Star Track tracking of astrophysical sources with neutrinos

Prof. dr. Maarten de Jong Nikhef – National institute for subatomic physics Amsterdam, the Netherlands www.nikhef.nl Project duration: 60 months

abstract Although a visit may seem the ideal way to study astrophysical sources in detail, the required travel time makes this idea impractical. The alterna- tive is to boldly view the sources with Earth-based telescopes. The prime objective of this proposal is the scientific capitalization of a new generation of telescopes. Unlike conventional telescopes, these telescopes will detect neutrinos and not light. The detection of neutrinos from the cosmos will break new grounds in the study of various frontier questions in science such as those related to the origin of cosmic rays, the mechanism of astro- physical particle acceleration and the birth of relativistic jets.

Section 1A: The Principal Investigator

Maarten de Jong CERN, Chorus experiment 1993–1997 Minervalaan 8" 1077 NX Amsterdam, the Netherlands Fell o w (CERN): Conducting the design, construc- Single, 43 years old, Dutch nationality tion and operation of a new detector. Proposed an e-mail: [email protected] upgrade of the detector to improve the charge and tel: +31 20 592.2121 momentum determination of hadrons. Approval of the project: nominated project leader (2 PhD students Work Experience and 5 technicians). Coordinated the design and the manufacture of three so-called honeycomb wire- Nikhef 1997–present chambers at Nikhef. Designed and implemented the necessary infrastructure (gas, cooling, low-voltage, EU Wo r k Pa c k a g e Co o r d i n a t o r : Responsible for high-voltage, slow control, DAQ, alignment). In the coordination of the information technology charge of the testing and final assembly and installa- aspects of the KM3NeT design study. Contact person tion. The new detectors have been successfully used. of the proposal “KM3NeT: The next generation neutrino telescope” submitted to the Dutch fund- Re s e a r c h As s o c i a t e (Nikhef): Participating in the ing agency NWO with a total investment budget of construction, calibration (~1 ns precision) and main- 8.8 M€. This proposal in under evaluation by NWO. tenance of the trigger system. Improved the neutrino oscillation trigger such that 10% more physics events Pr o f e s s o r : Appointed as the first professor in the in target were detected. Netherlands (Leiden University) in the field of experimental astro-particle physics. Lecturing one CERN, NMC & SMC experiments 1991–1993 semester per year. Re s e a r c h As s o c i a t e (University of Mainz): An t a r e s Depu t y Sp o k e s pe r s o n : Responsible for the Supervising an international group of 5 PhD stu- physics coordination of the Antares project and the dents. Established the key point in the data analysis dissemination of scientific results in refereed journals and developed a Monte Carlo simulation program and at conferences. Organising 2 workshops per year with enhanced statistical significance (up to a factor (about 50 participants) and 3 collaboration meetings of 10). Determined the radius of the J/ψ meson by a per year (about 100 participants). novel comparison of the cross section of J/ψ produc- tion to that of elastic π scattering. Te a m Le a d e r : Responsible for the coordination of the Dutch participation in the Antares project (5 PhD Re s e a r c h As s o c i a t e (Nikhef): Preparing the detector students, 1 post-doc, 4 senior physicists and 6 techni- and coordinating the startup of data taking. Adapted the cians), the project planning and management and inclusive muon trigger to the increased beam intensity. the spending of the investment budget. Supervising PhD students (subjects: Monopole, neutralino and Scientific Committees gamma-ray burst detection). Recruiting new people and building a team. Developed a new data filter pro- Peer Review Committee of the ApPEC 2002–2005. gram for the Antares neutrino telescope (2.5 × faster, CERN SPS-Committee 2001–2004. 100 × better signal-to-noise ratio). Scient. Adv. Committee of Nikhef 2003–present.

Se n i o r St a f f : Contact person of the proposal Outreach “Antares: a cosmic neutrino observatory”, sub- mitted to the Dutch funding agency NWO with a Contributing to public relations. total investment budget of 3.6 M€. Invented a new Citations in seven national newspaper articles. readout system. Nominated readout project leader One appearance on local TV. for the Antares experiment. Published a study of One appearance in a theatre play (AdHoc). neutrino physics at a future muon collider. Organised Giving seminars to the general public (1–2 per year). the Antares collaboration meeting at Nikhef (about 100 participants). Contributed to the Topical Education Lectures 1998 and 2001 at Nikhef, the European Graduate School in 2002 and the Fantom study week NIMO, Project Management, 2000, 6 days training. in 2003. Organised the Chorus collaboration meeting Vrije Universiteit, Amsterdam, 1986–1991, PhD. at Nikhef (about 100 participants). Supervising PhD CERN summer student, 1986. students (co-promotor). Developed new track fit Vrije Universiteit, Amsterdam, 1982–1986, Cum Laude. algorithms for the Chorus experiment which yielded Sweelinck Gymnasium, Amsterdam, (1976–1982). 20% more physics events found in target. 1 Scientific leadership profile

Having performed various particle physics experi- For the funding of scientific projects I have (co-) ments at CERN*, I have now focussed my career on written various proposals which have provided so the study of cosmic neutrinos. The study of cosmic far a total amount of about 3.6 M€ for (hardware) neutrinos is one of the key components in the field investments and about 4.8 M€ for personnel. In of astro-particle physics. This is a relative new addition, I am contact person of a proposal submit- field emerging at the interface between astrophys- ted to the Dutch funding agency NWO with a total ics and accelerator-based particle physics. In 2006, investment budget of 8.8 M€. At the moment of this I have been appointed as the first professor in the writing, this proposal is still under evaluation. Netherlands in the field of experimental astro-par- ticle physics. The election as deputy spokesperson I initiated the measurements and supervised the of the Antares¶ project and the nomination of work- analysis that has led to the publication “Measurement package coordinator of the KM3NeT† design study of nucleon structure functions in neutrino scattering”. speak for my leadership in this field. At that time, there were only two measurements made worldwide which yielded contradicting results. This My ideas can be considered as creative and effective. paper has been seminal in two ways. It has resolved Most recently, I have invented the ‘All-data-to-shore’ the long standing discrepancy and it is today the one concept for the readout of the Antares prototype and only measurement of all three structure functions neutrino telescope. In this, the rare neutrino signal is of the nucleon known to exist. filtered on shore from the background using a farm of commodity PCs and state-of-the-art software. I have been supervising master and PhD students The software has been realised by me and is now throughout my scientific career. About half of these fully operational in the Antares experiment. This students continued a career in science, the others achievement is remarkable (few considered this have found a job in industry relatively fast. One of possible) and changed the picture dramatically. Now, my students originally applied to a permanent post in my idea has become a seminal part of the astrophys- the computing department of Nikhef. I was member ics program with the next generation neutrino of the search committee and I persuaded this person telescope. Some of my ideas have led to new results to become a PhD student in my group. After a (e.g. measurement of the size of a very short lived successful PhD, this person has won two prestigious particle), others have disproved an initial claim for a awards in the Netherlands. new theory (e.g. colour transparency). In the following, I list a few quotes from a selection Both in particle physics and astro-particle physics, a of PhD theses of students whom I have supervised: key issue is the reconstruction of particle trajectories in a detector. For the Chorus experiment at CERN, “Maarten de Jong took a key role in all stages of I have developed a new track-fit algorithm. This has this work. He convinced the collaboration that a DIS

improved the final measurement of µν →ντ oscilla- program in 1998 was possible and thus obligatory. tions by such an amount that the most precise result In the analysis, he brought me back on track when I ever could be achieved. For the Antares experiment, was once again in panic and desperation. This text I have found a linear solution of the track-fit problem would have been unreadable without Maarten’s ex- by considering a subset of the phase space. Although tensive commenting.” R.G.C. Oldeman, PhD thesis, its impact has not been quantified yet, preliminary University of Amsterdam (2000). studies have shown that it can improve the angular resolution of a neutrino telescope significantly. “Maarten, to you I owe most thanks. I could not have wished a better supervisor! Your enthusiasm and I have coordinated the construction and opera- inexhaustible drive have continuously motivated me. tion of a new kind of particle detector (so-called Without your knowledge and bright ideas, I could ‘Honeycomb chamber’) for the Chorus experiment. never have written this thesis.” B.A.P van Rens, PhD This marked the first successful application of this thesis, University of Amsterdam (2006). kind of detector in a large scale experiment. For the Antares experiment I have coordinated the design “Maarten was so kind to become my thesis advisor efforts of the submarine power system. These efforts and read the whole manuscript very carefully.” have led to a system consisting of DC-DC convert- J. Uiterwijk, PhD thesis, University of Leiden (2007). ers with an unprecedented efficiency. By now, the system is operational for more than two years * www.cern.ch indicating that the high level of reliability required ¶ antares.in2p3.fr for submarine operations has been achieved. † www.km3net.org 2 Short list of selected publications Contributions to advanced international schools

th Final results from a search for νµ→ντ oscillations The 19 general FANTOM study week with the CHORUS experiment “Interplay between Theory and Experiment” Nucl. Phys. B 793, 326-343, 2008. 2003, Amsterdam, the Netherlands citations: not yet applicable. European Granduate School The data acquisition system for the ANTARES “Neutrino: masses, mixing and oscillations” Neutrino Telescope 2002, Giessen, Germany Nucl. Instrum. Meth. A 570, 107-116, 2007. citations: 5 Topical lectures “The Antares neutrino telescope” Measurement of nucleon structure functions in 2001, Nikhef, Amsterdam, the Netherlands. neutrino scattering Phys. Lett. B 632, 65-75, 2006. Topical lectures citations: 18 “Neutrino physics” 1998, Nikhef, Amsterdam, the Netherlands. Measurement of the Z/A dependence of neutrino charged-current total cross-sections Eur. Phys. J. C 30, 159-167, 2003. citations: 6

+ Measurement of Λc production in neutrino charged- current interactions Phys. Lett. B 555, 156-166, 2003. citations: 20

Cross-section measurement for quasi-elastic pro- duction of charmed baryons in neutrino-Nucleon interactions Phys. Lett. B 575, 198-207, 2003. citations: 14

Measurement of D0 production in neutrino charged- current interactions Phys. Lett. B 527, 173-181, 2002. citations: 26

The neutrino factory: Beam and experiments Nucl. Instrum. Meth. A 451, 102-122, 2000. citations: 82

Search for νµ→ντ oscillation using the τ decay modes into a single charged particle Phys. Lett. B 434, 205-213, 1998. citations: 53

*+ Observation of neutrino induced diffractive Ds *+ + + + production and subsequent decay Ds →Ds →τ →µ Phys. Lett. B 435, 458-464, 1998. citations: 21

3 Section 1B: Synopsis of the Project Proposal

The prime objective of this proposal is the scientific detector. This detector is now operational and data capitalization of a new generation of telescopes. are taken routinely 24 hours per day. After comple- Unlike conventional telescopes, these telescopes tion, KM3NeT will be the largest neutrino telescope will detect neutrinos and not light. The detection of in the world for the foreseeable future. The list of neutrinos from the cosmos will break new grounds targets includes active galactic nuclei, micro quasars, in the study of various frontier questions in science supernova remnants and gamma-ray bursts. such as those related to the origin of cosmic rays, the mechanism of astrophysical particle acceleration Active galactic nuclei (AGN) are objects associated and the birth of relativistic jets. The study of cosmic with the centres of far away galaxies. The amount of neutrinos is a key component in the field of astro- energy released by these objects exceeds that of any particle physics, a new interdisciplinary research other steady source. The energy is thought to be pro- field with strong links to its two progenitors: astro- vided by the gravitational force of a super-massive physics and accelerator-based particle physics. black hole. In some cases, relativistic jets have been observed. A relativistic jet can best be imagined as The common way to study the cosmos is by the a beamed outburst of matter moving at almost the detection of light. Light, however, is not in every speed of light. Many models predict neutrino pro- respect the most suitable probe for observational duction in these jets. astronomy. For example, there are regions in the Universe from which light cannot escape. Light can Micro quasars resemble AGN, but at a much smaller also be absorbed on its way to Earth. An alternative scale. They are believed to consist of a small black is to study the cosmos by the detection of cosmic hole or a neutron star that accretes matter from a rays. Cosmic rays are energetic particles originating companion star. Unlike AGN, micro quasars can be from space that bombard the Earth continuously. found in our galaxy. Because they are much closer, Their interactions with the atoms in the atmosphere they are easier to study. To some extent, they have give rise to extended showers of secondary parti- become ‘laboratories’ for revealing the physical cles which are observed on Earth. Although this processes that produce superfast jets. phenomenon has been known for almost a century, the origin of these cosmic rays remains unclear. It A supernova is a stellar explosion that creates an is suspected that the most energetic cosmic particles extremely luminous object. On average, supernovae are accelerated in astrophysical sources that are occur once every 50 years in a galaxy the size of located far away in the Universe. The intergalactic the Milky Way. The matter ejected in a supernova magnetic fields bend the trajectories of the cosmic explosion will collide with the interstellar medium, rays by an unknown and varying amount and so thus creating a shock wave. Particles accelerated impede the pin-pointing of the cosmic accelerators. in these shock waves are believed to represent the Even though the deflection is reduced with increas- bulk of the observed cosmic rays but no observa- ing energy, the difficulty remains, because at some tional evidence for this association has been found. point the cosmic rays are absorbed by the cosmic The detection of neutrinos from supernova rem- microwave background radiation limiting the depth nants could change this. of view to our galaxy. It is therefore planned to search for high-energy neutrinos that are expected Gamma-ray bursts (GRB) are very short and intense to materialise in the same cosmic accelerators. As flashes of gamma rays. These bursts occur at random neutrinos carry no electric charge, their trajectories times and at random places on the sky. GRBs were remain unperturbed despite the magnetic fields. The discovered by accident in the late 1960s by the weak interaction of neutrinos with normal matter U.S. Vela nuclear test satellites. As one of the most extends the panorama to the whole Universe. On energetic phenomena since the Big Bang, GRBs are the other hand, this same weak interaction neces- subject to many space- and ground-based observa- sitates an extremely large detector. tions. The prompt emission of the gamma rays lasts for a short time, typically less than one minute. It Following a sequence of pioneering projects, an is usually followed by a so-called afterglow. The international joint venture was initiated to build a afterglow is the emission of radiation at longer large neutrino telescope: KM3NeT. The main goal of wavelengths. From the observation of the afterglow KM3NeT is to detect neutrinos from astrophysical it has become possible to determine the redshift, and sources that exhibit the most violent processes in the hence the distance to the GRB. It has been found that Universe. Also exotic particles, like those constitut- GRBs occur at cosmological distances. The short ing the dark matter or magnetic monopoles, could be duration of the burst and the shape of the observed detected with this telescope. As a first step towards light spectra have led to the idea that a GRB is the realisation of the KM3NeT neutrino telescope, caused by a kind of ‘fireball’ explosion. The early the Antares collaboration has built a prototype phase of a GRB, however, cannot be studied directly 4 as no light can escape. But many models predict the telescope is needed to search for the neutrinos that production of high-energy neutrinos that do escape are related to dark matter. The observation of a neu- the initial fireball. The detection of neutrinos from tralino signal at the Large Hadron Collider (LHC) gamma-ray bursts will provide a unique image of at CERN and with the KM3NeT neutrino telescope the birth of relativistic jets. The neutrino signature would represent a major breakthrough as it would of GRB events will be particularly clean due to the establish a direct link between astrophysics (dark space and time correlation with the optical observa- matter) and particle physics (super-symmetry). tions provided by a network of satellites that detect the GRBs. The background is expected to be so The holy grail of particle physics is the unification small that a few events will be sufficient to claim of the fundamental forces of nature. The idea is discovery of a correlated neutrino signal. that, at sufficiently high energies, the strong force and the electroweak force have the same strength. A surprising but inevitable conclusion from a range In this case, a unique force should remain that of astronomical observations is that a large fraction results from a single symmetry. At lower energies, of the Universe consists of unknown matter. As it this symmetry is spontaneously broken into the is not luminous, this matter is referred to as dark apparent strong and electroweak forces. It has been matter. Although the existence of dark matter is shown that the existence of magnetic monopoles supported by several pieces of observational evi- appears as a generic prediction of this symme- dence, its nature is completely unknown. It has been try breaking. In practice however, no magnetic suggested that dark matter consists of a hypothetical monopole has been found so far. The mass of a particle that corresponds to one of the particles in monopole is related to the energy scale at which a new theory (super-symmetry model). Theoretical the unification occurs and the strength of the gauge support for this new theory comes mainly from the coupling. As the Universe expanded and cooled, observation that it gracefully unifies the electroweak phase transitions occurred that can be associated and strong forces at a single energy scale. The with the breakdown of symmetries. During these neutralino is the lightest super-symmetric particle. It phase transitions, magnetic monopoles could have is stable and the end-product of a decay chain of any been created. The density of monopoles in the massive super-symmetric particle which could have Universe that resulted from these phase transitions been created in the early Universe. The presence of is uncertain. In fact, simple assumptions lead to neutralinos in the Universe could explain the dark an abundance of monopoles that exceeds the total matter mystery. mass of the Universe. Many mechanisms have been proposed to overcome this problem. In all cases, The underlying physics of some neutrino point monopoles residing in the Universe will have been sources might well have its origin in super-symmetry. accelerated by large scale magnetic fields. An If neutralinos are indeed the dominant constituent of upper limit on the monopole flux can be estimated dark matter, they are expected to be found in large by assuming that all dark matter (see above) in quantities in the vicinity of galaxies. Because of their the Universe can be attributed to monopoles. This large mass and the small (but nonzero) interaction upper limit results in a measurable rate provided cross section with normal matter, neutralinos are that such monopoles can be detected efficiently in expected to gradually lose kinetic energy and gravi- a very large volume. Due to the large size of the tate towards the centre of stars and galaxies. The KM3NeT detector, the search for monopoles can be accumulation of neutralinos in e.g. the Sun and their pursued with an unprecedented sensitivity. subsequent annihilation can produce a significant flux of high-energy neutrinos. In direct searches for The scientific case for studying neutrinos from the dark matter, one tries to detect the (very small) recoil cosmos is compelling. The construction of a neu- energy deposited by a passing dark matter particle. trino telescope is, however, extremely challenging. The advantage of indirect searches for dark matter Neutrinos are best known for their reluctance to be based on the detection of cosmic neutrinos is that a detected. In short, the study of cosmic neutrinos directional correlation with a well defined source can requires a massive telescope with a size of at least be explored. The Sun is a good candidate to search one cubic kilometre. A solution to make such a large for dark matter in this way because its mass is large mass sensitive to neutrinos is to build a three-dimen- and the distance to Earth is small. The neutrinos sional array of very sensitive light sensors in the sea. coming from the annihilation of neutralinos have Neutrinos can then be detected indirectly through the much higher energies than those arising from the detection of the Cherenkov light emitted by charged nuclear fusion reactions that are the main energy particles emerging from a neutrino interaction. source of the Sun. The expected flux of high-energy The transparency of the water makes it possible to neutrinos is such that it cannot be detected by the distribute the light sensors in a cost-effective way. existing solar neutrino experiments. A large neutrino 5 The main deliverable of this proposal is the data-fil- ter system for the KM3NeT neutrino telescope. This system optimises the signal detection efficiency by using a large farm of standard PCs and state-of-the- art software. As a result, the value-for-money ratio of the entire project will be improved. The unique feature of this proposal is that for a moderate fraction of the total cost of the KM3NeT project, a key com- ponent can be accomplished. The foreseen analyses will be focused on neutrino-point sources and gamma-ray bursts. The sensitivity of the KM3NeT telescope to these potential neutrino sources will be greatly enhanced with the data-filter system devel- oped in the framework of this proposal. Hence, it is expected that these analyses will yield results within the foreseen time lines. With these analyses, one can become the first person to witness the birth of a relativistic jet and to unravel the mystery of particle acceleration in the cosmos.

With this proposal, the principal investigator (PI) will play a major role in the scientific capitalization of the KM3NeT neutrino telescope. This will position the PI and his group at the forefront of astro-particle physics for the foreseeable future.

6 Section 2: Project Proposal

Scientific motivation - detection of cosmic neutrinos The prime objective of this proposal is the scientific capitalization of the next generation neutrino tel- escope: KM3NeT. This telescope will break new grounds in the study of various frontier questions in science such as those related to the origin of cosmic rays, the mechanism of astrophysical particle ac- celeration and the birth of relativistic jets. The study of cosmic neutrinos is a key component in the field of astro-particle physics, a new interdisciplinary re- search field with strong links to its two progenitors: astrophysics and accelerator-based particle physics. The common way to study astrophysical sources is by the detection of electromagnetic radiation. Photons are, however, not in every respect the most suitable probe for observational astronomy. For ex- Figure 1: A TeV gamma ray image of the supernova ample, there are regions in the Universe from which remnant RX J1713.7-3946 as reported by the HESS photons cannot escape. Photons can also be absorbed experiment[1]. on their way to Earth. An alternative is to study the cosmos by the detection of cosmic rays. Association to have energies that extend orders of magnitude of the observed cosmic rays and any known astro- beyond the reach of any future machine. Cosmic physical source is easier said than done because the neutrinos have the potential of giving access to cosmic rays are either deflected by (inter-)galactic particle physics at the most extreme conditions. magnetic fields or absorbed by the cosmic micro- wave background radiation. This impasse has led to The scientific case for studying cosmic neutrinos is the consideration of a different cosmic messenger: compelling. Detection of these neutrinos, however, the neutrino. Indeed, the neutrino has no charge and is extremely challenging. With this proposal, the is therefore not deflected by magnetic fields. It is sensitivity of the KM3NeT neutrino telescope will be stable and interacts only weakly with matter. Thus enhanced significantly. it will travel in a straight line from the most remote places in the Universe to Earth unhampered by the Neutrino point sources cosmic microwave background radiation or (inter-) Various cosmic neutrino sources have been proposed galactic magnetic fields. in the literature. Among these are Active Galactic Nuclei (AGN), micro quasars, and supernova Traditionally, the particle accelerator has been the remnants. AGN are objects associated with the work horse of high-energy physics, but there is a centres of galaxies. The amount of energy released growing awareness that future experiments should by these objects exceeds that of any other steady also involve cosmic particles. This is not only source. The energy is thought to be provided by motivated by the ever increasing size and cost of the gravitational force of a super-massive (106–1010 man-made accelerators: cosmic particles are known solar masses) black hole. In some cases, relativistic

What is KM3NeT? KM3NeT (A km3 sized Neutrino Telescope) is a future deep-sea research infrastructure hosting a neutrino telescope with a volume of at least one cubic kilometre to be constructed in the Mediterranean Sea. The KM3NeT neutrino telescope is one of the major highways on the ASPERA roadmap. The European Strategic Forum for Research Infrastructure (ESFRI) has published a list of the most important large-scale infrastructures that should be built in the next decade. KM3NeT appears on this list confirming the strong level of support for this project Europe-wide.The design study for the infrastructure, funded by the EU FP6 framework, started in february 2006. The preparatory phase study of the infrastructure funded by the EU FP7 framework will start this year and finish in 2011. The construction phase will follow immediately and span five years. The total cost of KM3NeT is estimated at 220–250 M€.

According to the coordinator of the KM3NeT project, prof. dr. U.F. Katz: “The KM3NeT neutrino telescope will be unique in the world in its physics sensitivity and will provide access to scientific data that will propel research in different fields, including astronomy, dark matter searches, cosmic ray and high energy physics.”

7 jets have been observed. Models exist that predict neutrino production in these jets[2]. Micro quasars are 10 X-ray binary systems that are found in our galaxy. 9 They are believed to consist of a small black hole or 8 a neutron star that accretes matter from a companion 7 star. To some extent, micro quasars resemble AGN 6 5 but at a much smaller scale. A supernova is a stel- 4 lar explosion that creates an extremely luminous 3

object. On average, supernovae occur once every number of sources 2 50 years in a galaxy the size of the Milky Way. The 1 matter ejected in a supernova explosion will collide 0 1 2 3 4 5 6 7 8 9 10 with the interstellar medium, thus creating a shock years of operation wave. Particles accelerated in these shock waves are believed to represent the bulk of the observed cosmic rays but no observational evidence for this associa- Figure 2: The number of point sources that can be tion has been found. The detection of neutrinos from detected with the KM3NeT neutrino telescope at supernova remnants could change this. the 99% confidence level as a function of time. The results are based on the measurements of the HESS Recent observations from Cangaroo[3] and HESS[4] experiment and the assumption that pions are at the include TeV gamma rays from supernova remnants origin of the observed high-energy gamma rays[6]. The in the centre of our galaxy (see Fig. 1). It has been blue (bottom) bars correspond to the scenario with suggested that the observed high-energy gamma an energy cut-off at about 10 TeV. The green (top) rays are produced by inverse Compton scattering. bars correspond to the additional sources that can be In this process, a high-energy electron exchanges detected assuming no energy cut-off. energy with a low-energy photon. Due to the pres- ence of magnetic fields, the same electrons also emit limits[5]. In order to survey the galactic disk, includ- synchrotron radiation. As a result, the spectrum will ing its centre, a neutrino telescope on the northern have a characteristic broadband feature. Some of the hemisphere (relatively close to the equator) is neces- recently discovered TeV gamma ray sources, how- sary. The detector that is the basis of the present ever, do not have this broadband feature. It is gener- proposal will be located in the Mediterranean Sea. ally believed that in this case the TeV photons are the From this position, one can observe the galactic cen- result of the two-photon decay of neutral pions that tre for about 70% of the time. An evaluation of the are produced in interactions of high-energy protons discovery potential of KM3NeT for known sources with photons. The neutral pions are naturally accom- near the galactic centre is shown in Fig. 2. panied by charged pions which predominantly decay to muon neutrinos. These TeV gamma-ray emitters Dark Matter are found near the galactic centre and might there- A surprising but inevitable conclusion from a range fore be the nearest sources of high-energy neutrinos. of astronomical observations is that a large fraction of the Universe consists of unknown matter. As it An initial search for neutrino point sources away is not luminous, this matter is referred to as dark from the galactic centre resulted in a set of upper matter. Evidence for the existence of dark matter

v (km/s) observed

100

expected from 50 luminous disk

10 5 R (kpc)

M33 rotation curve

Figure 3: Three pieces of observational evidence supporting the existence of dark matter in the Universe. From left to right: The discrepancy between the observed and the expected rotational light curves, a picture of gravitational lensing, and the measured temperature anisotropy in the cosmic microwave background radiation by the COBE and the WMAP satellites. 8 comes from various sources and will be briefly sum- found in large quantities in the vicinity of galax- marised here. The observed rotational light curves ies. Because of their large mass and the small (but of (clusters of) stars in spiral galaxies do not comply nonzero) interaction cross section with normal mat- with Newton’s law when all the mass of the galaxy ter, neutralinos are expected to gradually lose kinetic is assumed to be in the luminous disk. By assuming energy and gravitate towards the centre of stars and a large amount of dark matter spread throughout galaxies. The accumulation of neutralinos in e.g. the these galaxies, this discrepancy can be resolved. The Sun and their subsequent annihilation can produce a relative motion of galaxies belonging to one cluster significant flux of high-energy neutrinos. and the thin arcs of light observed around the cluster centre due to gravitational lensing suggest that a In direct searches for dark matter, one tries to detect large fraction of the mass of the cluster is dark as the (very small) recoil energy deposited by a passing well. Measurements of the temperature anisotropies dark matter particle. The DAMA[10] experiment has in the 2.7 K cosmic microwave background radiation reported a so far unconfirmed and in fact disputed[11] indicate that 22 ± 4% of the total energy content of signal. The advantage of indirect searches for dark the Universe appears in the form of dark matter[8,9]. matter based on the detection of cosmic neutrinos is that a directional correlation with a well defined Although the existence of dark matter is supported by source can be explored[12]. The Sun is a good candi- several pieces of observational evidence (see Fig. 3), date to search for dark matter in this way because its the nature of dark matter is completely unknown. It mass is large and the distance to Earth is small. The has been suggested that the dark matter consists of neutrinos coming from the annihilation of neutral- a new hypothetical particle that corresponds to one inos have much higher energies (10–500 GeV) than of the particles in the super-symmetry model. In this those arising from the nuclear fusion reactions that model, a new symmetry between the two fundamental are the main energy source of the Sun (1–10 MeV). types of elementary particles (fermions and bosons) The expected flux of high-energy neutrinos is such is introduced. Theoretical support for this new theory that it cannot be detected by the existing solar comes mainly from the observation that it gracefully neutrino experiments. A large neutrino telescope is unifies the electroweak and strong forces at a single needed to search for the neutrinos that are related to energy scale. The neutralino is the lightest super- dark matter. symmetric particle. It is stable and the end-product of a decay chain of any massive super-symmetric Gamma-ray bursts particle which could have been created in the early Gamma-ray bursts (GRB) are very short and intense Universe. The presence of neutralinos in the Universe flashes of MeV gamma rays. These bursts occur at could explain the dark matter mystery. random times and at random places on the sky (see Fig. 4). The prompt emission of the gamma rays The underlying physics of some neutrino point- lasts for a short time, typically less than one minute. sources might well have its origin in super- It is usually followed by a so-called afterglow. symmetry. If neutralinos are indeed the dominant The afterglow is the emission of radiation at other constituent of dark matter, they are expected to be wavelengths. From the observation of the afterglow

Cosmic rays After the discovery of radioactivity by Henri Becquerel in 1896, it was generally believed that the atmospheric electricity (i.e. the ionisation of the air) was caused by radioactive elements in the ground or in the air. Then, in 1912, Victor Hess found that the ionisation rate at an altitude of 5000 meters was dramatically higher than that at the ground level. He concluded “The results of my observation are best explained by the assumption that a radiation of very great penetrating power enters our atmosphere from above”. Today, this phenomenon is known as cosmic rays. Charged particles accelerated in the expanding shock waves of supernova remnants are believed to represent the bulk of the cosmic rays up to energies of several thousands of TeV. However, no association between the observed cosmic rays and known supernova remnants can be made because the (inter-)galactic magnetic fields blur the image of the sky. This scientific deadlock could come to an end by the detection of cosmic neutrinos. Interactions of the accelerated particles with the matter surrounding the source are expected to produce neutrinos via the decay of short lived particles (mainly pions). As the cross sections of these interactions are well-known, the fluxes of the high-energy neutrinos can be calculated based on the measured flux of cosmic rays and the assumed densities of photons and nuclei. These calculations indicate that a detectable cosmic neutrino signal is within reach. On their way to Earth, these neutrinos are not deflected by (inter-)galactic mag- netic fields. Detection of such a neutrino signal would provide the experimental evidence that supernova remnants are indeed the sought after cosmic accelerators.

9 2704 BATSE Gamma-Ray Bursts Unfortunately, the start of the fireball and the early +90 phase of the explosion cannot be studied directly as no electromagnetic radiation can escape. Present fireball models predict the acceleration of charged particles (mainly protons) preceding the observable +180 –180 light flash. Interactions of these protons with the surroundings produce high-energy neutrinos that can escape the fireball. The detection of neutrinos associ- ated with a GRB will thus uncover the invisible core. The electromagnetic afterglow can be detected on –90 Earth provided that the accurate position of the GRB

10-7 10-6 10-5 10-4 is known instantaneously. In that case, ground-based Fluence, 50–300 keV (ergs cm-2) telescopes can be pointed in the right direction in time. For this purpose, satellite-based warning sys- Figure 4: Sky map of the gamma-ray bursts detected tems have been developed that distribute messages by the BATSE detector on board of the Compton around the world, containing the celestial positions Gamma Ray Observatory[21], launched in 1991. (The of the GRBs. The Swift[22] satellite was launched in horizontal centre line coincides with the galactic disk.) 2004 and is part of this system. It detects about 100 GRBs per year and distributes the corresponding it has become possible to determine the redshift, and warning messages within several seconds. hence the distance to the GRB16. It has been found that GRBs occur at cosmological distances: the mean The neutrino signature of GRB events will be redshift is about 1.3, which corresponds to a distance particularly clean due to the space and time correla- of about 2.5 Gpc or 7 billion light years. For com- tion with the optical observations provided by the parison, the size of our galaxy is 30 kpc across. previously mentioned warning systems. The back- ground is expected to be so small that a few events The short duration of the bursts and the non-thermal will be sufficient to claim discovery of a correlated character of the observed spectra have led to the neutrino signal. Based on the model in reference[14], idea that a GRB is caused by some kind of ‘fireball’ explosion[17]. By some unknown mechanism, a large Relativistic jets amount of energy is deposited into a small object. Another field where the observation of cosmic This object is then heated up and becomes a kind of neutrinos could lead to a breakthrough is related to fireball. Unable to cool or to keep its energy confined, relativistic jets. Relativistic jets are very powerful this object will release its heat by ejecting matter. beams of plasma which emerge from the centres The kinetic energy of this matter is such that it almost of some astrophysical objects. These jets can carry reaches the speed of light. The enormous amount of as much mass as the planet Jupiter and move at energy emitted by a single source made early models more than 99.9% of the speed of light. Relativistic that assumed an isotropic explosion questionable. jets have been observed in active galactic nuclei Nowadays, the outflow of matter from the fireball and micro quasars but it is not clear how they are is assumed to be collimated, which reduces the total produced. The formation of relativistic jets could energy output[18]. The kinetic energy of the matter be the key to explaining the production of gamma- and the collimation of the outflow make up a rela- ray bursts. Gamma-ray bursts were discovered tivistic jet. It is the conversion of the kinetic energy in the late 1960s by the U.S. Vela nuclear test of this jet into radiation that leads to the observed satellites. These satellites were built to detect gamma rays. Whether this is due to internal shocks gamma-radiation pulses emitted by possible secret or external shocks is not yet clear. In the internal nuclear weapons tests by the former U.S.S.R. shock model[19], many shells are ejected by the same after signing the nuclear test ban treaty in 1963. compact object. A later but faster shell can then As one of the most energetic phenomena since the collide with an earlier but slower shell. The external Big Bang, gamma-ray bursts are subject to many shocks[20] occur when a relativistic jet interacts with space- and ground-based observations. The early the ambient matter. The afterglow emission is a phase of gamma-ray bursts cannot be studied natural phenomenon in either model as the matter directly as no electromagnetic radiation can in the jets will gradually slow down anyway due to escape. Many models[13,14,15] predict the production interactions with the interstellar matter. While the of high-energy neutrinos that can escape the initial mechanism that initiates the fireball explosion is ‘fireball’. The detection of neutrinos from gamma- unknown, in some cases a supernova could be associ- ray bursts will provide a unique image of the birth ated with a GRB. This would indicate that the core of relativistic jets. collapse of a massive star is at the origin of a GRB. 10 about six GRBs have occurred during the last decade Although a neutrino telescope is meant to detect the that would have produced three (or more) detect- products of a neutrino interaction, it is also capable able events in the KM3NeT neutrino telescope. A of detecting the passage of a magnetic monopole. comparison of the time profiles of the neutrino signal When such a particle is in motion, its radial magnetic and the light curves will provide information on the field will excite and ionise the atoms in the medium creation and evolution of relativistic jets that are at that it traverses[31]. Phenomena related to this are the the origin of all GRBs. direct emission of Cherenkov light and the produc- tion of knock-on electrons (δ-rays). Both effects Exotic particles will lead to a detectable signal in deep-sea neutrino Theories aiming at the unification of the fundamental telescopes. Detection of a magnetic monopole would forces of nature are based on the hypothesis that at be considered a major discovery. On the other hand, sufficiently high energies, the strong coupling and experimental limits on the flux of magnetic monopo- the electroweak coupling have the same magnitude. les will constrain or even reject some cosmological In this regime, a unique force should remain that models and/or new theories. Due to the large size of results from a single symmetry. At lower energies, this the KM3NeT detector, the search for monopoles can symmetry is spontaneously broken into the apparent be pursued with an unprecedented sensitivity. strong and electroweak symmetries as formulated in the Standard Model of particles and fields. It has Detection principle been shown that the existence of magnetic monopoles The construction of a neutrino telescope is appears as a generic prediction of this symmetry extremely challenging. Neutrinos are best known breaking[26,27]. The mass of such a monopole is related for their reluctance to be detected. In short, a to the energy scale at which the unification occurs and systematic study of cosmic neutrinos requires the strength of the gauge coupling. It is commonly a massive telescope with a size of at least one expected that the broken symmetry was restored in the cubic kilometre. A solution to make such a large early Universe. As the Universe expanded and cooled, mass sensitive to neutrinos is to build a three- phase transitions occurred that can be associated with dimensional array of very sensitive light sensors the breakdown of symmetries. During these phase in the sea[4]. Neutrinos can then be detected transitions, magnetic monopoles could have been indirectly through the detection of the Cherenkov created. The density of monopoles in the Universe that light emitted by charged particles emerging from resulted from these phase transitions is uncertain. In a neutrino interaction (see Fig. 5). The transpar- fact, simple assumptions lead to an abundance of mo- ency of the water makes it possible to distribute nopoles that exceeds the total mass of the Universe. the light sensors in a cost-effective way. The

Many mechanisms have been proposed to overcome absorption length of the water (λabs) has been this problem[28,29,30]. In all cases, monopoles residing in measured at selected sites in the Mediterranean the Universe will have been accelerated by large scale Sea and was found to be about 50 metres. magnetic fields. An upper limit on the monopole flux can be estimated by assuming that all dark matter (see The angular resolution of such a detector is limited above) in the Universe can be attributed to monopoles. by the lever arm between the light sensors and the This upper limit results in a measurable rate provided measurement precision of their positions and the that such monopoles can be detected efficiently in a arrival times of the Cherenkov light. The mechani- very large volume[40]. cal structure that accommodates the light sensors

The quantum Universe At the interface between particle physics and astrophysics, several far-reaching questions persist which can be addressed with a neutrino telescope. These questions are motivated by the quest to explain the Universe in terms of quantum physics. Subtle asymmetries between particles and anti-particles –some of which have been observed experimentally– may explain the dominance of matter in the Universe. However, most of the matter in the Universe is ‘dark’: it appears not to consist of any known particle. But without dark matter, stars and galaxies would not have formed. The discovery that neutrinos have a non-zero mass could imply that they constitute the dark matter. But despite their omnipresence –there are almost as many neutrinos in the Universe as photons– their masses are simply too small to solve the dark matter mystery. An alternative candidate for dark matter is a new hypothetical particle: the neutralino. Neutralinos are weakly interacting with normal matter and therefore difficult to detect. But, if they exist, they will accumulate in massive celestial bodies, such as the Sun. There, they can annihilate and produce high-energy neutrinos. The observation of a neutralino signal at the Large Hadron Collider (LHC) at CERN[7] and with the KM3NeT neutrino telescope would represent a major breakthrough as it would establish a direct link between dark matter (astrophysics) and super-symmetry (particle physics).

11 Neutrinos In a famous letter to colleagues, Wolfgang Pauli In 1962, another kind of neutrino was discovered postulated in 1930 the existence of a new and by detecting interactions of neutrinos produced invisible particle as a desperate remedy to account with man-made particle accelerators. Experiments for the missing energy observed in radioactive at the LEP collider at CERN, Geneva revealed in decays. In the following years, Enrico Fermi took 1989 that there are three kinds of neutrinos. The up Pauli’s idea and developed the theory of weak first glimpse of the third kind of neutrino was interactions. He assigned the name neutrino to the caught in 2000 at Fermilab near Chicago. The three particle as a pun on neutrone, the Italian for neutron. different kinds in which a neutrino may appear (The neutron is the neutral companion of a proton; are usually referred to as neutrino flavours. Each together they are the constituents of the nucleus of neutrino is created with a well defined flavour all atoms.) In his theory, neutrinos can interact with (electron, muon or tau). However, in a phenomenon matter, but only very weakly. With the advent of known as neutrino flavour oscillations, neutrinos nuclear reactors, a powerful source of neutrinos was are able to oscillate between the three available at hand. This led to the first observation of neutrino flavours while they propagate through space. The interactions by Clyde Cowan and Frederick Reines existence of flavour oscillations implies a non-zero in 1956. This result provided the experimental neutrino mass. Despite their massive nature, it is evidence for the existence of a neutrino. not yet clear whether the neutrino and the antineu- trino are in fact the same particle, a hypothesis first proposed by Majorana in 1937.

does not form a static system due to changing sea and time (1 ns) resolutions of the light sensors make currents. Hence, their positions must be monitored it possible to reconstruct the direction of high-energy continuously through acoustic triangulation. Of the muons with an accuracy of about 0.2 degrees. This three neutrino species that exist in nature, the muon corresponds to about half of the apparent size of the neutrino yields the best angular resolution because Sun on the sky. Below the maximum penetration the muon that emerges from a neutrino interac- depth of daylight in the sea (about 1000 m), the light tion has the longest range. The range of the muon sensors can be operated 24 hours per day. Suitable increases linearly with its energy up to the so-called light sensors are photo-multiplier tubes. Such ‘critical energy’ of approximately 250 GeV. At this photo-multiplier tubes housed in pressure resistant energy, the range is about one kilometre. Above this glass spheres form the basic building blocks of the energy, Bremsstrahlung limits the increase of the detector (also referred to as optical modules). The muon range but does enhance the detectable signal. huge mass needed for a neutrino to interact requires The light transmission properties of sea water com- a large 3-dimensional array of such optical modules bined with the presently feasible position (10 cm) (about 10,000). The field of view of a neutrino telescope can in principle cover the full sky, but one must cope with the background from downward- going muons produced in cosmic ray interactions wavefront with the Earth’s atmosphere. The flux of these muons has been measured extensively and decreases with depth. At a depth of three kilometres, the flux of muons with energies in excess of 100 GeV is about muon 100 km–2s–1. A unique feature of a neutrino telescope is its ability to look downwards at neutrinos that 1–2 λabs traverse the Earth.

The concept of a neutrino telescope is based on the ~ 1 km detection of the products of a neutrino interaction that takes place in the vicinity of the detector. Hence, Figure 5: Schematic view of the production of Cherenkov the sensitivity of a neutrino telescope depends on the light by a muon. (A muon is one of the possible products interaction cross section of neutrinos and the effec- of a neutrino interaction.) According to the principle of tive volume of the detector. This contrasts sharply Huygens, spherical light waves are produced along the with the traditional picture that the sensitivity of a muon trajectory. The lightwaves interfere because the telescope is only proportional to a surface area, e.g. muon travels faster through the water than the light. As a that of a lens or a mirror. The equivalent expression result, a sharp wave front is formed that can be detected for the surface area of a neutrino telescope will be with very sensitive light sensors. given in the following. In general, the number of 12 muon detected neutrinos per unit time as a function of the νμ neutrino energy, Eν, can be expressed as: θ

W 1: dN(Eν ) = Φ(Eν )×σ(Eν )×NA×ρ×Veff (Eν )×dEν , other particles where Φ(Eν ) is the incident neutrino flux,σ (Eν ) the nucleon neutrino cross section, NA Avogadro’s number, ρ the density of the medium (i.e. water), and Veff (Eν ) Figure 6: Schematic view of an interaction of a muon the effective detection volume. In this equation, neutrino, νµ. A W boson is exchanged between the the dependence of the cross section and the effec- neutrino and a nucleon, transforming the neutrino into tive volume on the kinematics of the interaction is a muon. The angle θ is defined as the angle between the implicitly integrated. The cross section is known to incident neutrino direction and the muon direction. increase with the energy of the neutrino; up to about 10 TeV this increase is linear. The effective volume As the neutrinos are detected indirectly, it is impor- is defined as the volume in which a neutrino inter- tant to know the correlation between the direction of action produces a detectable muon, and basically the incident neutrino and the muon (see Fig. 6). The indicates the size of the telescope. neutrino interaction proceeds through the exchange of a W or Z boson, referred to as a charged and neu- Equation 1 can be rewritten as: tral current interaction, respectively. For a neutrino energy above 0.01 TeV, the interaction is dominantly

2: dN(Eν ) = Φ(Eν )×Aeff (Eν )×dEν , deep inelastic, i.e. a nucleon in the target breaks up into low energy particles (mainly pions). A muon is where Aeff (Eν ) is defined as the neutrino effective produced in the charged current interaction of a muon area. In this definition, the neutrino effective area neutrino. The energy of the muon (Eµ ) is on average includes both the neutrino cross section and the half of that of the neutrino. As long as the neutrino effective volume. As a result, the sensitivity of a energy is below 10 TeV, the value of Q2 (–Q2 is the neutrino telescope can be directly related to a surface four-momentum transfer squared) is small compared area. Due to the small cross section for a neutrino in- to the square of the W and Z mass. The cross sec- teraction with matter, this area is much smaller than tion is then proportional to the neutrino energy. The the geometrical cross section of the instrumented higher the neutrino energy, the smaller the value of volume. As both the neutrino cross section and to Björken x at a given Q2 (x can be interpreted as the some extent also the effective volume increase with momentum fraction of the nucleon carried by the the energy of the neutrino, so does the effective area. struck quark). Measurements at the HERA collider This is an important benefit as the signal-to-back- (DESY, Germany) determined the quark density ground ratio is also expected to increase with energy. in the proton down to very small values of x. The In conclusion, a huge instrumented volume (1 km3) results show that this density increases strongly with 2 2 is needed to achieve a reasonably large effective area decreasing x. Since Q ≃ 4EνEµ sin (θ / 2) ≤ 2MNEν 2 (100 m ) at the highest energies. (MN is the nucleon mass) and 〈Eµ 〉 = ½Eν , the

Neutrino telescope The neutrino is of great scientific interest because it can make an exceptional probe for places in the Universe that are otherwise concealed. An early fulfilment of this perspective has been the observation of neutrinos from the core of the Sun. Direct optical observation of the solar core is impossible due to the diffusion of the electromagnetic radiation by the huge amount of matter surrounding the core. While photons may require thousands of years to get to the outer layers of the Sun, neutrinos generated in nuclear fusion reactions escape the solar core unperturbed in a second or two. Neutrinos are also useful to probe astrophysical sources beyond our solar system. A prime example is the observation of neutrinos from a supernova. A supernova may briefly out-shine its entire host galaxy inferring that all its energy is released in the form of electromagnetic radiation. The observation of neutrinos from the supernova SN1987a revealed that most of the energy is actually released in the form of neutrinos and not light.

The next target in the list of neutrino sources is the centre of our galaxy which is known to host many powerful emitters of high-energy radiation. It is likely that neutrinos produced in the galactic centre will be observed for the first time by large Earth-based neutrino telescopes, such as KM3NeT, in the next decade.

13 average scattering angle θ reduces with increasing Is there an end to the cosmic ray spectrum? neutrino energy. Both the enhanced contribution The observations of cosmic rays with energies of small x and the effect of the W mass reduce the in excess of 1020 eV are a mystery as they should increase of the average Q2 with increasing neutrino have interacted with photons from the cosmic energy. Above 1 TeV, a safe limit on the average microwave background radiation before reaching scattering angle is: the Earth. This is known as the GZK[23,24] cut-off. Some scenarios to explain the paradox involve 22 2 3: 〈θ〉 ≤ 1.5 deg / √Eν [TeV] , very massive particles (about 10 eV/c ) which can decay nearby and produce the observed cos- i.e. the direction of the muon is closely related to that mic rays. Recent observations of the Pierre Auger of the neutrino, which is essential for the neutrino observatory in Argentina support, however, the telescope concept. existence of the anticipated cut-off[25]. Moreover, a correlation of the highest-energy cosmic rays with Even in the absence of daylight, a ubiquitous back- nearby extragalactic objects has been reported a ground luminosity is present in deep-sea water due to few months ago[41]. If all the high-energy cosmic decays of radioactive isotopes and bioluminescence. rays from far away sources are absorbed, plenty This background luminosity produces a relatively of neutrinos are produced (every lost proton high count rate of random signals in the light sen- yields a neutrino). One of the unique features of sors of the detector. Consequently, the data rate of a the neutrino is that it is not sensitive to the GZK deep-sea neutrino telescope exceeds any data storage cut-off. This leads to a detectable neutrino flux on capacity by several orders of magnitude. The rare Earth, provided that the detector is large enough. neutrino signal can be discriminated from the random KM3NeT will be the largest neutrino telescope in background utilising the time-position correlations the world, thus providing the best chance to find produced by a traversing particle. The main difficulty back the missing cosmic rays. is the real-time filtering of the neutrino signal from the continuous random background. As a conse- however, are decay products of short lived particles quence, the effective area of a neutrino telescope (mainly pions and K mesons) that are produced depends also on the quality of the data filter. in cosmic ray interactions with the Earth’s atmos- phere. The higher the energy of these particles, the Above a certain energy (1000 TeV), the neutrino longer they live and the farther they travel. It thus cross section has grown to such an extent that the becomes more and more likely that these particles Earth becomes opaque. In this regime, the neutrino re-interact with the Earth’s atmosphere (or hit the has a significant probability to interact with the Earth Earth) before they decay. Consequently, the energy before it reaches the detector. Hence, ultra-high spectrum of atmospheric muons is steeper than the energy neutrinos can only be detected if the tel- original cosmic ray spectrum, leading to an effective escope is also sensitive to downward going muons. reduction of the background at high energy. The In that case, one has to cope with the abundance background is further reduced by limiting the field of of atmospheric muons. The atmospheric muons, view of the neutrino telescope to angles close to the

… .… .. .. …… .. GRB .. .. …… warning remote Figure 7: Schematic operation view of the ‘All-data-to- shore’ concept and the real time link to the Swift satellite analysis that is part of the GRB warning system. The deep-sea neutrino tel- escope can be operated from anywhere in the data transmission world. The neutrino sig- software lter nal is filtered on shore by a farm of PCs and distributed real time. 14 horizon. Within this field of view, it will be possible warning systems in the manner for which they are to detect ultra-high energy cosmic neutrinos despite meant[35]. As all raw data are sent to shore, they can the background of atmospheric muons. be buffered before being processed. In case of a GRB alert, the data in memory and all data taken In the study of cosmic neutrinos, there is a small but during and following the burst can be saved on inevitable background due to the neutrinos that come disk and analysed offline. In this way, one can look together with the atmospheric muons (the pions and K before, during and after the GRB for a correlated mesons decay into a muon and a neutrino). For point neutrino signal. This offers a unique opportunity to sources, this background scales with the square of study the early phase of a GRB with the best pos- the angular resolution of the telescope and with the sible sensitivity. measurement time. For transient sources like gamma- ray bursts, the net result is an almost background-free A further development made recently is a direction data sample. For steady sources, the number of sensitive data filter. If the direction of the incident neutrinos needed to claim a discovery depends on the neutrino is known, one can in principle lower the location of the source on the sky and the energy of detection threshold. This idea is virtually impossible the neutrinos. For an observation period of 3 years, it to implement in hardware. A software solution has typically ranges between 5–10 events for a confidence been implemented by the PI for the Antares detector. level of 99%. This achievement shows that one can track an astro- physical source and thus detect a neutrino signal with Development plan the best possible sensitivity. As a first step towards a large neutrino telescope, the Antares collaboration has built a prototype detector. The concept of a software based data filter has been Major contributions of the principal investigator (PI) established for a small neutrino telescope. For the to this project include the novel ‘All-data-to-shore’ much larger KM3NeT neutrino telescope, massive readout system (see Fig. 7). In this system, the rare parallel computing will be necessary. The anticipated neutrino signal is filtered on shore from the back- data throughput poses tough requirements on the ground using a farm of commodity PCs and state- performance of the system. Therefore, simulations of-the-art software (the signal-to-noise ratio is about have to be made to study the data traffic through the 10–8 at the primary light sensor). The software devel- system. The rare neutrino signal needs to be filtered opments have been realised by the PI. This software real-time. Hence, all processes in the system have is now fully operational in the Antares experiment. to keep up with the high input rate. New pattern This achievement is remarkable (few considered recognition algorithms need to be developed to make this possible) and changed the picture dramatically: this possible. The results of these developments will the readout system has become a seminal part of the determine the success of the entire project. Analysis astro-particle physics programme with a neutrino of the data will start as soon as the first part of the telescope. It is obvious - but worth noting - that the KM3NeT detector is deployed. The foreseen analy- faster the software data filter is, the more physics can ses will be focused on neutrino-point sources and be done with the same instrument. gamma-ray bursts. The sensitivity of the KM3NeT telescope to these potential neutrino sources will be The Antares neutrino telescope is the first neutrino greatly enhanced with the data-filter system devel- telescope that uses the earlier mentioned GRB oped in the framework of this proposal. Hence, it is

Cosmic neutrinos The only cosmic neutrinos that have been detected so far are neutrinos from the Sun and from the supernova SN1987a. The detection of solar neutrinos not only confirmed the solar model[32], it unexpectedly led to the discovery of neutrino oscillations[33]. The neutrino signal from the supernova SN1987a was originally missed. It was only after the suggestion of the late J. Bahcall that a significant increase in the noise rate of the photo- multiplier tubes in the Kamiokande detector was noticed[34]. The neutrino signal appeared only a few minutes after a calibration run which would have wiped out the signal completely. A more detailed offline analysis of the data showed a dozen events that were coincident with the observed supernova explosion. This implied that most of the energy of the supernova explosion was released in the form of neutrinos and not light. It is as yet the one-and-only example of a time coincidence between a light signal and a neutrino signal from a single as- trophysical source. The cosmic neutrinos detected so far have energies of the order of a few MeV. They are not the subject of this proposal. The angular resolution of a neutrino telescope would then be limited to 10 degrees or so. With the advent of a new generation of high-energy neutrino telescopes, the study of cosmic neutrinos with an angular resolution of about 0.2 degrees becomes possible. The realisation of this perspective started with the successful commissioning and operation of the prototype detector built by the Antares collaboration.

15 expected that these analyses will yield results within Item description cost k€ the time lines of this proposal. With these analyses, i) personnel one can become the first person to witness the birth post-doc A (4 years) 240 of a relativistic jet and to unravel the mystery of post-doc B (4 years) 240 particle acceleration in the cosmos. PhD student A (4 years) 160 PhD student B (4 years) 160 Another key issue in neutrino astronomy is the travel 85 accuracy with which the direction of the neutrino can overhead 160 be determined. Due to the nature of the Cherenkov ii) data-filter system light and the sparse distribution of light sensors, this computer farm 820 problem is highly nonlinear. In the classical approach, Ethernet switch fabric 130 only one solution is searched for using a general op- running costs 420 timisation procedure. Such a procedure could lead to transport/installation 25 one of the degenerate solutions that is inherent to the iii) miscellaneous nonlinearity of the problem. The net result is a loss of office equipment 50 efficiency. It turns out that the algorithms used in the software licenses 10 data-filter software mentioned above can also be used to project the full parameter space of the problem Total 2,500 onto a sub-space in which the problem is (almost) Table 1: Budget break-down in k€ for the entire linear. All solutions can then be found by a systematic period covered by this proposal. scan of a minimal set of possible projections. The final result is determined by the selection of the best The total budget is 2.5 M€ for a period of five years. solution from all solutions found. In addition, a per- A breakdown of the budget is given in Table 1. The sistent ambiguity can be treated properly by keeping two post-docs will make the simulations and develop all corresponding solutions. This approach has never the required software. The two PhD students will been considered before but preliminary studies show make the physics performance studies and do the data that it is very promising. analyses. The obtained results will be presented at (international) conferences and published in refereed With this proposal, the PI will play a major role journals. The item ‘data-filter system’ includes cost of in the scientific capitalization of the KM3NeT prototyping, purchase of hardware (excluding VAT), neutrino telescope. This will position the PI and his transport and installation. The item ‘running costs’ group at the forefront of astro-particle physics for includes power, cooling and maintenance, for an the foreseeable future. assumed period of 4 years. The required computing power has been evaluated for the complete KM3NeT Budget detector, taking into account the results of the software The main deliverable of this proposal is the data- developments which are part of this proposal. The filter system for the KM3NeT neutrino telescope. item ‘miscellaneous’ covers the general running cost With this system it is possible to make the envisaged of this project. The overall costs of the data-filter scientific studies as soon as the telescope is opera- system includes the purchase and the operation during tional and obtain results within the foreseen time lines. the period covered by this proposal.

ID Task Name '05 '06 '07 '08 '09 '10 '11 '12 '13 '14 '15 '16 '17 '18 '19 '20 '21 1 KM3NeT FP6 design study 2 Conceptual design 3 Conceptual Design Report 30/11 4 Technical design 5 Technical Design Report 25/05 6 KM3NeT FP7 preparatory phase 7 Production modelling 8 Market analysis 9 Financial analysis 10 Human resources analysis 11 Funding model 16/02 12 This proposal 13 Intended starting date 01/07 14 Design & development 15 Purchase & testing 16 Installation 17 Data analyses 18 KM3NeT 19 Construction 20 Operation Figure 8: Time lines of the development plan of this proposal and the KM3NeT project. 16 90ϒ

80

60

Galactic latitude 40

20

180 ϒ 0

-20

-40

-60

-80

-150 -100 -50 050 100 150 Galactic longitude

Figure 9: Simulated sky map of neutrinos detected Reasons for the proposed investment by the KM3NeT neutrino telescope over a ten year The scientific case for studying neutrinos from the period. The simulation is based on the known cosmos is compelling. Detection of these neutrinos, diffuse background of atmospheric neutrinos and however, remains extremely challenging. Following includes the measurements of the HESS experiment. a sequence of pioneering projects of which Antares Clusters of neutrinos are clearly visible in the was one, an international joint venture was initi- galactic plane (horizontal line) and correspond to ated to build a large deep-sea infrastructure in the the HESS measurements. Mediterranean Sea. This infrastructure will accommo- date the next generation neutrino telescope. KM3NeT for the data-filter system of the the largest neutrino is supported through the 6th and 7th framework telescope in the world. In view of the time lines programmes of the EU. This will culminate in a tech- shown in Fig. 8 –the construction of the KM3NeT nical design report by the year 2009 and a Europe- neutrino telescope should start in 2011– this proposal wide funding model by 2010. The time lines of the is urgent. This planning makes it possible to profit KM3NeT project and the present proposal are shown maximally from the experience gained with the in Fig. 8. The installation of the computing hardware Antares project and to be ready in time for the opera- proceeds in two steps to comply with the actual tion of the KM3NeT neutrino telescope. construction of the KM3NeT neutrino telescope. The unique feature of this proposal is that for a The sheer size of the detector, the large number of moderate fraction of the total cost of the KM3NeT optical modules, the unavoidable background and the neutrino telescope, a key component can be ac- rare neutrino signal makes the filtering of the data complished. The cost of this project includes the a major challenge. This proposal aims at optimis- purchase of the hardware as well as the development ing the signal-detection efficiency by using a large of the software. These costs represent a significant farm of standard PCs and state-of-the-art software. fraction of the total budget. It should be noted, The necessary software does not exist and is not however, that the idea of transferring all data to commercially available. Although the project has a shore flies in the face of conventional wisdom which high risk (is it possible to process data at a rate of often dictates that the volume of data be reduced as 100-1000 Gb/s and detect a signal at a 10–8 level?) quickly as possible. The present proposal has in this the prospects are favourable. A proof of concept has sense no challenger. Hence there are no other budget been established in the framework of the Antares requests for the same purpose. The data processing prototype project. At this point, we propose to capi- on shore, where all signals can be searched simulta- talize on our experience gained within the Antares neously, is challenging and yet has the potential of project and set our ambitions to the next and highest great rewards. level: build the hardware and develop the software

17 National importance Neutrino astronomy commenced in the Netherlands 2006 this collaboration received funding from the 6th in the year 2000 when Nikhef –the Dutch national Framework Programme of the EU for a design study. institute for subatomic physics– joined Antares, the About 9 M€ was awarded to the KM3NeT design international effort to construct a prototype neutrino study, 0.65 M€ was allocated to the Netherlands. telescope. Within the Antares project, the PI has the The same collaboration submitted a proposal for full responsibility for the data acquisition system a Preparatory Phase Study in the 7th Framework and the coordination of the physics working groups. Programme of the EU, which was approved in August Since 2005 the PI is the Deputy Spokesperson of 2007. This type of support is only available for Antares. The participation of Nikhef in Antares has projects that appear on the ESFRI list. The KM3NeT been financed by the Dutch general scientific funding collaboration requested 6.8 M€, of which 5.0 M€ was agency NWO through an investment grant of 3.6 M€, granted. About 10% is earmarked for the Netherlands. and the Dutch physics funding agency FOM through a six-year scientific programme (6.7 M€). The involve- Existing underground neutrino detectors, such as ment of Nikhef in Antares has paved the way for the Super-Kamiokande[38], are optimised for the detec- participation of the Dutch astro-particle physics com- tion of lower energy neutrinos, typically in the range munity in the KM3NeT project. Within the KM3NeT 106–1010 eV. High-energy neutrino telescopes are project the PI is the coordinator of the ‘information radically different; their architecture is optimised to technology’ work package, which covers the comput- achieve a large detection volume and a good angular ing infrastructure described in this proposal. resolution. Worldwide, several large under-water or -ice neutrino telescope projects exist in various stag- In the ‘Strategic Plan for Astro-particle Physics in the es of development at the present time. The Amanda Netherlands’[36], written in 2005, searches for neutrino collaboration has built a 0.05 km3 size neutrino point sources with Antares (as a first generation neu- telescope in the ice cap of the South Pole. Due to the trino telescope) and KM3NeT (as the next generation light scattering in the ice, the angular resolution of neutrino telescope) were selected as key projects for this telescope is limited to about 1 degree. The U.S. the development of this field in the Netherlands. led IceCube collaboration has started building a large neutrino telescope around the Amanda detector[39]. International position The KM3NeT neutrino telescope is one of the major The KM3NeT collaboration plans to build a neutrino ambitions of the European astro-particle physics telescope in the Mediterranean Sea. Many of the community, as evidenced by the roadmap for astro- intricate technologies required for the operation of a particle physics, which is presently being prepared large detector in the deep-sea were developed by the by the FP6-funded ERA-NET project ASPERA[37]. Antares project. The analysis of the first data taken ASPERA, in which FOM is one of the leading has shown that the light scattering in the sea water is partners, aims at identifying and removing financial as small as expected and that the absorption length is and organizational barriers that could prevent the large. This implies that the angular resolution of the realization of large-scale European infrastructures in neutrino telescope will be better than 0.2 degrees. In the field of astro-particle physics. The KM3NeT neu- view of the competition with IceCube and the time trino telescope is one of the major highways on the needed to build such a large deep-sea infrastructure, ASPERA roadmap. Recently the European Strategic this proposal is timely. The present proposal is Forum for Research Infrastructures (ESFRI) pub- expected to start mid 2009 and will take five years lished a list of the most important large-scale infra- to completion. The KM3NeT neutrino telescope will structures that should be built in the next decade. complement IceCube in the sense that the telescope KM3NeT appears on this list confirming the strong will cover a different part of the sky, most notably level of support for the project Europe-wide. the centre of the galactic disk. The superior angular resolution of KM3NeT will yield the better back- More than 200 scientists from 9 different countries ground rejection. After its completion, KM3NeT will collaborate in the European KM3NeT project. In be the most sensitive neutrino telescope in the world.

18 References

1 http://www.mpi-hd.mpg.de/hfm/HESS 2 F.W. Stecker and M.H. Salamon, Space Sci. Rev. 75 (1996) 341. 3 http://icrhp9.icrr.u-tokyo.ac.jp/ 4 M.A. Markov, Proc. Of the Annual Int. Conf. On High Energy Physics, Rochester (1960). 5 J. Ahrens et al., Phys. Rev. Lett. 92 (2004) 071102. 6 A. Kappes, J. Hinton, C. Stegmann and F.A. Aharonian, J. Phys. Conf. Ser. 60 (2007) 243. 7 http://www.cern.ch/ 8 J.C. Mather et al., Astrophys. J. 420 (1994) 439. 9 M.R. Nolta et al., Astrophys. J. 608 (2004) 10. 10 R. Barnabei et al., Riv. N. Cim. 26 (2003) 1. 11 D.S. Akerib et al., Phys. Rev. Lett. 93 (2004) 211301. 12 A. Achterberg et al., Astropart.Phys. 26 (2006) 129. 13 E. Waxmann and J. Bahcall, Phys. Rev. Lett. 78 (1997) 2292. 14 S. Razzaque et al., Phys. Rev. Lett. 90 (2003) 241103. 15 A. Dar and A. De Rújula, astro-ph/0105094. 16 J. van Paradijs et al., Nature 386 (1997) 686. 17 M.J. Rees and P. Mészáros, Mon. Not. R. Astron. Soc. 258 (1992) L41. 18 J.E. Rhoads, Astrophys. J. 487 (1992) L1. 19 M.J. Rees and P. Mészáros, Astrophys. J. 430 (1994) L93. 20 P. Mészáros and M.J. Rees, Astrophys. J. 405 (1993) 278. 21 N. Gehrels et al., Scient. Am. 269 (1993) 68. 22 http://heasarc.nasa.gov/docs/swift/swiftsc.html 23 K. Greisen, Phys. Rev. Lett. (1966) 748. 24 G.T. Zatsepin and V.A. Kuz’min, JETP Lett. (1966) 78. 25 http://www.icrc2007.unam.mx/ 26 G. ’t Hooft, Nucl. Phys. B79 (1974) 276. 27 A.M. Polyakov, JETP Lett. 20 (1974) 194. 28 A.H. Guth and E.J. Weinberg, Nucl. Phys. B212 (1996) 321. 29 T.W. Kephart and T.J. Weiler, Astropart. Phys. 4 (1996) 271. 30 E. Huguet and P. Peter, Astropart. Phys. 12 (2000) 277. 31 S.P. Ahlen, Phys. Rev. D17 (1978) 229. 32 J.N. Bahcall and R. Davis, Science 191 (1976) 1494. 33 Q. R. Ahmad et al., Phys. Rev. Lett. 89 (2002) 011301. 34 K. Hirata et al., Phys. Rev. Lett. 58 (1987) 1490. 35 M.C. Bouwhuis, PhD thesis UvA, Detection of Neutrinos from Gamma-Ray Bursts (2005). 36 http://www.astroparticlephysics.nl/papers.php 37 http://www.aspera-eu.org/ 38 Y. Fukuda et al., Phys. Rev. Let. 81 (1998) 1562. 39 A. Achterberg et al., Astropart. Phys. 26 (2006) 155. 40 B.A.P. van Rens, PhD thesis UvA, Detection of Magnetic Monopoles below the Cherenkov Limit (2006). 41 The Pierre Auger collaboration, Science 318 (2007) 938.

19 20 Section 3: Research Environment

The National institute for subatomic physics Nikhef in the Netherlands. The fibre-optic data transmission –a joint venture of the Dutch funding agency FOM hardware was developed in close collaboration with and four universities– coordinates experimental Dutch industry. Similar opportunities for Dutch subatomic physics in the Netherlands. As such, industry will arise in the framework of KM3NeT. Nikhef forms the home base for Dutch experimental Neutrino astro–particle physics commenced in research and detector developments in this field the Netherlands in the year 2000 when, under the worldwide. Nikhef is host to about 225 persons, leadership of the principal investigator, physicists including about 60 fte permanent scientific staff, from Nikhef joined the development and construc- 60 fte PhD students and 15 fte post-docs. The tion efforts of Antares. These efforts have led to technical support is provided through an electronics the first fully operational neutrino telescope in the technology department (about 40 fte), a mechanical Mediterranean Sea. Building on the success of engineering department (about 12 fte), a mechanics Antares, it is now possible to consider the construc- workshop (about 25 fte), and a computer technology tion of the next generation neutrino telescope that department (about 20 fte). In addition, there is a is 20–50 times more sensitive: KM3NeT. The large ongoing GRID activity (about 10 fte) at Nikhef challenge of this project is to develop techniques which has been recently awarded 29 M€ to the BIG that have a substantially improved performance as GRID infrastructure proposal. compared to the presently employed techniques, but at a significantly reduced cost. Nikhef has put In the Nikhef ‘Strategic Plan 2007–2012’*, astro– forward a number of innovative designs and has the particle physics was identified as the major research ambition to build one of the assembly sites. In the prospect next to LHC. The relative budget for astro– longer term, Nikhef will become a major analysis particle physics will grow to approximately 30% by centre for KM3NeT. the year 2015. The strategic plan has been evaluated recently by an international panel under auspices of The present proposal will greatly profit from the the Dutch funding agency NWO. I quote from the excellent infrastructure at Nikhef. There will be report of this panel: “There is no doubt that Nikhef is strong support from the computer technology one of the leading laboratories in experimental par- department which will enhance the chances of ticle physics in the world, with an outstanding record success of this proposal. There is already a team of of achievement in detector and electronics design, scientists actively involved in the analyses of the construction and commissioning, physics analysis data obtained with the prototype Antares detector. As and advanced computing techniques, supported by a these scientists are literally next door, there will be strong phenomenology group.” and “Nikhef has very plenty of opportunity for fruitful exchanges of ideas. quickly established itself in the relatively new field of Last but not least, a significant Dutch contribution to astro-particle physics through its participation in the the construction of the KM3NeT neutrino telescope Antares project, where Nikhef made an immediate is foreseen, for which an investment proposal has and major impact, the benefits of which are just now been submitted. As the construction efforts will being realised.” In order to enable the astro-particle be concentrated at Nikhef, good knowledge of the physics programme to develop so that Nikhef can future detector is at hand. This knowledge is a key become a leader in this emerging field, building upon input to the optimisation of the data-filter system for its established reputation in particle physics and its the KM3NeT neutrino telescope, which is the prime achievement in Antares, the panel strongly endorsed deliverable of this proposal. the request for an increase in the budget. The overall assessment of the institute by this panel was excel- Whereas the scientific challenge of the present pro- lent and has led to a funding increase of the institute. posal forms the main reason for the funding request, there are additional strategic advantages. It permits Nikhef has a long-standing tradition to involve, the widening of the scientific scope of the institute. wherever possible, Dutch industry in the research The proposed research provides a natural extension and development for new facilities and detectors. of the scientific programme of Nikhef and will form Examples are the production of magnet systems for a bridge between particle physics and astrophysics. the HERA collider at DESY, the development of This proposal will link a variety of fields, including very large (10 meter in diameter) superconducting astronomy, dark matter searches, cosmic ray and magnets for the ATLAS detector at CERN and the high-energy physics, increasing the chances of a MEDIPIX project. For the Antares project, the sub- major discovery. marine power system was developed and produced

*K. Huyser et al. FOM publication FOM-07.0927 (2007). 21