Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xray.sai.msu.ru/~polar/html/publications/belt_aa/final.ps Дата изменения: Wed Apr 2 19:11:11 2003 Дата индексирования: Sat Dec 22 04:53:17 2007 Кодировка: Поисковые слова: local bubble

Astronomy & Astrophysics manuscript no.
(will be inserted by hand later)
Y o u n g i s o l a t e d n e u t r o n s t a r s f r o m t h e G o u l d B e l t
S.B. Popov 1;4 , M. Colpi 2 , M.E. Prokhorov 1 , A. Treves 3 , and R. Turolla 4
1 Sternberg Astronomical Institute, Universitetski pr. 13, 119992 Moscow, Russia
e-mail: polar@sai.msu.ru; mike@sai.msu.ru
2 Universita di Milano-Bicocca, Dipartimento di Fisica, Piazza della Scienza 3, 20126, Milano, Italy
e-mail: Monica.Colpi@mib.infn.it
3 Universita dell'Insubria, Dipartimento di Scienze, via Vallegio 11, 22100, Como, Italy
e-mail: treves@mib.infn.it
4 Universita di Padova, Dipartimento di Fisica, via Marzolo 8, 35131, Padova, Italy
e-mail: turolla@pd.infn.it; popov@pd.infn.it
Abstract. The origin of the local population of young, cooling neutron stars is investigated with a population
synthesis model taking into account the contribution of neutron stars born in the Gould Belt, in addition to those
originating in the Galactic disk. We estimate their emission in the soft X-ray band as a function of distance and
age and construct the Log N { Log S distribution. It is shown that the inclusion of neutron stars from the Gould
Belt provides a good t to the observed Log N { Log S distribution. As the Sun is situated inside the Gould
Belt, one can naturally explain the comparative local overabundance of massive progenitors and can remove the
diфculty of the decit of relatively bright ( >  0:1 ROSAT PSPC cts s 1 ) cooling neutron stars previously reported
from models where only neutron stars from the Galactic disk were accounted for.
Key words. stars: neutron { stars: evolution { stars: statistics { X-ray: stars
1. Introduction
Observations of isolated neutron stars (INSs) are impor-
tant for gaining deeper insight on their structure and ther-
mal evolution and ultimately might prove decisive in un-
veiling the physical properties of matter at ultra-high den-
sities. Up to a decade ago, the only known INSs were
active radiopulsars, with the addition of the -ray pul-
sar Geminga. Despite the fact that X-ray emission from
some radio-pulsars was already detected by Einstein, it
was in the '90s that ROSAT, thanks to its sensitivity in
the 0.1-2 keV band, gave a clearer picture of the faint X-
ray emission produced by the cooling surface of the closest
INSs. ROSAT, supplemented by more recent observations
by Chandra and XMM-Newton, revealed a variety of types
of behavior in the X-ray emission from INSs and its rela-
tion with radio activity.
In particular, a substantial contribution of ROSAT to
this eld has been the discovery of a group of seven radio-
quiet, close-by, thermally-emitting INSs, dubbed ROSAT
INSs (RINSs) and sometimes referred to as the \magnif-
icent seven" (see for recent references Treves et al. 2000,
Zampieri et al. 2001, Haberl & Zavlin 2002; see also Popov
& Prokhorov 2003). The nature of these sources has been
controversial and two interpretations were proposed, ei-
Send oprint requests to: S. Popov
ther as conventional middle-aged cooling NSs or as very
old NSs accreting the interstellar gas (see again Treves
et al. 2000). Although no compelling evidence has been
brought in favor of either picture as yet, it is now generally
believed that at least two of these objects (the brightest
ones, RX J1856-3754 and RX J0720.4-3125, e.g. Braje &
Romani 2002) are indeed relatively young cooling NSs.
Since the seven RINSs have remarkably similar ob-
served properties, it is quite natural to assume that they
belong to the same class, i.e. all of them are close-by, cool-
ing NSs. This, however, poses a major problem. A useful
and standard way to study a population of sources is to
compute the Log N { Log S distribution starting from a
model, and then compare it with observations. This has
been done for RINSs by Neuhauser & Trumper (1999) and
Popov et al. (2000b). The main conclusion of the latter in-
vestigation is that the typical spatial density of radio pul-
sars in our Galaxy is too low to explain RINSs. In other
words, the assumption that RINSs and ordinary radio-
pulsars derive from the same parent population underpre-
dicts the number of observed \coolers", i.e. NSs which are
hot and close enough for their thermal emission to be de-
tected in X-rays.
An obvious solution is to invoke a local (both in space
and time) overabundance of NSs with respect to those
originating in the Galactic disk and seen now as radiopul-

2 S.B. Popov et al.: Young isolated neutron stars
sars. The main goal of this paper is to investigate the pos-
sible origin of these objects. Here we suggest that the likely
birthplace for many of the INSs in the Solar proximity is
the Gould Belt (see Popov et al. 2002 for a preliminary dis-
cussion). The Gould Belt is a compound of young stellar
associations extending in a ring tilted at about 20 ф from
the Galactic plane, roughly centered at the position of the
Sun and with a radius of  1 kpc. Originally discovered
in the middle of the 19th century, the Belt was studied in
detail by Benjamin Gould (see Stothers & Frogel 1974 for
a historical overview). Most of the known close-by young
star associations (in Ophiuchus, Orion, Perseus, Scorpius
etc.) belong to the Belt. Since the Gould Belt is a rel-
atively young stellar system it contains a comparatively
large fraction of massive stars. As the Solar proximity is
embedded within the Belt, it is likely to have harbored
quite a number of massive young stars which are the pro-
genitors of NSs.
In Sect. 2 we perform a population synthesis of close-
by young INSs, assuming that they are born both in the
Gould Belt and in the Galactic disk with an assigned
mass spectrum. The dynamical evolution of the young
NSs population in the Galactic potential is then followed
together with the thermal evolution. This allows us to
compute the present surface temperature and spatial dis-
tributions of young, close-by NSs and nally their X-ray
count rate, once a model for interstellar absorption has
been prescribed. This is used to compute the Log N {
Log S distribution which is then compared with observa-
tions in Sect. 3. We refer to ROSAT PSPC in evaluating
the X-ray count rate and in confronting our model with
data. Even if Chandra and XMM-Newton provide much
better observations of individual sources and small elds,
ROSAT all-sky data are still the most complete by far in
soft X-rays. Discussion and conclusions follow in Sect. 4.
2. The Model
In this section we describe the method we used to com-
pute the local distribution of young, cooling neutron stars.
The main ingredients of our model are: the spatial distri-
bution of NS progenitors, the NS formation rate, the NS
cooling history, and the interstellar medium (ISM) distri-
bution (the latter determines the interstellar absorption
and hence aects the X-ray count rate). Even if we are
concerned with a young population ( <  10 6 yrs), its dy-
namical evolution can not be neglected if the number of
sources in a limited volume (size <  1 kpc) has to be as-
sessed. In fact, a NS with a typical velocity  300 km s 1
travels a distance  300 pc in its lifetime as a cooler. While
not much on a Galactic scale, such a displacement is non-
negligible in evaluating the cooler population in the Solar
proximity. For this reason we account for the dynamical
evolution of NSs in the Galactic potential in our model.
Our calculation proceeds in three steps. First, a spa-
tial distribution of the progenitors is selected and the
ratio of NSs born in the Galactic plane to those origi-
nating from the Gould Belt is xed, together with the
birth rates (Sect. 2.1). Second, the dynamical evolution
is followed assuming that NSs at birth receive a kick ve-
locity drawn from a prescribed velocity distribution (see
Sect. 2.2). Finally, we derive the X-ray ux as a function
of age and position on the basis of an updated set of cool-
ing curves (Sect. 2.3), and translate this in ROSAT count
rate for an assumed model of the interstellar absorption
(Sect. 2.4).
We varied the parameters of our model (NS forma-
tion rate, NSs mass spectrum, time of calculation, time
step, spatial distribution). In the following subsections we
focus on those parameters corresponding to the results
presented in Sect. 3.
2.1. Initial spatial distribution
In our picture NSs are born in the Galactic disk and in the
Gould Belt with a constant rate over the entire calculation
(see Fig. 1 for a schematic view). This is a reasonable ap-
proximation since we are dealing with young objects, less
than a few Myrs old. The Gould Belt is modeled as a very
thin disk of 500 pc in radius and inclination to the Galactic
plane of 18 ф with its center situated at 100 pc from the Sun
in the Galactic anticenter direction. The central region of
the Gould Belt is devoid of massive stars (see Poppel 1997
for detailed description of the Gould Belt, and Torra et al.
2000 for a shorter one). To mimic this, we assume that no
NS is born in the central region of the Belt up to a distance
of 150 pc from its center.
We assume that the birthrate per unit area is indepen-
dent of position, i.e. it is constant both in the Belt and in
the Galactic disk. The two (constant) values are dierent,
the one referring to the Belt being larger. To derive the
NS birthrate, we rely upon direct counting of massive stars
which are doomed to end in a supernova event. This gives
2:910 11 massive progenitors per yr per pc 2 (Tammann
et al. 1994) which implies a NS birthrate of 30 NSs Myr 1
in 0.6 kpc around the Sun, the region where data are avail-
able. In relating the NS and supernova birthrates we as-
sumed that core collapse produces a neutron star in 90%
of the cases. The above value gives a more reliable esti-
mate of the local NS birthrate, at present, as compared
with radio-pulsar and average supernova statistics.
To get the separate contributions to the NS birthrate
from the Belt and from the disk for the region inside 0.6
kpc we proceed as follows. In one Myr, out of a total of
thirty, twenty NSs are born in the Gould Belt accord-
ing to the estimate of 20{27 supernova events given by
Grenier (2000). The remainder is uniformly distributed
in the Galactic disk in a ring extending from 100 to 600
pc, to account for the known fact that the closest Solar
neighbourhood is underpopulated by massive stars (e.g.
Maz-Apellaniz 2001). This is in agreement with the re-
sults of Torra et al. (2000), who estimate that about 2/3
of the massive stars inside 0.6 kpc belong to the Gould
Belt, and the rest to the Galactic disk. Direct counting
of massive progenitors within 0.6 kpc from the Sun ac-

S.B. Popov et al.: Young isolated neutron stars 3
0 0
1 1
0000000000000000000000011111111111111111111111
100 pc
0000000000000000000000011111111111111111111111
3 kpc
500 pc
600 pc
18 o 50 pc
Fig. 1. R z projection of the initial spatial distribution of
NSs. The heavy lines mark the Gould Belt and the closer (< 3
kpc) part of the Galactic disk where NSs are assumed to be
born. Note the two regions devoid of NS formation, in the
Galactic disk with a radius of 100 pc around the Sun, and in
the Belt with a radius 150 pc around its center. The 0.6 kpc
sphere centered at the Sun is also shown for clarity.
counts for all NSs originating in the Belt. However, the
disk clearly extends well beyond 0.6 kpc from the Sun.
Although NSs born far away are very dim their contribu-
tion may become signicant at low uxes ( <  0:1 cts s 1 ).
For this reason we decided to include also NSs born in
the Galactic disk from 0.6 up to 3 kpc from the Sun. The
NS birthrate per unit area in this region is assumed to be
again constant and coincident with that in the disk inside
0.6 kpc. With this choice we nd an agreement with exist-
ing data on the supernova rate in a region of 1 kpc around
the Sun (Tammann et al. 1994).
In summary, we adopt a uniform birthrate of 20 NS
Myr 1 in the Gould Belt and 250 NSs Myr 1 in the
Galactic plane up to a limiting distance of 3 kpc from the
Sun. Our approach underestimates the number of distant
( >  1 kpc) newborn NSs in the direction of the Galactic
center because the rate of supernova events is likely to
increase there. However, interstellar absorption towards
the Galactic center is very high in the Galactic plane (see
Sect. 2.4), and this hinders the detection of sources at low
Galactic latitude. We note that the local overabundance
of young NSs is probably due not only to the enhanced
number of massive stars harbored in the Belt but re ects
also the fact that the star formation history in the Belt
favors an enhanced supernova rate in the present epoch.
2.2. Dynamical evolution
A detailed description of the evolutionary code may be
found in Popov et al. (2000a,b). Typically we calculate
about a thousand evolutionary tracks in each run (up to
 10 4 per production run) and then normalize our results
to the actual number of NSs born in the volume ( 1000)
during the 4.25 Myrs time interval, as discussed above. An
evolutionary track is calculated for specied initial posi-
tion and velocity of a NS. All tracks are used over their
whole duration and are applied to NSs of dierent masses,
i.e. for each track we have several dierent cooling histo-
ries, the number of which depends on the number of dier-
ent masses used in the run. As in Popov et al. (2000a,b),
each track actually represents a population of NSs of dif-
ferent masses continuously born at the specied location
with a prescribed spatial velocity distribution, during the
entire time interval of the calculation.
NSs are assumed to be born with a Maxwellian kick
velocity distribution with an average of  225 km s 1 .
Varying this parameter does not change our results signi-
cantly because we are probing young objects. The Galactic
potential is assumed to be axisymmetric, resulting from
the sum of three contributions: the disk, bulge and halo
(see Paczynski 1990). The Sun is in the Galactic plane at
a distance 8.5 kpc from the Galactic Center.
2.3. Cooling curves and ux calculation
To calculate the cooling of NSs we use the results of the
St. Petersburg group (see Kaminker et al. 2002, and the
review by Yakovlev et al. 1999). Cooling curves are for NSs
of masses from 1:1 M to 1:8 M with a step of 0:1 M
(see Fig. 2). Curves take into account all important pro-
cesses of neutrino emission. The equation of state (EOS)
used in Kaminker et al. (2002) was introduced by Prakash
et al. (1988). More precisely, it corresponds to Model I of
Prakash et al. for symmetry energy and compression mod-
ulus of saturated nuclear matter K = 240 MeV. The maxi-
mum NS mass in this model is 1.977 M . Since we base our
calculation on a denite EOS, the star radius R is known
once the mass M is xed. For 1:1M < M < 1:8M
it is 12:2 km < R < 13:2 km, where R is circumferential
radius. The EOS used here corresponds to matter com-
posed of neutrons, protons, and electrons (no muons, hy-
perons, and no exotic particles). Neutron super uidity was
ignored.
We start to compute the emitted ux when a NS has an
age of 10000 yrs. This choice is motivated by the fact that
the Vela pulsar (the youngest close-by thermally emitting
NS) is slightly older than that and also because the NS
birth rate used in our calculation corresponds to an event
every  10 000 yrs in the  1:7 kpc region around the Sun.
For each NS the calculation is stopped when the temper-
ature drops below 10 5 K. This happens at an age of 4.25
Myrs for the lightest NSs (M = 1:1 M ) or less for more
massive stars. NSs that cold could have been detected by
ROSAT PSPC within a distance of only 10 pc with a count
rate  5  10 3 cts s 1 . We assume that emission comes
from the entire star surface. This appears reasonable in
the light of the relatively low pulsed fractions ( <  15-20%)
detected so far in the majority of isolated NSs (see Haberl
& Zavlin 2002). We do not take into account any repro-
cessing of the blackbody surface emission in the NS atmo-
sphere.
NSs are expected to have a mass spectrum (see discus-
sion in Woosley et al. 2002). In the calculations we present
here NSs are taken to have a at mass spectrum between
1.1 and 1:8M . The mass spectrum (and clearly the mass
of each star) are assumed to be constant during the calcu-
lations. With the above values (maximum age 4.25 Myrs
and time step 10000 yrs) each evolutionary track repre-
sents 425  8 NSs of eight dierent masses born with a

4 S.B. Popov et al.: Young isolated neutron stars
Fig. 2. Cooling curves for dierent NS masses (Kaminker et al.
2002). Curves from top to bottom correspond to masses 1.1{
1.8 M (step 0:1M ). Here, and everywhere in the text, the
temperature is the red-shifted surface temperature (i.e. that
observed at innity).
time step 10 4 yrs at the same place with the same initial
velocity in a period 4.25-0.01 Myrs ago.
As one can see from Fig. 2, the main contribution to X-
ray bright sources comes from NSs with masses < 1:5 M .
Young isolated NSs which are observed in several SN rem-
nants as compact X-ray sources also should be relatively
low-mass objects (Kaminker et al. 2002). Observations of
binary radio pulsars suggest that NSs masses are strongly
peaked around 1.35 M (Thorsett & Chakrabaty 1999).
We repeated our calculations using a similar distribution
and found that it produces similar results, although the
number of observable isolated NSs increases by  30 %.
Calculations which take into account the realistic mass
function of massive progenitors in the Solar vicinity will
be the subject of a future paper.
2.4. ISM and absorption
Since young cooling NSs are expected to emit most of their
luminosity at UV/soft X-ray energies ( 20 200 eV,
corresponding to temperatures  10 5 {10 6 K), interstel-
lar absorption plays a crucial role with respect to their
observability. Any attempt to derive the number of ob-
servable cooling isolated NSs using the unabsorbed ux
would result in a substantial overestimate.
For the ISM distribution we use the same prescrip-
tion as in Popov & Prokhorov (1998). The Local Bubble
is modeled as a sphere with a radius of 140 pc and ISM
density of 0.1 cm 3 . Typical column densities for sources
inside the calculated volume are in the range NH  10 19 {
10 21 cm 2 . After the column density is evaluated for a cur-
rent NS position, we calculate the unabsorbed ux cor-
responding to the temperature, radius of the NS and its
distance from the Sun, and apply the standard procedure
to derive the ROSAT PSPC count rate. Outside the Local
Bubble in directions close to the Galactic plane, absorp-
tion starts to play a crucial role. That is why regions closer
to the Galactic center (< 7 kpc), where the NS formation
rate should be higher than in the Galactic disk in the
Solar vicinity, cannot add many sources to our sample. At
NH = 3  10 21 cm 2 even very young and hot low-mass
NSs with kT  0:1 keV cannot have been detected at a
distance 1 kpc in the ROSAT Bright Survey (RBS) at the
threshold of  0:2 cts s 1 .
Except for the Local Bubble (and even that in a quite
simplied way, see Sfeir et al. 1999 for a more complete
description), our model does not take into account small-
scale irregularities in the ISM distribution. They can be
important if one makes an attempt to produce a realistic
map of RINSs distribution on the celestial sphere, but in
the case of the all-sky averaged Log N { Log S distribution
our approximation is adequate.
3. Results
As discussed in the previous section, the dynamical evo-
lution of young neutron stars, together with the calcula-
tion of their X-ray ux, allows us to derive the Log N {
Log S for these sources. In particular, having xed the
number of NSs originating in the Gould Belt and in the
Galactic disk, we can compute the separate contributions
of these two sub-populations to the total Log N { Log S.
Our main results are presented in Fig. 3 where the total
(disk+Belt) and the disk-only distributions are compared
with observations. All curves refer to the whole sky, i.e. the
angular coverage used here is the entire solid angle. The
observed Log N { Log S has been derived from ROSAT
data of the seven RINSs and six other close-by young iso-
lated NSs (see Table 1). Contrary to RINSs, the latter
sources exhibit a composite spectrum with a non-thermal
high-energy tail superimposed on the thermal component.
The total count rate has been used in this case, but we
stress that the non-thermal contribution is sizeable only
for the Vela pulsar and PSR 1929+10. Poissonian errors
are assumed in both observations and simulations. The
computed curves were generated with a suфciently large
number of stars to reduce the statistical noise which is
particularly severe at large ux (see x2.2). Statistical er-
rors are shown for the observational points only, being too
small to be appreciated in the plot of the simulated dis-
tributions. Uncertainties arising from the assumed model
parameters are more diфcult to assess and could introduce
non-negligible deviations. Test runs with dierent choices
of key parameters show that dierences are of the order
of 50%. Symbols which show observational points (lled
diamonds or open circles) correspond to the type of the

S.B. Popov et al.: Young isolated neutron stars 5
Fig. 3. All-sky Log N - Log S distribution: lled diamonds are
the seven RINSs and open circles Geminga, the \three mus-
keteers", PSR 1929+10 and 3EG J1835+5918. We also show
the RBS limit (Schwope et al. 1999). Upper curve: NSs born in
the Gould Belt and in the Galactic disk (total birth rate 270
NS Myr 1 ). Lower curve: disk only (birth rate 250 NS Myr 1 ).
faintest object (RINS or not) which contributes to the to-
tal number at the specied count rate.
Previous investigations have convincingly shown that
cooling NSs with the same spatial density of ordinary
radio-pulsars cannot explain the observed Log N { Log S
distribution for RINSs (Neuhauser & Trumper 1999,
Popov et al. 2000b). In particular, it was stressed that an
additional contribution to the local population of young
cooling NSs should be invoked to explain the relatively
large number of bright sources at uxes >  0:1 cts s 1 . The
Gould Belt is a natural candidate to provide the missing
NSs in the Solar proximity. The population synthesis cal-
culations presented here strongly support this claim and
account not only for the observed distribution of RINSs
but also for that of other young, close-by isolated NSs
observed by ROSAT.
As can be seen from Fig. 3, the Gould Belt provides the
major contribution to the local population of young cool-
ing NSs and the theoretical prediction gives a very good
t to the data, within statistical errors. The contribution
of NSs born in the Galactic disk is not very important
at relatively large uxes (it leaves rooms for only a few
sources with count rate >  0.1 cts s 1 ), but, as expected,
becomes dominant at lower uxes ( <  0.01 cts s 1 ) where
far away stars contribute most. Fig. 3 clearly shows that
the curve referring to the Galactic disk alone is always be-
low the observed points with high statistical signicance
and below the RBS limit. The worsening of the agreement
at low uxes ( <  0.05 cts s 1 ) is to be attributed to the
incompleteness of the X-ray sample. The decit of very
bright objects ( >  1 cts s 1 ) may be attributed to chance
statistics.
Our calculations show that there can be about 10 50
unidentied isolated NSs in the ROSAT All-Sky Survey
(RASS) at a limiting ux of >  0:015 cts s 1 depending
on parameters of the model. The number of INSs in the
RASS/BSC has been recently investigated by Rutledge
et al. (2003), who found that at most 67 sources could
have been detected at the 0.05 cts s 1 level and have es-
caped identication. Our results are well within this limit.
Also there may be a few unidentied RINSs at uxes
>  0:1 cts s 1 at low Galactic latitudes (see also Schwope
et al. 1999). Most sources should be observed at  20 ф
from the Galactic plane towards the directions of lower
absorption. Some of them may be Geminga-like objects
with counterparts among unidentied gamma-ray sources
(also connected with the Gould Belt, see Grenier 2000).
Identication of these objects may prove important for
constraining cooling models and the NS mass spectrum.
Absorption, the at geometry of NS initial distribu-
tion and the nite extension of the Gould Belt naturally
explain the very at (slope atter than -1) Log N { Log S
curves in Fig. 3. Our model contains some degrees of free-
dom (e.g. the NS mass spectrum, details of the formation
rate and surface emission, a simple blackbody was used
here) that can be varied. We nevertheless believe that even
this simple picture (albeit based on realistic assumptions)
gives a quite satisfactory explanation of the observed prop-
erties of isolated NSs in the Solar proximity. Preliminary
calculations which take into account realistic mass spec-
trum of NS progenitors, atmospheric eects and dierent
spatial and velocity distributions are in progress and show
no qualitative changes.
4. Discussion and conclusions
At present about 20 nearby (distance < 1 kpc), young
(age < 4:25 Myrs) isolated NSs are known (see Table
1). The local NS population includes objects with dier-
ent properties: radio-quiet, thermally-emitting NSs (the
seven RINSs), Geminga and the Geminga-like object 3EG
J1835+5918 (these are probably active pulsars with ra-
diobeams missing the Earth), radio-pulsars with observed
thermal X-ray emission (PSR 1929+10 and the \three
musketeers": the Vela pulsar, PSR 0656+14, PSR 1055-
52) and seven other pulsars which are not detected in
X-rays. The latter objects are relatively old in compar-
ison with the others and lie further away, as can be seen
from Table 1 where the wide gap in the estimated age
of PSR 1055-52 and PSR J0056+4756 is apparent. For
ages of a few Myrs, even very low-mass (and hence slow-
cooling) NSs are too cold by now to have been detected
with ROSAT at distances >  0:5 kpc. The case of PSR
1929+10 is intermediate since it is  3 Myrs old but rel-
atively close ( 300 pc) and its X-ray emission is mainly
non-thermal.
As has been discussed by Neuhauser & Trumper (1999)
and Popov et al. (2000b), the number of X-ray bright INSs

6 S.B. Popov et al.: Young isolated neutron stars
Table 1. Local (r < 1 kpc) population of young (age < 4:25 Myrs) isolated neutron stars
Source name Period, Count rate, _
P Distance Age a Ref.
s ROSAT cts s 1 10 15 ss 1 kpc Myrs
RX J185635-3754 | 3.64 | 0.117 d  0:5 [1,2]
RX J0720.4-3125 8.37 1.69  30 60 | | [1,3]
1RXS J130848.6+212708 (RBS 1223) 10.3 0.29 < 10 4 ? | | [1,4]
RX J1605.3+3249 (RBS 1556) | 0.88 | | | [1]
RX J0806.4-4123 11.37 0.38 | | | [1,5]
RX J0420.0-5022 22.7 0.11 | | | [1]
1RXS J214303.7+065419 (RBS 1774) | 0.18 | | | [6]
PSR B0633+17 (Geminga) 0.237 0.54 c 10.97 0.16 d 0.34 [7]
RX J1836.2+5925 (3EG J1835+5918) | 0.015 | | | [8]
PSR B0833-45 (Vela) 0.089 3.4 c 124.88 0.294 d 0.01 [7,9,10]
PSR B0656+14 0.385 1.92 c 55.01 0.762 e 0.11 [7,10]
PSR B1055-52 0.197 0.35 c 5.83  1 b 0.54 [7,10]
PSR B1929+10 0.227 0.012 c 1.16 0.33 d 3.1 [7,10]
PSR J0056+4756 0.472 | 3.57 0.998 e 2.1 [10]
PSR J0454+5543 0.341 | 2.37 0.793 e 2.3 [10]
PSR J1918+1541 0.371 | 2.54 0.684 e 2.3 [10]
PSR J2048-1616 1.962 | 10.96 0.639 e 2.8 [10]
PSR J1848-1952 4.308 | 23.31 0.956 e 2.9 [10]
PSR J0837+0610 1.274 | 6.8 0.722 e 3.0 [10]
PSR J1908+0734 0.212 | 0.82 0.584 e 4.1 [10]
a ) Radio-pulsar ages are estimated as P=(2 _
P ),
for RX J1856-3754 the age estimate comes from kinematical considerations (Walter & Lattimer 2002).
b ) Distance to PSR 1055-52 is uncertain ( 0.9-1.5 kpc)
c ) Total count rate (blackbody + non-thermal)
d ) Distances from parallactic measurements
e ) Distances from the dispersion measure
(1) Treves et al. (2000), (2) Kaplan, van Kerkwijk & Anderson (2002), (3) Zane et al. (2002),
(4) Hambaryan et al. (2001), Haberl (2003), (5) Haberl & Zavlin (2002),
(6) Zampieri et al. (2001), (7) Becker & Trumper (1997), (8) Mirabal & Halpern (2001),
(9) Pavlov et al. (2001), (10) ATNF Pulsar Catalogue, (http://wwwatnf.atnf.csiro.au/research/pulsar/catalogue/)
is too large to be explained in terms of the average density
of radio-pulsars in the Solar neighborhood. This motivated
the suggestion that a sizeable fraction of NSs never be-
come active radio emitters, as suggested at the same time
by considerations on young NSs in supernova remnants
(Gotthelf & Vasisht, 2000) albeit the origin of the local
NS population was left open. In this paper we have shown
that thermally-emitting INSs are naturally explained as
cooling NSs born in the Gould Belt. The relatively high
local spatial density of young NSs is then due to the large
number of massive progenitors in the young stellar associ-
ations which constitute the Gould Belt. Our analysis lends
support to the idea that RINSs (the \magnicent seven")
represent the slowly cooling NSs and hence a clean sample
of NSs with M <  1:3M (Kaminker et al. 2002).
The computed Log N {Log S distribution for X-ray
thermally-emitting INSs born in the Gould Belt and (to a
lesser extent) in the Galactic disk accounts for all bright
INSs in the Solar vicinity (see Fig. 3) and leaves room
for at most 1-2 undetected sources above  1 cts s 1 .
The absence of sources brighter than RX J1856 (Log S>
0:56) is consistent with our calculations at the 2  level.
Having neglected very young NSs (age < 10000 yrs) does
not change our conclusions signicantly. Only young stars
with M < 1:35 M (about 1/2 of their total number)
have temperatures about 2  10 6 K (see Fig. 2), which
corresponds to a luminosity  10 34 erg s 1 . With our

S.B. Popov et al.: Young isolated neutron stars 7
NS formation rate we expect to have one such NS within
a distance of  2:5 kpc. About half of the NSs born in
that region should originate inside 1.7 kpc. Placing the
source at this distance its unabsorbed ux is  3  10 11
erg cm 2 s 1 . With a typical column density of 10 21 cm 2 ,
Web-PIMMS gives  1:3 cts s 1 for ROSAT PSPC. We
can then conclude that stars younger than 10000 yrs can
contribute only to uxes < 1:3 cts s 1 . This translates
into a  20% dierence in the Log N - Log S curve shown
in Fig. 3 in the worst case. Typical uncertainties of the
present model are at about the same level.
The local ( <  1 kpc) INS population was estimated to
comprise total about one hundred objects younger than
4 Myr, taking into account that some NSs born inside 1
kpc can leave the local space in their lifetime. These NSs
are not detected as radio pulsars, but tens of them could
be identied in ROSAT surveys as dim sources. The beam-
ing eect can be responsible only for part of these young
NSs to be radiosilent. According to Tauris & Manchester
(1998) in fact, about 50-70% of young pulsars are not visi-
ble from the Earth (see also Brazier & Johnston 1999). In
our model this would imply the presence of at least  30
active pulsars with an age <  4 Myrs in the Solar vicinity.
However, only about 1/3 of such pulsars are observed (see
again Table 1). A bimodal velocity distribution, with the
high charcteristic velocity >  500 km s 1 (Arzoumanian,
Cherno & Cordes 2002), would reduce the discrepancy
since a number of young pulsars leave the local volume
considered here. However, even barring observational bias
against low-luminosity radio-pulsars, our model seems to
support the argument by Gotthelf & Vasisht (2000), that
at least half of the observed young neutron stars follow
an evolutionary path quite distinct from that of the Crab
pulsar.
Acknowledgements. We want to thank Dmitry Yakovlev for
putting his cooling model to our disposal and for his invalu-
able help with it. We are also indebted to Vasily Beskin, Matteo
Chieregato and Andrea Possenti for many useful discussions.
The work of SP was supported by the Russian Foundation
for Basic Research (RFBR) grant 02-02-06663 and by RSCI.
SP thanks the Universities of Insubria at Como and Milano-
Bicocca, where part of this investigation was carried out, for
hospitality. The work of MC, AT and RT was partially sup-
ported by the Italian Ministry for Education, University and
Research (MIUR) under grant COFIN-2000-MM02C71842.
References
Arzoumanian, Z., Cherno, D.F., & Cordes, J.M. 2002,
ApJ, 568, 289
Becker, W. & Trumper, J. 1997, A&A, 326, 682
Braje, T.M., & Romani, R.W. 2002, ApJ 580, 1043
Brazier, K.T.S., & Johnston, S. 1999, MNRAS, 305, 671
Gotthelf, E.V., & Vasisht, G. 2000, in: Proc. of IAU Coll.
177 \Pulsar astronomy - 2000 and beyond", ed. M.
Kramer, N. Wex, & N. Wielibinski, ASP Conf. Series
202, 699
Grenier, I.A. 2000, A&A, 364, L93
Haberl, F., & Zavlin, V.E. 2002, A&A, 391, 571
Haberl, F. 2003, COSPAR Symposium on High Energy
Studies of Supernova Remnants and Neutron Stars (in
press) [astro-ph/0302540]
Hambaryan, V., Hasinger., G., Schwope, A.D., & Schulz,
N.S. 2001, A&A, 381, 98
Kaminker, A.D., Yakovlev, D.G., & Gnedin, O.Yu. 2002,
A&A, 383, 1076
Kaplan, D.L., van Kerkwijk, M.H., & Anderson, J. 2002,
ApJ, 571, 447
Maz-Apellaniz, J. 2001, AJ, 121, 2737
Mirabal, N., & Halpern, J.P. 2001, ApJ, 547, L137
Neuhauser, R., & Trumper, J. 1999, A&A, 343, 151
Paczynski, B. 1990, ApJ, 348, 485
Pavlov, G.G., Zavlin, V.E., Sanwal, D., Burwitz, V., &
Garmire, G.P. 2001, ApJ, 552, L129
Poppel, W. 1997, Fund. Cosm. Phys., 18, 1
Popov, S.B., Colpi, M., Treves, A., Turolla, R., Lipunov,
V.M., & Prokhorov, M.E. 2000a, ApJ, 530, 896
Popov, S.B., Colpi, M., Prokhorov, M.E., Treves, A., &
Turolla, R. 2000b, ApJ, 544, L53
Popov, S.B., Prokhorov, M.E., Colpi, M., Treves, A., &
Turolla, R. 2002, Grav. & Cosmol., 8 (suppl. II), 133
[astro-ph/0201030]
Popov, S.B., & Prokhorov, M.E. 1998, A&A, 331, 535
Popov, S.B., & Prokhorov, M.E. 2003, Trudy GAISH (in
press) [astro-ph/0205298]
Prakash, M., Ainsworth, T.L., & Lattimer, J.M. 1988,
Phys. Rev. Lett., 61, 2518
Rutledge, R.E., Fox, D.W., Bogosavljevic, M. & Mahabal,
A. 2003, ApJ (in press) [astro-ph/0302107]
Schwope, A. D., Hasinger, G., Schwarz, R., Haberl, F., &
Schmidt, M. 1999, A&A, 341, L51
Sfeir, D.M., Lallement, R, Crifo, F., & Welsh, B.Y. 1999,
ApJ, 346, 785
Stothers, R., & Frogel, J.A. 1974, AJ, 79, 456
Tammann, G.A., Loer, W., & Schroder, A. 1994, ApJS,
92, 487
Tauris, T.M., & Manchester, R.N. 1998, MNRAS, 298, 625
Thorsett, S.E., & Chakrabarty, D. 1999, ApJ, 512, 288
Torra, J., Fernandez, D., & Figueras, F. 2000, A&A, 359,
82
Treves, A., Turolla, R., Zane, S., & Colpi, M. 2000, PASP,
112, 297
Walter, F.M., & Lattimer, J.M. 2002, ApJ, 576, L145
Woosley, S.E., Heger, A., & Weaver, T.A., 2002, Rev.
Mod. Phys. 74, 1015
Yakovlev, D.G., Levensh, K.P., & Shibanov, Yu.A. 1999,
Phys. Uspekhi, 42, 737
Zampieri, L., Campana, S., Turolla, R., Chieregato, M.,
Falomo, R., Fugazza, D., Moretti, A., & Treves, A. 2001,
A&A, 378, L5
Zane, S., Haberl, F., Cropper, M., Zavlin, V., Lumb, D.,
Sembay, S., & Motch, C. 2002, MNRAS, 334, 345