and the quest for gravitational waves A Vicer

and the quest for gravitational waves A. Viceré – INFN Firenze/Urbino for the Virgo Collaboration

Plan of the talk • • • Few words about gravitational waves Working principles of GW detectors The large interferometers in the world, and Virgo A personal choice of science results LSC-Virgo joint observation perspectives Towards GW astronomy: multimessenger opportunities

Ripples in the Cosmic Sea Linearized Einstein eqs. (far from big masses) admit wave solutions (perturbations to the background geometry) GW: transverse space-time distortions propagating at the speed of light, 2 independent polarization Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 3/63

Coupling constants strong e. m. 0. 1 1/137 weak 10 -5 gravity 10 -39 GW emission: very energetic events but almost no interaction In SN collapse withstand 103 interactions before leaving the star, the gravitational waves instead leave the core undisturbed Very early GW decoupling after Big Bang – GW ~ 10 -43 s (T ~ 1019 Ge. V) – ~ 1 s (T ~ 1 Me. V) – γ ~ 1012 s (T ~ 0. 2 e. V) Ideal information carrier, Universe transparent to GW all the way back to the Big Bang!! Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 4/63

PSR 1913+16: GW do exist Pulsar bound to a “dark companion”, 7 kpc from Earth. Relativistic clock: vmax/c ~10 -3 GR predicts such a system to loose energy via GW emission: orbital period decrease Radiative prediction of general relativity verified at 0. 2% level P (s) 27906. 9807807(9) d. P/dt -2. 425(10)· 10 -12 dw/dt (º/yr) 4. 226628(18) Mp 1. 442 ± 0. 003 M Mc 1. 386 ± 0. 003 M Nobel Prize 1993: Hulse and Taylor Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 5/63

Plausible target GW amplitude Luminosity: Amplitude: Compactness C 1 for BH 0. 3 for NS 10 -4 for WD Efficient sources of GW must be asymmetric, compact and fast GW detectors sensitivity expressed in amplitude h : 1/r attenuation Example target amplitude: coalescing NS/NS in the Virgo cluster (r ~10 Mpc) h ~ 10 -21 Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 6/63

Synopsis of sources LONG DURATION SHORT DURATION MATCHED FILTERING Rotating NS Coalescing compact binaries Stochastic GW Supernovae TEMPLATE-LESS METHODS Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 7/63

• • • Few words about gravitational waves Working principles of GW detectors The large interferometers in the world, and Virgo A personal choice of science results LSC-Virgo joint observation perspectives Towards GW astronomy: multimessenger opportunities

Principle of Detection GW induce space-time deformation Measure space-time strain using light Interference fringes Target h ~ 10 -21 (NS/NS @Virgo Cluster) Feasible L ~ 103 m Credit: M. Lorenzini Need to measure: DL ~ 10 -18 m Big challenge for experimentalists! Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 9/63

How much is 10 -18 m? Size of the Universe Virgo cluster Galactic center 1 Light-year Target GW wavelengths Neutron star radius Wavelength of YAG laser Size of an atom Proton radius Virgo sensitivity log 10 r Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 10/63

A real detector scheme Virgo optical scheme Input Mode Cleaner 3 km long Fabry-Perot cavities: to lengthen the optical path to 100 km Laser 20 W Output Mode Cleaner Power recycling mirror: to increase the light power to 1 k. W Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 11/63

Laser Master laser, 1 W F. I. E. O. F. I. Main Beam Path Slave laser, 22 W Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 12/63

Super mirrors Fused silica mirrors Bologna – February 19 th, 2009 35 cm diam, 10 cm thick, 21 kg Scattering losses: a few ppm Substrate losses: 1 ppm Coating losses: <5 ppm Surfacedideformation: l/10013/63 A. Viceré – Università Urbino & INFN Firenze

VIBRATION ISOLATION Superattenuator: filters off the seismic vibrations. Bologna – February 19 th, 2009 The rma l no ise A. Viceré – Università di Urbino & INFN Firenze 14/63

Vacuum enclosure 7000 m 3 Requirements Bologna – February 19 th, 2009 10 -9 mbar for total pressure 10 -14 mbar for hydrocarbons A. Viceré – Università di Urbino & INFN Firenze 15/63

Plan of the talk • • • Few words about gravitational waves Working principles of GW detectors The large interferometers in the world, and Virgo A personal choice of science results LSC-Virgo joint observation perspectives Towards GW astronomy: multimessenger opportunities

The GW detectors network LIGO – Hanford, WA LIGO – Livingston, LA A network of 4 (5) GW detectors GEO 600, Hannover, D VIRGO, Pisa, Italy Virgo and the LIGO Scientific Collaboration have signed a Mo. A for full data exchange and joint data analysis and publication policy Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 17/63

• • • LAPP – Annecy NIKHEF – Amsterdam GPG - Birmingham RMKI - Budapest INFN – Firenze-Urbino INFN – Frascati INFN – Genoa LMA – Lyon INFN – Napoli • • • Virgo: 3 km arms OCA - Nice LAL – Orsay APC – Paris INFN – Padova-Trento INFN – Perugia INFN – Pisa INFN – Roma 1 INFN – Roma 2 POLGRAV – Warsav

h (Hz-1/2) Design sensitivity curves -18 10 Pulsars hmax – 1 yr integration -19 10 1 st generation detectors LIGO Credit: P. Rapagnani Virgo -20 10 GEO BH-BH Merger Oscillations @ 100 Mpc -21 10 Core Collapse @ 10 Mpc QNM from BH Collisions, 100 - 10 Msun, 150 Mpc QNM from BH Collisions, 1000 - 100 Msun, z=1 Resonant antennas BH-BH Inspiral, 100 Mpc -22 10 NS-NS Merger Oscillations @ 100 Mpc BH-BH Inspiral, z = 0. 4 NS, =10 -6 , 10 kpc -23 10 NS-NS Inspiral, 300 Mpc -24 10 100 1000 4 Hz 10

LIGO Commissioning started in 1999. Design sensitivity achieved end 2005 Detector technology demonstrated ! • S 1 - Aug 23, 2002 – Sep 9, 2002 • S 2 - Feb 14, 2003 – Apr 14, 2003 • S 3 - Oct 31, 2003 – Jan 9, 2004 • S 4 - Feb 22, 2005 – March 23, 2005 • S 5 - November 2005 – Fall 2007 1 year of 3 det. coincident data LIGO Scientific Collaboration

Virgo sensitivity evolution

FIRST SCIENCE RUN (VSR 1) From May 18 to Oct 1 2007 Joined LIGO S 5 The detector demonstrated excellent stability Sensitivity improved during the run, exploiting short interruptions BNS Inspiral Range (Mpc) Bologna – February 19 th, 2009 Duty cycle: 84% Longest lock: 94 hours Avg. Lock Duration: 11. 2 hrs Lock Recovering Time: ~30 min A. Viceré – Università di Urbino & INFN Firenze 22/63

Recent progress Reaching the design sensitivity and being limited by fundamental noises is all but simple One has to fight with many little technical noises, often unpredicted and unmodelled Most relevant: scattered light triggered by environmental noise, eddy currents, magnetic couplings, thermal transients Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 23/63

VIRGO vs LIGOs & GEO 600 Same HF sensitivity LIGO slightly better in the mid range Virgo much better at LF GEO 600 not competitive A factor 2 -3 still missing at low frequency. WHY? The big step forward in the last decade has been the demonstration of the interferometers technology. The design sensitivity has been (almost) reached and stability is so good (unexpectedly) that an efficient network could be created. Virgo, now, has opened the road to very low frequency region. Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 24/63

• • • Few words about gravitational waves Working principles of GW detectors The large interferometers in the world, and Virgo A personal choice of science results LSC-Virgo joint observation perspectives Towards GW astronomy: multimessenger opportunities

Science Runs So Far 368 days of triple-coincident LIGO data 2002 2003 2004 2005 2006 2007 LIGO: S 1 S 2 S 3 S 4 S 5 GEO: Since end of S 5 / VSR 1 : –► Upgrading LIGO 4 -km interferometers and Virgo –► GEO and LIGO 2 -km interferometer taking data whenever possible for “Astro. Watch” vigil Bologna – February 19 th, 2009 Virgo: VSR 1 A. Viceré – Università di Urbino & INFN Firenze 26/63

Unmodeled burst searches Supernova collapse: dynamics and waveform badly predictable – Estimated rate: several /yr in the VIRGO cluster, but the efficiency of GW emission is strongly model dependent – Simulations suggest EGW~10 -6 Mʘc 2, but NS kick velocities suggest possible strong asymmetries GW emitted secs hrs [Zwerger, Muller]

All-Sky Burst Searches The most-recent published results use S 4 data LIGO-only – – search[ Classical and Quantum Gravity 24, 5343 (2007) ] ► Searched 15. 53 days of triple-coincidence data (H 1+H 2+L 1) for short (<1 sec) signals with frequency content in range 64 -1600 Hz ► No event candidates observed ► Upper limit on rate of detectable events: 0. 15 per day (at 90% C. L. ) ► Sensitive to GW energy emission as small as ~10 7 M at 10 kpc, or ~0. 25 M at the distance of the Virgo Cluster LIGO-GEO joint search [ CQG 25, 245008 (2008) ] – First use of fully-coherent network analysis for burst signals S 5 / VSR 1 all-sky search is currently under internal review – Factor of ~2 better amplitude sensitivity, and much longer observation time – Doing coherent network analysis using LIGO and Virgo data Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 28/63

Gravitational Waves from Soft Gamma Repeaters SGRs are believed to be magnetars – Occasional flares of soft gamma rays – May be associated with cracking of the crust that excites vibrational f-modes of the neutron star LIGO searched for GW signals associated with SGR flares – Dec. 2004 “giant” flare of SGR 1806– 20 – 190 flares from SGR 1806– 20 and SGR 1900+14 during first year of S 5 – Placed upper limits on GW signal energy for each flare – [ PRL 101, 211102 (2008) ] – Within the energy range predicted by some models LIGO also searched for GW signals matching the quasiperiodic oscillations seen in X-rays in the tail of the Dec. 2004 giant flare – Placed upper limits [ PRD 76, 062003 (2007) ] Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 29/63

Coalescing binaries Pairs of compact stars, like PSR 1913+16, but close to the final “coalescence” – PBH: Primordial Black Holes (in the galactic halo): M in [0. 2, 0. 9] – BNS: Binary neutron stars: M in [0. 9, 3. 0] – BBH: Binary black holes: M in [3, 20] Inspiral signal accurately predictable – Newtonian dynamics – Post-Newtonian corrections (3 PN, (v/c)11/2) [L. Blanchet et al. , 1996] Recent big progress in merger 3 D simulation [Baker et al 2006, Praetorious 2006] – Crucial to extract physics, mostly encoded in the merger phase [Campanelli et al. , PRL, 2006] chirp

Binary Inspiral Searches New result from first year of S 5 data No inspiral signals detected Using population models, calculated 90% confidence limits on coalescence rates: For binary neutron stars: 100 Mpc 3. 8× 10– 2 per year per L 10 For 5+5 M binary black holes: 2. 8× 10– 3 For BH-NS systems: 1. 9× 10– 2 (Slightly tighter limits if BHs are assumed to have no spin) [ Preprint ar. Xiv: 0901. 0302 ] Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 31/63

GRB 070201 Short, hard gamma-ray burst – Leading model for short GRBs: binary merger involving a neutron star Position (IPN) consistent with being in M 31 LIGO Hanford detectors were operating – Searched for inspiral & burst signals Result from LIGO data analysis: No plausible GW signal found; therefore very unlikely to be from a binary merger in M 31 [ Ap. J 681, 1419 (2008) ] Hundreds of GRB occurred during the live time of LIGO and Virgo detectors: still under analysis Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 32/63

Spinning Neutron Stars Non-axisymmetric rotating NS emit periodic GW at f=2 fspin but…weak SNR can be increased by integrating the signal for long time (months) 109 NS in the galaxy, 163 known in LIGO/Virgo band Doppler correction of Earth motion needed ( f/f 10 -4): blind search limited by computing power Data from ATNF Pulsar Catalogue (www. atnf. csiro. au/research/pulsar/psrcat)

Searches for Periodic Signals from Known Radio/X-ray Pulsars Demodulate data, correcting for motion of detector – Doppler frequency shift, amplitude modulation from antenna pattern – For a triaxial star, expect GW signal at twice the spin frequency S 5 preliminary results (using first 13 months of data): – Place limits on strain h 0 and equatorial ellipticity e ► e limits as low as ~10– 7 It’s plausible that an ordinary neutron star could sustain an ellipticity as large as ~10– 6 ; Some models allow larger Bologna – February 19 th, 2009 Crab A. Viceré – Università di Urbino & INFN Firenze 34/63

Searches for a Stochastic Background of Gravitational Waves Weak, random gravitational waves should be bathing the Earth – Left over from the early universe, analogous to CMBR ; or due to overlapping signals from many astrophysical objects / events Results from LIGO S 5 data analysis – Searched for isotropic stochastic signal with power-law spectrum – For flat spectrum, set upper limit on energy density in gravitational waves: – Preliminary result from ~half of S 5 data: 0 < 1. 3 × 10– 5 – Starts to constrain cosmic (super)string and “pre-Big-Bang” models – Final S 5 result to be released soon, with factor of ~2 better sensitivity – will dip below Big Bang Nucleosynthesis bound Or look for anisotropic signal: [ PRD 76, 082003 (2007) ] (S 4 data) Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 35/63

A broader look at the stochastic background Credit: B. Sathyaprakash Laser Interferometer Space Antenna - LISA -2 LIGO S 3, 2 wk data Nucleosynthesis h 1002 < 8 x 10 -4 PRL (2005) Initial LIGO, 1 yr data Expected h 1002 < 2 x 10 -6 100 2 ( 0 h ) 0 Log LIGO S 1, 2 wk data h 1002 < 23 PRD (2004) -4 Cosmic strings Pulsar -6 -8 -10 CMB Advanced IFOs, 1 yr data Expected h 1002 < 7 x 10 -10 Pre-big bang model -12 Inflation EW or SUSY Phase transition -14 Slow-roll Cyclic model -18 -16 -14 -12 -10 -8 -6 -4 -2 Log ( f [Hz]) 0 2 4 6 8 10

• • • Few words about gravitational waves Working principles of GW detectors The large interferometers in the world, and Virgo A personal choice of science results LSC-Virgo joint observation perspectives Towards GW astronomy: multimessenger opportunities

1 st generation detection chances 1 ST GENERATION INTERFEROMETERS CAN DETECT A NS-NS COALESCENCE AS FAR AS VIRGO CLUSTER (15 MPc) LOW EXPECTED EVENT RATE: 0. 01 -0. 1 ev/yr (NS-NS) Bologna – February 19 th, 2009 FIRST DETECTION: POSSIBLE BUT UNLIKELY A. Viceré – Università di Urbino & INFN Firenze 38/63

First step to improve: Virgo+ Important technology progress achieved in the last years. It is possible to upgrade Virgo now, enhancing the sensitivity by 2 -3 (and the rate by one order of magnitude) The Virgo+ package: – more laser power (20 50 W) – compensation of mirror thermal lensing – better electronics – change of input mode cleaner mirror – monolithic suspensions LIGO also undergoing similar upgrades, towards Enhanced LIGO Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 39/63

2 nd generation detectors Virgo and LIGO have achieved the design sensitivity. Data still under analysis, but the expected event rate is low (0. 1 -0. 01 ev/yr) Enhanced LIGO/Virgo+ 2009 Virgo/LIGO 108 ly To increase the chance of first detection and to open the way to GW astronomy we want to enhance the amplitude sensitivity by x 10 (hence the detection rate by x 1000!) Advanced LIGO (USA): funded by NSF. In construction Adv. Virgo/Adv. LIGO 2014 Credit: R. Powell, B. Berger Advanced Virgo: Conceptual Design e Preliminary Project Execution Plan submitted to funding agencies (INFN and CNRS). Facing a project review process to get approved. Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 40/63

Advanced LIGO Projected Sensitivity Factor of ~10 in amplitude sensitivity 10– 21 – 22 10– 23 10– 24 10 Factor of ~1000 in volume 1000 Hz Advanced LIGO is approved and funded; construction has begun Expect to be operational starting in 2014 or 2015 Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 41/63

Adv design features tilt control heavier mirrors low dissipation coating larger spot high finesse cavity compensation of thermal lensing moreover… better vacuum environmental noise signal recycling fused silica suspension fibers reduction low noise electronics … Bologna – February 19 th, 2009 DC detection high power laser (200 W) A. Viceré – Università di Urbino & INFN Firenze 42/63

Binary NS sight distance in Ad. V: ~150 Mpc Advanced LIGO: ~170 Mpc each detector

Virgo upgrade plans Mpc 16 BNS inpiral range – expected progress 103 102 v Ad 101 ed c n a 13 o rg i V 14 15 Advanced Virgo Science Run 1 Advanced Virgo commissioning 100 08 10 12 14 16 18 12 yr Advanced Virgo installation 11 + go Virgo Science Run 3 Installation of monolithic suspensions (? ) r Vi 10 Virgo Science Run 2 09 Virgo+ commissioning Virgo+ installation 08 Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 44/63

Benefits by the LIGO-Virgo network LIGO VIRGO False alarm rejection thanks to coincidence Triangulation allowing to pinpoint the source A network allows to deconvolve detector response and regress signal waveform --> measure signal parameters, including source distance for BNS signals Joint operation yields a longer observation time, and a better sky coverage

BNS events: will we ever see them? Empirical models Use observed (4) galactic binary systems coalescing on timescales comparable to Universe age Infer # of events/Milky Way Equivalent Galaxy Assume galactic density 0. 01 Mpc-3 Population synthesis models Use galactic luminosity to deduce star formation rate Alternatively, use supernova events to calibrate the number of massive stars Model binary formation and evolution to deduce # of systems coalescing in less than Hubble time

BNS: Ad. V predictions Empirical model rather uncertain Small number of systems observed, little statistic Population synthesis still unconclusive Strong dependence on models Ad. V alone sees from O(1) to O(10) events/year Ad. V to operate together with Advanced LIGO! Combined sight distance may exceed 300 Mpc Network will see from O(10) to O(100) events/year

BBH sight distance Ad. V: ~ 700 Mpc

BBH: pop. synth. predictions Notes Sight distance is effective: takes into account the distribution of masses in the population synthesis Only masses < 10 M are simulated BBH population synthesis very uncertain Merger rates vary by factors of hundreds If model A is true, prospects of detection are dim! However. . .

BBH: empirical prediction IC 10 X-1 Binary system in local group (~ 700 kpc) Includes a BH, m~24 Mo, and a massive Wolf-Rayet star, m~ 35 Mo Allows to predict a rate (Bulik et al. ) The WR will evolve in BH, without disrupting the binary system The resulting system should have Mchirp~14 Mo Such systems are detectable by Ad. V up to 1. 1 Gpc. . . Rate for Ad. V should be ~ 250 /year Rate for combined Advanced LIGO – Ad. V ~ 2500/year

Known pulsars: Ad. V limits on h Dots: spin down limits. Beaten by Ad. V for about 40 known objects

Stochastic background limits with Ad. V H 1 - L 1 H 1 - V 1 One year of operation of Ad. V – Ad. LIGO Will improve over nucleosynthesis bounds by several orders For comparison, LIGO S 5 results should be just below BBN limit Ad. V contribution depends on the exponent n of the stochastic background model, and is more relevant for larger n

Astrophysical backgrounds A network can locate point sources of random GW signals Such could be objects of astrophysical interests, for instance very large black holes in active galaxies LIGO – Virgo network, with multiple baselines, improves sensitivity by 25% at equator and by 42% at poles, over LIGO only Source localization is improved by a factor O(10)

Advanced Detectors will see GWs The technology of interferometric detectors has been demonstrated A further step in sensitivity appears necessary to open the way to physics and astronomy. Some sources appear certain, unless astrophysical assumptions are wrong To make science, a multimessenger approach will be mandatory Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 54/63

• • • Few words about gravitational waves Working principles of GW detectors The large interferometers in the world, and Virgo A personal choice of science results LSC-Virgo joint observation perspectives Towards GW astronomy: multimessenger opportunities

Targeting SNe; low energy 's. . . Boost detection confidence Neutrino and GW expected within a few ms delay Very tight coincidence can be required Constrain mass strongly 1 ms accuracy: m < 1 e. V constrain

High energy 's KM 3 Net and Ice. Cube will see with E up to 100's Ge. V Coverage of Southern and Northern sky Reconstruction capabilities in the 1° range Common targets: GRB's, SGR giant flares, etc. . .

Targeting GRB events Swift now, Fermi (GLAST) keep looking at rays from GRB powered by accretion disks on newly formed objects • Neutrino and GW expected within a few ms delay Short GRB (< 2 s) potentially related to BNS, BH-NS Long GRB (>2 s, average 30 s) related to (classes of) SNe Again, boost detection confidence Provide insight in the fireball mechanism

Other messengers. . . Radiotelescopes Crucial, f. i. to “lock” on pulsar signals (Automated) Optical telescopes To alert GW detectors of interesting events To follow up triple coincidences observed in GW detectors X-ray telescopes Privileged eyes on the hot material falling into compact objects For instance, in LMXB Another eye at GRB events . .

Beyond the 2° generation? Where and how can we reduce the detector noise? Seismic Underground detectors Thermal New materials Cryogenic interferometers Shot High power laser Better optics New optical configuration QND techniques

ET, THE “ULTIMATE” DETECTOR Underground facility to minimize seismic noise Mirrors held at cryogenic temperature Longer arms, new geometry E. T. - Einstein gravitational-wave Telescope Design Study Proposal funded by EU within FP 7 Large part of the European GW community involved (EGO, INFN, MPI, CNRS, NIKHEF, Univ. Birmingham, Cardiff, Glasgow) Credit: H. Lück Bologna – February 19 th, 2009 A. Viceré – Università di Urbino & INFN Firenze 61/63

Credit: B. Sathyaprakash h (1/√Hz) 10 -22 Current detectors LISA 2015 10 -23 10 -24 Adv detectors 2008 2013 3 rd generation 2020 10 -25 0. 1 m 10 m 1 Hz 100 10 k frequency f / binary black hole mass whose freq at merger=f 4 x 107 4 x 105 4 x 103 M 40 0. 4

Conclusions? I recall lessons at a summer school in theoretical physics in Parma in 1997: detectors still mostly on paper. General skepticism This seminar in 2009: 1° generation detectors demonstrated! LIGO and Virgo upgrading towards 2° generation We hope that 2° generation will allow to start GW ASTRONOMY! To make the most science of it, close cooperation with the astrophysical community will be a must.