La rivelazione diretta di onde gravitazionali: tecniche sperimentali e implicazioni scientifiche Gianluca Gemme Istituto Nazionale di Fisica Nucleare - Genova LIGO Livingston Observatory Louisiana, USA LIGO Hanford Observatory Washington, USA Virgo, Cascina, Italy

New Era in Astronomy! 2 14 Sep 2015: First detection of Gravitational Waves! Phys. Rev. Lett. 116, (2016) 229,000 paper downloads from APS in the first 24 hours, servers down! Binary black holes do exist! and we can listen to them coalesce Binary black holes do exist! and we can listen to them coalesce This is the birth of gravitational wave astronomy Torino - Apr 14, 2016Gianluca Gemme

Gravitational Waves 3 the waves have two components, rotated by 45º from each other What are GWs?  a consequence of General Relativity  ripples in space-time due to cosmic cataclisms  quadrupolar distortions of distances between freely falling masses strain Torino - Apr 14, 2016Gianluca Gemme

Tiny interaction with matter: Extremely difficult to detect Ideal messengers from remote space-time regions Can bring a whole new view of the Universe h VIRGO cluster Binary NS coalescence h Virgo Cluster 50 million light-years away (15Mpc) NS Binary Time Gravitational waves are hard to measure, they are small... 4 h =  L/L ~ even with test masses L~km far apart, displacement is  L~ m h =  L/L ~ even with test masses L~km far apart, displacement is  L~ m Torino - Apr 14, 2016Gianluca Gemme

DETECTING GW S WITH INTERFEROMETERS

Interferometer: a Gravitational Wave Transducer 6 Laser used to measure relative lengths of two orthogonal arms …causing the interference pattern to change at the photodiode Torino - Apr 14, 2016Gianluca Gemme

How to improve sensitivity? 7  Very long arms to get a larger displacement Measure difference in length to one part in or meters  Fabry-Perot cavity in each arm To increase phase change  Recycle injected power To increase input power  Recycle outgoing signal To amplify the output Torino - Apr 14, 2016Gianluca Gemme

GW Detectors - Noise 8 Thermal noise (coating + suspension) Thermal noise (coating + suspension) Radiation pressure fluctuation Residual gas (phase noise) Seismic vibration Newtonian noise Seismic vibration Newtonian noise Stray-light EM field quantum noise Residual laser noise Torino - Apr 14, 2016Gianluca Gemme

GW Detectors - Noise Limiting noises at different frequency ranges: High-freq: quantum shot-noise 9Torino - Apr 14, 2016Gianluca Gemme Class. Quantum Grav. 32 (2015) Low-freq: newtonian noise, seismic noise, residual technical noises Mid-freq: thermal noise

 Dominated by seismic noise  Managed by suspending the mirrors from extreme vibration isolators (attenuation > )  Technical noises of different nature are the real challenge in this range  Ultimate limit for ground-based detectors: gravity gradient noise Low frequency noise 10Torino - Apr 14, 2016Gianluca Gemme

Credit: M.Lorenzini Newtonian Noise 12Torino - Apr 14, 2016Gianluca Gemme

 Mid frequency range: Dominated by thermal noise of mirror coatings and suspensions  Reduced by: Larger beam spot (sample larger mirror surface) Test masses suspended by fused silica fibers (low mechanical losses) Mirror coatings engineered for low losses Intermediate frequency noise 12Torino - Apr 14, 2016Gianluca Gemme

High frequency noise  High frequency range: Dominated by laser shot noise. Improved by increasing the power: >100W input, ~1 MW in the cavities  Requires: New laser amplifiers (solid state, fiber) Heavy, low absorption optics (substrates, coatings) Sophisticated systems to correct for thermal aberrations 13Torino - Apr 14, 2016Gianluca Gemme

Advanced Virgo in a nutshell  Advanced Virgo (AdV): upgrade of the Virgo interferometric detector of gravitational waves  Participated by scientists from Italy and France (former founders of Virgo), The Netherlands, Poland and Hungary  First science data scheduled in 2016 APC Paris ARTEMIS Nice EGO Cascina INFN Firenze-Urbino INFN Genova INFN Napoli INFN Perugia INFN Pisa INFN Roma La Sapienza INFN Roma Tor Vergata INFN Trento-Padova LAL Orsay – ESPCI Paris LAPP Annecy LKB Paris LMA Lyon NIKHEF Amsterdam POLGRAW(Poland) RADBOUD Uni. Nijmegen RMKI Budapest 5 European countries 19 labs, ~200 authors

Advanced detectors  10X better amplitude sensitivity –Event rate ∝ (reach) 3 ~ 1000X greater 1 day of observation with Advanced detectors 1 year with Initial detectors

THE EVENT GW150914

time analyzed to determine the significance of GW (Sept 12 - Oct 20, 2015, 39 days, 16 days of obs data)

O1 sensitivity average measured strain-equivalent noise, of the Advanced LIGO detectors during the time analyzed to determine the significance of GW (Sept 12 - Oct 20, 2015) Torino - Apr 14, 2016Gianluca Gemme18

September 14, 2015 – 12:56 CET Torino - Apr 14, 2016Gianluca Gemme19

September 14, 2015 – 11:50:45 CET LIGO Livingston Observatory LIGO Hanford Observatory Initial detection made by a low latency searches for generic GW transients: Coherent WaveBurst Reported within 3 minutes after data acquisition Torino - Apr 14, 2016Gianluca Gemme20

GW150914: the signal PRL 116, (2016)  Top row left – Hanford  Top row right – Livingston  Time difference ~ 6.9 ms with Livingston first  Second row – calculated GW strain using Numerical Relativity Waveforms for quoted parameters (solid) compared to reconstructed waveforms (shaded)

Why black holes? Binary neutron stars excluded Binary made by one BH and one NS? If so, M BH very large ⇒ Coalescence takes place at lower frequencies NS-BH binary excluded ~ 210 km ~

Source Parameters for GW  Median values with 90% credible intervals, including statistical errors from averaging the results of different waveform models. Masses are given in the source frame: to convert in the detector frame multiply by (1 + z)  Source redshift assumes standard cosmology  Total energy radiated in gravitational waves is 3.0±0.5 M o c 2 The system reached a peak ~3.6 x10 56 ergs, and the spin of the final black hole < 0.7

Transient noise  Detectors were operating in their nominal state at the time of GW  Still contain non-Gaussian transients, examples:  Anthropogenic noise  Seismic noise  “Blip” transients  Mitigate noise by “vetoing” times of elevated noise, measured in auxiliary channels.  Data are clean and stationary around time of event arXiv: Torino - Apr 14, 2016Gianluca Gemme24

Transient Event Searches Binary Coalescence search  Targets searches for GW emission from binary sources  Component masses 1 to 99 solar masses; total mass, up to 100 solar masses dimensionless spin < 0.99  ~250,000 wave forms, calculated using analytical and numerical methods, are used to cover the parameter space arXiv: Torino - Apr 14, 2016Gianluca Gemme25

Matched filtering Calculate matched filter signal/noise as function of time  (t) and identify maxima and calculate   to test consistency with matched template, then apply detector coincidence within 15 msec Calculate quadrature sum of the signal to noise of each detector Background: Time shift and recalculate 10 7 times equivalent to 608,000 years 26

GW has = 23.6 (largest signal), corresponding to false alarm rate less than 1 per 203,000 years or significance > 5.1  Torino - Apr 14,

IMPLICATIONS OF GW150914

Astrophysics implications Key facts  Binary black holes do exist! Form and merge in time scales accessible to us Predictions previously encompassed [0 – 10 3 ] / Gpc 3 / yr Now we exclude lowest end: rate > 1 Gpc 3 / yr  Masses (M > 20 M ☉ ) large compared with known stellar mass BHs  Progenitors are Likely heavy, M > 60 M ☉ Likely with a low metallicity, Z < 0.25 Z ☉  Measured redshift z ~ 0.1  Low metallicity models can produce low-z mergers at rates consistent with our observation ApJL, 818, L22, 2016 Torino - Apr 14, 2016Gianluca Gemme29

A bright future?  GW BBH could have been born either Recently, with a short merger time Earlier, with a long merger time We cannot distinguish with a single observation  Depending on models, at higher redshifts the rate increases! Potential for a very bright aLIGO, AdV future! Dominik 2013, adapted Torino - Apr 14, 2016Gianluca Gemme30

Testing GR  Most relativistic binary know today : J Orbital velocity  GW : Higly disturbed black holes Non linear dynamics  Access to the properties of space-time Strong field, high velocity regime testable for the first time  Tests : Waveform internal consistency check Deviation of PN coefficients from General Relativity Bound on graviton mass  All tests are consistent with predictions of General Relativity arXiv: Torino - Apr 14, 2016Gianluca Gemme31

Waveform internal consistency 1.Predict final black hole mass and spin from the inspiral signal 2.Predict final black hole mass and spin from the ring-down phase 3.Compare to check consistency of GR in different regimes arXiv: Torino - Apr 14, 2016Gianluca Gemme32

Deviation of PN coefficients from GR  Post Newtonian formalism  Phase of the inspiral waveform -> power series in  Nominal value predicted by GR  Allow variation of the coefficients  Is the resulting waveform consistent with data ?  No evidence for violations of GR arXiv: Torino - Apr 14, 2016Gianluca Gemme33

Upper bound on the graviton mass  If  gravitational waves have a modified dispersion relation  Findings : at 90 % confidence, or equivalently arXiv: Torino - Apr 14, 2016Gianluca Gemme34

Any neutrino background ? arXiv: Torino - Apr 14, 2016Gianluca Gemme35

EM follow-up key facts  LVC called for EM observers to join a follow-up program –LIGO and Virgo share promptly with astronomers interesting triggers; up to a few at current sensitivity –Provide limited directional information, promptly estimated  Big participation to GW observation: –24 groups carried out observations –Challenging! Source location with large uncertainty ~ 600 deg 2 Torino - Apr 14, 2016Gianluca Gemme36

Sky location Source location with large uncertainty ~ 600 deg 2 Torino - Apr 14, 2016Gianluca Gemme37

Why is our error box so large?  Two interferometers (HL), each with poor directionality, determine by time delay an annulus in the sky  Folding in also amplitude information, we can do a bit better (in the RHS, a simulation with a BNS event) Torino - Apr 14, 2016Gianluca Gemme38

A sky map produced by LIGO and Virgo is tiled with multiple observations, searching for transients Looking for fading objects, repeat observations after days How do we cope? With telescope time.. Torino - Apr 14, 2016Gianluca Gemme39

.. and smart algorithms Torino - Apr 14, 2016Gianluca Gemme40

Torino - Apr 14, 2016Gianluca Gemme41

Torino - Apr 14, 2016Gianluca Gemme42 In the design LIGO-Virgo network, GW could have been localized to ~20 deg 2

43 The advanced GW detector network: GEO600 (HF) 2011 Advanced LIGO Hanford 2015 Advanced LIGO Livingston 2015 Advanced Virgo 2016 LIGO-India 2022 KAGRA 2017 Torino - Apr 14, 2016 Gianluca Gemme 43

In the future, we’ll be more precise As sensitivity progresses, so does the localization… Torino - Apr 14, 2016Gianluca Gemme44

OUTLOOK AND CONCLUSIONS “Was that you I heard just now, or it was two black holes colliding?” © The New Yorker (2016)

Torino - Apr 14, 2016Gianluca Gemme46

Rate estimates  With only one event, can’t measure rates accurately  Estimates also depend upon astrophysical assumptions Star Formation Rate Delay between formation and merger, in turn depending on initial eccentricity, spin ….  Rate: Gpc -3 yr -1 Consistent with former predictions: Gpc -3 yr -1 (Abadie et al arXiv: )arXiv:  Including LVT – 400 Gpc -3 yr -1  Overall 4 – 600 Gpc -3 yr -1  At high redshifts, could be many more sub-threshold Not detectable as individual signals Potentially detectable as a correlated noise arXiv: Torino - Apr 14, 2016Gianluca Gemme47 How many BBH merger in future data?

Expectations for future runs Torino - Apr 14, 2016Gianluca Gemme48 Probability of observing N > 0 (blue) N > 5 (green) N > 10 (red) N > 35 (purple) highly significant events, (FARs <1/century) as a function of surveyed time-volume arXiv:

Conclusions Binary black holes do exist! Form and merge in time scales accessible to us and we can listen to them coalesce Binary black holes do exist! Form and merge in time scales accessible to us and we can listen to them coalesce We built audio telescopes to sense the vibrations of spacetime that these events send out This is the birth of gravitational wave astronomy We built audio telescopes to sense the vibrations of spacetime that these events send out This is the birth of gravitational wave astronomy We look forward to upcoming science runs with Advanced Virgo online!