A SEARCH FOR FAST RADIO BURSTS WITH THE GBNCC SURVEY PRAGYA CHAWLA - PowerPoint PPT Presentation

A SEARCH FOR FAST RADIO BURSTS WITH THE GBNCC SURVEY PRAGYA CHAWLA McGill University (On Behalf of the GBNCC Collaboration)

P. Chawla 1 , V. M. Kaspi 1 , A. Josephy 1 , K. M. Rajwade 2 , D. R. Lorimer 2,3 , A. M. Archibald 4 , M. E. DeCesar 5 , J. W. T. Hessels 4,6 , D. L. Kaplan 7 , C. Karako-Argaman 1 , V. I. Kondratiev 4,8 , L. Levin 9 , R. S. Lynch 3 , M. A. McLaughlin 2 , S. M. Ransom 10 , M. S. E. Roberts 11 , I. H. Stairs 12 , K. Stovall 13 , J. K. Swiggum 7 and J. van Leeuwen 4,6 1 Department of Physics & McGill Space Institute, McGill University 2 Department of Physics and Astronomy, West Virginia University 3 National Radio Astronomy Observatory, Green Bank 4 ASTRON, the Netherlands Institute for Radio Astronomy 5 Department of Physics, Lafayette College 6 Anton Pannekoek Institute for Astronomy, University of Amsterdam 7 Department of Physics, University of Wisconsin-Milwaukee 8 Astro Space Center, Lebedev Physical Institute, Russian Academy of Sciences 9 Jodrell Bank Centre for Astrophysics, School of Physics and Astronomy, The University of Manchester 10 National Radio Astronomy Observatory, Charlottesville 11 New York University, Abu Dhabi 12 Department of Physics and Astronomy, University of British Columbia 13 National Radio Astronomy Observatory, Socorro 2017, ApJ, Submitted Pragya Chawla, McGill University December 9, 2016 2

FAST RADIO BURSTS 𝒆 𝒐 𝒇 ⅆ𝒎 DM = ׬ 𝟏 • Millisecond-duration events • 1.4 DM MW < DM < 35 DM MW • Origin unknown, likely extragalactic • 18 FRBs discovered till date in the freq. range of 700 – 1500 MHz. 𝐮 𝐭𝐝𝐛𝐮𝐮 ∝ 𝛏 −𝟓.𝟓 • Scattering, spectral index and free-free absorption could be possible reasons for non- detection at low frequencies. Thornton et al. 2013 Pragya Chawla, McGill University December 9, 2016 3

FAST RADIO BURSTS • Millisecond-duration events • 1.4 DM MW < DM < 35 DM MW • Origin unknown, likely extragalactic • 18 FRBs discovered till date in the freq. range of 700 – 1500 MHz. • Scattering, spectral index and free-free absorption could be possible reasons for non- detection at low frequencies. DM MW Contours Based on NE2001 Model (Cordes & Lazio 2002) Pragya Chawla, McGill University December 9, 2016 4

FAST RADIO BURSTS • Millisecond-duration events • 1.4 DM MW < DM < 35 DM MW • Origin unknown, likely extragalactic • 18 FRBs discovered till date in the freq. range of 700 – 1500 MHz. • Scattering, spectral index and free-free absorption could be possible reasons for non- detection at low frequencies. Pragya Chawla, McGill University December 9, 2016 5

THE GBNCC PULSAR SURVEY • Operates at 350 MHz with a bandwidth of 100 MHz . • The analysis pipeline (Stovall et al. 2014) based on PRESTO (Ransom 2001) includes: • RFI Removal • Dedispersion • Single Pulse Search • Dispersion delays are corrected for by dedispersing the data at a large number of trial DMs. • Single pulses are searched for by convolving the dedispersed time series with boxcars of varying duration. An Example Plot Generated by the Analysis Pipeline Pragya Chawla, McGill University December 9, 2016 6

THE GBNCC PULSAR SURVEY GBNCC Pointings Searched to DM = 3000 pc cm -3 • Maximum DM for FRB search = 3000 pc cm -3 (for 45000 pointings). • 17000 pointings searched to a DM of 500 pc cm -3 have maximum Galactic DM along LOS < 100 pc cm -3 . • Total observing time for all pointings = 84 days. Pragya Chawla, McGill University December 9, 2016 7

THE GBNCC PULSAR SURVEY GBNCC Pointings Searched to DM = 500 pc cm -3 • Maximum DM for FRB search = 3000 pc cm -3 (for 45000 pointings). • 17000 pointings searched to a DM of 500 pc cm -3 have maximum Galactic DM along LOS < 100 pc cm -3 . • Total observing time for all pointings = 84 days. Pragya Chawla, McGill University December 9, 2016 8

THE GBNCC PULSAR SURVEY All GBNCC Pointings Searched for FRBs • Selected pointings with at least one pulse with S/N > 10 . • Single pulse events at DM > 2 DM MW were processed with the grouping and rating algorithm, RRATtrap. • RRATtrap (Karako-Argaman et al. 2015) was developed to detect Rotating RAdio Transients (RRATs). • Enabled discovery of 10 new RRATs in GBNCC data. • RRATtrap output was inspected and no FRBs were detected . Pragya Chawla, McGill University December 9, 2016 9

CALCULATION OF FRB RATE • Assuming Poisson statistics, field of view = 0.41 sq. deg for the GBT beam and a threshold flux density of 0.63 Jy for a pulse of intrinsic width = 5 ms . • We place a 95% confidence upper limit on FRB rate = 3.6 x 10 3 FRBs sky -1 day -1 at 350 MHz. Pragya Chawla, McGill University December 9, 2016 10

CONSTRAINING SPECTRAL INDICES OF FRBs • Assuming a power-law flux density model for FRBs: • Varying the index γ of the log N – log S function of the FRB population ( γ = 0.8, 1.2 and 1.5) such that: • Performing Monte Carlo simulations of FRB flux density distribution consistent with the 1.4-GHz rate estimate reported for the Parkes surveys by Crawford et al. (2016) (3.3 x 10 3 FRBs sky -1 day -1 ) Pragya Chawla, McGill University December 9, 2016 11

ABSENCE OF SCATTERING AND FREE-FREE ABSORPTION Generate 1.4-GHz Scale distribution to Compute 350-MHz flux density 350 MHz by sampling FRB rate above distribution α from a normal S = 0.63 Jy from the ( γ = 0.8, 1.2 or 1.5) distribution (σ = 0.5) resulting distribution For constraining spectral index, α lim, computed 350-MHz FRB rate = 95% confidence GBNCC upper limit γ α lim 0.8 +0.19 1.2 +0.28 1.5 +0.35 Pragya Chawla, McGill University December 9, 2016 12

SCATTERING Scale distribution to Assign t scatt drawn from Generate 1.4-GHz Is t scatt for TRUE 350 MHz by sampling a lognormal distribution flux density all FRBs α from a normal to each detectable FRB distribution > 100 ms? (S > 0.65 Jy) distribution FALSE Increase mean of Mean scattering time lognormal dist. for spectral index α = mean of lognormal dist. Given the observed range of scattering times and FRB rate, R, at 1.4 GHz, γ α lim 0.8 -0.9 1.2 -0.6 1.5 -0.3 13

PREDICTING FRB DETECTION RATES FOR OTHER SURVEYS • Based on flux density distribution consistent with 1.4- GHz rate estimate for Parkes surveys. • Distribution scaled to each survey’s center freq. by sampling α from a normal distribution . • Scattering time for each FRB drawn from a lognormal distribution . Pragya Chawla, McGill University December 9, 2016 14

PREDICTING FRB DETECTION RATES FOR OTHER SURVEYS • Based on flux density distribution consistent with 1.4- GHz rate estimate for Parkes surveys. • Distribution scaled to each survey’s center freq. by sampling α from a normal distribution . • Scattering time for each FRB drawn from a lognormal distribution . Pragya Chawla, McGill University December 9, 2016 15

PREDICTING FRB DETECTION RATES FOR OTHER SURVEYS • Limits on bursts per hour arise because of range of spectral indices considered ( α lim < α < +2 ). • No free-free absorption assumed. γ = 0.8 Pragya Chawla, McGill University December 9, 2016 16

PREDICTING FRB DETECTION RATES FOR OTHER SURVEYS • Predictions consistent with upper limits reported for FRB searches with MWA (Tingay et al. 2015, Rowlinson et al. 2016), LOFAR (Coenen et al. 2014, Karastergiou et al. 2016), AO327 (Deneva et al. 2016) and UTMOST (Caleb et al. 2016). γ = 1.2 Pragya Chawla, McGill University December 9, 2016 17

PREDICTING FRB DETECTION RATES FOR CHIME • CHIME bandwidth divided into 4 parts centered at 450, 550, 650 and 750 MHz. • For γ = 1.5 , simulations predict detection of 1-60 bursts a day for the freq. range of 400-500 MHz . • For 700-800 MHz, simulations predict 5-42 bursts a day . γ = 1.5 Pragya Chawla, McGill University December 9, 2016 18

SUMMARY • No FRBs above a S/N of 10 were detected in GBNCC pointings amounting to a total observing time of 84 days . • We place a 95% confidence upper limit on the FRB rate of 3.6 x 10 3 FRBs sky -1 day -1 above 0.63 Jy at 350 MHz . • Non-detection with GBNCC is consistent with 1.4-GHz rate estimate for α > +0.35 in the absence of scattering and free-free absorption and α > -0.3 in the presence of scattering, for γ = 1.5. • We predict CHIME to detect 3-54 bursts per day assuming the Crawford et al. rate estimate, for γ = 1.5. Pragya Chawla, McGill University December 9, 2016 19