Search and X spectroscopy of Compton Thick Active Galactic Nuclei in the 4 Ms Chandra observation of the CDFS

Chapter 1
Introduction
1.1 Active Galactic Nuclei: toward a definition
The term “Active Galactic Nuclei” (AGN) is used to describe a wide variety of
phenomena of non-stellar origin observed in the nuclei of many galaxies. About 1%
of the known galaxies in the local universe shows some of the following properties:
(a) nuclear luminosity greater than or equal to that of the host galaxy
(L
bol
= 10
41
− 10
48
erg s
−1
)
1
;
(b) fast variability (Δt∼seconds-years) that entails compactness
(R∼ 10
12
− 10
18
cm);
(c) broad band emission (from the radio waves to the gamma rays);
(d) strong emission lines in the optical/ultraviolet/X ranges;
(e) powerful radio emission (P
5GHz
> 10
25
W Hz
−1
sr
−1
) for about 10% of the
AGN.
The reason for the reference to the galactic nucleus is that the non-stellar emission
comes from the central galactic regions, point-like when observed
2
. Galaxies that
shows at least some of these features are called "active galaxies". In addition to
general features, the nuclear activity is shown in several ways. Different peculiarities
are used to classify AGN. Every class is distinguished thanks to particular properties
in different bands of the spectrum. The subclasses set up the so-called AGN
"taxonomy".
1.1.1 Taxonomy
AGN are classified according to the radio emission, the luminosity, the features
of the optical (OPT) and/or the ultraviolet (UV) spectrum and the X spectrum
morphology. The first AGN division was made by the intensity of radio emission,
due to the first AGN identification as powerful radio-emitter (Baade & Minkowski
1
For comparison, standard galaxies have L
bol
< 10
42
erg s
−1
.
2
Angular dimensions lesser than the more powerful available telescopes resolution: ∼ 0.5
00
for
Chandra (X-Ray),∼ 0.1
00
for Hubble Space Telescope (Optical),∼ 0.001
00
for the Very Long Baseline
Interferometer (Radio).
1
2 1. Introduction
Figure 1.1. Composite quasar spectrum on logarithmic scale. The spectrum is fitted with
two different power-laws according to the considered range (see § 1.3 in this
work). Data from the Sloan Digital Sky Survey.
(Figure from Vanden Berk et al. 2001)
1953): defining the radio-loudness parameter as R
L
= log(F
5GHz
/F
B
)
3
, if R
L
< 1
the object is classified as Radio Quiet, otherwise Radio-Loud (Kellermann et al.
1989). Moreover, under a radio point of view, another classification is assumed
according to the ratio R of the distance between the two brightest spots and the
overall size of the radio image (Fanaroff & Riley 1974). Consequently, FRI indicates
radio-loud galaxies with R < 0.5 and FRII stands for radio-loud galaxies with
R> 0.5. Luminosity classification is based on two parameters: the magnitude in
the B band (m
B
) and the X luminosity (L
x
). In the local Universe, AGN with
m
B
>−23 eL
x
< 10
44
erg s
−1
, are identified as Seyfert galaxies (Sey), while objects
with larger L
x
are called Quasi-Stellar Objects (QSOs). AGN with typical Seyfert
galaxies luminosities and observed atz> 1 are classified as Seyfert-like objects. AGN
with luminosities lesser than Seyfert galaxies ones are called Low Luminosity AGN
(LLAGN). A further subdivision is based on the OPT-UV spectral lines observation.
The AGN spectrum in the OPT-UV band shows either permitted lines (Hα,Hβ,Hγ)
and forbidden ones ([OIII], [OII], etc.). Permitted lines are observed either as broad
lines (FWHM = 10
3
− 10
4
km s
−1
) and narrow lines (FWHM < 10
3
km s
−1
), while
forbidden lines are observed as narrow lines only (see Figure 1.1). If the OPT-UV
spectrum shows only narrow lines, either permitted and forbidden, the AGN is
classified as Type 2; instead, if either narrow and broad lines are present, the AGN
is classified as Type 1 (i.e., see the Figure 1.2). An intermediate classification can be
considered (Osterbrock 1981). The classification according the X spectrum is based
on the amount of matter along the line of sight. The presence of gas and dust, indeed,
causes absorption. This is proportional to the matter column density N
H
of neutral
molecular hydrogen. WhenN
H
< 10
21.5
cm
−2
the AGN is classified as “unobscured”;
3
F5GHz is the radio flux at 5GHz and FB is the optical flux in the B band, centered on the
wavelength λ = 4400Å.
1.1 Active Galactic Nuclei: toward a definition 3
Figure 1.2. Spectra for the two types of Seyfert galaxies. Emission lines for the Type I
Seyfert galaxies are wider than that for Type II ones.
(Image adapted from Bill Keel’s website)
if 21.5<logN
H
< 24 the AGN is classified as “obscured” (Compton thin); finally, if
N
H
> 10
24
cm
−2
the AGN is classified as "Compton thick (CT)". In fact, in the last
case, the column density reaches values of the order of the inverse of the Thompson
cross section (σ
T
= 1.5× 10
24
cm
2
). This returns the maximum probability of
Compton scattering, blocking the hitting radiation to propagate outward.
More exotic AGN
For the sake of completeness, a short explanation about other classes of AGN
will be given. These are indicated as the more exotic AGN due to their scarcity, and
even to their extreme manifestation of activity.
• Low-Ionization Narrow-Lines Region (LINER): they are characterized by [O
II]3727Å/[OIII]5007Å≥ 1 and [O I]6300Å/[OIII]5007Å > 1/3 (Heckman
1980). Most of the nuclei of nearby galaxies are LINER. They are the weakest
form of activity in the whole AGN taxonomy.
• Broad Absorption Lines (BAL) QSOs: they are otherwise standard QSOs
with deep blue-shifted absorption lines in their spectrum. These particular
absorption lines correspond to resonance lines of [C IV], [Si IV], [N V]. The
entire population of BAL is observed at high redshift (z≥ 1.5): at those
redshift they constitute about 10% of the observed AGN population.
• Blazars: they are the most violent manifestation in the AGN classes. Their
spectra show emissions extending to gamma-rays and very brief variability time-
scale. This class in turn encompass two main subfamilies: the Optically Violent
Variable (OVV) and theBL Lac objects (from the constellation where the
prototype of these objects was observed). Whole Blazar population accounts
for only∼few% of the entire AGN population.
4 1. Introduction
(a) Images of the AGN NGC4261 from
ground based facilities and Hubble Space
Telescope. Radio jets and the galactic
nucleus are showed.
(b) Same of (a) for the AGN NGC4261.
(c) A composite image of the giant elliptical ac-
tive galaxy Centaurus A.
(d) Images of the barred active galaxy
NGC1097 from 2MASS Atlas Im-
age Mosaic.
(e) Cygnus A radio source at 6 cm wavelength; the central nucleus
and radio jets with lobes are showed (Figure from Perley,
Dreher, & Cowan 1984).
Figure 1.3. Exposure of various Active Galactic Nuclei in different observatve bands.
1.2 The AGN engine 5
1.2 The AGN engine
Observing the nuclear emission variability on very short time scales (through
Reverberation Mapping techniques, Peterson 1994) implies very compact
objects in the middle of AGN. Measures of the characteristic quantities, e.g.,
bolometric luminosity (L
bol
), time variability (Δt), and nuclear region sizes
(R), gives the following results (in order of magnitude):
L
bol
= 10
41
− 10
48
erg s
−1
Δt∼s−yr
R =cΔt∼ 10
12
− 10
18
cm
If a central SMBH is hypothesized, than the characteristic radius can be chosen
as the Schwarzschild radius, that is:
R = 2
GM
c
2
(1.1)
so it is possible to invert (1.1) to obtain:
M =
R
2
c
2
G
that can be written in terms of the measured quantity R =cΔt to return a
valuation of the central mass:
M =
(cΔt)
2
c
2
G
= 10
6
− 10
9
M
 (1.2)
In 1964 Salpeter used informations about dimensions (compactness) and mass
characteristic of these exotic sources to elaborate the so-called “Supermassive
Black Hole (SMBH) paradigm”. According to this paradigm, the central engine
of every AGN is a SMBH accreting matter through an accretion disk around
the compact object (Shakura & Sunayev 1976). The fundamental process in
producing luminosity is the transformation of gravitational potential energy.
As a consequence, it’s possible to write the extractable energy from a given
mass M as E =Mc
2
. So the energy emission rate (that is, the luminosity L)
is given by:
L =
dE
dt
= ˙
Mc
2
with
˙
M =
dM
dt
(1.3)
The secret of accretion as powerful engine is the process capability to convert
potential energy in luminosity more efficiently with respect to other energy
production mechanisms (e.g., the thermonuclear processes). For that, the
potential energy of a mass m placed at distance r from the source of the field
(of mass M), is:
U =
GMm
r
(1.4)
As before, luminosity can be written as the time derivative of energy, that is:
L∼
dU
dt
=
GM
r
dm
dt
=
GM
˙
M
r
(1.5)
6 1. Introduction
Figure 1.4. Illustration of BLR sizes derivation from causality arguments. Time variability
observed in BLR clouds (supposed circular in shape) implies the derivation of
the emitting region. Surfaces of constant time delay, or isodelay surfaces are
also showed, each one labelled with the time delay (in units of the shell radius
r) observed with respect the continuum source: points along the line of sight
to the observer are seen to respond with zero time delay, while the farthest
point on the shell responds with a time delay 2r/c. (Figure from Peterson et
al. 2006)
From these equations comes out that the conversion rate  is proportional to
the term (M/r), that measures of the system compactness. Black holes comes
to play as the most compact known objects. The characteristic dimension of a
black hole, that is the Schwarzschild radius, can be written asR
s
∼ 10
13
M
8
cm
(see 1.1) where M
8
is the black hole mass in 10
8
M
 unit. Considering that
the last stable orbit for a black hole is at 3R
s
, the extractable energy from a
particle of mass m that goes down until the last stable orbit is
4
:
U =
1
2
GMm
3R
s
=
1
2
GMm
(6GM/c
2
)
∼ 0.083mc
2
The SMBH hypothesis is even more interesting when rotating black holes are
considered, because it results in a huge efficiency ( ∼ 0.47, around 50%!)
when the Kerr metric is used.
Nowadays, there are a number of strong observative evidences of SMBH in the
central regions of galaxies (either active and quiescent):
(a) measures of the star motion around massive nucleus in the Galaxy (i.e.,
Eckart & Genzel (1996));
(b) relativistically distorted emission lines profiles (Fabian et al. 2002);
(c) flux measures able to distinguish between accretion on compact objects
with solid surface (e.g., neutron star) and accretion of matter that crosses
an event horizon (Broderick & Narayan (2006)).
1.2.1 The lines Regions
The strong emission lines observed in the OPT/UV range of the AGN spectra
(with the exception of BL Lac objects) show widths really larger than the
4
For comparison, it’s useful to consider that the efficiency of hydrogen thermofusion in helium is
∼ 0.007.
1.2 The AGN engine 7
Figure 1.5. Diagnostic diagram for spectral classification of AGN and star-forming galaxies.
The orange continuous and dashed lines represent the empirical and the
theoretical division between different galaxies types, respectively. The gray
scale presents the locus of∼ 400, 000 SDSS galaxies. (Figure from Kriek et al.
2007)
usual observed ones (e.g., in star-forming galaxies). Chemical species identified
by these lines suggest that there is matter in orbit around the active nucleus,
and that it’s processed by the incident radiation. If the Doppler effect is
the only reason for broadening of emission lines (so that the emitting region
velocity is known), and the central mass is known (see §1.2), by virial arguments
(M∼v
2
R)it’spossibletoobtainthedistanceofthese“lines-productor”regions
from the nucleus. Following this scenario, the observed lines are emitted from
clouds of matter orbiting around the massive nucleus at different distances from
thecentreandwithdifferentdensities(suchthateitherpermittedandforbidden
lines are justified), velocities (to justify different widths) and ionization degrees
(because lines of the same element are observed in various ways). The general
scheme defined by the observations (see above) places the so-called Broad Lines
Region (BLR) very close to the galactic centre (∼ 1pc), with electron density
n
e
∼ 10
[9,10]
cm
−2
and velocities between [10
2
− 10
4
] km s
−1
, while the narrow
prohibited lines are produced in a more distant (∼ 100 pc) and much less
dense (n
e
∼ 10
5
cm
−2
) region, orbiting with smaller velocities [10−10
2
] km s
−1
,
called Narrow Lines Region (NLR). Until now the size of the BLR could not
be directly measured and the only available information is provided by the
so-called reverberation mapping technique (e.g., Peterson 1994): under the
assumption of causality arguments, the BLR size is estimated as cΔt (see
Figure 1.4), where Δt is the time variation observed in continuum and lines
spectra. However, the observed lines in AGN spectra play not just a vital
8 1. Introduction
Figure 1.6. Spectral Energy Distribution for different classes of non-blazar AGN.
A "vacuum" is present in the UV range due to Galactic obscuration. (Figure
from Koratkar & Blaes 1999)
role in the historical introduction of the AGN in modern astronomy, but also
they are useful tools for the AGN identification and classification by using the
so-called diagnostic diagrams (see Figure 1.5), in which the characteristic lines
ratio allows discriminating practically to an optimal selection campaign.
1.3 Spectral Energy Distribution
The usual way to plot the multiwavelength AGN emission is through the
so-called Spectral Energy Distribution (SED). This is a wide band spectrum
that covers many decades in frequency (from radio waves to gamma rays),
generally plotted as logν− logνF
ν
. The advantage of this method consists in
highlighting the band where the majority of energy is emitted. Some bands
have been studied better than others, of course, as the radio, the optical and
the X-ray, while others bands are not measurable by actual instruments, as
the extreme UV beyond the limit of 912 Å, owing to the high opacity of the
Galactic interstellar medium to those wavelengths. Apart from the extreme
SED bands (the radio and the gamma bands), where some AGN subclasses are
characterized as powerful radio/gamma emitters (e.g., FR/Blazar), the AGN
emission results nearly constant in the whole SED. The nearly constant SED
can be fitted quite well by an underlying power-law with flat spectral index,
plus some important spectral features (see Figure 1.6). So the flux density in
the SED will be of the form
f(ν)∝ν
−α
( erg cm
−2
s
−1
Hz) (1.6)
1.3 Spectral Energy Distribution 9
whereα is called the spectral index. For high energy astrophysics, an equivalent
parametrization is used for AGN spectra: concerning to photon counts, the
spectrum is described by
N(E)∝E
−Γ
(counts s
−1
cm
−2
keV
−1
) (1.7)
where N(E) is the number photon counts for a given energy. The Γ index is
called the photon index. Spectral and photon indices are related by the law
Γ =α + 1.
This SED description allows to highlight many features. Through these
spectral features, the SED is subdivided in various components according to
the considered band:
1. Big Blue Bump (BBB): it’s thermic emission in the OPT/UV range. It’s
generated by the "blackbody spectrum sum", each one produced by the
accretion disk at different distances and temperatures (§ 1.3.1);
2. Soft-X Excess: in the soft X band an emission excess is present with
respect to that if the low energy tail power-law was extrapolated. This
effect is generally attributed to the high energy tail from the thermic disk
emission (§ 1.3.2);
3. X-peak: it’s the emission from the hot corona when comptonized by the
primary disk emission (§ 1.3.2);
4. Iron Kα fluorescence line at 6.4 keV (§ 1.3.2);
5. InfraRed Bump: it’s the emission resulting from elaboration by circum-
nuclear dust of the primary emission coming from the disk. About 1/3 of
the bolometric luminosity is measured in this range (§ 1.3.3). .
The main SED features will be outlined in the following.
1.3.1 The thermal component: the Big Blue Bump
The derivating accretion disk emission is “thermic” in origin and the emission
frequency changes according to the distance of the emitting region from
the central engine. The resulting spectrum will be a so-called “multicolor
blackbody”, that is the convolution of different blackbody spectra, each one
with different characteristic temperature, emitted from disk annulus at different
distances from the nucleus. From the theoretical analysis of accretion disks
results that it’s possible to obtain the temperature in which the total spectrum
peaks, that is, for typical central masses (10
6
− 10
9
M
 ) corresponds to UV
energies (T∼ 10
5
− 10
6
K). Although the multicolor blackbody approximation
isquitecrudeandsimplistic, it’sausefultooltoparametrizetheAGNdominant
emission. A drastic deviation from the blackbody spectrum is observed at
high frequencies, in the X band. In fact, it’s not possible to justify the
observed X emission simply through accretion around such massive objects (as
T
disc
∝ 1/M
BH
). Another component is necessary, hotter than the disk, able
to energize the emitting particles. The main ingredient of this component is
the so-called “Hot Corona”.
10 1. Introduction
Figure 1.7. Average total spectrum (thick black line) and main components in the X-ray
spectrum of a type I AGN. The main primary continuum component is a power
law with an high energy cut-off at E∼ 100− 300 keV. A prominent iron line
is also showed. (Figure from Risaliti & Elvis 2004)
1.3.2 A two-phase disk model: the X component
The X-ray emission in AGN generally accounts for∼ 10% of the bolometric
luminosity. As the accretion disk surrounding a SMBH can not account for
this emission another mechanism is necessary.
The basic idea for high energy photons production is the primary radiation
comptonization. X-rays production can be explained by the interplay of a hot
state with a colder accretion flow (Haardt & Maraschi 1991, 1993). Assuming
that a fraction f of the gravitational power P
G
is spent in the hot zone (phase
2) while the rest (1−f) is emitted as thermic radiation from the optically thick
cool
5
disk (phase 1), then a thermal balance equation can be written for each
of the two phases. The hot phase produces hard photons, that are isotropically
re-emitted. Isotropic emission is emitted from source away for an half, and
the other half is re-emitted toward the disk (in the plane parallel limit, see
beyond). The hard emission against the disk is reprocessed in different ways
(in Figure 1.7 the main components for an average spectrum in the X-ray range
are showed). Some of it (10%− 20%) is reflected, resulting in the characteristic
spectral hump between 10− 30 keV. Another fraction, the smaller one, is
re-emitted as fluorescence iron line. The most (80%− 90%) is absorbed and
re-emitted from the disk as blackbody radiation in the soft X-ray. According
to this process, a model where the optically thick accretion disk is embedded in
a hot optically thin hot medium (a simple “sandwich-geometry”) can accounts
for the X-ray generation. This medium is called “Hot Corona”. However, this
geometry should request that the most of gravitational energy is dissipated
in the corona, rather than in the disk. It would implies that the AGN X-ray
5
The term "cool" in astrophysics is referred to objects with T < 10
6
K.
1.3 Spectral Energy Distribution 11
Figure 1.8. Reflection of a power law X-ray spectrum from an optically thick material.
The incident continuum is shown as a dashed line and the solid line shows
the reflected spectrum. The Fe Kα line is the more prominent feature in the
X-ray spectrum. (Figure from Reynolds et al. 1999)
emission is bigger than the UV emission. Differently, the opposite is observed:
UV luminosities are often very bigger than the X-ray ones. To resolve that,
a "lesser isotropic" geometry was supposed. The solution can be a plasma of
energetic electrons placed close to the SMBH, and not distributed along the
whole accretion disk (Haardt et al. 1994). This kind of disk model is called
“two phase” model, and it is held accountable for the X-ray emission coming
from AGN. The major disadvantage of this model is the energetic plasma
reservoir. The inverse Compton cooling, indeed, is extremely efficient, so the
plasma would cool if energization mechanisms were not acting, killing the
prolonged observed emission. A very efficient mechanism integrating to the
two phase model is the "magnetic reconnection" (Liu et al. 2002): magnetic
filed lines with opposite polarities reconnect themselves nearby the corona
releasing thermic energy.
The Fe Kα line
The so-called “smoking gun” of CT AGN is the Fe Kα line at 6.4 keV in the
X-ray spectra (e.g., Matt et al. 2000). This is a fluorescence line, that is, the
photon emitted by the electron that fills the K-shell vacancy leaved by the
electron ejected away from the atom after the incidence of an X-ray photon.
The fluorescence yield is a rising function of the atomic number Z, and iron is
by far the most abundant among the heavy elements (e.g., Anders & Grevesse
1989), because it represents the final stage in thermonuclear stellar fusion. For
these reasons, it’s not surprising thatKα iron lines are very strong in a number
of astrophysical objects, CT AGN included. However, the line is composed by
a pair: Kα
1
and Kα
2
, with energy lines of 6.391 and 6.404 keV, respectively,
for neutral iron (while energy shifts to higher values for ionized iron). As the
12 1. Introduction
Figure 1.9. The X-ray spectrum of a Seyfert galaxy for different column densities of the
absorber, assumed as a torus. Larger the column density, stronger the iron
line. (Figure from Matt et al. 2003).
separation of these values is smaller than the best energy resolution available
in present X-ray missions, it is customary to adopt the value of 6.4 keV for the
unique Fe Kα line. More specifically, the key parameter for characterization
of CT AGN is the Equivalent Width (EW) of the iron line. Generally, the EW
of an emission line is an important value, as the carrier of many informations.
The importance of the iron line EW in CT AGN is tightly related to the
continuum spectra. The EW, indeed, is a parameter that depends on the
underlying continuum, being defined as:
EW =
Z
F
λ
−F
cont
(λ)
F
cont
(λ)
dλ (1.8)
where F
λ
is the flux measured at the line’s wavelength, and F
cont
(λ) is the
corresponding continuumlevel. So, the absorption process of the incident power
law emphasizes the strongness of the iron line EW (Figure 1.9). This explains
the observation of EW in CT AGN up to a rest-frame value of∼ few keV.
1.3.3 The InfraRed bump
It’s believed that the origin of infrared (IR) observed emission in the SED is
thermal mainly. The characteristic IR Bump, with a minimum at∼ 1μm (see
Figure1.6), is interpreted as emission from dust, that reprocesses the primary
UV incident radiation. These features are ubiquitously observed in AGN (e.g.,
Sanders et al. 1989). The dust temperature is calculated to be no more than
∼ 2000 K, because dust sublimates for higher temperatures. Concerning to
the Far-Infrared (FIR) tail, the so-called “submillimiter break” is justified by
the rapid loss efficiency of dust grains in reprocessing the long wavelengths
radiation. It’s worth to note that also the reprocessing dust model is supported
1.4 The Nature/Geometry of the absorbing medium 13
by the IR variability. This model, indeed, predicts that the variability in the
OPT/UV spectrum is followed by an IR emission variability, with a delay
depending upon the distance between the two emitting regions. So this is
what is observed (see Clavel et al. 1989 for the particularly well studied case
of Fairall 9). This scenario is completed by the picture where the radiation
comes out from the central engine "burning" the dust until a typical distance
called sublimation radius. Beyond this radius, dust does sublimates no more;
instead, it warms to different temperatures depending on the distance from the
radiative source, so producing the observed IR bump. However, it’s necessary
to consider that a a significant contribution to the IR band in AGN comes
from ordinary starlight.
1.4 TheNature/Geometryoftheabsorbingmedium
From what just mentioned, it’s clear that the circumnuclear matter properties
are key elements to understand AGN. More specifically, obscuration deriving
from gas and dust along the line of sight affects the observation and the
classification of AGN. Physical state (temperature, density and grains shape),
gas metallicity, column density, chemical dust composition are all elements
affectingtheobservedSED.Soacarefulanalysisofthecircumnuclearproperties
is necessary to understand the AGN phenomena in a global sense. However,
this kind of analysis is not straightforward because, even if the presence of
obscuring matter it’s surely verified, the exact location and distribution of
this matter is not yet known. Obscuring matter was already inferred in the
circumnuclear region starting from X-ray spectra analysis of type 1 AGN with
the EXOSAT satellite (Perola et al. 1986, Elvis & Lawrence 1988), and then
from type 2 objects obtained with Ginga (Warwick et al. 1989, Awaki et
al. 1990, 1991, Iwasawa et al. 1992). On the other hand, direct measures
have been obtained from IR observation starting from the IRAS satellite
(Sellgren 1984, Leger & Puget 1984). The circumnuclear matter, indeed,
absorbs radiation coming from the central engine and then it re-emits as IR
radiation. To properly explain the IR emission it’s necessary a deep knowledge
of the absorbing dust properties. When size, shape and composition of the
dust grains are known, Mie theory (1908) permits to obtain the absorption and
the scattering cross sections for wavelengths ranging from FIR to X-ray bands.
This returns the UV/OPT/NIR
6
extinction as a function of the wavelength:
that is, the so-called extinction curve. Extinction is the combined effect of
absorption and scattering from dust grains. As grains extinction is more
efficient for wavelengths equal to grains dimension, the extinction curve can
constrain grains size. In the range 0.125≤ λ≤ 3.5μm, the Galactic curve
can be characterized by a single parameter, the total-to-selective extinction
ratioR
V
=A
V
/E
(
B−V ), that is the ratio between the optical extinction and
the color reddening (Cardelli, Clayton & Mathis 1989). For the Galaxy, the
extinction curve is showed in Figure 1.10. The curve grows from the NIR to
the near-UV bands, with a bump at λ = 2175 Å, and then growing again in
6
NIR is the shortening for Near-InfraRed.