1 / 16

CW signals: plans for S2

CW signals: plans for S2. June 16-17 the group held a f2f meeting to discuss plans for analyses of S2 data (and beyond). Slides from the various presentations can be found at www.lsc-group.phys.uwm.edu/pulgroup/ The content of this presentation is mostly cut-and-paste from these.

deidra
Download Presentation

CW signals: plans for S2

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. CW signals: plans for S2 June 16-17 the group held a f2f meeting to discuss plans for analyses of S2 data (and beyond). Slides from the various presentations can be found at www.lsc-group.phys.uwm.edu/pulgroup/ The content of this presentation is mostly cut-and-paste from these. Next f2f scheduled 1 day before LSC meeting.

  2. Coherent search methods • Time domain Bayesian analyses: • extend to all known isolated pulsars emitting above ~ 50 Hz (Glasgow). • special search for Crab (Glasgow) • Metropolis-Hastings Markov Chain Monte Carlo method for parameter estimation and small parameter space searches (Carleton) • Frequency domain analyses: • extend to inspect a larger parameter space (non-targeted search) (AEI+UWM) • extend to search for signals from a pulsar in a binary system, will be applied to Sco X-1 (Birmingham)

  3. Non-coherent search methods • Unbiased searches (Michigan) • Hough search (UIB, AEI) • Stack-slide (Hanford)

  4. TDS: Known pulsars (R. Dupuis, G. Woan, M. Pitkin GLASGOW) • Perform similar statistical analysis as S1 for all known isolated radio pulsars with fGW > 50 Hz. • Total of 34 pulsars • Computational cost minimal:processing time on Medusa should be around 15 min per pulsar. Another 15 minutes on a workstation to calculate posterior probability. • Automatic characterization of noise in small freq. band • Crab pulsar • young pulsar with large spindown and timing noise: dedicated effort. Would like to perform a coherent analysis using all IFOs for every pulsar. • Preliminary results should be available by August LSC.

  5. Isolated pulsars with fGW > 50 Hz #PULSARFREQUENCY (HZ) 1 B0021-72C 173.708218966053 2 B0021-72D 186.651669856838 3 B0021-72F 381.15866365655 4 B0021-72G 247.50152509652 5 B0021-72L 230.08774629150 6 B0021-72M 271.98722878893 7 B0021-72N 327.44431861755 8 J0024-7204V 207.90 9 J0030+0451 205.53069927493 10 B0531+21 30.2254370 11 J0537-6910 62.055096 12 J0711-6830 182.117237666390 13 J1024-0719 193.71568669103 14 B1516+02A 180.06362469727 15 J1629-6902 166.64990604074 16 B1639+36A 96.362234564 17 J1709+23 215.94 18 J1721-2457 285.9893480119 19 J1730-2304 123.110289179797 20 J1744-1134 245.4261237073059 21 J1748-2446C 118.5382530563 22 B1820-30A 183.82343541777 23 B1821-24 327.40566514989 24 J1910-5959B 119.6487328457 25 J1910-5959C 189.48987107019 26 J1910-5959D 110.6771919840 27 J1910-5959E 218.7338575893 28 B1937+21 641.9282611068082 29 J2124-3358 202.793897234988 30 B2127+11D 208.211688010 31 B2127+11E 214.987407906 32 B2127+11F 248.321180760 33 B2127+11H 148.293272565 34 J2322+2057 207.968166802197 Black dots show expected upper limit using H2 S2 data. Red dots show limits on ellipticity based on spindown.

  6. Time Domain Searches MCMC's (N. Christensen) Definitely useful to nail down signal parameter values around candidates Might be useful for small parameter space searches, e.g. the Middleditch optical pulsar in SNR1987A. On-going work to efficiently search over additional parameters (frequency, fdot, wobble angle) in low signal-to-noise conditions. Automation of TDS (G. Santostasi) Plan to automate the data processing done in the time domain search making it an on-line analysis (perhaps LDAS?) Just starting to look at this.

  7. FDS: parameter space searches (M.A. Papa, X. Siemens – AEI , UWM) • Moving towards a hierarchical search. We now expand the coherent search to inspect a larger parameter space. (At the same time the incoherent stage is being developed and tested) • The limiting factor is computational time need to decide how to spend the computing cycles that we have • We have decided to pursue two types of searches: • short observation time (~1/2 day), no spin-down params, wide band (~300 Hz) centered at ~ 300 Hz, all sky search • perhaps longer observation time, 1 spin-down param, small area search (galactic plane/SNRs), small bands. There is a delicate trade-off between sensitivity, observation time (spanned and effective), resolution in parameter space and class of sources that one chooses to target. Calibration info needed to finalize choice. Note: different choices could be made in order to produce the best ULs.

  8. FDS: Sco X1 (C. Messenger, A. Vecchio – Birmingham) • Sco X-1: • Frequency: not known a priori – search on a wide band (~ 100 Hz) • Doppler effects: • Source position: known (no search over sky position) • Orbit: circular (e < 3x10-4), therefore 3 search parameters (projected semi-major axis, orbital period, initial phase on the orbit) • For T < 1 month period is not a search parameters • Search over only 2 parameters (flat parameter space) • Spin-down/up: number of spin-down parameters depend on the integration time, but for T < 10 days signal monochromatic • Physical scenarios: • Accretion induced temperature asymmetry (Bildsten, 1998; Ushomirsky, Cutler, Bildsten, 2000) • R-modes (Andersson et al, 1999; Wagoner, 2002)

  9. PSR J1939+2134 Sco X-1 (Rejean’s plot) Upper-limit on Sco X-1 Sco X-1 and LIGO/GEO sensitivity PULG F2F Meeting, UWM, 16th - 17th March, 2003

  10. Unbiased All-Sky Search (Michigan)[ D. Chin, V. Dergachev, K. Riles ] Analysis Strategy: • Measure power in selected bins of averaged periodograms • Bins defined by source parameters (f, RA, d) • Estimate noise level & statistics from neighboring bins • Set “raw” upper limit on quasi-sinusoidal signal on top of empirically determined noise • Scale upper limit by antenna pattern correction, Doppler modulation correction, orientation correction • Refine corrected upper limits further with results from explicit signal simulation Unbiased, All-sky CW Search -- Michigan Group

  11. What will be ready by August LSC meeting? • Expect bare-bones baseline limits over samples of clean frequency ranges with semi-analytical efficiency corrections • Primary goal in next two months What is unlikely to be ready by August: • Graceful handling of problematic frequency ranges • Efficiency corrections based on full MC simulations

  12. Hough Transform (B. Krishnan, A.M. Sintes, M.A. Papa – AEI-UIB) • start with SFTs with baseline of ~ 30 minutes (or more if stationarity of data allows it). • for every SFT (a) select frequency bins i such |SFT|ia2/<|SFT|2>a exceeds some threshold  time-frequency plane of zeros and ones. • Hough transform: track patterns of “ones” consistent with what one would expect from a signal.

  13. 1st Driver code – in final validation phase • use of SFT input data (non-demodulated data) • peak selection on a time-frequency plane • Hough transform for the different template parameters:sky location,f0, fk, i.e. build histograms for all points in the template parameter space. • We make use of an efficient implementation of a HT: • stereographic projection is used to tile the sky (to avoid unnecessarydistortions and facilitate calculations) • use of LUT • many computational savings if spin-down search parameters are considered. • Future plans include performing an incoherent search using F statistic as input data (but not for S2). The (LAL) Hough routines are general and can already handle this case.

  14. Plans for S2 • This driver code will analyze a large sky area (on a coarse grid) in agiven frequency band (to be determined),with possible spin-down parameters. • Perform an incoherent Hough search using S2 data,with the main purpose ofdetermine a detailedimplementation using real IFO data and then settingan strategy for signal detection or setting an upper limit • Issues to be addressed mainly concern automatic handling on noise features and robust psd estimation.

  15. Stack slide (M . Landry, G. Mendell – Hanford) • new proposal presented in May at a meeting devoted to new data analysis proposals • based on P.R. Brady, T. Creighton Phys.Rev. D61 (2000) 082001 • it is a hierarchical search with a several passes of coherent and incoherent searches. The incoherent search consists in summing (stacking) appropriately (by choosing the appropriate search frequency values – this is the sliding) the outcome of the previous stage. • new effort, a lot of work. Probably can take advantage of some of the lessons learned while developing the Hough code.

  16. S2 and S3, papers • The group has discussed whether to eventually consider the S2 and S3 data sets together. The discussion was inconclusive, depending on too many unknowns like the character of the S3 data and the outcome of the S2 analyses. For now we’ll concentrate 100% on S2 data. • Papers: • there will likely be more than 1 paper • a possible scenario is to group searches according to the method that was used – common statistical analysis: • ULs on 34 known pulsars • FD methods: search for Sco X1 and for parameter space searches (all-sky,wide band, no spin-down & small area search, narrow band, 1 spin-down param) • incoherent methods: unbiased, hough, stack

More Related