nu^+1.0: 500 muJy/beam
      nu^+1.5: 1100 muJy/beam
      nu^+2.0: 2500 muJy/beam
    • Configuration:
      B array
      (BnA probably desired for southern sources)
    • Theta_res:
      2.7"
      (Note EWG states 2”, which indicates 25% better centroiding compared to below)
    • Centroiding:
      5 sigma detected centroiding 0.54” @ 500 muJy
    • S. Myers Survey Speed:
      16.53 sq. deg / hr / epoch
    • Total cost (w 25% overhead): 2580 hr (2350 hr for dec > -30 only)
    • Potential Alternatives:
      A. Request A array for 3.2x better Theta_res, although it will likely only be ~1.6x better for southern sources (assuming a factor 2 fudge factor)
  • Tier 2: S band Galactic Plane Survey (dec >-40 ,|b| < 5) / Bulge (dec >- 40, |l| < 14, |b| < 10) Deep Survey
    This was designed to match Tier 1, adding transients and getting deeper
    Also, regardless of the array chosen, we want to make sure that we work with the Technical Working Group to develop ideas to do parallel processing that will sacrifice point-source sensitivity for a larger beam and thus probe extended sources. Also note, that parts of the bulge area and anticenter plane will provide good windows for extragalactic science.

    • Galactic Science Driver(s):
      Pulsars (Chatterjee), Quiescent Black Hole (& Nearby Neutron Star) X-ray Binaries, transients, radio flaring stars, SNR survey
    • Area:
      3220 sq. deg (2880 sq. deg for dec > -30 only)
    • Cadence:
      15 epochs in addition to Tier 1, logarithmic spacing
    • Single Epoch RMS:
      100 muJy/beam
    • Stacked Epoch RMS:
      25 muJy/beam
    • Stacked Epoch RMS equivalent at K_u band:
      nu^-1.5: 2 muJy/beam
      nu^-1.0: 4 muJy/beam
      nu^-0.5: 11 muJy/beam
      nu^+0.0: 25 muJy/beam
      nu^+0.5: 56 muJy/beam
      nu^+1.0: 125 muJy/beam
      nu^+1.5: 280 muJy/beam
      nu^+2.0: 625 muJy/beam
    • Configuration:
      B array
      (BnA probably be desired for southern sources)
    • Theta_res:
      2.7"
      (Note EWG states 2”, which indicates 25% better centroiding compared to below)
    • Centroiding:
      5 sigma detected centroiding 0.54” @ 125 muJy
      10 sigma detected centroiding 0.27” @ 250 muJy
      20 sigma detected centroiding 0.14” @ 500 muJy
    • S. Myers Survey Speed:
      16.53 sq. deg / hr / epoch
    • Total cost (w 25% overhead): 3650 hr (3270 hr for dec > -30 only)
    • Potential Alternatives:
      A. Request A array for 3.2x better Theta_res, although it will likely only be ~1.6x better for southern sources (assuming a factor 2 fudge factor)
      B. Limit to |b| < 4, reducing total cost (w 25% overhead): 3040 hr (2680hr for dec > -30 only)
  • Tier 3: Ku band Galactic Plane Survey (dec >-40 ,|b| < 5) / Bulge (dec >- 40, |l| < 14, |b| < 10) Deep Survey (GUTS+)
    This was designed to match Tier 2 footprint, and would give spectral index constraints in concert with Tier 2.
    Also, regardless of the array chosen, we want to make sure that we work with the Technical Working Group to develop ideas to do parallel processing that will sacrifice point-source sensitivity for a larger beam and thus probe extended sources. Also note, that parts of the bulge area and anticenter plane will provide good windows for extragalactic science.

    • Galactic Science Driver(s):
      Thermal sources such as Hyper compact HII regions, Massive YSOs, and non-termal broadened recombination lines
    • Area:
      3220 sq. deg (2880 sq. deg for dec > -30 only)
    • Cadence:
      1 epochs performed close to Tier 2 for same coordinates
    • Single Epoch RMS:
      100 muJy/beam
    • Stacked Epoch RMS:
      100 muJy/beam
    • Stacked Epoch RMS equivalent at S band:
      nu^-1.5: 1100 muJy/beam
      nu^-1.0: 500 muJy/beam
      nu^-0.5: 220 muJy/beam
      nu^+0.0: 100 muJy/beam
      nu^+0.5: 45 muJy/beam
      nu^+1.0: 20 muJy/beam
      nu^+1.5: 9 muJy/beam
      nu^+2.0: 4 muJy/beam
    • Configuration:
      C array
      (CnB desired for southern sources???)
    • Theta_res:
      1.8"
      (Note that our 1" resolution does not match S. Myers number for Ku Band at C array)
    • Centroiding:
      5 sigma detected centroiding 0.36” @ 500 muJy
    • S. Myers Survey Speed:
      1.45 sq. deg / hr / epoch
    • Total cost (w 25% overhead): 2780 hr (2480 hr for dec > -30 only)
    • Potential Alternatives:
      A. Request B array for 3.3x better Theta_res, although it will likely only be ~1.7x better for southern sources (assuming a factor 2 fudge factor)
      B. Limit to |b| < 4, reducing total cost (w 25% overhead): 2310 hr (2040 hr for dec > -30 only)

gsivakoff
Posts: 17
Joined: Tue Mar 25, 2014 11:28 am

Re: Galactic WG Summary so far (comments please!)

Postby gsivakoff » Fri Mar 28, 2014 9:28 pm

I still am concerned about one thing with GUTS (or GUTS+ as I have made above by putting the RMS down to 100muJy and added some Galactic Bulge coverage).

How well can we justify GUTS compared to deep S Band unless f_nu is propto nu^(>1.2)?

How far out would it be sufficient to detect massive YSOs out to?
How far out would it be sufficient to detect UCHIIs out to?

(I know I need to answer similar questions for the low frequency science)

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

Re: Galactic WG Summary so far (comments please!)

Postby lsjouwerman » Fri Mar 28, 2014 10:32 pm

Greg,

note that if you want to be sensitive to thermal sources you should not use the center of the band survey speed as quoted by Steves paper as that does not do right to the homogeneous sensitivity coverage at these high freqs. You should use the speed at 18 GHz, which should be about 0.8 sqdeg/hr for 130 uJy/bm, (not 100) and which would take your proposal to 5000+ hours for the Ku band part at 130 uJy/bm.

Why don't we just take the outcome of the all sky as given and just add what we need for galactic science, ie propose a full Ku band on the plane as in GUTS and hope for some extra S band in the bulge.

I would not be surprised that some common ground would be sought for so let the ExGal take the 3000h hit for the all sky in low frequency, of which part covers the plane/bulge with their parameters, and the GalWG pick up 3000h for Ku/GUTS, with the remainder third for multi-epoch obs of the transients in their preferred band (hope for S, or Ku if they need sensitivity and resolution over area) for the bulge as the starting point for the total VLASS proposal to keep everyone happy. Not that I think 10k hours will be allocated, but it will show common grounds and that everyone is willing to compromize for the greater good - also for the non-radio astronomy community (which has a large thermal interest for which no real radio survey exists) that I do not notice in this forum and not that each group is after something that should be done as PI science. You want to do something unique that cannot be done anywhere else (now or in the future), and that won't have a chance to be allocated if it were PI science. I don't think anything low frequency fits these descriptions (I'd say requirements) for a multi-thousand hour VLA survey

Of course, when a low freq all sky (ie FIRST repeat of NVSS) is deemed not fruitful, or when ExGal prefers a deep sky over all sky, then things may change for low freq gal plane

gsivakoff
Posts: 17
Joined: Tue Mar 25, 2014 11:28 am

Re: Galactic WG Summary so far (comments please!)

Postby gsivakoff » Sat Mar 29, 2014 11:38 pm

Lorant,

That's a great point about using the upper band. I guess, what we really would want to do would be to fold in the expected spectral index, which we don't seem to have agreed upon in this forum, and adjust the time appropriately. Unfortunately, I was unable to do such a calculation quickly while I was at a conference. Do you think you can take a crack at it?

I fully agree that my proposal was meant to be one that all three teams would find palatable, and which would have overlap with nearly every community for each part. For what it's worth, the bulge addition to GUTS is relatively small, less than 10% for dec>-40; however, I think it is very worthwhile to put forward the bulge and disc together from the GWG. Your quote of 2800 square degrees seems a bit low for dec>-40, but pretty close to what I get for dec >-30, including a bulge extension.

I think that the VLA is the only instrument that can do deep high resolution surveys in the Northern Hemisphere, now or for quite some time in the future. In the end, I think the key to the VLASS will be its relatively unbiased nature over as large an area as possible, similar to SDSS. I also think it needs to be done to sufficient depth for a wide range of sources, both non-thermal and thermal. Any such project done to the appropriate depth and wide area is not something that can be sold as PI science and will create a strong legacy.

The key issue for this GWG is what is sufficient depth for non-thermal and thermal sources. On the non-thermal end, Shami made it clear that 500 mJy detections don't really cut it for pulsar science. They'd also be fairly limiting for most other compact object science I can think of. This is what is driving a request for 125 muJy detection depth in the plane and bulge, which is a compromise from Shami's requested 100 muJy depth. This would be substantially better than any other existing survey (and would surpass glostar by a factor of 2 sensitivity at a lower frequency, be much more sensitive to steep sources, and (likely) cover a wider area). Plus, to get that depth, we automatically get the transient science for free. There has been quite some disagreement on this forum about the spectral index for the "thermal" sources. My big concern is that if one use nu^1, then there appears to be to first order, little difference between an S Band survey at 25 uJy RMS and Ku Band survey at 130 muJy rms. If one uses nu^1.5, then the difference in depth is only a factor of ~2. My questions are meant to ensure that we clearly overcome such a possible interpretation in our proposal.

Finally, I don't believe it is fair to state that this forum has been ignoring the non-radio astronomy community. My perspective has been influenced most by taking the community view I got considering the proposals I have seen in the past on the ETP panel. Moreover, I also have been gathering input from non radio-astronomers at the conference I was at. Yes, the pulsar crowd (of which I am not really a member) tends to be a radio astronomy crowd, but identifying a BH/NS pulsar has much wider impact for fundamental physics. The MSPs have a clear connection with the gamma-ray crowd. I've talked to optical astronomers about CVs. We've largely ignored the SNRs on this forum, but an unbiased survey would have connections across the multi-wavelength band. My point being that non-thermal emission has a wide impact beyond radio astronomy. Moreover, it is one of the strengths of radio astronomy. I don't feel we need to run away from it to make connections with the entire astronomical community. It does mean however, that we must do a good job in showing the entire community the importance of studying both non-thermal and thermal emission sources and why the VLASS will accomplish both to an unprecedented depth and area.

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

Re: Galactic WG Summary so far (comments please!)

Postby lsjouwerman » Wed Apr 02, 2014 3:58 pm

gsivakoff wrote:Lorant,

That's a great point about using the upper band. I guess, what we really would want to do would be to fold in the expected spectral index, which we don't seem to have agreed upon in this forum, and adjust the time appropriately. Unfortunately, I was unable to do such a calculation quickly while I was at a conference. Do you think you can take a crack at it?


I'm about to step on a plane, but I guess the point I want to make here is that is why we ask for Ku band. That is, because you don't know what the spectral index is, you want to observe there and deduce it from the measurements in combination with the low frequency measurements that are already available.

gsivakoff wrote:I fully agree that my proposal was meant to be one that all three teams would find palatable, and which would have overlap with nearly every community for each part. For what it's worth, the bulge addition to GUTS is relatively small, less than 10% for dec>-40; however, I think it is very worthwhile to put forward the bulge and disc together from the GWG. Your quote of 2800 square degrees seems a bit low for dec>-40, but pretty close to what I get for dec >-30, including a bulge extension.


Not sure what I used for cutoff in Dec, probably not as low as -40 as observing so close to the horizon may not yield great data. I'm in favor of extending into the bulge and other interesting regions, or going deeper overall, but honogeneous coverage of the plane is my main concern so I'd leave the bulge out if that otherwise means cutting the plane - but these things could be discussed of course.
I guess that in the three WGs, we need to say what we want to do as a WG, and not come up with a generalized grand scheme VLASS (which we however both did) as I think that is for the SSG and WG chairs to sort out.

gsivakoff wrote:I think that the VLA is the only instrument that can do deep high resolution surveys in the Northern Hemisphere, now or for quite some time in the future. In the end, I think the key to the VLASS will be its relatively unbiased nature over as large an area as possible, similar to SDSS. I also think it needs to be done to sufficient depth for a wide range of sources, both non-thermal and thermal. Any such project done to the appropriate depth and wide area is not something that can be sold as PI science and will create a strong legacy.


I guess the key phrase here is "for quite some time in the future". I don't think we should use a VLASS allocation of a few thousand hours to do - and potentially can be done much better - what is in the pipeline with SKA, etc, even if that has to wait another 10 years. Better try to go for something that otherwise never ever will happen anywhere.

elisabethmills
Posts: 13
Joined: Tue Mar 11, 2014 3:58 pm

Re: Galactic WG Summary so far (comments please!)

Postby elisabethmills » Wed Apr 02, 2014 4:31 pm

Hi Greg et al.--

A few assorted thoughts

(1)
There has been quite some disagreement on this forum about the spectral index for the "thermal" sources. My big concern is that if one use nu^1, then there appears to be to first order, little difference between an S Band survey at 25 uJy RMS and Ku Band survey at 130 muJy rms.


For thermal sources it is not just the sensitivity to nu^2 sources, but also sensitivity to recombination lines which should drive the choice of frequency if one is interested in (compact) thermal sources. Recombination lines from such higher-density sources will be stronger and less pressure broadened at 15 GHz compared to 3 GHz (see my earlier post, though I am not an expert on this!).

(2)
If part of the goal for Tier 2 is an SNR survey, wouldn't we want better sensitivity to extended emission?

(3)
As with the discussion in the extragalactic groups, I think it is valuable to discuss both synergy and overlap with other suggested surveys. I notice that one of the top surveys planned with MeerKAT is a 7860h Radio Pulsar Timing experiment, and there are also two other Galactic pulsar/transient surveys -- ThunderKAT (3000h + commensal observations) and TRAPUM (3080h). TRAPUM in particular proposes a 5780 square degree survey of the plane (b<15) which they suggest will discover 1000 new pulsars, >100 of which will be millisecond pulsars (considering that one of the PIs argues for pulsar searches with phased ALMA, I am not actually sure what frequency they are going for here).

I think it is important to specify the science that will be uniquely done here by a VLA survey in the face of other upcoming surveys (why do we need higher resolution, for example). Of course, this should also be done for the Ku tier (for which the complementary-in-frequency MeerGal is planned in the south, leading to some overlap).

(4)
It seems like some of the MeerKAT drivers for the above surveys (TRAPUM?) are high-frequency (Ku) pulsar studies. Perhaps the depth proposed for GUTS is simply not useful, but it seems like there are some interesting questions about their spectral behavior at these frequencies, as well as the benefit that at higher frequencies, observations are not limited by propagation effects (like dispersion) which is better for probing intrinsic source properties (and of course better for observations in the highly dispersed Galactic center environment). There are also suggestions that by going higher in frequency one may probe more deeply into the magnetosphere. Can anyone comment on these points?

gsivakoff
Posts: 17
Joined: Tue Mar 25, 2014 11:28 am

Re: Galactic WG Summary so far (comments please!)

Postby gsivakoff » Wed Apr 02, 2014 10:41 pm

And now for the first two of Elisabeth's points. I'm getting a little tired and want to look at the Meerkat surveys tomorrow in (clearer) detail.

For thermal sources it is not just the sensitivity to nu^2 sources, but also sensitivity to recombination lines which should drive the choice of frequency if one is interested in (compact) thermal sources. Recombination lines from such higher-density sources will be stronger and less pressure broadened at 15 GHz compared to 3 GHz (see my earlier post, though I am not an expert on this!).


If I'm reading your plots it looks like having both Ku and S Band would help us contrain emission measure. Would this motivate studying recombination lines in S and Ku band?

(2)
If part of the goal for Tier 2 is an SNR survey, wouldn't we want better sensitivity to extended emission?


This is largely an add-on science goal here assuming we can do a tapered process of the same data. If we can, we obviously would not have the same 25 muJy/Bm sensitivity for the larger (tapered) beam.

gsivakoff
Posts: 17
Joined: Tue Mar 25, 2014 11:28 am

Re: Galactic WG Summary so far (comments please!)

Postby gsivakoff » Wed Apr 02, 2014 11:00 pm

Looks like the post I thought I submitted I did not. So here's try two.

lsjouwerman wrote:I'm about to step on a plane, but I guess the point I want to make here is that is why we ask for Ku band. That is, because you don't know what the spectral index is, you want to observe there and deduce it from the measurements in combination with the low frequency measurements that are already available.


I don't see how we can answer this question if we don't have deep low frequency data. NVSS is way too shallow and FIRST does not cover large chunks of the plane.
So my question is do we already have (or will soon get) the low frequency measurements needed to make these comparisons. If we don't, this suggests a strong synergy between S and Ku band. I'm not 100% sure, but suspect we would want ~matching resolution for the comparison as well. Deep surveys at worse resolution could run into confusion issues; unless we don't expect many sources within ~10" beams. And if not, how are we motivating the 1" needed resolution?

lsjouwerman wrote:Not sure what I used for cutoff in Dec, probably not as low as -40 as observing so close to the horizon may not yield great data. I'm in favor of extending into the bulge and other interesting regions, or going deeper overall, but honogeneous coverage of the plane is my main concern so I'd leave the bulge out if that otherwise means cutting the plane - but these things could be discussed of course.


I'm not advocating bulge coverage vs. plane, rather their combination. I think it would be a community disservice to just do the plane and ignore the bulge, or vice versa. The dec limit sets our coverage. NVSS looked at dec>-40. We can choose to go down that far, or we can make a decision to do less. Either way, I can run code to determine the area of each component of the survey.

lsjouwerman wrote:I guess the key phrase here is "for quite some time in the future". I don't think we should use a VLASS allocation of a few thousand hours to do - and potentially can be done much better - what is in the pipeline with SKA, etc, even if that has to wait another 10 years. Better try to go for something that otherwise never ever will happen anywhere.


I think the high resolution aspect would keep this data's legacy value for longer than 10 years. I have to take a look at the Meerkat planned surveys, but the real SKA (and pathfinders) survey instrument will be ASKAP and its improvement, SKA1_SURVEY. EMU will be done on ASKAP (http://www.atnf.csiro.au/people/Ray.Norris/emu/index.html) likely covering dec < 20 to RMS=10 muJy at L band, but only with a 10" beam. Thus, it will not cover the entire plane, nor at the 1" resolution that has been suggested. If we need 1" resolution in both a low and high frequency band, then we'd have to wait until SKA1_SURVEY. And even then there will be parts of the plane where we are missing data.

I'm beginning to see a complementary science case for both S and Ku band for thermal sources.


Return to “Galactic Working Group”

Who is online

Users browsing this forum: No registered users and 1 guest