Skip to main content
European Commission logo print header

Einstein gravitational-wave Telescope

Final Report Summary - ET (Einstein gravitational-wave telescope)

Executive summary:

The Einstein gravitational-wave telescope (ET) will be an observatory of the third generation aiming to reach a sensitivity for Gravitational wave (GW) signals emitted by astrophysical and cosmological sources about a factor of 10 better than the advanced detectors currently being built. An observatory with such a level of sensitivity will open the era of routine GW astronomy. The main purpose of the ET project is the realisation of an infrastructure (an 'observatory') capable of hosting more than one GW detector. This infrastructure will be usable for many decades, while the implemented detectors will undergo successive upgrades or replacements according to the current state of the art of interferometer technologies. To reduce the effect of the residual seismic motion, thus allowing a better sensitivity at low frequencies (between 3 and 100 Hz), ET will be located underground at a depth of about 100 m to 200 m and, in the complete configuration, it will consist of three nested detectors, each in turn composed of two interferometers (xylophone configuration) having arms of about 10 km of length. Thanks to such infrastructure, Europe will lead the GW astronomy in the next decade. In the next sections the impressive science potential of the ET observatory is draft, describing the targets in fundamental physics, General Relativity (GR), astrophysics and cosmology investigated during the design study, the design achievements, with the selected topology and geometry of the observatory, the technology needs identified and the cost and timing evaluations. All that aspects are deeply investigated and described in the design study document, freely available online.

Project context and objectives:

GW, according to the Einstein GR theory, are generated by stellar masses under huge accelerations. As GR is a metric theory of gravity, GW can be thought to be deformations of the space-time-geometry propagating at the speed of light. However, GW were found to be extremely faint signals. For instance, the amplitude of a GW, described as the strains in space, in the vicinity of the Earth created by a highly dynamical astrophysical event, typically searched for by a GW detector, is of the order of 10-21 (m / Hz) or lower.

The direct detection of GW is one of the most important open issues of the theory of gravity which has impacts on various fields of science and on modern technology. The first direct detection of them will open a totally new window to the universe as what we know about our universe is based on traditional astronomy using various frequency bands of electromagnetic radiation. Recall that according to the conclusion drawn by recent observations on the base of the standard cosmological model 96% of our universe does not emit electromagnetic radiation at all as it is supposed to be composed of either dark energy or dark matter. Despite its weakness gravity is believed to be the dominant force of the Universe, governing the evolution of astrophysical objects and the entire cosmos.

Although the direct detection is still missing, an indirect confirmation of the existence of GW has been obtained. The discovery and precise timing measurements of a binary pulsar system by Hulse and Taylor in the 70 showed that the decay of the orbit of this system over 10 years could be accurately accounted for by the emission of GW, as predicted by Einstein's theory. A Nobel Prize was awarded in 1993 to Hulse and Taylor for the discovery of this binary system and for the long lasting stringent measurement they made.

The first generation of interferometric GW detectors (GEO600, LIGO, TAMA, Virgo) have reached or approached their design sensitivities, and thus demonstrated the effectiveness of the working principle. Advanced detectors (like Advanced LIGO and Advanced Virgo), also called the second generation, will show a sensitivity improved roughly by a factor of ten with respect to the initial interferometers.

The advanced detectors are based on technologies currently available, sometimes tested in reduced scale prototypes, but still to be implemented in full scale. According to the current models of GW sources, the sensitivity of the advanced interferometers is expected to guarantee the detection of signals generated by astrophysical sources within months to a year at most. For example, at the nominal sensitivity of the advanced detectors, the expected detection rate of the GW signal generated by a binary system of coalescing Neutron star (NS)s is about a few tens per year.

But apart from extremely rare events, the expected signal-to-noise ratio (SNR) of these detections obtained with the advanced detectors is too low for precise astronomical studies of the GW sources and for complementing optical, X-ray and X-ray observations in the study of fundamental systems and processes in the Universe.

These considerations led the GW community to start investigating a new (third) generation of detectors. In particular, the European Commission supported the institutions to realise the ET conceptual design study within the Seventh Framework Programme (FP7) 'Capacities'. With a considerably improved sensitivity, such new machines of the third-generation will open the era of routine GW astronomy and with the ET project Europe will lead this scientific revolution. Since the first detection of a GW signal is expected to occur in the advanced interferometers, the evaluation of the scientific impact of ET it is especially focused on the observational aspects rather than on the detection capabilities.

To realise a third-generation GW observatory with a significantly enhanced sensitivity (we defined a target of a factor of ten improvement in sensitivity for ET with respect to the advanced detectors over a wide frequency range), several limitations of the technologies adopted in the advanced interferometers must be overcome and new solutions to be developed are proposed in the ET design study to reduce the fundamental and technical noises that will limit the next generation machines.

However, the first and main target of the ET conceptual design study is the definition of the requirements and of the main characteristics of the site hosting ET, the design of the key elements of the research infrastructure, the rough evaluation of the costs and of the timeline of its implementation. To understand the importance and the need of the site and infrastructures in the ET design it is worth to recall the history of the current GW detector infrastructures.

Current and advanced detectors are using the same infrastructure that will be about 20 years old when the second generation will be online. Further improvements of the sensitivity will be limited by the constraints of the site and infrastructure (arm length, local seismic noise, absence of cryogenic apparatus, vacuum pipe size). Indeed, in order to realise a third-generation GW observatory like ET, an infrastructure hosting several detectors that evolve for many decades is crucial.

Project results:

The research activities performed in the ET design study have been addressed to the full understanding of the scientific potential of a 3rd generation GW observatory, to the definition of the requirements of a new site for ET, to the drafting of the detector design and to the identification of the needed new technologies. All these aspects have been investigated and coherently combined to realise the design of the future ET observatory.

Understanding the scientific potential of ET

Interferometric detectors that are currently taking data, and advanced detectors that will be built over the next five to ten years, will be the first steps in establishing the field of gravitational astronomy. Advanced Virgo and LIGO are expected to observe several tens of inspiralling and merging binaries of NSs and Black hole (BH)s each year. They could also detect occasional Galactic sources such as transients associated with supernovae, glitching pulsars, or soft gamma-ray bursts. This phase of observation will, for the first time, test Einstein's theory in the dissipative regime beyond the basic quadrupole approximation, verify the existence of Binary BHs (BBH)s, measure the speed of gravitational radiation relative to the speed of light and map the expansion rate of the Universe on scales of hundreds of megaparsec (Mpc), providing a completely independent estimate of the Hubble parameter. Advanced Virgo and LIGO will be sensitive to Binary NSs (BNSs) at a distance of 200Mpc and to stellar mass BBHs at a redshift of z circa 0.5: The lower frequency cut-off of a detector places an upper limit on the total mass of binary systems they can detect. A lower frequency cut-off of 20 Hz in the case of Virgo and LIGO limits the total mass to be less than about 200 solar masses. ET plans to improve the amplitude sensitivity by an order of magnitude and extend the frequency sensitivity down to 1 Hz. This will allow astronomers to explore binaries at cosmological distances and unveil new sources: ET would observe BNS up to a redshift of z circa 2; stellar-mass BBH population at the edge of the Universe (z circa 15) and intermediate-mass (102-104) BBH out to a typical redshift limit z circa 5. ET will be sensitive to supernovae out to a distance of 15 million light years within which one might expect to observe an event every year. An observatory with the capability of ET will produce tremendous scientific payoffs. ET will make it possible to observe a greater variety of phenomena and provide a new tool for expanding our knowledge. In the ET design study have been identified the crucial questions at which such as observatory will be able to answer:

Fundamental physics and gravity

Is the nature of gravitational radiation as predicted by Einstein's theory?

ET will allow a test of the wave generation formula beyond the quadrupole approximation and check whether there are only two polarisations as predicted by Einstein's theory or six as in scalar tensor theories. It could accurately measure the GW propagation speed by coincident observation of GW and electro-magnetic (EM) radiation from BNS coalescences at z circa 2 and constrain the graviton mass.

Are BH spacetimes uniquely given by the Kerr geometry?

By measuring different quasi-normal modes, ET will test if the spacetime geometry of a BH is uniquely described by its mass and spin. Additionally, ET can measure the multipole moments of a source from the radiation emitted as a stellar-mass BH spirals into an intermediate-mass BH and confirm if the different moments depend only on the massive BH's mass and spin.

What is the physics of gravitational collapse?

ET can study supernovae and explore if they leave behind a massive object that is trapped inside an event horizon or lead to a naked singularity, or some other exotic object. ET could well reveal a new class of objects and phenomena, for instance silent supernovae and other gravitationally unstable transients.

What is the equation of state (EoS) of matter at supra-nuclear densities as might be found in NS cores?

The EoS of NSs affects the late-time evolution of BNS and NS-BH (NSBH) binaries. By matching the observed radiation from the coalescence of such sources to theoretical predictions, ET will deduce the EoS of NS cores.

What is the maximum mass of a NS?

The maximum mass of a white dwarf is circa 1:4 M as determined by the electron degeneracy pressure. The maximum mass of a NS is an additional test of the nature of matter at extremely high densities; it is currently unknown and should be determined by accurately constructing their mass function from millions of BNS systems observable in ET.

Cosmology

What are the true luminosity distances of cosmological sources?

BBH and BNS binaries are an astronomer's ideal standard candles or, more appropriately, sirens. GW observations alone can determine both the apparent and absolute luminosity of a source. With ET these standard sirens can be used to calibrate the cosmic distance ladder.

What is the EoS of dark energy and how does it vary with redshift?

ET could observe thousands of coalescing BNS and NSBH systems in coincidence with optical or X-ray observations and hence measure both the luminosity distance and redshift. ET will, therefore, facilitate precision measurement of the dark energy EoS and its variation with redshift.

How did BHs at galactic nuclei form and evolve?

ET can verify if seeds of galaxy formation were intermediate BHs of hundreds to thousands of solar masses and map their merger history up to redshifts of z circa 5-15 depending on the total mass and mass ratio of progenitor binaries.

What were the physical conditions in the primeval Universe and what phase transitions occurred in its early history?

Stochastic GW backgrounds could be produced by quantum processes in the primordial Universe or during phase transitions in its early history. ET will be sensitive to background densities GW 10-12.

Astrophysics and multimessenger astronomy

What is the mass function of BHs and NSs and their redshift distribution?

ET will measure masses and spins of millions of NSs and BHs in binary systems and will thereby obtain a census of these objects as a function of redshift. This will be a very valuable database for understanding a host of questions in astronomy related to redshift evolution of compact objects.

What are the progenitors of ?-ray bursts?

Gamma-ray burst (GRB)s are the most luminous electromagnetic sources in the Universe. While advanced detectors might provide some clues as to their origin, ET will provide a large statistical sample of events that could be used to understand GRB progenitors and to test their astrophysical models.

How do compact binaries form and evolve?

The process by which main sequence binary stars evolve into compact binaries (that is, BNS and BBH) could be understood by ET's observation of millions of coalescing binaries with different masses, mass ratios and spins and mapping the observed population to astrophysical models.

What is the physical mechanism behind supernovae and how asymmetric is the gravitational collapse that ensues?

Supernovae are complex processes whose modelling requires many different inputs, including relativistic magneto-hydrodynamics, GR and nuclear and particle physics. ET's observation of supernovae in coincidence with EM afterglows and neutrinos could provide the data necessary to constrain models and help understand the process by which stars collapse to form NSs and BHs.

Do relativistic instabilities occur in young NSs and if so what is their role in the evolution of NSs?

Non-linearities of GR could cause instabilities in NSs that lead to parametric amplification of GWs. ET's observations of the formation of NSs can explore if such instabilities occur in young NSs and how that might affect their spin frequencies.

Why are spin frequencies of NSs in low-mass X-ray binaries bounded?

ET will verify if gravitational radiation back-reaction torque is responsible for the observed upper limit on NS spin frequencies in low-mass X-ray binaries.

What is the nature of the NS crust and its interaction with the core?

ET should detect NS ellipticities that are few 10-10 or larger depending on their spin frequency. This can be used to deduce the property of NS crusts. ET might also detect GWs that are expected to be emitted when pulsars glitch and magnetars are and thereby help understand crust-core interaction that is believed to transfer angular momentum from the core to crust.

What is the population of GW sources at high redshifts?

A large population of point sources would produce a confusion background that would be detectable by ET if the energy density of the background is large enough. Detection of confusion backgrounds can be used to understand the nature and population of GW sources in the Universe.

Drafting the observatory design

The sensitivity of GW detectors improved considerably from the bar detectors to the first generation of interferometric detectors, which are currently being upgraded to the advanced generation.

In order to achieve the scientific goals, the sensitivity in comparison to the second generation of GW detectors must be improved by about an order of magnitude over the entire detection frequency band ranging from 10 Hz to about 10 kHz. Frequent observation of low-frequency sources, e.g. intermediate mass BHs, requires an extension of the detection range towards lower frequencies.

The initial sensitivity goal for the ET, estimated at the start of the design study, was driven by the need to get frequent high SNR events for routine GW astronomy. The high-frequency sensitivity was given by the maximum power feasible, the mid-frequency range was governed by thermal noises and the low-frequency range by either thermal or seismic noises. The initial estimates have been refined during the design study.

Size, shape and layout

The conceptual design of a so large project has to be based on well proven and experimentally tested techniques. To achieve the sensitivity that the ET project aims for, on the other hand, it will be necessary to exploit all state-of-the-art technologies and drive them to their physical limits.

This sensitivity can only be reached by significantly increasing the size of the detector beyond the size of currently available instruments (i.e. 3 km for Virgo and 4 km for LIGO) and going to an underground location, where the seismic noise is lower than at the surface. Only by increasing the arm length to 10 km can the influence of unavoidable displacement noises be lowered to a tolerable level. In its final construction stage the ET will consist of three nested detectors, which will be arranged in a triangular pattern. In contrast to the traditional L-shaped geometry of the first and second generations of GW detectors this arrangement is equally sensitive for both polarisations of the GW. Additionally it shows a more isotropic antenna pattern compared to the L-shaped detectors. The overall frequency range covered will reach from a few Hertz to about 10 kHz. Each individual detector in turn will comprise two interferometers, one specialised for detecting low-frequency GWs and the other one for the high-frequency part. Each individual interferometer has a classical dual-recycled Michelson topology with arm cavities. This is a mature technique, well tested in laboratory experiments, and currently being set up for the second-generation detectors, Advanced LIGO and Advanced Virgo. More elaborate topologies like Sagnac interferometers or optical bars using Quantum non-demolition (QND) techniques do not promise significant advantages and have not yet reached the level of maturity required for a project of this scale.

Quantum noise

In order to achieve a sufficient sensitivity at high frequencies the light power in the arms of the high-frequency interferometer needs to be increased to the megawatt range. Thermal noise considerations on the other hand require cryogenic optics to reach the sensitivity goal at low frequencies.

Operating cryogenic optics at a level of several megawatt of light power presents a serious technological challenge which is extremely hard to master. Even for the best mirrors that state-of-the-art coating technology can produce, the residual absorption of only about one ppm leads to an absorbed power of several Watt at a circulating light power level in the megawatt range. The resulting thickness of the suspension fibres, which would be needed to remove the heat, would spoil the performance of the ultra-low loss suspension. The ET will therefore be realised in what we call a xylophone arrangement, where each single detector is split into two interferometers. The one dedicated for detecting high-frequency GWs in the range from about 30 Hz to 10 kHz will be operated at room temperature, use fused silica optics with a diameter of about 60 cm and a mass of about 200 kg each, have a light power of about 3 MW in the interferometer arms, and run with broadband tuned signal recycling. The cryogenic, low-frequency one, operated at a temperature of 10 K and aimed at the frequency range from 1.5 Hz to 30 Hz, will use detuned signal recycling, have a light power of 18 kW in the interferometer arms, and silicon mirrors with a diameter of about 40 cm and a mass of about 200 kg. The cryogenic optics will either be made of sapphire or, more likely, of silicon. The dimensions will partly be determined by the maximum available bulk material size and otherwise be comparable to the room temperature ones. A summary of the main parameters for the high and low temperature interferometers is given. The high mirror mass will not only keep radiation pressure effects low but also allow larger sized beam spots on the mirror surfaces for lowering thermal noise effects. This split detector arrangement also offers an elegant solution for the challenge that radiation pressure noise and shot noise scale in opposite ways with light power and cannot individually be optimised in a single interferometer. In an interferometer using classical states of light the so-called Standard quantum limit (SQL) determines the lower limit for the quantum noise. For each frequency there is a different optimal compromise between shot noise and radiation pressure noise, meaning that in a single interferometer the SQL cannot be reached for all frequencies simultaneously. It can only be overcome if non-classical light with correlations between the phase and the amplitude quadratures is used, so-called squeezed light. In the shot noise dominated frequency range squeezed light is used, which shows lower phase fluctuations at the cost of the amplitude fluctuations in comparison to classical laser light in the interferometer arms. In the low-frequency, radiation pressure dominated range the fluctuations need to be lowered in the amplitude quadrature. In the ET design study it has been investigated how to achieve this goal and a solution has been proposed by reflecting squeezed light off a filter cavity. The usage of squeezed light is currently tested in the existing GW detectors and is foreseen as an add-on to the second generation. Squeezing levels over the full observation band width of up to 10dB, and stable long-term operation and best squeezing values of almost 13 have been demonstrated. For the ET we assume 15 dB initial squeezing level at the squeezing source and an effective squeezing level of 10 dB to be available (equivalent in shot noise reduction to a laser power increase of a factor of 10). The squeezing level, and with it the sensitivity improvement that can be reached, depends on the optical losses in the squeezer, the filtering optics, the interferometer, and all optical devices on the way to the photodetector, including the photodetector efficiency itself. Optical losses easily add vacuum fluctuations to the squeezed quadrature and hence reconvert squeezed light into classical light. It will therefore be essential to keep the optical losses as low as possible. Optical losses of 75 ppm per round trip are currently achievable with state-of-the-art techniques and are used as a conservative estimate for the filter cavities.

Thermal noise

Reaching the sensitivity goal at low frequencies requires a significant reduction of thermal noises compared to the first and second generations of gravitational-wave detectors, which can be achieved by operating the mirrors at cryogenic temperatures as low as 10 K.

Cryogenic operation is also foreseen for the final stage of the planned Japanese GW detector LCGT. At these low temperatures fused silica has a low mechanical quality factor and becomes unusable. Silicon and sapphire show excellent low-temperature and are good candidates for cryogenic GW detectors. Its availability in large quantities and good purity through the semiconductor industry makes silicon a promising candidate for the ET cryogenic optics. Some quantities such as the temperature dependence of the refractive index at low temperatures and the residual optical absorption in ultra-pure silicon, although assumed to be good enough for use in ET, are currently not known and need to be investigated in Research and development (R&D) activities. Removing the heat generated by laser light being absorbed at the mirror surfaces without introducing excess vibration levels poses another technical challenge. As thermal radiation does not provide sufficient coupling at cryogenic temperatures this heat removal has to be done by thermal conduction of the suspension fibres. The resulting requirement for the thickness of the silicon suspension fibres needs to be balanced against good seismic isolation properties of thin fibres. The vibration level of cryo coolers, which could be placed close to the optics, threatens the low-frequency. R&D in active and passive vibration suppression is still required to sufficiently cut the remaining noise level down for use in the ET. Cryo fluids, like superfluid Helium II, which are cooled down at a remote location above ground, are available as a seismically quieter alternative. The final operating temperature for ET remains to be fixed in a technical design phase. The cooling capabilities foreseen so far will allow mirror temperatures as low as 10 K.

Seismic Isolation

The ET main optics need to be very well isolated against seismic ground motion. For the second generation of GW detectors both active and passive isolation strategies are being pursued. In the advanced LIGO detectors a two stage system actively isolates a platform from ground motion, from which the main optics are suspended by quadruple pendulums. The passive strategy employed at the Virgo detector demonstrated an excellent performance over the full frequency range and is foreseen as the reference solution for the ET. The horizontal isolation is achieved with a six-stage pendulum system, whereas for the vertical degree of freedom cantilever springs are used. The pendulum suspension system itself is supported by a platform resting on an inverted pendulum, which provides additional horizontal seismic isolation. All mechanical resonances of the whole structure are actively damped to avoid resonant mechanical amplification of ground motion. The overall height of the suspension system is about 17 m, requiring correspondingly tall vacuum chambers and caverns.

Newtonian gravity gradient noise

Newtonian gravitational interactions between the optics and the surrounding soil provide a direct coupling mechanism of ground motion to the interferometer test masses. As the resulting, so-called gravity gradient noise cannot be shielded from the mirrors, a location has to be found where this seismic motion is minimal and the surrounding soil as homogeneous as possible. This goal can be achieved in an underground location in a seismically quiet region. Preliminary measurements show that a depth of 100 to 200 m in a remote location with low population density provides sufficiently low seismic motion. The potential of measuring the ambient seismic motion, feeding it into a gravity gradient noise model, and then subtracting the predicted effect from the interferometer output signal has been investigated. Initial results are promising, and are interesting also for the second generation of GW detectors, but investigations need to be continued in an R&D programme.

Vacuum

The space between the mirrors in the interferometer arms has to be evacuated to very low residual partial gas pressures to keep the apparent length changes caused by fluctuations of the refractive index sufficiently low. The tolerable maximum pressure is on the order of 10-8 Pa

Some technical parameters will remain to be defined in this design study document. At this stage of the conceptual design these parameters are not important and will be worked out in a future technical design phase.

Noise budget

The xylophone strategy, i.e. the division of each detector into a low-frequency and a high-frequency interferometer, allows to pursue different strategies in optimising the noise for each frequency range. The noise budget for the high-frequency interferometer is shown. In the frequency range from about 7 to 30 Hz the sensitivity is limited by suspension thermal noise, resulting from the interferometer being operated at room temperature. At frequencies above 500 Hz the dominating noise source is photon shot noise. Between these two frequency ranges mirror thermal noise is limiting the overall sensitivity. In the noise budget for the low-frequency interferometer, quantum noise is limiting the sensitivity over the entire frequency range above 7 Hz. Due to the operation at cryogenic temperatures the influence of suspension thermal noise in the frequency range above 7 Hz can be cut down to a sufficiently low level. Below 7 Hz the sensitivity is limited by comparable amounts of quantum noise, gravity gradient noise, and suspension thermal noise. Due to the good performance of the multistage pendulum suspensions the influence of seismic noise can be limited to the frequency range below 2 Hz. Investigations of new QND techniques in a planned R&D programme will show whether it is possible to cut down the quantum noise contributions to an even lower level in the frequency range from 7 to 30 Hz.

Layout of the observatory

As a consequence of the extremely demanding seismic requirements, ET will be located underground at a depth of about 100 to 200 m and will, in the complete configuration, consist of three nested detectors, each in turn composed of two interferometers. Selecting the geometry of an equilateral triangle, where each side of the triangle is simultaneously used by two detector arms, allows to determine the polarisation of the GW and optimises the usage of the tunnels. The topology of each interferometer will be the dual-recycled Michelson layout with Fabry-Perot arm cavities. An artist's impression of ET is shown. Underground seismic measurements at eight different European locations have been performed within this design study and additional measurements from external sources have also been evaluated. Satisfactory seismic performance has been found in several locations. For the final site selection long-term seismic noise measurements including seasonal effects like variable wind speeds and ocean wave height need to be made, and other non-scientific factors of influence (e.g. political, financial, interest of local parties, vicinity to research institutions) have to be included in the decision process. The sensitivity curve gives the sensitivity for a single detector with 10 km arm length and an angle of 90 degrees between the arms. This is done for better comparison with the existing detectors and their advanced versions. ET will in fact have three 10 km detectors and the angles between the arms will be 60 degrees. The resulting sensitivity in comparison with a single 90 degree detector depends on the source location in the sky and its orientation, as the angular antenna pattern and the polarisation dependence (independent in the triangular case) influence the signal strength differently for different detector layouts. On average the sensitivity of the triple 60 degree detector is slightly better than a single, optimally oriented 90 degree one. For the desired sensitivity an overall side length of the triangle of about 10 km is required. More specifically in the ET design document we assume 10 km for the arm cavity length and an additional 300 m of tunnel length for telescopes for mode matching the large beam size from the interferometer arms to smaller beams in the beam splitter area. This gives a total triangle side length of 10.3 km and an overall tunnel length for the ET observatory of 30.9 km. This length of 300 m from the vertex station to the satellite station is also used for the filter cavities for the high-frequency interferometer housed in a separate tunnel, which simultaneously serves as a safety escape route from the satellite caverns. The main circa 10 km tunnels that connect two satellite stations will have an inner diameter of 5.5 m, which will locally be increased to 6.0 m wherever the insulation for cryogenic operation requires more space. The three vertex caverns and six satellite caverns will house the vacuum vessels and must be about 25 m high. Access to the underground detectors is foreseen via vertical shafts. It remains to be explored in a technical design phase after site selection whether horizontal access is favourable. This option may, for instance, be advantageous if the observatory is built inside a mountain. The ET infrastructure will house the observatory for many decades, during which the interferometers will receive upgrades as technology advances. Some of these changes may result in the necessity to change mirror positions and with it vacuum tank positions as we now see in the upgrades from the initial to the advanced generation of GW detectors. Hence we plan to build large caverns where the tanks can be placed at arbitrary positions, instead of building an inflexible system of short tunnels connecting narrow but tall caverns housing a single vacuum tank each. Above ground, at the entrance to the vertical axis shafts, facilities housing workshops, offices, apartments, technical facilities providing cryogenic fluids, air conditioning and venting, emergency electricity generators will be set up. One major aspect of the design of the total infrastructure is to provide an environment able to house not only the basic initial version of the ET that we describe in the design study document but also be versatile enough to accommodate upgraded versions in the following decades to come.

ET timeline

The evolution toward the ET observatory has been, is and will be a long path. The idea of a third generation GW detector born within the ILIAS project (see http://www-ilias.cea.fr/ online, 2004--2008), supported by the European Commission within FP6. In fact in that 'Integration activity', both a Joint research activity (JRA3 or STREGA), supporting R&D activities addressed to the thermal noise reduction, and a networking activity (N5-WP3), addressed to the studies for the third generation of GW detectors, have been implemented. In 2005, the European Science Foundation (ESF) supported an exploratory workshop (http://esf-gw.pg.infn.it/) specifically devoted to an European third generation GW observatory. This has been the right kick-off meeting to prepare the organisation and clarify the targets of the successful proposal that bring to the funding, by the European Commission, under FP7, of the ET conceptual design study, object of this document. The main target of the ET conceptual design is the demonstration of the feasibility of the ET project, through the definition of the requirements and main characteristics of the hosting site, the design of the key components of the research infrastructure, the indication of the possible technologies and the presentation of the detector design main elements. A rough cost evaluation is then reported to evaluate the financial feasibility of the ET project. After the conceptual design phase, a series of difficult steps are expected. The first step is the consolidation of the design solving the several question marks currently present in many technologies needed for ET. In fact the detectors in the ET observatory will adopt technologies that aren't explored in the advanced detectors, like cryogenics, silicon mirrors, different wavelength lasers, optical squeezing, gravity gradient noise subtraction techniques. The ET design consolidation needs, then, an intense R&D activity devoted to these technologies. The ET community must find the (human and financial) resources to support these activities; these resources should be a mix of national and international funds. In parallel to the progresses of the technologies, the conceptual design needs to evolve in a technical design, describing in detail the components of the observatory. This phase of the project will need a correct framework and, because of the fact that the national funding agencies in Europe are fully engaged with the realisation of the advanced detectors, a possible choice is a networking or integration tool at European level (probably in FP8). To support this phase and to give a future to the ET project is, then, strongly recommended to be inserted in the list of the major research infrastructures (the so-called ESFRI roadmap) recommended by the European strategy forum on research infrastructures (ESFRI). Already part of the activities, necessary to achieve this target, has been performed inserting ET in the specialised roadmaps, like the GWIC roadmap (see http://gwic.ligo.org/roadmap/ online) and the ASPERA roadmap (see http://www.aspera-eu.org/images/stories/files/Roadmap.pdf online), but intense outreaching and proposition actions are still needed. In Figure 17 the worldwide roadmap for the GW detectors, produced by GWIC, is shown; although it is a bit obsolete, being produced before the outcomes of this design study, it is useful to show the overall scenario where the ET project will evolve. The expected steps of the ET project are shown, as elaborated within this conceptual design. Hence, after the current conceptual design, as previously mentioned, a technical design phase, supported by an intense R&D activity, is necessary. Then the ET project will need to define the best site location and to start a funding search activity. An a priori condition for the funding of the construction phase is the detection of a GW signal in the advanced detectors. To base the ET construction funding and start-up on a so intrinsically unpredictable event seems rather hazardous, but considering the installation and commissioning schedule of the advanced detectors and the promised sensitivity evolution, it is possible to predict a realistic time window for the GW signal from a BNS system in the 2015-2017 period. It will follow a site preparation phase, where all the legal and preliminary aspects of the land ownership acquirement are exploited. In parallel the production and test of the first detector hardware components will start in the laboratories participating to the ET project. The site and infrastructures realisation will last for about four years, followed by the first ET detector installation. To save time, it is expected that part of the first detector installation (i.e. vacuum pipes deployment) could overlap with the latest infrastructure realisation activities. Thanks to the experience acquired with the initial GW detectors (that will be improved with the next commissioning of the advanced interferometers) it is possible to state that the commissioning phase will last at least for more than 3-4 years, with some early science data taking interlaced with the commissioning periods. The ET observatory infrastructures and detectors are designed having as requirement the modularity of the components; this will allow sequential installation phases, interlaced with periods of data taking for the detectors, already operative in the ET site. In this way it will be also possible to upgrade the installed detectors when the technological progresses will make it convenient, maximising the duty cycle. This possibility underlines the main target of this study: to provide the design of a research infrastructure operative for decades and will be able to host evolving detectors.

Potential impact:

Scientific impact: Fundamental physics

Astronomical sources of GWs are essentially systems where gravity is extremely strong, and are often characterised by relativistic bulk motion of massive objects. The emitted radiation carries an uncorrupted signature of the nature of the space-time geometry, and is thus an invaluable tool to understand the behaviour of matter and geometry in extreme conditions of density, temperature, magnetic fields and relativistic motion. Here are some examples of how GW observations can impact fundamental physics. In Einstein's theory, gravitational radiation travels at the speed of light and has two polarisation states. In alternative theories of gravity one or both of these properties might not hold, owing to the presence of massive gravitons, or a scalar field mediating gravity in addition to the tensor field. Current experimental tests of gravity, as well those afforded by the data from the Hulse-Taylor binary, are consistent with both Einstein's theory and one of its alternatives called the Brans-Dicke theory. GW detectors will bring these theories vis-a-vis observations that could decisively rule out one or the other. According to Einstein's gravity the space-time in the vicinity of BHs is described by a unique geometry called the Kerr solution. Observation of the radiation from the infall of stellar-mass BHs into intermediate mass BHs will make it possible to test such uniqueness theorems. X-ray astronomy has provided from indirect evidence that intense sources of X-rays may well host a BH. An unambiguous signature of the BH geometry, however, could eventually be provided by the detection of BH quasi-normal modes: gravitational radiation with characteristic frequencies and decay times that depend only on the mass and spin angular momentum of the BH. Failure to detect such radiation from, for example, a newly formed BH would mean that gravity is more exotic than we currently believe, and might reveal new phases of matter at extremely high densities. The most attractive versions of string theory require a ten or eleven-dimensional spacetime, far above those that we perceive. In certain phenomenological models at the interface of string theory and cosmology, what we perceive as a four-dimensional Universe could be one part, or brane, within a higher dimensional bulk Universe. The extra spatial dimensions may be compact and sub-millimetre scale, or even macroscopically large, if their geometry has properties, known as warping. The key feature of brane-world theories is that gravitational interactions, and in particular GWs, propagate in the bulk, while other interactions are restricted to the brane, which partly explains why gravity is so weak. The ET offers the exciting possibility of observing radiation from the bulk and thereby exploring whether the Universe has large extra dimensions.

Scientific impact:

Relativistic astrophysics

Astronomy has revealed a Universe full of diverse and exotic phenomena that remain enigmas decades after their discovery. Supernovae are the end-states of stellar evolution, resulting in gravitational collapse followed by a huge explosion of the formerly infalling matter. Gamma-ray bursts are intense sources of gamma radiation that last only a few seconds to minutes yet emit more energy than a star does in its entire lifetime. Radio pulsars are compact objects as massive as the Sun but only about 10 km in size, and the regularity of their radio pulses and occasional glitches in that regularity have remained puzzles for a long time after their discovery three decades ago. Transient radio sources thousands of light years away are associated with magnetic fields so strong that the emitted radiation could break down terrestrial radio stations. The source in question in each case is believed to be couched in dense environs and strong gravitational fields and, therefore, a potential source of gravitational radiation. For example, gamma-ray bursts could be produced by colliding NSs which are electromagnetically invisible for most of their lives but are very powerful emitters of GWs. Transient radio sources could be the result of quakes in NS with concomitant emission of GWs. Observing such `multi-messengers' (sources that are strong emitters of both EM and GW radiation) will help understand phenomena that have remained puzzles for decades. The centre of every galaxy is believed to host a compact object a million to a billion times as massive as our Sun, a powerful emitter of optical, radio and other radiation. What is the nature of this object? How and when did it form? Did it form from small 100 solar-mass seeds and then grow by accreting gas and other compact objects? What is its relation to the size of the galaxy as a whole? These are some of the questions which a model of the formation of structure in the Universe must answer. While electromagnetic observations have provided valuable data, ET can explore the population of stellar mass and intermediate mass BHs as a function of redshift and shed light on BH demographics, their mass distribution and growth. ET will also be sensitive to a population of sources at very high redshifts, helping us study cosmological evolution of sources, the history of star formation and its dependence on the matter content of the Universe, and the development of large-scale structure in the Universe.

Scientific impact:

Cosmology

The twentieth century was the golden age of cosmology. With the advent of radio and microwave astronomy it was possible to finally address key questions about the origin of the Universe. The Cosmic microwave background (CMB) is a relic radiation from the hot Big Bang that is believed to have been the initial condition for primordial nucleosynthesis. Since the early Universe was very dense, this radiation was in thermal equilibrium with matter for about 380 000 years after the Big Bang and cannot directly reveal the conditions in the very early phases of the Universe's history. The most direct way of observing the primeval Universe is via the gravitational window with a network of two or more detectors. From fairly general assumptions one can predict the production of a stochastic background of GWs in the early Universe, which travel to us unscathed as a consequence of their weak coupling to matter. The most amazing aspect of the Universe is that only about 4 % of its energy density appears in the form of visible matter, the rest being dark matter and dark energy. In order to understand the behaviour of these 'dark' contents it is necessary to have a standard candle - a population of sources whose distance from Earth can be inferred from their luminosity. Compact binaries are an astronomer's ideal standard candle. By measuring the signature of the gravitational radiation they emit, it is possible to infer their intrinsic parameters (e.g. the masses and spins of the component objects) and accurately deduce their luminosity distance. In fact, compact binaries eliminate the need to build a cosmic distance ladder-the process by which standard candles at different distances are calibrated in astronomy since there is no source that is suitable at all distances. The synergy of multi-messenger astronomy is nowhere more apparent than in the use of standard sirens of gravity to test the concordance model of cosmology. ET might detect several hundred compact binary coalescence events each year in coincidence with short-hard gamma-ray bursts, provided, of course, the two are related. While gravitational observations would provide an unambiguous measure of the luminosity distance, the host galaxy of the GRB could be used to measure the redshift. By fitting the observed population to a cosmological model, it will be possible to measure the Hubble parameter, dark matter and dark energy densities, as well as the dark energy equation-of-state parameter. The early history of the Universe may have witnessed several phase transitions as the temperature decreased through the energy scales of a Grand Unified Theory (GUT) and electroweak symmetry-breaking, and eventually to the current state in which we see four different fundamental interactions. During some phase transitions, cosmic strings form as one- dimensional topological defects at the boundaries of different phases. Vibrations of these strings at the speed of light can sometimes form a kink that emits a burst of gravitational radiation. The spectrum of such radiation has a unique signature that can help us detect cosmic strings and measure their properties, and thus provide a glimpse of the Universe as it underwent phase transitions. Perhaps the most exciting discovery of the new window will be none of the above. If astronomical history is any example, gravitational astronomy should unveil phenomena and sources never imagined in the wildest theories--a possibility of any new observational tool.

Socioeconomic impact

The European scientists involved in the experimental search of GWs have been divided in two separated collaborations, working into the GEO600 detector and into the Virgo experiment. These two communities had obviously contacts and exchanges in the past, but they never be focused in a single common target. Furthermore, the astrophysicists and the numerical relativity experts had, in the past, a marginal role in the definition of the strategies of the two main detectors in Europe. The ET project completely changed the scenario, bringing together, for the first time in Europe, all the main actors in the experimental search for GW (the Virgo and GEO600 collaborations) and deeply involving in the definition of the ET science case the most brilliant astrophysicists in Europe and the major European teams working in numerical relativity simulation (an important step toward the European Research Area, ERA). All these components formed the ET science team (see http://www.et-gw.eu/science-team online), more than 220 scientists from all Europe (and worldwide) collaborating to the ET design study. This has been possible thanks to the open structure of the ET design study project; since the beginning, the ET design study has been considered open to all the scientists interested to the science behind ET. In cascade, the creation of the ET science team and the successful progress of the ET design study affected the entire field of the GW research. R&D activities have been focused toward subjects of interest of ET and new collaborations and synergies between European (and worldwide) research groups have been structured. Technological aspects of the ET design requires the development of new technologies of high societal impact, like optics, coatings, lasers, cryogenics, precision mechanics; the larger synergy between the groups that, are working, until now independently, on these items will speed-up the progresses. For example, a new multi-annual program of collaboration has been set-up between the European groups and the Japanese team working respectively on the cryogenics for ET and LCGT.

Dissemination activities

The ET project is far more than a simple conceptual design study of a new research infrastructure; it heralds a new idea, 'the 3rd generation of GW observatories' proposing to the scientific community to open a new observational window of the Universe: the GW astronomy.

For this reason the outreaching and dissemination activity has been of primary importance and several have been the targets of the dissemination activity:

1. the GW community;
2. neighbouring scientific communities;
3. institutions and funding agencies;
4. students.

To act on the GW community, ET science and technology has been proposed in the major events (conferences and workshops). At the GWADW 2009 (GW Advanced Detectors Workshop, Miami, USA, 2009) only two talks have been as subject ET, but one year later,, a full session devoted to ET has been organised in occasion of the GWADW 2010 (Kyoto, Japan, May 2010), a plenary talk and a full session during the 8th Amaldi conference (the major conference of the GW community, New York, US, 2010), a full session and many technological talks scattered in the whole workshop at the GWADW 2011 (Isola d’Elba, 2011). Several talks have been presented during the LIGO-Virgo meetings and other restricted workshops (i.e. GW workshop, Minneapolis, 2010). ET talks were everywhere in the programme of the 9th Amaldi conference, Cardiff, UK, 2011. To transmit the ET idea to the neighboring scientific communities (Astroparticle, General / Numerical relativity, Astrophysics, Underground laboratories) ET talks have been presented during the ASPERA events (and ET entered in the ASPERA roadmap as one of the Magnificent seven), at the Marcel-Grossmann meetings, at the ASTRONET meetings, during the TAUP conferences. To stimulate the interest of the funding agencies and the research institutes actions have been done on ASPERA, on the national committees; ET has presented during the national events, like the Hungarian Academy of Sciences annual meeting (May 2010), In particular, the 20th of May 2011 ET organised at EGO the 'ET design study document presentation', inviting representative of the beneficiaries, of the ESFRI committee and of the governments (see http://www.et-gw.eu/design-study-presentation online). This event had a worldwide echo, stimulating articles in several newspapers in the world (a list of about 70 links is available here: entries >11). Finally, the ET design study document has been publically released and it is available in the ET codifier: see reference (1) and https://tds.ego-gw.it/ql/?c=7954

Students have been trained on the ET science and technologies in several occasions; for example ET has been presented during the annual Virgo-EGO scientific forum (VESF) international school in 2010 and 2011 and ET collaborated with ISAPP (International School on Astroparticle Physics) to organise at EGO the annual (2010) school. Furthermore, a special edition of the GR and Gravitation (publisher Springer Netherlands, see http://www.springer.com/physics/journal/10714 online) has been published by Springer Netherlands, containing a series of articles, peer reviewed, written by the scientists of the ET science team, fully dedicated to ET and its possible science reaches (Journal General Relativity and Gravitation, ISSN 0001-7701 (Print) 1572-9532 (Online), (e.g.) DOI 10.1007/s10714-010-1010-8 Subject Collection Physics and Astronomy, SpringerLink Date Tuesday, May 25, 2010).

ET stimulates a series of scientific articles devoted both to the technical aspects and to the scientific implications; an updated list is linked in the ET web site and stored here: http://virgo-bwulf.pg.infn.it/~punturo/ETtriggeredPapers.html. Finally, the ET design study will be published in a special issue of Classical and Quantum Gravity and the wider societal implications of the project so far and the main dissemination activities and exploitation of results. ET is in the conceptual design phase and then, obviously, the contact with industries is not a key action.

Nevertheless a number of meetings with industries have been organised mainly to discuss the evolution of interesting technologies:

- Associazione Italiana di Scienza e Tecnologia, XIX congresso AIV (Senigallia 19-21 Maggio 2009), Vacuum technologies.
- Triangular meeting at Nishitokyo-city, Tokyo, among Sumitomo Heavy Industry / Cryogenics group LCGT project ( KEK - High Energy accelerator organization ) ET - project (INFN Sezione di Roma).
- ASPERA R&D Meeting (Amsterdam, 2008).
- Excavation Companies meeting (Amsterdam, 2009).

In October 2011, a special ASPERA meeting focused on oprics for GW detectors and ET will be organised at the EGO site: http://www.et-gw.eu/events/asperatf

Project website: http://www.et-gw.eu