2022-11-10

Owned by Gemma Elizabeth Anderson
Last updated: Nov 11, 2022

Minutes for MWA Transient meeting on the theme: FRB detection and follow-up

Attendees: Gemma Anderson, Jun Tian, Clancy James, Bryan Gaensler, Danny Price, Xu Zhijun, Kat Ross, Laura Driessen, Ziteng Wang

Declined due to timezone or other commitments:  Bradley Meyers, Tara Murphy, George Heald


Agenda (see email sent on 9 Nov 2022)

GW-themed specific discussions:

  • Clancy James (5 min + 5-10 min): Overview of Alex Moroianu et al. paper on the possible GW-FRB association
  • Jun Tian (5 min + 5-10 min): MWA repeater follow-up and future plans
  • Bryan Gaensler (5 min + 5-10 min): CHIME FRB plans and possible synergies with MWA
  • Gemma Anderson (5 min + 5-10 min): MWA triggering on FRBs and blind FRB searches (will include content from Bradley Meyers)

Further MWA FRB discussions.

Meeting minutes

Clancy James:

  • GW 190425 BNS merger - only one in O3 out of 25 GW sources, very poor localisation
  • Looking for FRBs in the vicinity of O3 events
  • One association spatially and temporally coincident FRB 20190425A just 2.5 hr later detected by CHIME (774 one off FRBs in catalogue)
  • Product of time and spatial coincidence is not super interesting. However....
  • 150 +/- 50 Mpc, fold in the amount of matter in Universe and uncertainties.
  • 0.5% chance of product time, spatial and distance = 2.8 sigma so very interesting but not super strong
  • It places very interesting upper limit on mass of NS remnant - EoS implications

Questions:

  • Bryan - Many discussions related to the significance of this potential association where this has been presented at meetings. has this been resolved?
  • Clancy - You could count 21 other GW sources as trial sources. Is an FRB association plausible for GW BBH? It comes down to what is your prior and what models are plausible.


Jun Tian:

  • Broad emission range of FRBs 300-8GHz
  • LOFAR detection FRB <200 MHz only low-frequency FRB detection
  • VCS archival search of 6 repeating GRBs - 4 CHIME, 1 ASKAP, 1 Parkes
    • No detection of first 5 - submitted MNRAS
  • Parkes repeater FRB 180301A (pointed out by Danny, also VLA detection) 1-2 GHz,
    • very active in 2021 and 2022
    • 0.7-10 Jyms Fluence so not very bright
    • 1.4 GHz 20 sigma
    • 9 hours VCS data, signal pulse search
    • Four candidates found, the candidate with highest significance is SNR 6.55
  • 36 candidates with S/N > 6, none above 7

Questions:

  • Danny - This is a low S/N. Have you done a subband search to see if the significance increases S/N >7?
  • Jun - Tried splitting band but no huge improvement in S/N so still <7.
  • Danny - worth checking how many false positives are expected with S/N > 6 based on noise statistics and DM trials.
  • Clancy - You mentioned candidates consistent with FRBs. Did you look at other DMs to see how many candidates are found that do not have a similar DM?
  • Jun - DM step size 0.1 so 1000 trials in total over a 100 pc/cc range centred at known DM.
  • Clancy - Create a plot of candidates with DM to see if it peaks at the known DM of the repeaters. this will help demonstrate it is unlikely to be RFI.
  • Xu - What software to search for FRBs
  • Jun - PRESTO
  • Xu - Might be worth trying my software - https://github.com/Xu-Zhijun/STEP
  • Xu - Have you tried processing VCS data from MWAX?
  • Jun - No, the MWA Fast Transients group are still developing software and pipelines for this. The data I have searched are between 2016-2019.
  • Gemma - How would you go about finding more repeating FRBs?
  • Jun - CHIME has more repeaters that will be published soon that we can follow up
  • Jun - Given MWA's sensitivity is low compared to ASKAP and Parkes, it would be best to focus on bright repeaters in the future for a better chance of detection.


Bryan Gaensler:

CHIME FRB update:

  • CHIME finds 3/day
  • Already published 500 FRBs in a catalogue that is public 
  • Published 3 papers on 18 repeaters including one with a 16-day activity window
  • The next repeater catalog will be submitted soon.
    • How many repeaters you ask?
      • Complicated as localisation not great and many FRBs could have similar DMs, characteristics, etc
      • Gold sample physical repeaters ~20
      • Silver sample >20
      • Extra odds and ends
  • CHIME outriggers
    • Baby CHIMES in different places in Canada and USA 
    • Detect FRBs at same time at multiple sites so will achieve mas localisation to bursts
    • Do not have resources for optical/IR/multi-wavelength(MW) follow-up
      • Extensive MW follow-up will likely occur for the first few
      • Following those, MW follow-up only on interesting FRBs
      • After this CHIME will start releasing FRBs as they go when the statistical regime is reached.
      • Outriggers due to come online during the first half of 2023 - good tining for O4!
  • Backlog of baseline processing of thousands of events
    • Getting towards ~4000 bursts detected by CHIME but
    • Most have not been vetted. CHIME is completely overwhelmed as there are not enough people to process the data
  • Online triggering system released as VOEvent if it "thinks" something is FRB-like
    • Most turn out to not be FRBs
    • Subset are real, about ~1 in 4 or 1 in 5 informally
    • We could rigger MWA on VOEvents
  • Next project is to use CHIME to search for objects the vary on timescales of minutes and days
    • Next step is for CHIME to be sensitive to intermediate times
    • Timescales mins
    • Timescales of days
    • Efforts currently on the backburner as CHIME have prioritised repeaters and outriggers.
  • Clancy - outriggers good for host galaxy association. On what timescale might we expect better information on a localised FRB?
    • Bryan's response (from email follow-up with direct quotes from https://chime-frb-open-data.github.io/voevents-faq/):
      • VOEvents carry almost all the header information, including the probability that it is astrophysical. The false alarm rates at various S/N is also quantified for the VOEvent users: see https://chime-frb-open-data.github.io/
      • Real-time VOEvents are verified by humans only after a VOEvent has been published. Following human verification, an event may be found to be a false positive signal, for example, due to RFI contamination. While the real-time FRB detection pipeline performs multiple levels of RFI excision, it is not a perfect filter. 
      • Under the current regime, once per day around 22:00 Pacific Standard Time we will publish retraction VOEvents in bulk for all false positives classified in the previous 24-hour period.
  • Citizen science project, hand classifying FRBs, take results and use as a training set for current auto-detection of FRBs.
  • There is also a person in the loop that currently rates an event as good <10 mins. This could result in sending out additional VOEvent for confirmation or retraction.

MWA Synergies: 

  • Challenges, CHIME is a transit telescope  so best sensitivity at hour angle of 0. This is close to the horizon in WAL ~8 hr difference in LST (far North near horizon)
    • Sensitivity increases at Zenith, at +10, +20 is not great
    • CHIME can formally see down to Dec -10 but sensitivity gets much worse. 
    • Some sky overlap formally but very difficult.
    •  Occasionally see a repeater go crazy 
      • This is currently the case for one at +48 Dec - CHIME can only see it for 15 mins a day but several hundred bursts/hr seen by FAST, been going on for week. 
      • ATel released for such instances
      • Skip hoops of vetting 
    • Gemma - MoU with MWA for knowing repeaters?
      • Bryan Prob not necessary
    • Just look at all southern CHIME repeaters
    • Ideally, MWA looking at following-up CHIME repeaters should look at events with low scattering. 
      • Real-time or near-real-time scattering info is not really possible but is in catalogues
      • Clancy - if CHIME sees them despite being scattered then presumably very bright anyway.
      • Bryan - Depends on where they are detected in the CHIME band.

Gemma Anderson and Bradley Meyers: MWA plans

  • Plans to ingest VOEvents from ASKAP, UTMOST-2D, Parkes, CHIME when they become available
    • CHIME has already developed FRB VOEvent triggers based on the telescope-agnostic VOEvent standard developed by Petroff et al. (2017). We should base all FRB alerts on this. 
    • If the response is <40s between UTMOST-2D event and MWA being on target, can detect events with DMs between 30-3000 pc/cc.
    • Could point MWA to UTMOST-2D sky and list/use buffers.
  • Follow-up of CHIME repeaters that are currently active
  • Do FRB searches as part of the 2nd phase of the SMART survey. Could include SETI searches too.