1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{emulateapj}
3:
4: %% manuscript produces a one-column, double-spaced document:
5: %\documentclass[manuscript]{aastex}
6: %% preprint2 produces a double-column, single-spaced document:
7: %% \documentclass[preprint2]{aastex}
8: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
9: \newcommand{\axpa}{\hbox{CXOU\,J010043.1-721134}}
10: \newcommand{\axpb}{\hbox{4U\,0142+61}}
11: \newcommand{\axpc}{\hbox{1E\,1048.1-5937}}
12: \newcommand{\axpd}{\hbox{1RXS\,J170849-400910}}
13: \newcommand{\axpe}{\hbox{XTE\,J1810-197}}
14: \newcommand{\axpf}{\hbox{1E\,1841-045}}
15: \newcommand{\axpg}{\hbox{AX\,J1845-0258}}
16: \newcommand{\axph}{\hbox{1E\,2259+586}}
17: \newcommand{\sgra}{\hbox{SGR\,0526-66}}
18: \newcommand{\sgrb}{\hbox{SGR\,1627-41}}
19: \newcommand{\sgrc}{\hbox{SGR\,1806-20}}
20: \newcommand{\sgrd}{\hbox{SGR\,1900+14}}
21: \newcommand{\rxb}{\hbox{RX\,J0720.4$-$3125}}
22: \newcommand{\rxc}{\hbox{RX\,J0806.4$-$4123}}
23: \newcommand{\rxd}{\hbox{1RXS\,J130848.6+212708}}
24: \newcommand{\rxe}{\hbox{RX\,J1605.3+3249}}
25: \newcommand{\rxf}{\hbox{RX\,J1856.4$-$3754}}
26: \newcommand{\rxg}{\hbox{1RXS\,J214303.7+065419}}
27: \newcommand{\rbsd}{\hbox{RBS1223}}
28:
29: \usepackage{natbib}
30: \bibliographystyle{apj}
31:
32: \shorttitle{The impact of magnetic field on NS cooling}
33: \shortauthors{Aguilera, Pons, \& Miralles}
34:
35:
36: %
37: \begin{document}
38: %
39: \title{The impact of magnetic field on the thermal evolution of neutron stars}
40: % \subtitle{ }
41: \author{Deborah~N.~Aguilera\altaffilmark{1,2},
42: Jos\'e A. Pons\altaffilmark{1},
43: Juan A. Miralles\altaffilmark{1}}
44: %\affil{Departament de F\'{\i}sica Aplicada, Universitat d'Alacant,
45: % Apartat de Correus 99, E-03080 Alicante, Spain}
46: \altaffiltext{1}{Departament de F\'{\i}sica Aplicada, Universitat d'Alacant,
47: Apartat de Correus 99, E-03080 Alicante, Spain}
48: \altaffiltext{2}{Theoretical Physics,
49: Tandar Laboratory, National Council on Atomic Energy, (CNEA-CONICET), Av. Gral. Paz 1499,
50: 1650 San Mart\'in, Pcia. Buenos Aires, Argentina}
51: \begin{abstract}
52: The impact of strong magnetic fields $B>10^{13}$G on the thermal evolution
53: of neutron stars is investigated, including crustal heating by magnetic field
54: decay. For this purpose, we perform
55: 2D cooling simulations with anisotropic thermal conductivity
56: considering all relevant neutrino emission processes for realistic neutron stars.
57: The standard cooling models of neutron stars are called into question by showing that the magnetic
58: field has relevant (and in many cases dominant) effects on the thermal evolution.
59: The presence of the magnetic field significantly affects the thermal surface distribution and the
60: cooling history of these objects
61: during both, the early neutrino cooling era and the late photon cooling era.
62: The minimal cooling scenario is thus more complex than generally assumed. A consistent
63: magneto-thermal evolution of magnetized neutron stars is needed to explain
64: the observations.
65:
66: \end{abstract}
67:
68: \keywords{Stars: neutron - Stars: magnetic fields - Radiation mechanisms: thermal}
69:
70: \maketitle
71:
72: %________________________________________________________________
73:
74: It has been long hoped that the comparison of theoretical models
75: for the cooling of neutron stars (NSs) with the direct observation of their thermal emission
76: would help to unveil the physical conditions in the interior of these fascinating objects
77: \citep{Page2004,Yakovlev2004}.
78: Our knowledge of the cooling history of a NS has been improving as we
79: were refining the physical ingredients that play a
80: key role on the thermal evolution of NSs.
81: In the past, the field has become more
82: exciting every time that a new relevant idea was introduced
83: (direct Urca, superfluidity in dense matter,
84: fast processes due to exotic matter, etc.). However, despite the fact that a number of
85: NSs are known to have large magnetic fields, most studies assumed
86: weak magnetic fields. The main reason for
87: this simplification was that
88: the observed distribution of magnetic fields in radio-pulsars peaks in a region where
89: its effect were thought not to be relevant.
90:
91: The increasing evidence that most of the nearby NSs with reported thermal emission
92: in the x-ray band of the electromagnetic spectrum have anisotropic
93: surface temperature distributions \citep{Zavlin2007,Haberl2007},
94: the striking appearance of {\it magnetars} \citep{Kaspi2007},
95: and the discovery of thermal emission
96: from some high field radio-pulsars \citep{Gonzalez2005}, are indicating that most
97: NSs which can be potentially used to contrast theoretical
98: cooling curves have actually large magnetic fields ($B>10^{13}$ G).
99: The conclusion is that a realistic NS cooling model must not avoid
100: the inclusion of high magnetic fields.
101:
102: The so-called {\it minimal cooling} scenario
103: \citep[see e.g.][for recent reviews]{Page2004,Yakovlev2004}
104: defines the cooling model in which the emissivity is given by
105: slow processes in the core, such as modified Urca and nucleon--nucleon Bremsstrahlung,
106: and enhanced by the neutrino emission from the formation and breaking
107: of Cooper-pairs of superfluid neutrons.
108: On the other hand, if fast neutrino processes (i.e. direct URCA) take place,
109: the evolution of a NS changes
110: dramatically, resulting in the {\it enhanced} or {\it fast cooling} scenario.
111: Nevertheless, direct URCA only operates in the inner core of high mass NSs
112: for some equations of state.
113:
114: In this letter, we want to revisit the {\it minimal cooling}
115: model considering the effects of magnetic field. If a {\it minimal model} must include
116: the minimum number of ingredients (but all the necessary ones) to explain the observations, the magnetic field should be taken into account as well.
117:
118: The effect of the magnetic field on the surface temperature distribution caused by the
119: anisotropic heat transport in the envelope was studied in
120: a pioneering paper by \citet{Greenstein1983}.
121: The observational consequences of these models were analyzed for the pulsars
122: Vela and Geminga among others \citep{Page1995}.
123: \cite{Potekhin2001} calculated the angular distribution of temperatures
124: in magnetized envelopes
125: taking into account the quantizing effect of the magnetic field on the electrons,
126: and the suppression (enhancement) of the electron thermal conductivity
127: in the direction perpendicular (parallel) to magnetic field lines.
128: Nevertheless, the anisotropy generated in the envelope is not strong enough to be
129: consistent with the observed thermal distribution of some isolated NSs,
130: and it should be originated deeper in the NS crust.
131: The understanding of kinetic properties of matter
132: in NS crusts and envelopes has also been recently improved,
133: with special attention received by
134: the role of ions and phonons \citep{CH2007}, which can be relevant at low
135: temperatures and densities. In addition, the effect of impurities
136: on the heat conduction in a non--perfect
137: lattice is also an open problem that must be considered in the near future.
138:
139:
140:
141: More recently, crustal confined magnetic fields were considered to be responsible for the
142: surface thermal anisotropy observed in some isolated NSs.
143: Temperature distributions in the crust were obtained as
144: stationary solutions of the diffusion equation with axial symmetry \citep{Geppert2004}.
145: The approach assumed an isothermal core
146: and a magnetized envelope as inner and outer boundary condition, respectively.
147: The results showed important deviations from the crust isothermal case for crustal confined magnetic
148: fields with strengths $B>10^{13}$ G and temperature $T< 10^{8}$ K.
149: Same conclusions have been obtained
150: considering not only poloidal but also toroidal components of the magnetic field
151: \citep{Azorin2006, Geppert2006}. These models succeeded in explaining simultaneously
152: the observed x-ray spectrum, the optical excess, the pulsed fraction, and other
153: spectral features of some isolated NS, such as RX J0720.4-3125 \citep{Perez2006}.
154:
155: Although former studies about anisotropic temperature distributions
156: on the cooling history of NSs \citep{ShibYak1996,Potekhin2001}
157: provided very useful information, detailed
158: investigations of heat transport in the non spherical case for different
159: magnetic field geometries have not been available until recently.
160: In \cite{Aguilera2007} we have revisited the cooling of NSs combining the
161: insulating effect of strong non radial fields with the additional source of heating
162: due the Ohmic dissipation of the magnetic field in the crustal region.
163: We have shown that, during the neutrino cooling era and the early stages of the
164: photon cooling era, the thermal evolution is coupled to the magnetic field evolution,
165: and both processes (cooling and magnetic
166: field diffusion) proceed on a similar timescale ($\simeq 10^{6} $ yr).
167: The energy released by magnetic field decay (Joule heating) in the crust
168: is an important heat source that modifies or even controls the thermal evolution of a NS.
169: There is indeed observational evidence of this fact.
170: As shown in \cite{PonsLink2007}, there is a strong correlation between the inferred
171: magnetic field and the surface temperature in a wide range of magnetic fields:
172: from magnetars ($B \geq 10^{14}$ G), through radio-quiet isolated NSs
173: ($B \simeq 10^{13}$ G)
174: down to some ordinary pulsars ($B \leq 10^{13}$ G). The main conclusion is that,
175: rather independently from the stellar structure and the matter composition,
176: the correlation can be explained by heating from dissipation of currents
177: in the crust on a timescale of $\simeq 10^{6}$ yrs.
178: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
179: \begin{deluxetable}{lcccr}
180: \tablecolumns{5}
181: \tablewidth{0pc}
182: \tablecaption{Properties of neutron stars with reported thermal emission, ordered
183: by decreasing magnetic field strength.}
184: \tablehead{
185: \colhead{Source}& \colhead{$T^{\infty}$} & \colhead{$t_{sd}$} & \colhead{$B_{d}$} & \colhead{Ref.} \\
186: \colhead{ } & \colhead{($10^6$ K)} & \colhead{($10^3$ yr)} & \colhead{(10$^{13}$ G)} & \colhead{}
187: }
188: \startdata
189: \sgrc &$7.56^{+1.6}_{-0.7}$& 0.22& 210& 1 \\\noalign{\smallskip}
190: \sgra &$6.16^{+0.07}_{-0.07}$& 2.0 & 73 & 1 \\\noalign{\smallskip}
191: \axpf &$5.14^{+0.02}_{-0.02}$& 4.5 & 71 & 1 \\\noalign{\smallskip}
192: \sgrd &$5.06^{+0.93}_{-0.06}$& 1.1 & 64 & 1 \\\noalign{\smallskip}
193: \axpd &$5.30^{+0.98}_{-1.23}$& 9.0 & 47 & 1 \\\noalign{\smallskip}
194: \axpc &$7.24^{+0.13}_{-0.07}$& 3.8 & 42 & 1 \\\noalign{\smallskip}
195: \axpa &$4.44^{+0.02}_{-0.02}$ & 6.8 & 39 & 1 \\\noalign{\smallskip}
196: \axpe &$7.92^{+0.22}_{-5.83}$& 17 & 17 & 1 \\\noalign{\smallskip}
197: \axpb &$4.59^{+0.92}_{-0.40}$& 70 & 13 & 1 \\
198: \hline\noalign{\smallskip}
199: PSR J1718-3718 &$1.69^{+0.62}_{-0.23}$&34 &7.4 &2 \\\noalign{\smallskip}
200: \axph &$4.78^{+0.34}_{-0.89}$&230 &5.9 &1 \\\noalign{\smallskip}
201: CXOU J1819-1458 &$1.40^{+0.47}_{-0.47}$&117 &5.0 &3 \\\noalign{\smallskip}
202: PSR J1119-6127 &$2.40^{+0.30}_{-0.20}$&1.7 &4.1 &4 \\\noalign{\smallskip}
203: \rbsd &$1.00^{+0.0}_{-0.0}$&1461 &3.4 &5 \\\noalign{\smallskip}
204: \rxb &$1.05^{+0.06}_{-0.06}$&1905 &2.4 &5 \\
205: \hline\noalign{\smallskip}
206: PSR B2334+61 &$0.65^{+0.13}_{-0.34}$&40.9 &0.99 &6 \\\noalign{\smallskip}
207: PSR B0656+14 &$1.25^{+0.03}_{-0.03}$&111 &0.47 &7 \\\noalign{\smallskip}
208: PSR B0531+21 (Crab) &$<1.97$ &1.24 &0.38 &8 \\\noalign{\smallskip}
209: PSR J0205+6449 &$<1.02$ &5.37 &0.36 &9 \\\noalign{\smallskip}
210: RX J0822-4300 &$1.75^{+0.15}_{-0.15}$&7.96 &0.34 &10 \\\noalign{\smallskip}
211: PSR B0833-45 (Vela) &$0.68^{+0.03}_{-0.03}$&11.3 &0.34 &11 \\\noalign{\smallskip}
212: PSR B1706-44 &$0.82^{+0.01}_{-0.34}$&17.5 &0.31 &12 \\\noalign{\smallskip}
213: PSR J0633+1748 &$0.50^{+0.01}_{-0.01}$&342 &0.16 &7 \\\noalign{\smallskip}
214: PSR B1055-52 &$0.79^{+0.03}_{-0.03}$&535 &0.11 &7 \\
215: \enddata
216: \tablerefs{(1) SGR/AXP Online
217: Catalogue\footnote{http://www.physics.mcgill.ca/~pulsar/magnetar/main.html};
218: (2) \cite{Kaspi2005}; (3) \cite{Reynolds2006}, \cite{McLaughlin2007};
219: (4) \cite{Gonzalez2005}; (5) \cite{Haberl2007}; (6) \cite{McGowan2006};
220: (7) \cite{DeLuca2005}; (8) \cite{Weisskopf2004}; (9) \cite{Slane2004};
221: (10)\cite{Zavlin1999}, \cite{Hui2006}; (11) \cite{Pavlov2001};
222: (12) \cite{McGowan2004}.
223: For the PSRs, $t_{sd}$ and $B_{d}$ from Pulsar Online
224: Catalogue\footnote{http://www.atnf.csiro.au/research/pulsar/psrcat/}.
225: }
226: \end{deluxetable}
227: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
228:
229: In order to investigate if there is observational evidence that supports our models for
230: the crustal field evolution, we have compared theoretical cooling curves including magnetic
231: field decay with a sample of NSs with reported thermal emission. To obtain the cooling curves
232: we have performed two--dimensional simulations by solving
233: the energy balance equation that describes the thermal evolution of a NS
234: \begin{equation}
235: C_{v} \frac{\partial T}{\partial t} - \vec{\nabla} \cdot
236: ( \hat{\kappa} \cdot \vec{\nabla} T) = - Q_{\nu} + Q_{\rm J}~,
237: \label{eneq}
238: \end{equation}
239: where $C_v$ is the specific heat per unit volume, $Q_{\nu}$ are energy losses
240: by $\nu$-emission, $Q_{\rm J}$ the energy gains by Joule heating,
241: and $\hat \kappa$ is the thermal conductivity tensor, in general anisotropic in presence of
242: a magnetic field. In this equation we have omitted relativistic factors for simplicity.
243: A detailed description of the formalism, the code, and results can be found in
244: \cite{Aguilera2007}. The geometry of the magnetic field is fixed during the evolution.
245: As a phenomenological description of the field decay, we have assumed the following law
246: \begin{equation}
247: B= B_0\frac{\exp{(-t/\tau_{\rm Ohm})}}
248: {1+(\tau_{\rm Ohm}/\tau_{\rm Hall})
249: (1-\exp{(-t/\tau_{\rm Ohm})})}~,
250: \end{equation}
251: where $B$ is the magnetic field at the pole, $B_0$ its initial value,
252: $\tau_{\rm Ohm}$ is the Ohmic characteristic time,
253: and $\tau_{\rm Hall}$ the typical timescale of the fast, initial Hall stage.
254: In the early evolution, when $t\ll \tau_{\rm Ohm}$,
255: we have $B \simeq B_0 (1+t/\tau_{{\rm Hall}})^{-1}$
256: while for late stages, when $t \geq \tau_{\rm Ohm}$,
257: $B \simeq B_0 \exp(-t/\tau_{{\rm Ohm}})$.
258: This simple law reproduces qualitatively the results from more complex simulations
259: \citep{PonsGeppert2007} and facilitates the implementation of field decay
260: in the cooling of NSs for different Ohmic and Hall timescales.
261:
262: Our sample of objects includes magnetars, isolated radio-quiet NSs,
263: and radio-pulsars, and it covers
264: about three orders of magnitude in magnetic field strength.
265: Before we discuss our results several caveats should be mentioned.
266: For most NSs, $B$ is estimated by assuming that the lose of angular momentum
267: is entirely due to dipolar radiation. If the magnetic field is constant
268: along the evolution, the dipolar component can be estimated by
269: $B_d = 3.2 \times 10^{19} ({P \dot{P}})^{1/2}$ G, where $P$ is the
270: spin period in seconds, and $\dot{P}$ is its time derivative.
271: Alternatively, for a few radio-quiet isolated NSs,
272: $B$ is estimated assuming observed x-ray absorption features are proton cyclotron
273: lines. This association is still controversial and, if true, it
274: should be read as an estimate of the surface field, which is usually larger than the
275: external dipolar component. In order to work with a more homogeneous sample,
276: we have included in our study only those objects for which
277: $\dot{P}$ is available. The quoted magnetic fields are always the spin-down
278: estimate for the dipolar component.
279:
280: Ages are also subject to a large uncertainty.
281: If the birth spin rate far exceeds the present spin rate and $B$ is considered constant,
282: the age can be estimated by the {\it spin-down age} ($t_{sd}=P/2\dot{P}$).
283: In the cases that an independent estimate is available (e.g. a kinematic age), the spin-down age
284: does not necessarily coincide with the other estimation.
285: If one considers magnetic field decay, however, $t_{sd}$ seriously
286: overestimates the {\it true age} ($t$).
287: Correcting the spin-down age to include the temporal variation of the magnetic field
288: may help to reconcile the observed discrepancy between
289: the spin-down ages and independent measures of the ages of some of the objects we have
290: included in this analysis.
291: This makes the comparison with observations more meaningful.
292: Other sources of error do probably exist but it is not
293: the purpose of this letter to discuss each observation separately.
294:
295:
296: As for the temperatures, in general there is no clear way to define error-bars. We have adopted
297: the range of values found in the literature, and we also indicate those cases in which the
298: estimate should be interpreted as an upper limit, more than a measure.
299: This is the case of the Crab pulsar, or even of some magnetars, which show
300: large variations in the flux in the soft x-ray
301: band on a timescale of a few years, indicating that the {\it real} thermal component may be
302: that measured during quiescence and that the x-ray luminosity during their active periods
303: might be a result of magnetospheric activity. Following the criteria adopted
304: in \cite{Page2004}, most reported temperatures are blackbody temperatures, except for
305: low field radio-pulsars in which the blackbody fit results in an unrealistic
306: small radius of the NS. In these four cases, we took the temperature consistent with
307: Hydrogen atmospheres.
308: We include a list of the sources considered in Table 1, with the corresponding references.
309: %%%%%%%%%%%%%%%%%%Figure%%%%%%%%%%%%%%%%%%%%%%%%%
310: \begin{figure}[thb]
311: \centering
312: \vspace{0.8cm}
313: \includegraphics[width=0.45\textwidth]{fig1a.eps}\\
314: \vspace{0.95cm}
315: \includegraphics[width=0.45\textwidth]{fig1b.eps}
316: \caption{Cooling curves. Surface temperature at the pole as a function of $t$
317: (upper panel) and $t_{sd}$ (lower panel) for different $B_0$.
318: The magnetar region (red) contains curves for
319: NSs born with $B_0 > 5 \times 10^{14}$G.
320: Intermediate field NSs ($5\times 10^{13}$G$\,<B_0< 5\times10^{14}$G)
321: evolve within the brown area and the blue region corresponds to NSs with
322: $B_0 < 5 \times 10^{13}$G.
323: Observations: squares for sources with $B_{d}>10^{14}$G,
324: triangles with $10^{13}$G$\,<B_{d}<10^{14}$G,
325: and circles with $B_{d}<10^{13}$G. Open circles are temperatures obtained from
326: fits to Hydrogen atmospheres.}
327: \label{Joule}
328: \end{figure}
329: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
330:
331: In Fig.~\ref{Joule}, we show two types of theoretical curves: temperature versus
332: true age (according to the simulation) and temperature versus the spin-down age consistent
333: with the assumptions made on each model about the magnetic field evolution.
334: We have sampled the objects in three groups
335: according to their measured magnetic field: {\it high field} NSs
336: with $B_{d}\ge10^{14}$~G,
337: {\it intermediate field} NSs with $10^{13}$~G$<B_{d}<10^{14}$~G and
338: {\it low field} NSs for which $B_{d}\le10^{13}$~G.
339: We found that this three samples could be explained qualitatively by cooling curves
340: in three different regimes: high, intermediate and low magnetized NSs, in all cases with
341: Joule heating included. Each of these regimes is represented by a set of curves with a
342: given order of magnitude of the initial field, $B_0$. We have taken
343: ${\tau_{\rm Ohm}}=10^6$ yr and ${\tau_{\rm Hall}}=10^3 {\rm yr}/B_{0,15} $,
344: where $B_{0,15}$ is $B_0$ in units of $10^{15}$ G. The dependence of the results on the decay rates is
345: discussed in \cite{Aguilera2007}.
346: These three regions are depicted in Fig.~\ref{Joule} as colored zones. For comparison,
347: the cooling curve corresponding to a non-magnetized NS is represented by a dashed
348: line.
349:
350: Focusing in the high field (red) region, we see that the effect of the magnetic field is visible
351: from the very beginning of the NS evolution. The initial equilibrium temperature reached
352: in a non--magnetized model may be increased up to a factor of 5 and it is kept nearly
353: constant for a much longer time, up to $10^4$ years.
354: The effect of Joule heating is very significant and may help to understand
355: the high temperatures observed in magnetars \citep{Kaspi2007},
356: although other physical processes could contribute as well.
357: For instance, the initially higher temperatures result in higher electrical
358: resistivity, therefore accelerating
359: the magnetic field dissipation in an early epoch. We have not yet considered the
360: consistent temperature dependence of the resistivity, in combination with the
361: evolution of the magnetic field geometry. A fully coupled magneto-thermal evolution
362: code should be employed for this purpose (work in progress).
363: The observed temperatures of radio-quiet, isolated NSs
364: could be explained either if they are old NS ($\approx 10^6$ years) born as magnetars
365: with $B_0=10^{14}$-$10^{15}$~G, or by middle age NSs born with fields in the range of
366: $B_0=10^{13}$-$10^{14}$~G, as plotted in the intermediate brown zone of Fig.~\ref{Joule}.
367: For intermediate field strengths, the initial effect is not so pronounced but the star
368: can be kept much hotter than non-magnetized NSs during the period from $10^4$ to $10^6$~yrs.
369: Finally, the effect of the magnetic field in stars born with initial fields of the order of
370: $10^{13}$ G is moderate, but still visible. These would be the case of some radio-pulsars,
371: for which the detection of a thermal component in their spectrum has been reported.
372: For weakly magnetized NSs with $B \simeq 10^{12}$~G, the effect of the magnetic
373: field is small and they can satisfactorily be explained by non-magnetized models,
374: except for very old NSs ($t> 10^6$ years), as discussed in \cite{Miralles98}.
375: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
376: \begin{figure}[thb]
377: \vspace{0.8cm}
378: \centering
379: \includegraphics[width=0.45\textwidth]{fig2.eps}
380: \caption{Surface temperature (at the pole) as a function of the
381: magnetic field. Theoretical curves are shown for different $B_0$
382: and varying $\tau_{\rm Ohm}$ and $\tau_{\rm Hall}$.
383: The grey band corresponds to the HBL \citep{PonsLink2007}. Observations correspond
384: to Fig.~\ref{Joule}.}
385: \label{FigTB}
386: \end{figure}
387: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
388:
389:
390:
391: The fact that the magnetic field plays a relevant role in the thermal evolution of neutron
392: stars at least during the first million years of its life is supported but another observational
393: fact. The magnetic field distribution observed in radio-pulsars is peaked in $3\times 10^{12}$
394: G. Therefore, assuming no correlation between $B$ and $T$, one would expect to find
395: thermal emission from neutron stars below and above
396: that value in approximately the same number.
397: However, almost all nearby objects for which its temperature is known have fields
398: far exceeding that value. This mismatch between the pulsar field distribution and the
399: thermal emitters field distribution can be easily explained by our assumption: magnetic
400: field decays significantly during the first $10^6$ years of a NS life and this effect governs its
401: thermal evolution. As a consequence, NSs with larger fields have higher temperatures
402: and cool slower,
403: increasing the chances to be detected. This hypothesis has been recently outlined by
404: \cite{PonsLink2007} who reported a strong correlation between
405: the surface temperature and the magnetic field, well approximated by the expression
406: \begin{equation}
407: T_{s,6}^4 \simeq C B_{d,14}^2
408: \label{eqHBL}
409: \end{equation}
410: where $T_{s,6}$ is $T_s$ in units of $10^6$ K, $B_{d,14}$ is $B_d$
411: in units of $10^{14}$ G, and $C$ is a constant that depends on the
412: thickness of the crust, the Ohmic dissipation
413: timescale, and the ratio between the unknown internal field and the observed external
414: dipole. For typical numbers, $C \simeq 10$. This straight line in a logarithmic
415: plot of temperature versus magnetic field has been named the {\it heat balance line}
416: (HBL).
417:
418: To contrast our results with this hypothesis based on a simple energetics argument,
419: we plot in Fig.~\ref{FigTB} the variation of the polar surface temperature $T_s$
420: (usually identified with the hot component of the thermal spectrum)
421: as a function of $B$ for different theoretical curves, and we compare
422: them with the observational data from Table 1.
423: In this $T$-$B$ diagram the thermal history of a NS proceeds as follows.
424: A NS begins its life high on the figure with some initial field $B_0$. As it cools it moves
425: vertically downward, until decay of its field provides an energy source able to
426: counterbalance the thermal loses. This causes the trajectory to
427: bend to the left. Then it
428: continues moving down but the temperature and the field evolution are coupled
429: since heating produced by magnetic field depends on the field strength and the
430: cooling mechanisms depend on the temperature.
431: For young magnetars the initial stage is very short because the high field prevents the star to cool down further and the surface temperature remains relatively high. A further evolution proceeds only when the field is also decaying. For low magnetic field stars the drop in the initial temperature is more
432: pronounced and they reach the equilibrium region much later.
433: The region in the diagram where all cooling curves tend to converge is in agreement with
434: the HBL defined in \cite{PonsLink2007} (indeed it is a band, since not
435: all NSs are necessarily identical).
436: The grey band corresponds to the choice $C=5$-$50$.
437:
438: Since the initial vertical trajectory is very fast, NSs do not spend much
439: time in that region of the diagram.
440: A clustering of sufficiently old objects in the grey region is expected.
441: Low magnetic field stars reach undetectable low temperatures more quickly, thus they
442: are more difficult to detect than highly magnetized NSs.
443: In the late photon cooling era, since the main mechanism to radiate energy is
444: photons from the star surface, one expects that the correlation
445: given by Eq. (\ref{eqHBL}) is more clear.
446: Although it was a very simplified approximation, we show that our evolution curves
447: approach the HBL slope asymptotically.
448: Finally, it is worth to notice that some objects are crossed by curves
449: belonging to different initial $B_0$, so there is no unique
450: determination of the initial conditions.
451:
452: As a result of this investigation we conclude that consistent magneto-thermal evolution
453: simulations are needed before we can disentangle the properties of the interiors
454: of neutron stars by studying their cooling history. A first step in that direction
455: has been given in this work, showing that the effect
456: of magnetic field decay in the highly resistive crust (as opposed to the highly
457: conductive core)
458: could be very large. It certainly has a significant impact on the thermal evolution of stars
459: with $B \gtrsim 10^{13}$ G.
460: The thermal and magnetic evolution of neutron stars is (at least) a
461: two parameter space, in which the evolutionary tracks of NSs born with
462: different initial conditions (the mass of the object, the initial magnetic
463: field) cannot be properly described by a unique temperature versus age
464: curve. It becomes clear that the {\it minimal model} that hopefully will be used
465: to understand neutron stars needs to include the magnetic field structure and evolution.
466:
467:
468: \acknowledgements
469: D.N.A is supported by the VESF fellowship EGO-DIR-112/2005.
470: This work has been supported in part by the Spanish MEC grant AYA 2004-08067-C03-02.
471:
472: %\bibliography{letterIMF}
473: \begin{thebibliography}{29}
474: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
475:
476: \bibitem[{{Aguilera} {et~al.}(2007){Aguilera}, {Pons}, \&
477: {Miralles}}]{Aguilera2007}
478: {Aguilera}, D.~N., {Pons}, J.~A., \& {Miralles}, J.~A. 2007, preprint (ArXiv
479: 0710.0854)
480:
481: \bibitem[{{De Luca} {et~al.}(2005){De Luca}, {Caraveo}, {Mereghetti},
482: {Negroni}, \& {Bignami}}]{DeLuca2005}
483: {De Luca}, A., {Caraveo}, P.~A., {Mereghetti}, S., {Negroni}, M., \& {Bignami},
484: G.~F. 2005, \apj, 623, 1051
485:
486: \bibitem[{{Chugunov} and {Haensel}}(2007)
487: {{Chugunov} and {Haensel}}]{CH2007}
488: {Chugunov}, A.~I. and {Haensel}, P. 2007, \mnras, 381, 1143
489:
490: \bibitem[{{Geppert} {et~al.}(2004){Geppert}, {K{\"u}ker}, \&
491: {Page}}]{Geppert2004}
492: {Geppert}, U., {K{\"u}ker}, M., \& {Page}, D. 2004, \aap, 426, 267
493:
494: \bibitem[{{Geppert} {et~al.}(2006){Geppert}, {K{\"u}ker}, \&
495: {Page}}]{Geppert2006}
496: ---. 2006, \aap, 457, 937
497:
498: \bibitem[{{Gonzalez} {et~al.}(2005){Gonzalez}, {Kaspi}, {Camilo}, {Gaensler},
499: \& {Pivovaroff}}]{Gonzalez2005}
500: {Gonzalez}, M.~E., {Kaspi}, V.~M., {Camilo}, F., {Gaensler}, B.~M., \&
501: {Pivovaroff}, M.~J. 2005, \apj, 630, 489
502:
503: \bibitem[{{Greenstein} \& {Hartke}(1983)}]{Greenstein1983}
504: {Greenstein}, G. \& {Hartke}, G.~J. 1983, \apj, 271, 283
505:
506: \bibitem[{{Haberl}(2007)}]{Haberl2007}
507: {Haberl}, F. 2007, \apss, 308, 181
508:
509: \bibitem[{{Hui} \& {Becker}(2006)}]{Hui2006}
510: {Hui}, C.~Y. \& {Becker}, W. 2006, \aap, 454, 543
511:
512: \bibitem[{{Kaspi}(2007)}]{Kaspi2007}
513: {Kaspi}, V.~M. 2007, \apss, 308, 1
514:
515: \bibitem[{{Kaspi} \& {McLaughlin}(2005)}]{Kaspi2005}
516: {Kaspi}, V.~M. \& {McLaughlin}, M.~A. 2005, \apjl, 618, L41
517:
518: \bibitem[{{McGowan} {et~al.}(2004){McGowan}, {Zane}, {Cropper}, {Kennea},
519: {C{\'o}rdova}, {Ho}, {Sasseen}, \& {Vestrand}}]{McGowan2004}
520: {McGowan}, K.~E., {Zane}, S., {Cropper}, M., {Kennea}, J.~A., {C{\'o}rdova},
521: F.~A., {Ho}, C., {Sasseen}, T., \& {Vestrand}, W.~T. 2004, \apj, 600, 343
522:
523: \bibitem[{{McGowan} {et~al.}(2006){McGowan}, {Zane}, {Cropper}, {Vestrand}, \&
524: {Ho}}]{McGowan2006}
525: {McGowan}, K.~E., {Zane}, S., {Cropper}, M., {Vestrand}, W.~T., \& {Ho}, C.
526: 2006, \apj, 639, 377
527:
528: \bibitem[{{McLaughlin, M.~A. et al.}(2007)}]{McLaughlin2007}
529: {McLaughlin, M.~A. et al.} 2007, preprint (ArXiv 0708.1149)
530:
531: \bibitem[{{Miralles} {et~al.}(1998){Miralles}, {Urpin}, \&
532: {Konenkov}}]{Miralles98}
533: {Miralles}, J.~A., {Urpin}, V., \& {Konenkov}, D. 1998, \apj, 503, 368
534:
535: \bibitem[{{Page}(1995)}]{Page1995}
536: {Page}, D. 1995, \apj, 442, 273
537:
538: \bibitem[{{Page} {et~al.}(2004){Page}, {Lattimer}, {Prakash}, \&
539: {Steiner}}]{Page2004}
540: {Page}, D., {Lattimer}, J.~M., {Prakash}, M., \& {Steiner}, A.~W. 2004, \apjs,
541: 155, 623
542:
543: \bibitem[{{Pavlov} {et~al.}(2001){Pavlov}, {Zavlin}, {Sanwal}, {Burwitz}, \&
544: {Garmire}}]{Pavlov2001}
545: {Pavlov}, G.~G., {Zavlin}, V.~E., {Sanwal}, D., {Burwitz}, V., \& {Garmire},
546: G.~P. 2001, \apjl, 552, L129
547:
548: \bibitem[{{P{\'e}rez-Azor{\'{\i}}n}
549: {et~al.}(2006{\natexlab{a}}){P{\'e}rez-Azor{\'{\i}}n}, {Miralles}, \&
550: {Pons}}]{Azorin2006}
551: {P{\'e}rez-Azor{\'{\i}}n}, J.~F., {Miralles}, J.~A., \& {Pons}, J.~A.
552: 2006{\natexlab{a}}, \aap, 451, 1009
553:
554: \bibitem[{{P{\'e}rez-Azor{\'{\i}}n}
555: {et~al.}(2006{\natexlab{b}}){P{\'e}rez-Azor{\'{\i}}n}, {Pons}, {Miralles}, \&
556: {Miniutti}}]{Perez2006}
557: {P{\'e}rez-Azor{\'{\i}}n}, J.~F., {Pons}, J.~A., {Miralles}, J.~A., \&
558: {Miniutti}, G. 2006{\natexlab{b}}, \aap, 459, 175
559:
560: \bibitem[{{Pons} \& {Geppert}(2007)}]{PonsGeppert2007}
561: {Pons}, J.~A. \& {Geppert}, U. 2007, \aap, 470, 303
562:
563: \bibitem[{{Pons} {et~al.}(2007){Pons}, {Link}, {Miralles}, \&
564: {Geppert}}]{PonsLink2007}
565: {Pons}, J.~A., {Link}, B., {Miralles}, J.~A., \& {Geppert}, U. 2007, Physical
566: Review Letters, 98, 071101
567:
568: \bibitem[{{Potekhin} \& {Yakovlev}(2001)}]{Potekhin2001}
569: {Potekhin}, A.~Y. \& {Yakovlev}, D.~G. 2001, \aap, 374, 213
570:
571: \bibitem[{{Reynolds, S.~P. et al}(2006)}]{Reynolds2006}
572: {Reynolds, S.~P. et al}. 2006, \apjl, 639, L71
573:
574: \bibitem[{{Shibanov} \& {Yakovlev}(1996)}]{ShibYak1996}
575: {Shibanov}, Y.~A. \& {Yakovlev}, D.~G. 1996, \aap, 309, 171
576:
577: \bibitem[{{Slane} {et~al.}(2004){Slane}, {Helfand}, {van der Swaluw}, \&
578: {Murray}}]{Slane2004}
579: {Slane}, P., {Helfand}, D.~J., {van der Swaluw}, E., \& {Murray}, S.~S. 2004,
580: \apj, 616, 403
581:
582: \bibitem[{{Weisskopf} {et~al.}(2004){Weisskopf}, {O'Dell}, {Paerels}, {Elsner},
583: {Becker}, {Tennant}, \& {Swartz}}]{Weisskopf2004}
584: {Weisskopf}, M.~C., {O'Dell}, S.~L., {Paerels}, F., {Elsner}, R.~F., {Becker},
585: W., {Tennant}, A.~F., \& {Swartz}, D.~A. 2004, \apj, 601, 1050
586:
587: \bibitem[{{Yakovlev} \& {Pethick}(2004)}]{Yakovlev2004}
588: {Yakovlev}, D.~G. \& {Pethick}, C.~J. 2004, \araa, 42, 169
589:
590: \bibitem[{{Zavlin}(2007)}]{Zavlin2007}
591: {Zavlin}, V.~E. 2007, preprint (ArXiv astro-ph/0702426)
592:
593: \bibitem[{{Zavlin} {et~al.}(1999){Zavlin}, {Tr{\"u}mper}, \&
594: {Pavlov}}]{Zavlin1999}
595: {Zavlin}, V.~E., {Tr{\"u}mper}, J., \& {Pavlov}, G.~G. 1999, \apj, 525, 959
596:
597: \end{thebibliography}
598:
599: \end{document}
600:
601: