Figure 1. History of gamma-ray missions
The key features of the Integral satellite will allow us to study in great detail these, and other sites of key interest in today's astrophysical research, with high sensitivity combined with very good energy resolution and very good imaging capabilities (in order to avoid source confusion) using two main instruments, an 'Imager' and a 'Spectrometer'.
Gamma-ray photons carry the highest form of energy of any electromagnetic radiation known today. Space detectors have measured gamma-rays with energies more than 10 to 12 times that of the photons of visible light. Because of their penetrating nature,i.e.low interaction with matter, these high-energy photons reach us from the most distant (and oldest) arts of the Universe,and they also carry information from closer regions like the Centre of our Galaxy, which are obscured by dense regions of molecular clouds and gas that are opaque at most other wavelengths.
Owing to the high-energy processes that are closely connected with the creation of gamma-ray photons, gamma-ray astronomy provides us with information about some of the most exotic objects in the Universe, such as black holes. Although radiation cannot escape from such an object, we 'see' a black hole by observing radiation from matter that is accelerated to extreme energies by the enormous gravitational field near a black hole, emitting gamma-rays during the process. Massive black holes, more than a million times the mass of our Sun, are believed to power the enormous energy outputs of quasars and other active galactic nuclei.
One of the most striking characteristics of the sky at gamma-ray energies is its dynamic variability. Two exposures of a given region of the sky will, in many cases, give different results. Variability of the gamma-ray emission provides information about the parent object. X-ray novae, for instance, have energy spectra that extend into the lower part of the gamma-ray spectrum. These novae can be observed several times per year and - at their peak brightnesses - are the brightest X-ray sources in the sky. Their intense, hard emission is believed to originate from an instability in the disk surrounding a black hole with stellar mass.This disk contains matter accreted by the black hole from a nearby companion star. Recently, gamma-ray line emission has been detected from such an X-ray nova.
The explosion of a supernova (type II) is another example of a very dynamic event: a star with more than 10 times the mass of the Sun explodes because it has collapsed under its own gravitational pressure after its nuclear fuel has been exhausted. The end result of such an explosion is either a black hole or a neutron star at the centre of the supernova explosion.The layers of the star undergo extreme temperature increases, thereby creating new heavy chemical elements, some of which are unstable isotopes that are characterised during radioactive decay by the emission of gamma-ray line photons.
More than seven years ago, a supernova was discovered in our nearest 'galactic neighbour', the Large Magellanic Cloud. The short (by astronomical standards) distance of 180 000 light years and the availability of modern detectors made it possible, for the first time, to detect gamma-ray lines from such a supernova, providing concrete evidence that new elements are indeed created in this explosion.
The creation of elements in our Galaxy can be traced via the gamma-ray emissions of radioactive isotopes. For example, the isotope '26' of aluminium, having a lifetime of 10 to 6 years, emits a characteristic line with a photon energy of 1.8 x 10 to 6 eV (1.8 MeV), or approximately 1 million times more energy than visible light. The High Energy Astronomical Observatory (HEAO-3) detected this line in 1980, and recently the Comptel instrument on the Compton Gamma-Ray Observatory generated a first map of this mission. Future investigations with Integralmay allow us to clearly identify some of these sites of element building within our Galaxy.
The history of gamma-ray space-science missions is summarised in Figure 1.
In general, the development of new space-science instruments has to fulfil basic boundary conditions, the most important of which are scientific requirements and technical feasibility, the latter also taking launcher capabilities (mass) and associated costs into account.
Detection of gamma-ray photons in the energy range above 1 keV - Integral's energy range extends from 10 keV up to 10 MeV - involves the measurement of the energy exchange (or energy loss) between the photons and the mass of the detector. The scientific advantage of the 'penetrating power' (low interaction with matter) of gamma-ray photons leads directly to difficulty in measuring them with a suitable detector. Mainly, it is the kinetic energy of the incoming photon, transferred to charged particles, which is dissipated and measured in the detector.
The most important energy-loss processes - in the energy range of interest to Integral - are the photoelectric effect (interaction of gamma-ray photons with bound electrons of the detector crystal), pair production (annihilation of gamma-ray photons in the vicinity of a nucleus producing pairs of electrons and positrons), and Compton scattering (elastic scattering of gamma-ray photons with free electrons). The energy dependency of the total cross-section for all three energy-loss processes is dependent on the detector material: for CsI and Ge (as proposed for Integral), the photoelectric effect dominates below about 100 - 300 keV, the pair production process above about 10 MeV, while the Compton effect is the dominating process in between.
Typical detector materials are scintillation substances (e.g. caesium iodide: CsI), where the amount of energy loss is measured through ionisation and scintillation of the detector material (via photodiodes or photomultipliers), and semiconductor materials (e.g. germanium: Ge), where electron-hole pairs are created and registered through external electrical fields. CsI is fairly cheap in large-volume production and has a good light output with 40 keV energy resolution at 1 MeV, thereby making it a good material for imaging telescopes with fairly large detector pixel arrays, where each pixelis produced from a say 1 cm x 1 cm x 3 cm CsI bar. Ge has a superb energy resolution (actively cooled at a temperature of 80 K) of 2 keV at 1 MeV, and is the prime material used for spectroscopy detectors.
Key design parameters for high-energy telescopes have to take into account that the best sensitivity with a minimum of radiation background is required for a mission that will produce first-class scientific results during the first decade of the next century. In the Integral energy range, the signal-to-noise ratio of typical gamma-ray measurements is of the order of 1%, indicating the need for very careful detector design and background reduction in order to extract the maximum of information.
Figure 1. History of gamma-ray missions
The sensitivity of gamma-ray measurements is basically proportional to the background counting rate and inversely proportional to the detector area and length of observing time. Reduction of background noise, increased detector area, and long observations are therefore required to achieve good sensitivity. Long, and uninterrupted, observations can be achieved using high eccentric orbits (see below). The detector area can be maximised, but a minimum thickness of typically a few centimetres of high-atomic-number (i.e.high Z) material is needed to achieve sufficient stopping power: with a diameter of the order of 1 m, this translates immediately into a rather heavy detector array weighing approximately half a ton (including shield; see below).
For background noise, which has to be minimised to achieve good sensitivity, an anti-coincidence (or 'veto') system is usually employed using BGO (Bismuth Germanate Oxide) crystals, a material with high mean Z and high density. The system flags (vetoes) charged particles, but is also used to minimise the effects of gamma-ray radiation not coming from the source under investigation.
For the imaging detector, the key components of the background noise are cosmic diffuse gamma-rays, atmospheric gamma-rays, gamma rays produced in the material of the spacecraft and instrument, cosmic-ray-induced spallation products within the material of the detector, interactions within the detector between locally produced photons and atmospheric albedo neutrons, and events derived from protons trapped within the Earth's radiation belts. Elastic scattering from neutrons is a small effect due to the large mass of the caesium nuclei. Gamma-rays from diffuse cosmic radiation or locally created gamma-rays enter the detector either through the aperture (viewing direction) or leak through the shield.Clearly, a thick and massive shield reduces many unwanted components, but such a massive structure exposed to particles and radiation from space is itself a source of (often time-delayed) secondary photon and particle emission too. In other words, it is important to optimise the shield thickness. For the imager, the anti-coincidence system will be designed using 3 cm-thick BGO and employing an outer ring of the CsI detector pixels as a guard ring.
The key background components affecting the spectrometer detector are: the cosmic diffuse background (mainly through the aperture), neutrons which elastically scatter off germanium nuclei in the detector and transfer recoil energy to the nuclei, neutrons and protons interacting in the detector and producing beta-unstable radio nuclides which decay (delayed), and shield leakage (background from cosmic diffuse gamma radiation and locally produced photons). To optimise the sensitivity of the 350 square centimetres Ge detector area (approx. 8 cm thick), a BGO shield of approximately 7 cm thickness will be used for the spectrometer, resulting in a mass of some 800 kg for the detector assembly (including shield and structure).
To these detector assemblies, coded-mask subsystems (each about 150 kg) need to be added to allow imaging in the high-energy domain. Coded masks in the high-energy range consist of high-Z material (e.g.tungsten) a few centimetres thick, in which holes are arranged in a special pattern in order to produce a 'shadowgram' of the source on the detector array (see below).
Detailed feasibility studies have shown that, in order to achieve the scientific goals set for Integral, two separate instruments need to be designed, one optimised for imaging and the other for spectroscopy. This means, however, that each instrument needs its own aperture and shielding subsystems, which add to the total mass. The combined weight of the two main instruments for Integral- Imager and Spectrometer, including electronics boxes, coolers and harness - has been estimated during the study phase to be about 2000 kg.
Comparison with experiments being flown in space today, namely the Comptel instrument aboard CGRO and the French Sigma telescope on the Russian Granat spacecraft, shows that those single instruments (using NaI crystals) are in the 1300 - 1500 kg range. Earlier studies (e.g. the ESA Grasp candidate mission submitted in 1988) proposed saving some mass by combining the imaging and spectroscopy detectors into one common (and smaller) detector assembly and using common shield and aperture systems. This resulted in a total mass for Grasp of about 1000 kg, but with significant reductions in scientific capabilities, stemming from poor individual optimisation for imaging and spectroscopy (mask pattern, detector pixel size, shielding), and much more complex technical and programmatic interfaces.
In summary, a scientific mission with the objectives that have been set for Integral, and one which must make significant progress vis-a-vis existing or previous missions in the field of 10 keV to 10 MeV astronomy, requires an instrument complement that results in a total payload weight of around 2 tons.
The Integral payload, at about 2 tons,will be the heaviest ever flown on an ESA scientific mission. As overall programme cost tends to be correlated with overall spacecraft mass, it is a considerable challenge to accommodate Integral as a medium size mission within ESA's 'Space Science: Horizon 2000' long-term plan. In order to meet the budgetary constraints of a medium-size mission, the programme policy chosen was one of concentrating development effort on areas specific to Integral and to share that effort with partners.
This approach has resulted in the following programme set-up:
Each Integral scientific instrument is provided, as for all ESA scientific missions, by a collaboration of scientists headed by a Principal Investigator (PI) with national-organisation funding. Although Integral is an observatory-type mission for the science community at large,a certain portion of the observation time is guaranteed to these PIs in return for the instruments that they are providing.
Russia, eager to continue its gamma-ray astronomy activities beyond its Granat mission, have offered to participate in the Integral Programme by providing the launcher (a Proton) in return for observation time. Use of the Proton launcher would enhance the scientific return from the mission as it could place Integral into an orbit with unconstrained observing time. However, due to uncertainty about whether the Proton system will still be operational in 2001, Ariane-5 is being retained as the alternative launcher for Integral.
The USA may contribute one or two additional ground stations to enhance spacecraft cover age. They also intend to participate in the instrument collaborations.
The service module of the Integral spacecraftis a re-build of that developed for the XMM project. This is feasible because the requirements of the two missions in terms of orbit, power, data rates and pointing are very similar. Moreover, the time span of only 1.5 years between the XMM and Integral schedules will allow the necessary continuity in hardware procurement from industry.
As XMM will be launched on an Ariane-5, re-use of the XMM service module also facilitates keeping it as the alternative to the Proton launcher for Integral.
Gamma-ray astronomy can be performed out side the Earth's atmosphere either below or above the Earth's radiation belts. For Integral, a High Eccentric Orbit (HEO ) allowing observations above the radiation belts, i.e.above a height of 40 000 km, has been chosen. The orbit must be optimised for:
The Proton orbit has a period of 72 h, a perigee of 48000 km and an apogee of 115000 km. As this orbit lies entirely outside the radiation belts, the observation time is 100% and the radiation dose is minimal. Fullground coverage can be achieved with three ground stations, and about 85% coverage with two stations.
The Ariane-5 orbit has a period of 24 h, a perigee height of 4000 km and an apogee height of 68000 km. The apogee is in the Northern Hemisphere, to give visibility from a European ground station. The observation time above 40 000 km is 16.6 h per 24 h orbit (69% ). One station provides up to 21 h of coverage. Because the spacecraft is passing through the radiation belts at each perigee, the radiation dose is significant- about one order of magnitude higher than that in the Proton orbit. The high inclination of 65 degrees minimises exposure to the Earth's proton belt, whilst giving good orbit stability. The resulting trapped proton fluence is low compared to that originating from solar flares.
Proton orbit Ariane 5 orbit
Apogee 115 000 Km 68 000 Km
Perigee 48 000 Km 4 000 Km
Inclination 51.6 deg 65 deg
Argument of perigee 270 deg 270 deg
Orbital period 72 h 24 h
Time above 40 000 Km 100% 69%
Ground station Villafranca Villafranca
Goldstone (TBC)
Canberra
Radiation dose 10 krad 150 krad
(4mm Al shielding)
Integral's orbit and mission characteristics are summarised in Table 1.
Figure 2. Integral orbits
The Integral science payload complement, as established during the mission study phase (so-called 'Phase-A'), consists of two main instruments, the Imager and the Spectrometer, and two monitoring instruments, the X-Ray Monitor and the Optical Transient Camera.
These instruments have been carefully chosen to complement each other:
All four instruments are co-aligned and will observe the same region of the sky simultaneously. The measurements of all four instruments will be made available to users as comprehensive data sets for each target.
The Imager, Spectrometer and X-Ray Monitor share a common principle for producing images - all three are coded-mask telescopes. At moderate X-ray energies, it is a still possible to form images using mirrors and standard optics principles, but above an energy of about 10 keV it becomes difficult or even impossible to make suitable mirrors. Coded-aperture imaging offers a means of producing images which avoids this limitation.
With coded-mask imaging, a mask with opaque and transparent regions (the 'coded mask') is placed in front of a position-sensitive photon detector (Fig. 3). A coded-mask telescope is basically a pin hole camera, but with a larger aperture (i.e.many pinholes) in order to cope with the low gamma-ray fluxes. A beam of photons incident on the mask will be absorbed where it strikes opaque areas, thereby casting a shadow onto the detector. For a point source of photons, the registration of the shadowon the detectormay be used to determine the direction of the source. A collection of point sources casts overlapping shadows and, provided the mask pattern is suitably chosen, the directions of the point sources may be determined by correlation techniques. Similarly, a complex field can be imaged by determining the contribution of each pixel in the source to the combined 'shadowgram'.
The coded-mask technique offers a second important advantage for gamma-ray astronomy; namely, in addition to imaging, it provides close-to-ideal background subtraction. In the reconstruction of the intensity from a particular region of the sky, the difference is taken between the signal in all those detector elements that have a clear view of a source at the position under consideration, and that from detector elements for which the source is obscured by opaque regions of the mask. Thus, the detector background is measured exactly contemporaneously, using a subset of the detector elements intermingled with those observing the source.
The detectors must have a spatial resolution matching the resolution of the coded masks and the corresponding shadowgram:
As Figure 4 shows, the payload module consists basically of an equipment platform accommodating the detector assemblies and an empty box supporting the 'upper floor' with the masks at a height of about 4 m.
Figure 3. The operating principle of a coded-mask telescope
Figure 4. Integral's payload configuration
The overall spacecraft configuration resulting from the Phase-A study is shown in Figure 5. It consists of two separate modules, the Service Module (SVM) and the Payload Module (PLM). ESA will be responsible for the overall spacecraft design and the procurement of the SVM and PLM, as described below.
The Service Module provides the following functions:
The configuration of the Integral SVM, developed from that for XMM, is very much driven by the fact that the central area has to house the XMM mirrors.
The Payload Module has to accommodate the Integral payload, but must also interface with the SVM. The resulting overall PLM configuration is shown in Figure 6. It consists of:
Simplicity of the interface between the SVM and the PLM has been a major design driver. The electrical interface is basically reduced to a power bus and a data-handling bus. The modular approach has been conceived to allow parallel development, assembly, integration and testing of the SVM and PLM.
The main characteristics of the Integral spacecraft are summarised in, Table 2.
Figure 5. Overall Integral spacecraft configuration
Figure 6. The Integral Payload Module - Exploded view
Table 2. Integral spacecraft's characteristics
As shown in Figure 7, the ground segment consists of three main elements: the Integral Science Operations Centre (ISOC), the Mission Operations Centre (MOC), and the Integral Science Data Centre (ISDC):
ESA is responsible for the ISOC and the MOC. The ISDC will be the responsibility of the ISDC Principal Investigator and his collaborators.
The various phases in the implementation of the Integral Programme are shown in Figure 8.
The Integral project is presently issuing the Announcement of Opportunity (AO) for the payload instruments, the Science Data Centre and the mission scientists. In about a year's time, immediately after final payload selection, the Invitation-to-Tender (ITT) will be issued to Industry. Phase-B is planned to start in the first half of 1996, leading to the commencement of Phase-C/D in mid-1997.
The first half of Phase-C/D will concentrate on the structural/thermal model, in parallel with the engineering model. The second half will be devoted to the flight model, culminating in the launch of Integral in the first half of 2001.
The nominal mission duration will be two years. Consumables and life-limiting items will be sized to allow extension of the mission to five years in total.
Figure 7. The Integral ground segment
Figure 8. The various phases in the Integral Programme
ESA Bulletin Nr. 79.