Searching for cosmic strings With the LIGO-Virgo gravitational-wave - PowerPoint PPT Presentation

Searching for cosmic strings With the LIGO-Virgo gravitational-wave experiments Florent Robinet On behalf of the LIGO Scientific Collaboration and Virgo Collaboration

Topological strings Topological strings Cosmic strings were first introduced by Kibble in 1976. Their formation relies on the Higgs mechanism and on the fact that the universe can transition between different ground states when expanding and cooling. Ground states with different phases When a symmetry is broken in multiple and uncorrelated space-time locations, the field phase can take different values. Stable topological defects may form as a result of the universe expansion. Broken U(1) symmetries → one-dimensional topological defect: Cosmic strings 2 2 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Superstrings Superstrings In the early 2000s, it was realized that cosmic strings can also form in the context of string theory (or M-theory) They are called superstrings CMB measurements show that cosmic strings contribute very little to the primordial density perturbation → Loss of interest for cosmic string Cosmic string rebirth with string theory: – Fundamental strings can grow to macroscopic scales – D-branes partly wrapped in compact dimensions could appear as strings – Brane collisions could result in cosmic strings Cosmic strings could provide observational tests for string theory 3 3 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Cosmic string parameters: G μ , , α α , , p Cosmic string parameters: G μ p Cosmic strings are characterized by the dimensionless mass per unit length or string 2 G μ/ c c = 1 tension From now on, − 6 G μ< 10 CMB data → 2 strings 1 string Cosmic string loops are produced when 2 string segments meet or when a single string loops back → reconnection Kinks are left behind. We note p , the reconnection probability – it is equal to 1 for topological strings – it can be much smaller than 1 for superstrings l ∼α t The loop size is a fraction of the horizon ( ) α Note that is not a "true" parameter. It only reflects our Cosmic string ignorance about the loop size loops 4 4 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Cosmic string signatures Cosmic string signatures Primordial density Keiser-Stebbins inhomogeneities effect Loop oscillation/decay Gravitational Formation of cusps lensing effect → GW radiation GW bursts Kinks are straightened as the strings stretch Stochastic GW → GW radiation background Other possible effects: Electromagnetic radiation form superconducting strings (GRB engines...) Matter wakes behind moving strings Cosmic rays 5 5 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Cosmic string signatures Cosmic string signatures Galaxy surveys CMB line discontinuities Primordial density Keiser-Stebbins inhomogeneities effect CMB non-Gaussianities Fits of the CMB angular power spectrum Loop oscillation/decay Gravitational Formation of cusps lensing effect → GW radiation GW bursts Kinks are straightened Image as the strings stretch patterns Stochastic GW → GW radiation background Interferometers Pulsar timing (LIGO/Virgo) 6 6 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Cosmic string gravitational-wave signatures Cosmic string gravitational-wave signatures Loop oscillation/decay Formation of cusps → GW radiation GW bursts Kinks are straightened as the strings stretch Stochastic GW → GW radiation background Interferometers Pulsar timing (LIGO/Virgo) 7 7 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Cosmic string gravitational-wave signatures Cosmic string gravitational-wave signatures CMB angular power spectrum (WMAP / Planck) Indirect bounds on the stochastic GW background can be obtained from CMB data: Loop oscillation/decay A large GW background at the time of Formation of cusps the decoupling would alter the observed → GW radiation CMB and matter power spectra GW bursts Kinks are straightened as the strings stretch Stochastic GW → GW radiation background Interferometers Pulsar timing (LIGO/Virgo) 8 8 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Cosmic string GW radiation Cosmic string GW radiation The gravitation radiation represents the main mechanism of energy loss of a cosmic string network: ● GW are emitted when kinks straighten ● GW are emitted when loops decay ● GW are emitted by string cusps produced by oscillating loops (~1 cusp per oscillation) G μ Observational constraints are often given as bounds on . Here we adopt the model and conventions described in Damour, T. and Vilenkin, A. Phys. Rev. D71 (2005) 063510 l ∼α t =ϵ Γ G μ t Γ≈ 50 A second loop size parameterization is used: where − 1 t − 3 n ( t )=(Γ G μ) p = 1, ϵ= 1 Standard scenario: Loop density: − 1 t − 3 δ( l −ϵΓ G μ t ) ϵ≪ 1 n ( t )=(Γ G μ p ) Small loop scenario: Loop density: The loop size is given by the gravitational back reaction − 1 t − 3 n ( t )=(Γ G μ p ) ϵ≫ 1 Large loop scenario: Loop density: d ρ GW Ω GW ( f )= f − 1 (Γϵ G μ f t 0 ) − 1 / 3 ∝ G μ p Density of GW background of unresolved sources: ρ c df 9 9 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Cosmic string current limits (GW background only) Cosmic string current limits (GW background only) The searches for a stochastic GW background provide upper limits on the density of GW. These limits can be converted into bounds on the cosmic string parameters. ) c The pulsar timing bound mostly t i c t s e constrains large loop scenarios a r i h d c Astrophys.J 653 (2006) 1571 n o Pulsar timing i ( t s B - M The LIGO bound mostly constrains O C G very small loop scenarios I L Nature 460 (2009) 990 p = 10 − 3 ̃ NB: The (WMAP) CMB indirect bound only applies to backgrounds generated before the CMB decoupling 10 10 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Cosmic string GW bursts from cusps Cosmic string GW bursts from cusps Damour and Vilenkin [1] have shown that the stochastic GW background is strongly non Gaussian. Occasional sharp bursts of GW are expected to be produced by cusps Our search follows the cosmic string model described in [2]. We only consider the small loop regime since the large loop regime is already well-constrained In the frequency domain, the gravitational waveform is: − 4 / 3 h cusp ( f )= A ( z ;G μ , ϵ)× f z With a high frequency cutoff f h ( = redshift ) The expected rate of GW events is derived in [2]. It makes use of a generic cosmology including effects from a late time acceleration This model includes ignorance factors of O(1). We define tilde parameters to absorb these factors ( G ̃ μ , ̃ ϵ , ̃ p ) [1] Damour, T. and Vilenkin, A. Phys. Rev. Lett. 85, 3761 (2000) [2] Siemens et al. Phys. Rev. D.73 105001 (2006) 11 11 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Our search with the LIGO-Virgo network Our search with the LIGO-Virgo network We used the 2005-2010 data set (S5/S6 – VSR1/2/3) at LIGO-Hanford USA detector design sensitivity (4&2 km) We require the coincidence of at least 2 detectors This represents a total of 625 days of observation LIGO-Livingston USA (4 km) Virgo Italy(3 km) 12 12 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Our analysis Our analysis Cosmic string burst events are extracted using match-filtering techniques. We use a template bank with waveforms predicted by Damour and Vilenkin - only one free parameter: (range: 75-4096 Hz) f h We perform a time coincidence with at least 2 detectors (from 2 to 4) We use a multivariate likelihood method to rank our events from the less likely to the most likely to be a GW event. Our ranking statistic is tuned using: ● Background events obtained from fake coincidences using time-shifted data ● Simulated cusp events injected in the data with random amplitude, frequency and sky position 4-interferometers network (H1-H2-L1-V1) = 27 discriminative parameters 13 13 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Did we detect a cosmic string event? Did we detect a cosmic string event? PRELIMINARY 1-sigma stat. error All candidates are less than 1 sigma away from the expected background 14 14 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

What is the sensitivity to cosmic string cusp signals? What is the sensitivity to cosmic string cusp signals? The search sensitivity is measured using fake signals injected in the data. The detection efficiency counts how many injections are recovered and ranked above the loudest event of the search PRELIMINARY Previous LIGO results Phys. Rev. D 80 (2009) 062002 15 15 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10

Cosmic string parameter exclusion results Cosmic string parameter exclusion results The model we use predicts the expected rate of cosmic string events as a function of the cosmic string parameters Knowing the sensitivity of our search we can constrain the model. PRELIMINARY 16 16 Florent Robinet GR20 / Amaldi 10 Florent Robinet GR20 / Amaldi 10