Title: Magnetic field structure and evolution in NSs: Some open problems and questions
1Magnetic field structure and evolution in
NSsSome open problems and questions
José A. Pons University of Alicante, Spain
- Motivation. Why worrying about MF ?
- Observational issues. How do we see MFs ?
- A brief history of magnetized NSs.
- Structure of magnetized NSs
- Reminder about thermal evolution.
- Core vs. Crustal magnetic field evolution.
- Coupled magneto-thermal evolution. Feedback.
- Population synthesis studies vs. single NS
fitting.
2Motivation
- Despite the observers tendency to name a new
class every 1-2 newly discovered objects, or
theorists to make use all sorts of exotic matter
(kaons, quark matter, axions ) to explain some
phenomena, the most basic question may be - What is the NS model that includes the minimum
reasonably well known physics and can explain or
connect all (as many as possible) different
classes ? - And there is one thing we are sure NSs do have
MFs. - Despite MF-related issues are usually overlooked
for simplicity, it is a - Necessary ingredient in any NS model.
- Question Is magnetic field the missing link that
explains the variety of NS and - their connexions ?
3Observations The P-Pdot diagram
4 B and age estimates magnetic dipole
5 B and age estimates magnetic dipole
6 B and age estimates magnetic dipole
- Typical values R ?106 cm, I?1045 g cm3
Uncertainty ? a factor of a few. But this simple
thing is what gives 99 of B field measures.
7Alternative Emission models
How does a magnetized hot body radiate ?
How are emitted photons processed through the
atmosphere and magnetosphere ? If you know it,
comparing your model spectra with obervation can
be used to estimate B fields
- Magnetic atmospheres ?
- Magnetospheric processes ?
- Solid/liquid surface. Gas undergoes a phase
transition when TltTcrit - Lai (2001) estimates
- Tcrit 27 B122/5 (Fe)
- Revisited by Medin Lai
8Cyclotron resonant absorption or condensed
surface ?
T106 K, B1013 G
a angle between magnetic field and normal to the
surface
9Example RX J0720 spectral fitting
- BBGaussian absorption line.
- But what is the absorption ?
A condensed surface model with a dipolar field
with B132.5 and polar temperature of 100 eV.
Several models can explain the same spectrum.
10Observational problems Summary
- Magnetic dipolar emission model provides most of
B field estimates. Not too bad, but a factor of a
few (angles ?) uncertainty. Measuring Pdot not
always possible (easy in radio, harder in
X-rays). - Spectral fitting is strongly model dependent
(emission model ?). Much to be improved in
atmospheric/magnetospheric modelling. - Luckily, both measures sometimes possible.
11Evolution A brief history of magnetars
- A neutron star is born hot and liquid (melting T
approx 1e10 K). - Hydrodynamics is appropriate, and if a strong
magnetic field is present we can use MHD - (large electrical conductivity).
- Stable MHD solutions are complex and require a
toroidal component -
MHD equilibrium must be established in few
dynamical timescales (seconds, minutes)
Braithwaite and Spruit 2004,2005
12Structure of (proto-)magnetars
- A perturbative approach fro equilibrium MHD
reduces the equation describing the magnetic
field structure to the Grad-Shafranov equation
Simplest case Decoupled multipoles
Toroidal field
- But is MHD valid ?
- Composition
- Stratified medium
- Ambipolar diffusion
- (Reiseneger 2009)
Lorentz force
13Structure of (proto-)magnetars
- Perturbative models can also explain this
geometry (e.g. Lander and Jones, 2009, Ciolfi et
al. 2009).
Toroidal field
In any case, very small ellipticities, 10-6, (not
promising for GWs). What is the most
energetically favoured configuration
? Conservation of helicity ?
14Evolution A brief history of magnetars
- But a NS cools fast, and in a few hours or days
after birth two things happen - The crust freezes
- Neutrons and protons become superfluid/superconduc
tor - If you were happy with MHD, I am sorry, but MHD
is not valid in a superconductor or in a solid
SC
SOLID
Temperature profiles at different ages
from Aguilera et al. 2008
Not clear how much flux penetrates into the core,
and what is the evolution of a SC fluid (fluxoids
drift and interact with vortices ?)
15Magneto-thermal evolution of NSsIngredients
- Neutron star model (structure, EOS).
- Thermal evolution (energy balance equation)
standard cooling of NSs. (Similar timescales,T-B
coupling) - Magnetic field evolution in the crust Hall
induction equation. Field decay and Joule
heating. - Magnetic field evolution in the core
superconducting fluid dynamics, interaction
between fluxoids and vortices ??? (no formalism
yet) - Microphysics ingredients thermal conductivity,
electrical resistivity, neutrino emission
processes
16Thermal Diffusion (Energy balance
equation) Effects of magnetic field
17Cooling of weakly magnetized NSs
Intensively studied (Page et al., Yakovlev
Pethick)
18Thermal structure of magnetized NSs
- F - k . ÑT - k b (ÑT . b) - k b (ÑT b)
- kL (b ÑT) - Isothermal surfaces aligned with B Strong
dependence on B field geometry !
(Geppert, Küker, Page, 2004,2007,
Perez-Azorin et al. 2005, 2006a, 2006b, Henderson)
19Magnetized envelope models
- Meridional heat transfer important for large B
fields - Former 1D (plane-parallel) models revisited.
Improved Tb-Ts relations. - Significant differences when B tangential to the
surface
Pons, Miralles, Geppert AA 2009
20Joule heating ? Do the easy thing first energy
balance
Prediction slope1/2 in a logT-logB plot We
have about 30 NSs (7 magnificents, 3 musketeers,
RRATs, 7 AXPs, 2 SGRs, some radio-pulsars ) with
reported thermal emission and B.
21Joule heating effective in many NSs ?
Crust size 1 km Bint 10-15 x Bdip B decay
time 1 Myr
22Joule heating masquerades fast cooling ?
High B
B0
23Joule heating masquerades fast cooling ?
Mass dependence vs. B field dependence
All NSs with fast cooling ? not ruled out !
24Crustal B field evolution
- In a real NS the crust is not a fluid, so the MHD
approximation is not valid. It is more
appropriate to describe it as a Hall plasma,
where ions have very restricted mobility and only
electrons can move freely through the lattice. - The proper equations are Hall MHD. If ions are
strictly fixed in the lattice, the limit is known
as EMHD (electron MHD) - There are two basic wave modes in the
homogeneous limit (constant electron density),
whistler or helicon waves, and also Hall drift
waves in the inhomogeneous case.
Hall induction equation
Electrical resistivity depends strongly on T
25Crustal B field evolution
Problems
- Conductivity varies many orders of magnitude
- Magnetization parameter varies with time and can
get very large (Hall term dominates) - Back-of-the-envelope estimates vary in a range of
5-6 orders of magnitude
26B field evolution weak field
27B field evolution intermediate field
28B field evolution strong field
29B field evolution asymmetric
30Coupled B-T evolution
- maximum B field for old NSs !!
- higher fields more heating higher
resistivity faster decay
31Crustal B field evolution Summary
- The first Hall stage (few kyrs) is very active.
Whistler and Hall waves stress the crust,
resulting in frequent glitches and flares. The
timing anomaly is always present, but only when
the stresses break the crust or fast magnetic
reconnexion releases enough energy there will be
outbursts. - After the Hall stage, the system reaches a
quasi-equilibrium configuration (not simply
dipolar) and the field has dissipated in about a
factor of 10. Ohmic dissipation dominates during
1 Myr. All NSs born as magnetars end up with
similar fields. Look like isolated NSs or high
field radio-PSRs. A chance of rare transient
phenomena (less energetic). - When Joule heating is not efficient any more, the
star cools down and dissipation proceeds much
slower. A second Hall stage may happen for NSs
older than 1Myr and B fields of the order of 1e12
(timing noise with large positive and negative
braking index ?) - Effect of B field on observed temperature large
enough to masquerade fast cooling. Is rapid
cooling going on in all NSs but we can only see
it in some low field NSs ?
32Population Synthesis studies Motivation
- Question we all agree we must compare
models/theory with data/observations, but how do
we do that ? - Given the large uncertainty in many of the
physical ingredients of a magnetized NS model,
and the limited quality of data (temperatures and
B fields are not entirely reliable), can one
really trust constraints based on fitting
particular models on individual objects ? - But we have something else physics (EOS,
composition, processes on B fields) must be the
SAME in all NSs. Whichever model that works for
an INDIVIDUAL NS, must pass the test of being
consistent with the WHOLE population. - Population synthesis offers an interesting way
simulate a large populations of NS in the whole
galaxy (or an interesting region) and do an
statistical analysis of general properties. It
has lots of problems, but luckily you wont be
biased by one particular data point or anomalous
behaviour.
33Population synthesis I nearby thermally
emitting NS
- LogN-LogS study of known NSs at dlt3 kpc
- Same underlying physical model, same magnetic
field geometry, only varying strength. - Only ROSAT all sky survey with flux gt 0.1 counts
per second is complete.
34Population synthesis I nearby thermally
emitting NS
For Log-normal B field distributions, constraint
on the number of high field NSs 10 with Bgt1e14
G
35Population synthesis II galactic magnetars
Same distributions are consistent with magnetar
population. Degenaracy in parameter space not
broken Maybe some extra luminosity needed for
young objects (lt1 kyr)
(magnetar data from McGill online catalogue, Muno
et al. estimates in shaded box) )
36Population synthesis III radio-pulsars
Evolution with field decay affects mainly to
highly magnetized objects and the first Myr of
evolution. Spin-down ages overestimated Can we
find statistically acceptable results for these
models ?
37Population synthesis III radio-pulsars
Faucher-Guiguere and Kaspi (2006), no field decay
Popov et al. (2009)
38Why population synthesis ? Summary
- After introducing magnetic fields in the game,
even more parameters are needed to fit individual
objects (and NS do have magnetic fields). Highly
degenerate parameter space. - Simultaneous population synthesis studies of
different classes are a promising method to
constrain the initial field distribution and its
evolution, together with the internal physics of
NSs. - Both, explaining individuals and populations are
needed. It is important to think also globally.
If a model works (or seems to be requested) for a
particular object, can I prove that it does not
contradict properties of many others ?