INTRODUCTION
X-ray astronomy primarily involves the study of plasmas with thermal
temperatures in the range of 1E6 to 1E8 K. Such plasmas radiate the
bulk of their energy in the X-ray regime, from ca. 0.1 to 10 keV.
XMM works in the energy range from 0.1 to 15 keV. Apart from X-ray
continuum emission, produced through processes such as e.g. thermal
bremsstrahlung, a significant fraction of the total emissivity of hot
thermal plasmas can arise from line emission. At the high temperatures
mentioned above, abundant elements in cosmic gases, such as hydrogen
and helium, are stripped of all their electrons. Only heavier elements
can, depending on the temperature, retain their K or L shell electrons.
Amongst the most prominent X-ray line emitting elements in the XMM
passband are Iron, Oxygen, Magnesium, Sulfur, Silicon, Sodium, Calcium,
Argon, Neon and Nickel. The study of transitions from these elements
which are primarily in a hydrogenic or helium-like state (i.e. with
either all or all but one electrons left in their outermost shell),
represents an important diagnostic tool for an understanding of the
physics of cosmic X-ray sources.
XMM's science objectives are described here in two parts:
GENERAL SCIENCE OBJECTIVES
XMM is designed specifically to investigate in detail the X-ray emission
characteristics, i.e. the emission distributions, the spectra and the
temporal variability, of cosmic sources down to a limiting flux of order
1E-16 erg/(s cm²). With its high throughput and moderate angular
resolution, XMM is extremely sensitive to low surface brightness X-ray
emission. Some astronomical sources (see below) are prominent X-ray
emitters, but faint or even invisible in other parts of the electromagnetic
spectrum. Therefore, high-quality X-ray observations of these objects are
very important and cannot be replaced by data obtained through other
observing techniques. Instead, X-ray observations supplement data from
other wavebands, leading to a more complete picture of the universe.
Other objects are bright not only in the X-ray, but also in other parts
of the spectrum, e.g. the optical or UV. Due to internal reprocessing,
some sources emit both X-rays and other photons, but the optical or UV
emission sometimes lags the X-ray light. Therefore, to further broaden
the scope of the investigations, the Optical Monitor (OM) onboard XMM
offers the possibility to simultaneously study the optical/UV properties
of the observed X-ray sources. The basic characteristics of XMM's X-ray
telescopes are a 6" (FWHM) point-spread function, a 30' field of view,
spectroscopic resolution (E/dE) in the range from a few ten to several
hundred and a large effective area of 4650 cm². A more detailed
description of XMM is provided in the XMM
Users' Handbook.
X-ray observations can be conducted in different ways, depending on the
scientific goals of the investigator. XMM has different science instruments,
each of which can be operated in different modes so that the observations
can be tuned to the scientific needs. The basic observing techniques are:
In addition, with its Optical Monitor (an optical/UV telescope mounted
parallel to the three X-ray telescopes), XMM can perform another basic
task:
Imaging
XMM carries telescopes with CCD cameras in their focal surfaces which
can image the X-ray sky with very high sensitivity and good angular
resolution. This way, "pictures" of the sky as seen at X-ray energies
are created. Similarly, images of the optical/UV emission from the
same regions on the sky can be obtained contemporaneously with the OM.
X-ray spectroscopy
The same X-ray CCD cameras that are used for imaging can also register
the energy of incoming X-ray photons. Therefore, radiation can also be
analysed with respect to its spectral characteristics within the XMM
passband (from 0.1 to 15 keV energy). The spectral resolution of the
CCDs is only moderate (but as good as ASCA's!) and does not reveal the
full complexity of many X-ray spectra. Therefore, XMM carries a different
type of spectrometer, with much higher spectral resolution for very
detailed studies in the 0.35 to 2.5 keV energy range, so-called
"Reflection Grating Spectrometer"s (RGSs). The OM offers grisms for
simultaneous low-resolution optical or UV spectroscopy.
Photometry
The time of each photon's detection within the X-ray detector, in
addition to the direction and energy (as in imaging mode), can be
registered. This allows observers to perform studies of the homogeneity
or variability of X-ray sources over time by counting how many events
were registered over short time intervals. Operating the OM in its
fast mode, the arrival times of individual optical or UV photons can
be registered in the same fashion, thus allowing for comparative
timing studies.
Optical identification of X-ray sources
It is a general goal of X-ray missions to detect and identify X-ray
sources on the sky. XMM has the Optical Monitor (OM) onboard for
contemporaneous X-ray and optical/UV observations. Both, the X-ray
telescopes and the OM, are very sensitive and capable of detecting
faint sources. However, since the pointing of satellites is not
always perfectly accurate, it is sometimes difficult to determine
unambiguously which X-ray emitting objects that might
be visible on optical images of the sky have actually been observed.
XMM has good X-ray imaging capabilities, with a width of the point-spread
function's core of only 6". Together with good pointing reliability,
this will make sure that the X-ray and optical images will be
well-aligned, making it easy to identify sources in the field of
view by comparing the images from the different instruments.
PROMINENT X-RAY EMITTING CELESTIAL SOURCES
The above observing techniques can be used to study many different types
of X-ray sources on the sky with XMM. We outline below, in very general
terms, some of the areas in which XMM can help us make progress in the
observation - and thereby our understanding - of celestial objects.
Cosmic X-ray background radiation
The question how much of the observed "diffuse" X-ray background
radiation comes from discrete sources and which fraction of it is
truely diffuse is the matter of a long-lasting debate. The technique
to investigate the nature of the extragalactic X-ray background is to
obtain extremely sensitive observations of an "empty" field, i.e. a
"deep field" pointing.
In this example we show a simulation of a 200 ks XMM EPIC observation,
performed by D. Lumb.
With extremely long XMM observations a limiting flux of order 10E-16
erg/(s cm²) can be reached (with a PSF of 6"), compared to ca.
10E-15 erg/(s cm²) (over a 35" beam) with ROSAT. Therefore, a
much lower confusion limit will be reached on the search for discrete
sources in the early Universe. This will render it possible to extend
current log(N)-log(S) plots (of the number of sources, N, found in a
certain flux interval [S to S+dS]) for population studies over cosmological
time scales to much fainter flux limits than before.
Typical XMM pointings will be quite long (of order several ks to several
10 ks). This, together with XMM's large photon collecting area, will make
average X-ray pointings very sensitive observations. Thus, significant
results on the different sources contributing to the X-ray background
will also come from studies of serendipitous sources in XMM pointings.
Like this, good source count statistics can be achieved over large areas
on the sky. In this context it is also important to note that XMM will
observe up to 15 keV, where the optical depth for incoming X-ray radiation
is lower than, for example, in the ROSAT band.
Elliptical galaxies and clusters of galaxies
The distribution of the gravitationally heated hot gas in elliptical
galaxies and clusters, under the assumption that it is approximately
in hydrostatic equilibrium, reflects the distribution of mass, i.e.
the shape of the gravitational potential of these systems; the gas
temperature is a measure of the depth of the potential well in which
it is confined. The typical temperatures, namely just below 1 keV
(indicating equilibrium temperatures of order 10E7 K) for single
ellipticals, kT = 1...2 keV (1 to 2x10E7 K) for poor clusters and
values in the 2 to 10 keV range for rich clusters (2x10E7 to 10E8 K)
result in a maximum emissivity in the XMM band. Typical X-ray
luminosities range from ca. 10E41 erg/s for individual elliptical
galaxies to 10E45 erg/s for rich clusters. Thus, rich clusters belong
- together with AGNs and QSOs - to the most luminous X-ray sources in
the Universe.
XMM observations can be used to study several key properties of the hot
intracluster medium. Spatially resolved spectroscopy will allow the
determination of the radial variations of the gas density, temperature
and metallicity. The knowledge of metallicities is important in the
context of the chemical evolution of galaxies and gas in clusters.
Central cooling flows, with decreasing gas temperatures towards the
centres of the clusters, might lead to the accretion of relatively
cool gas in the nuclear regions of elliptical galaxies and clusters.
Such material potentially
feeds the central engine and/or relates to present-day star formation
in these systems. X-ray observations can also be used to map the
distribution of hot gas and thereby the distribution of matter in
elliptical galaxies and clusters. Up to 30% of the total mass of
galaxy clusters has been identified as X-ray emitting intracluster
gas. This is a significant fraction of the formerly "missing" mass
and XMM offers a possibility to trace even fainter emission than
previous satellites (and thus more gas and thereby mass).
In some cases, e.g. the Perseus cluster, X-ray observations can also
be used to study the interaction of a jet emanating from a radio core in
a massive elliptical galaxy with the ambient hot gas.
XMM's sensitivity will allow us to observe clusters out to cosmologically
relevant redshifts (z>1), rendering possible investigations on the
cosmological evolution of galaxy clusters and their central galaxies.
As an example, we display here a simulation
of EPIC spectroscopy of the galaxy cluster Abell 2199, calculated
by H. Siddiqui with the XMM Science
Simulator (SciSim) software.
Normal and starburst galaxies
External galaxies host several kinds of objects that are bright enough
in the X-ray regime to be observable for modern X-ray observatories such
as XMM. Among these, Seyfert-1 nuclei are the potentially brightest sources
(see below). Other bright sources are high-mass X-ray binaries (HMXRBs;
also below).
HMXRBs are the most likely candidates for the brightest unresolved sources
detected outside the nuclei of external galaxies so far. XMM offers a
sensitivity that will allow for spectroscopic studies of individual bright
X-ray binaries in galaxies out to the distance of the Virgo cluster (ca.
15 Mpc). Another prominent contributor to the X-ray emission of galaxies
is the hot phase of their ISM. Starburst galaxies in particular have
relatively high X-ray luminosities (of order 10E41 erg/s), of which a
substantial fraction comes from diffuse hot gas. Detailed spectroscopy
of this gas will allow us to assess the abundance of various elements
in different areas within these galaxies and thereby get a handle on
their chemical composition and evolution. XMM, with its spatial resolution
of 6", allows the observer a much better discrimination between compact
sources and diffuse emission in galaxies than had been possible before.
The above implies that XMM will lead us away from integral spectra of
entire galaxies in the past to dedicated X-ray studies of various
individual X-ray sources that they comprise.
Investigations of the hot ionised medium in galaxies are not restricted
to their disks - soft X-ray emission has been detected from the halos of
up to now about 20 nearby edge-on galaxies. Most of these are starburst
galaxies (with very high star formation rates), but also galaxies not
classified as starbursts can have enough power to heat their halos to
X-ray temperatures (e.g. NGC 891). The hot gas produced in large numbers
of type II supernovae can in some cases be shown to reach escape velocity.
Therefore, starburst superwind outflows are a likely source for metal
injection into intergalactic space. Since many galaxies in the past
(at z = 0.6 to 0.7) show extremely blue colours indicative of massive
star formation, the rate at which metals were expelled into the
intergalactic medium might have been much higher in the past than it
is now. This could contribute significantly to the detected metal
content of absorption line systems on the sightlines towards distant
quasars. XMM will help us to put much tighter constraints on various
ascpects of this scenario, including e.g. the metallicity of gas leaving
galaxies.
Active galactic nuclei and quasars
As in many other parts of the electromagnetic spectrum, most AGNs and QSOs
are also bright in the X-ray regime. In the framework of the unified scheme
of AGNs, it is believed that hard X-ray emission (typically
with a power law spectral slope) from Seyfert 1 systems arises from the
nuclei. Additional radiation might come from the broad-line regions (BLRs)
encompassing the cores. In the same theory, the emission of Seyfert 2
nuclei is thought to be composed of heavily absorbed continuum radiation
from the nuclei and, superimposed on this, scattered nuclear continuum
flux from material close to the rotation axis of the massive central
object that is re-emitted at lower energies. The difference between the
two types of Seyfert nuclei in this picture is that Seyfert 1 cores are
directly visible (in viewing geometries close to face-on), while Seyfert
2's are heavily obscured by the molecular tori around their central black
holes (i.e. viewed at higher inclination angles). BL Lacs are believed to
be face-on Seyfert 1 systems with their beams directed exactly towards us,
which gives rise to the observed hard and highly variable X-ray radiation.
BL Lac spectra lack line emission, probably because little reflected
radiation is observed in these viewing geometries.
Current X-ray observations suffer from weaknesses in spectral resolution,
limited width of energy passband, lack of simultaneous optical/UV data,
insufficient timing resolution or combinations of these. XMM will allow
for spectroscopy of X-ray sources with a spectral resolution of ca. 3.5
eV at 1 keV energy (RGS, -1. order) and even ca. 1.5 eV in the -2. grating
order. This, in conjunction with XMM's large effective area and its
well-sampled line spread function (LSF), will enable users to obtain high
signal-to-noise spectra in the energy range from 0.35 to 2.5 keV (together
with moderate resolution imaging spectroscopy over the 0.1 to
15 keV band) for detailed investigations of the diagnostic lines in the
spectra of AGNs. The fact that the XMM bandpass reaches to an energy of
15 keV is important, because one can penetrate very deep into the central
area of an AGN at high energies due to the low absorption cross section
of high-energy X-rays in interstellar matter.
For research on AGNs and QSOs it is of particular importance that sources
can be observed simultaneously with the Optical Monitor (OM), which allows
for studies of the time lag between the emission of (primary) X-ray and
(secondary) optical/UV radiation, a quantity that will provide crucial
information on the internal radiation transport, i.e. the re-processing
of the emission coming from the nuclear area.
Stellar black holes, neutron stars, pulsars, binary stars
Accreting systems in general are potentially bright X-ray sources. This
is not restricted to massive black holes in the centres of galaxies, but
applies to stellar accretion systems as well, as e.g. cataclysmic variable
stars (where mass accretion from a low-mass companion is occurring onto a
white dwarf) and X-ray binaries. In X-ray binaries, the accreting object
is either a neutron star or a black hole and the mass flow comes from a
companion star. The X-ray luminosities can be extreme, up to of order
10E39 erg/s. This number applies to high-mass X-ray binaries (HMXRBs),
i.e. black hole candidates, which thus can produce a substantial fraction
(up to some 10%) of the total observable X-ray flux of an entire galaxy.
The spectra of X-ray binaries have approximately power law shapes with a
high energy cutoff around 10-20 keV (which distinguishes them from AGN
power law spectra) and 6.4 keV Fe K alpha line emission. The Fe line
emission is a diagnostic to investigate the matter encompassing a binary
star (typical column densities, i.e. numbers of atoms/ions along the
line of sight to an emitting source, lie in the range from 10E22 to
10E23/cm²). Fe 6.4 keV
fluorescence emission with a high equivalent width but low line width is
expected from relatively cool matter and Fe at relatively low ionisation
stages, while 6.7 keV emission arises from radiative recombination of
H-like Fe followed by cascade processes in relatively hot matter. XMM,
with its good energy resolution and simultaneous high sensitivity, will
be very well-suited to conduct such studies into great detail, thus
revealing the internal geometry of binary systems.
Another potentially interesting area of future XMM research lies in X-ray
timing analyses of binary stars. X-ray emission is observed e.g. from
pulsars in binaries. Pulse periods range from the millisecond regime up
to ca. 15 minutes. One extraordinary feature of X-ray pulsars, e.g., is
that they are "spinning up", i.e. decreasing the length of their pulsation
period over time, while radio pulsars are invariably spinning down. The
pulse profiles of pulsars contain information not only on the geometrical
configuration of these systems, but also on the accretion column near the
magnetic poles of the neutron stars. The EPIC p-n camera offers extremely
high time resolution for studies of rapidly variable sources. Since the
changes in pulse profiles are accompanied by changes in the energy spectra,
it is argued that anisotropic radiation transfer (Thomson scattering) must
play an important role in pulsars. Again, much information will also be
contained in the Fe emission lines, because the 6.4 keV radiation is caused
by fluorescent re-emission of continuum emission absorbed by relatively
cool matter surrounding the central source. Thus, X-ray observations can
help in various ways to properly constrain radiation transfer calculations.
New, improved X-ray observations will also contribute to more accurate
mass determinations of neutron stars, which can be compared with radio
measurements to see whether X-ray data can corroborate the "canonical"
value of 1.4 solar masses.
Supernova remnants
Among the most prominent Galactic X-ray sources are supernova remnants
(SNRs). X-rays are emitted from the hot gas itself and also when the
expanding gaseous shell of the SNR hits and shock-ionises the ambient
medium. X-ray observations, in particular imaging spectroscopy with
good spatial and spectral resolution and high-resolution
spectroscopy, will provide details of the ongoing line emission processes
(thermal vs. nonthermal emission), thus tracing the interaction of expanding
shell and ambient medium. Earlier ROSAT observations have, for example,
for the first time been used to investigate so-called "bullets" of gas
preceding the general shell on their way into the surrounding medium.
It is most likely that the expansion of this material has been locally
faster than in its surrounding due to the existence of cavities in the
ambient gas.
With temperatures in the range of a few tenths of a keV to a few keV,
SNRs emit the maximum of their thermal radiation in the XMM bandpass.
A wealth of lines from elements, such as Fe, O, Mg, S, Si, Na, Ca, Ni,
Ne and Ar, can be expected, primarily in the energy range from ca. 0.5
to 2 keV, but also at higher energies (Fe K alpha at 6.4 keV). Detailed
studies of these lines, revealing the temperature and ionisation
structure of the SNRs, are first-class diagnostics of the nuclear fusion
processes going on in the SN precursor stars, of the metal enrichment
of interstellar matter via SNRs and of the importance of shock heating.
SNRs are also believed to contribute measureably to the total soft X-ray
spectra of external galaxies. Extragalactic supernovae most likely to be
seen in the X-ray regime (types II and V) have massive progenitors.
However, X-ray observations of extragalactic supernovae and their remnants
are still sparse. This is caused by a lack of both angular resolution and
sensitivity of previous missions, two areas in which XMM will make
significant progress.
The hot phase of the Galactic ISM
Early soft X-ray all-sky surveys, in particular that conducted by the
University of Wisconsin in the 1970'ies, indicated that the intensity of
diffuse soft X-ray radiation received from high Galactic latitudes is
higher than within the disk. This led to the discovery of a soft X-ray
halo in our Galaxy. Most of the halo emission is irradiated at an energy
of about 0.25 - 0.3 keV, similar to the temperatures of halos in external
galaxies. Galactic studies, despite our unfavourable location and viewing
geometry from within the disk (and therefore also within the X-ray emitting
medium), offer the chance to study the properties of the hot gas in great
detail, like e.g. its volume filling factor, total energy content and its
distribution with respect to the other phases of the ISM.
ROSAT for the first time provided a detailed view of the Galactic X-ray
halo. However, many questions remained unanswered. XMM offers us a chance
to revisit particularly interesting objects with much improved sensitivity
and resolution in both the spatial and the spectral regime. Such data
will provide better constraints for radiation transport calculations,
which will shed new light on the composition of the ISM and its total
energy balance, including the sources of ionisation. The metallicity and
energy balance of hot gas in the Galactic halo are key parameters in
investigations of the chemical evolution of our Galaxy via metal enrichment
of the Galactic ISM by stellar winds and supernovae and the distribution of
metals via disk-halo interactions (so-called Galactic "fountains" or
"chimneys", depending on outflow velocity).
Cool gas
Not only the hot phase of the ISM can be studied, but also the properties
of cooler gas (with temperatures far below 1E6 K). This material
absorbs X-rays quite efficiently. The absorption characteristics of the
gas can be used to study, e.g. its metal content. XMM will be able to
measure absorption edges due to intervening gas on the sightlines to
background X-ray emitters, thus helping us to understand the physics
and chemistry of the foreground material. The depth of an absorption
edge is a direct measure of the abundance of the element causing the
edge. The total amount of gas along the sightline can also be determined
by fitting an equivalent HI column density to the low energy part of
observed X-ray spectra, which suffer from absorption losses on the way
to the observer, primarily at energies below 0.4 keV. This is one kind
of investigation in which sources are not studied in emission,
but absorption.
Stellar coronae
While it had been known for quite a while that massive stars, primarily of
spectral types O and B, are X-ray sources, it has been found only by recent
satellite missions that stars of almost all spectral types emit X-rays.
The bulk of the X-ray emission is thought to come from their million
degree coronae, which are thin plasmas approximately in collisional
equilibrium. However, stellar X-ray luminosities are low, making them
invisible outside our own Galaxy. Stellar X-ray luminosities are of order
10E26 to 10E31 erg/s for late-type stars (spectral types F to M) and ca.
10E29 to 10E34 erg/s for the above-mentioned early-type stars (O and B).
With its extreme sensitivity, XMM will make many more stars (probably
several thousand) accessible to X-ray observations than any other previous
mission. Important questions to be addressed by such observations are, e.g.
the heating mechanism of the coronae (which is as yet unknown), and various
aspects of the interaction of the hot coronal plasma with the stellar
magnetic field. XMM will enable both more detailed studies of individual
stars as well as better statistical studies of much larger samples than
available at present.
A SciSim simulation of an RGS spectrum of the nearby star Capella gives
an impression of the wealth of lines and the spectral resolution and
sensitivity to be expected.
Comets
ROSAT and SAX observations for the first time revealed that comets are
sources of X-ray emission. In fact, it is currently not understood which
mechanism dominates the unexpectedly high observed fluxes observed, e.g.
from comets Hyakutake and Hale-Bopp. In the case of Hale-Bopp, evidence
has been found that the X-ray emission arises from an interaction of
Solar X-rays with cometary dust. In other cases, however, scattering of
Solar X-rays and fluorescent emission in the coma cannot account for the
observed flux densities (falling short of the observed values by about
factors of 10³ or more) and the source geometries. Some X-ray
emission was found to be offset from the nuclei of the comets, along a
line normal to the Sun-nucleus line, which is inconsistent with models
of pure reflection of Solar radiation. Therefore, although scattering of
Solar X-rays by dust appears to play an important role in one observed
event in Hale-Bopp, no general solution as to the origin of
X-rays from comets is known yet.
XMM observations will help in the future to observationally confirm which
of the several currently controversial theoretical models describes the
X-ray emission of comets correctly. Given the extremely limited amount
of solid information available in the literature at the moment, one can
expect XMM to perform pioneering work in this field. One should bear in
mind, however, that XMM cannot track the fast movement of comets once
they are nearby. Instead, one will need to let the source pass through
XMM's field of view.
Summary
The above indicates that, within the rapidly evolving field of X-ray
astronomy, XMM will be a major milestone. Its expected versatility in
astronomical research is based on its very high sensitivity, good optics
and innovative detectors. Compared to the current generation of X-ray
satellites, XMM will offer improved capabilities in all three general
observing techniques (imaging, spectroscopy and photometry), with the
additional advantages of a wide energy passband and simultaneous
optical/UV observations.