Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
On the Evolution of Pulsars V.S. BESKIN, A.V. GUREVICH, YA.N. ISTOMIN Lebedev Physical Institute Recently Lyne and Manchester (1988) presented data on the angle X between the axis of rotation and the magnetic dipole awns, determined from polarization observations. Such a complete catalogue makes it possible to carter out a detailed comparison of our theoretical results with the observed distribution of radio pulsars over the angle X. Before making such a comparison, we recall the main features of our theory (for more details see Beskin e' al. 19861. The results of such calculations of the generation of particles and of directed pulsar radio emission flux suggest that the properties of pulsars depend substantially on the parameter (Is) (10 ) (1) So, pulsars with Q ~ 1 generate particles practically on the entire surface of the polar cap whereas pulsars with Q > 1 may generate only within a thin ring, and the generation itself may be of nonstationary, irregular character. In other words, pulsars with Q > 1 are located near We death line on the diagram PP. Indeed, all pulsars exhibiting pulling, subpulse drifts, and mode switching have Q > 1, whereas pulsars with Q < 1 are characterized by stable radio emission. The theory of neutron star evolution associated with current losses is different for pulsars with Q < 1 and Q > 1. It is developed more thoroughly for pulsars with Q < 1. It is just for this reason that in all our preceding papers we distinguished between pulsars with Q < 1 and those with Q > 1 and compared theory with experiment separately for ~em. The same should be done in the analysis of pulsar distribution over the slope X. 9
10 AMERICAN AND SOVIET PERSPECTIVES Let us first see how the distribution of observed pulsars with Q < 1 over X will loot To this end we shall make use of the distribution function of pulsars NQ<1 (P. X, B12) over the period P (in seconds), over the magnetic field B:2 = B/10~2 Gauss and over the slope X Of the axes (Beskin e! al. 1986~. N Npv+iB,-io/7(l+B~2)-~-~-§F(x)~[Coso5X-B:2 P ~ (2) Here N is the total number of observed radio pulsars, ~ = -.1, ~ = 3, ,B = 0~75, F(x) = Amp cos-° sx and the 8-function just separates pulsars with Q < 1. As shown by Beside et al. (1986), the distribution function (23 is in agreement with the observed distn~ution of pulsars (with Q < 1!) over period P and magnetic field B:2. Integrating now the distribution function NQ < 1 over the magnetic field B. we can find distn~ution function NQ < 1 (P. X) by which one can estimate, for example, the mean value of the angle X as a function of period P: X(P) = ~ NQ<1(P, x)xdx (3) The results are presented in Figure 1. We see that the mean slope X(P) is in perfect agreement with results, decreases as P increases, in spite of the fact that for each pulsar the angle X increases with time due to the current losses. The reason for this is that pulsars having Q < 1 and large periods P cannot have angles X close to 90°. The region Q < 1 on the diagram P -sin X is shown in Figure 2. One can see that the generation of particles and radio emission by pulsars with a large enough period P is only possible when X are small Thus, the observed decrease of the mean angle X with increasing P cannot be regarded as an argument in favor of the decrease of X with time. This is rather an indication of the fact that the observed radio emission is indeed associated with particle generation near the star surface. Similarly, going over in (2) to the variables B12, X and to dynamical age ED = P/2P, we obtain NQ<1(~15, B12, X) ~ NT15+1B112-1°/7~+1(1 + B12)-1-~-~8(COSX)1 5(~ 2) F(X)~[1-ri517B92/49cosx, (4) where 715 = {D/15 million years. The analysis of (4) shows mat the distribution function NQ<1 (~15, X) = ~ NQ<1 (715 ,B12, X)dBl2, for pulsars
HIGH-ENERGY ASTROPHYSICS 75 Ox 50 25 ~1 OIL - . , , iLy,_ l .2 .4 .6 P(S) .8 1. FIGURE 1 The mean value of the slope angle X as a function of period P for pulsars with Q < 1. Ibe curve corresponds to formula (33, the points correspond to the observations reported by Lyne and Manchester (1988~. vvith ED < 15 million years tie., practically for all pulsars with Q < 1) must depend wealthy on the AD value, and the distribution over X must be approximately equiprobable. Indeed, if we consider only pulsars with Q < 1, then, as shown in Figure 3, their distn~udon over X for ED < 2 million years and word > 2 million years practically coincide, the mean value of X even somewhat increasing with time. X(~< 2mln.yr) = 43° ~ 24° X(~TD> 2mln.yr) = 48° ~ 29° (~5') But this increase cannot be regarded as statistically significant. Hence, here too, we see close agreement between our theory and experiment. Let us now consider pulsars with Q > 1. The distribution function of such pulsars over the magnetic field B and the angle X, according to Beskin et al. (1986), has the form NQ>1.(BI2, X) = NQ<~EP(X, Bt2), X, Bt2iPr (6) where P = 10-~5B'2~°/7cos: 5x is the rotation deceleration velocity in the mechanism of current losses, ~ = 3 Unix B~2-~°/7 minion years is the characteristic life-time of radio pulsars at the stage Q < 1, and the relation P = P(x, B:2) corresponds to the condition Q = 1 which describes transition of pulsars from ache region Q < 1 to the region Q > 1 as in formula (4~. As a result, for the distribution function over the angle X for pulsars with Q ~ 1, we obtain
12 AMERICAN AND SOVIET PERSPECTIVES if 3~10 \ \3,1042 \ \1012 Q < 1 \ at> 1 \ .5 1. 1.5 P(s) FIGURE 2 The region Q < 1 on the diagram P - sin X. The boundary is shown for three characteristic magnetic fields 31011 Gauss, 1012 Gauss, 31012 Gauss. NQ>1(X) = 2 Xcos1 5X (7) As shown In Figure 4, here we also see agreement between our theory and the experimental results reported by Lyne and Manchester (1988~. Concluding, we would like to emphasize that an unquestionable con- clusion of the modern theory is that for rotating pulsar magnetosphere filled with plasma, the law of vacuum deceleration dW 1B2Q4R6 . 2 -= 6 ,~ sin X (8) is not obeyed. The presence of plasma in pulsar magnetosphere follows from the very fact of the existence of observed powerful coherent radio emission of pulsars, and the rotation deceleration is determined by the ponderomotive action of current leading to an increase with time of the slope of axes for each individual pulsar. We can see that the conclusions of this theory are in satisfactory agreement with observations. Thus, the observational data presented by Lyne and Manchester are
HIGH-ENERGY ASTROPHYSICS ED ~ 2 million years 3 a) ILL I 1 1 30 60 90 At x° 13 ~ b) 3 ED > 2 million years -I l l 30 60 90 x° FIGURE 3 Distnbution of pulsars with Q ~ 1 over the slope angle X for a) ED < 2 million years; by TD > 2 million years. of 20 10 . ~ _ ~ . 1 1 1 90 x° FIGURE 4 Distribution of pulsars with Q > 1 over the slope angle X. The come corresponds to formula (73. in good agreement with our theory of neutron star evolution. New ob- se~vations which would promote direct estimation of the value dX/dt for individual radio pulsars would be, of course, of great interest. REFERENCES Beskin, V.S., NV. Gurevith, and Ya.N. Istomin. 1986. Sov. Phys. Usp. 29:946. Lyne, JOG., and RN. Manchester. 1988. MADRAS. 234:477.
