Amazing properties of giant pulses and the nature of pulsar's radio emission

0.16 MB·English

by  V. Soglasnov

Proceedingsofthe363.WE-HeraeusSeminaron:“NeutronStars and Pulsars”(Postersand contributedtalks)
Physikzentrum BadHonnef, Germany,May.14−19,2006, eds.W.Becker,H.H.Huang, MPEReport291,pp. 68- 71
Amazing properties of giant pulses and the nature of
pulsar’s radio emission
V. Soglasnov
Astro SpaceCenter of theLebedev Physical Institute, Profsoyuznaya str. 84/32, 117997 Moscow, Russia
7
0
0
2
n Abstract. For comprehensive studying of giant pulses Canadian S2 and Japanese K5 as high performance and
a (GPs) from the Crab pulsar and the original millisec- largecapacitydataacquisitionsystems.Theyprovidetime
J ond pulsar (MSP) B1937+21 (J1939+2134), we con- resolution8–16nsand6–12hourscontinuousrecord.Raw

8 ducted multifrequency observations over the last few data werethen encodedandprocessedforcoherentdedis-

years. They show that giant pulses may be improbably persion.
1
bright, 105,106 Jy and more, they have extra ordinal In section 2 we describe briefly main results of these
v
0 spectraandpolarization.EMenergyconcentratedinsuch observations, the most important is that the peak flux
9 strong pulse is high enough to accelerate particles up to density of giant pulse may reach improbable huge value
1 Lorenz factor γ ∼104−106, since giant pulses may play 105−106 Jyandmore.TheinteractionofsuchstrongEM
1
an important role in physics of pulsar’s magnetosphere. wavewithplasmaparticlesisveryspecific(section3.1).In
0
particular, a strong wave may work as effective particles
7
0 accelerator. In the sections 3.2, 3.3 we discuss a probable
/ origin of giant pulses and their role in physical processes
h 1. Introduction
inside magnetosphere.
p
- Giantpulses (GPs)areaveryspecific classofsingleradio
o
r pulses observed in several pulsars which have power law
2. Observational properties
t intensity distribution, while normally the distribution is
s
a of exponential type. Therefore, sometimes one can detect 2.1. Waveform and time duration
:
v a pulse exceeding by many times a normal pulse inten-
i sity. However, the time of occurrence of such events is Giant pulses from MSP B1937+21 initially are ex-
X
unpredictable, it makes very difficult their search and in- tremely short, less than 10 ns (Kondratiev et al. 2006,
ar vestigation. At the moment giant pulses are detected or Soglasnov et al. 2004).Theirapparentwaveformandtime
suspected in nearly a dozen of pulsars, but only two, the duration are caused completely by interstellar scattering.
Crab pulsar and the original millisecond pulsar (MSP) Only several pulses from the analyzed few thousands can
B1937+21(J1939+2134),havethe rate of GP occurrence be suspectedashavingsomestructureotherthanscatter-
sufficiently highfor goodstatisticsandmoreorless detail ing waveform(Fig. 1).
study of this phenomenon. During few last years we con- Waveform and duration of the Crab giant pulses de-
ducted many observations of giant pulses from these pul- pend on their strength. Weak and medium GPs gen-
sars over a wide frequency range from 20 MHz to 5 GHz, erally have complex structure. The total duration may
singleandmultifrequency/station.Theobservationswere reachfewdecadesofmicroseconds(Fig2).However,giant
made in collaboration with many observatories and ob- pulseswithpeakfluxdensityexceedingsomecriticalvalue
servers:Kalyazinobservatorywith 64-mdish, Yu. Ilyasov (300 kJy at 1.4 GHz) are also very short. They consist of
V. Oreshko; Arecibo 305-m telescope, T. Hankins; 100- oneortwonarrowpeaks,sometimeswithweakerpedestal
m GBT, Yu. Kovalev,F Ghigo (NRAO GB); ARO (40-m of 1−2 µs duration (Fig. 3, 4).
dishinCanada),N.Bartel,W.Cannon,A.Novikov(York GiantpulsesfromMSPandthosestrongfromtheCrab
University); WSRT, B. Stappers (NFRA); UTR-2 deca- are very similar, in spite of a great difference of magnetic
metric telescope, O. Ulyanov, V. Zakharenko (Ukrainian field strength and other physical conditions. In both pul-
Institute of Radio Astronomy). An important feature of sars strong events intrinsically are extremely short, we
these observations is a long time continuous record with couldnotresolvethem.Perhaps,weakextentgiantpulses
hightime resolution(the latter is necessarybecausegiant from MSP also exist but they are below detection thresh-
pulses are very short, see below, paragraph2.2). We used old,becausethepulsarB1937+21ismoredistantandini-
standard VLBI terminals Mk5A (Kovalev et al. 2005), tially weaker.
V.Soglasnov: Amazing properties of giant pulses and the natureof pulsar’s radio emission 69
2.2. Intensity may change the sign.It seems improbablethat the condi-
tions change considerably at so small scale (∼1−10 m).
Obviously power law distribution of giant pulse intensity
The only explanation we can propose at the moment is
can not be continued up to zero as well as to the infinity,
that the emission may be produced by stratified particles
it must have cutoff both at low and high energies (or, at
of opposite sign of charge (e.g. electrons and positrons),
least, strong break in power index: flattening at low and
as a result, few spikes alternatively RCP and LCP polar-
steepeningathighenergies).Lowintensityfatteningofthe
ized are generating, many of them are undistinguishable
distributionforCrabgiantpulseswasrecentlydetectedby
with our time resolution.Similar picture was observedby
Popov & Stappers (2006):E =1000−3500Jy·µsat
break Hankins et al. (2003) at higher frequencies.
1.2 GHz 1.
In case of MSP the limit is far below GP detection
threshold, since it cannot be observed directly. However, 2.4. Spectra
it may be derived under adoption that power law distri-
Inaverage,giantpulseshavenearlypowerlawspectra.For
bution is valid up to this limit (Soglasnov et al. 2004):
MSP the spectral index equals -2.8. For the Crab pulsar,
S =16 Jy (main pulse), S =5 Jy (interpulse) at 1.2
min min thespectrumbecomesslightlyflattenedathigherfrequen-
and 1.6 GHz.
cies,it changesfrom−2.7(at frequencies below 1GHz)to
As for high intensity limit, the situation seems to be −1.8 (between 1.4−2.2GHz).
ratherintriguing.Now there are no indications onthe de-
Instantspectraofsinglepulses,obtainedinsimultane-
viationfrom power law distribution up to the improbable
ous multi frequencysessions,are verydifferent.The spec-
high intensity. Giant pulses from the Crab pulsar with tral index change within wide limits from +0.4 to −4.0
peak flux density exceeding million jansky (1 MJy) were
(Popov et al. 2006b). Only one third of events occur si-
detected at 2.2 and 1.4 GHz in observations conducted
multaneouslyatwidelyspacedfrequencies.Partiallyitcan
in 2005 year (Fig. 3, 4). Such events are not something
be explained by interstellar scintillations, but it is not a
exclusive, they normally occurred in each session if the
main reason,as it follows from simultaneous observations
observing time was sufficiently long, in accordance with
oftheCrabat600and111MHz,wherethereceiverband-
power law distribution. At the moment we have caught
pass exceed greatly decorrelation band, since interstellar
a half of hundred “millionaires” over total 30 hours. It
scintillations do not affect considerably on the apparent
is difficult to provide much more longer observing time
strengthofGPs.Detailanalysisshowsthatspectraofsin-
which is needed to detect more intense pulses, because
glegiantpulsescoverawidefrequencyrangebutnotcon-
of huge volume of data. We tested at Kalyazin the sys-
tinuously. They represent chains of spots or bands, which
tem forlong time multi frequencymonitoring ofthe Crab
parametersareinconsistentwithusualISS;probablythey
total power without any dispersion removal, which can
areintrinsicgantpulsestructure.Themostremarkableex-
provide detection the strongest giant pulses. Few days of
amplesarequiteexclusivespectraoftheCrabgiantinter-
probe observations show tat the system works properly.
pulseswithdiscreteregularstructureobservedatArecibo
Weplantostartregularmonitoringat5,1.4and0.6GHz
atfrequencieshigherthan5GHz (Eilek & Hankins 2006).
in September 2006.
3. Giant pulses and pulsar’s physics
2.3. Polarization
3.1. Giant pulses as strong EM waves
The emission of narrowgiantpulses is strongly polarized,
Megajansky intensity of GPs detected in the Crab pul-
up to 100 % of linear, circular (may be of both signs),
sar looks rather impressive, however, are these pulses re-
elliptical, either pure orvariously mixed.In itself it is not
ally giant with physical point of view? There are sev-
very surprisingly, because their extremely short duration
eral different criteria for the strong EM wave. The top
means thatthe emissionoriginatesfromverycompactre-
is Schwinger’s quantum limit, when the field strength of
gion, since particles are emitting under the same physical
EM wave exceeds a critical value 4.4 × 1013 Gs. Such
conditions. What is really marvelous, that giant pulses
wave can create e+e− pairs. It is convenient to measure
show rapid changes of polarizationover a very short time
the wave field strength E in (angular) frequency units
interval. Thus, two peaks separated by only few decades w
ω ≡ eE /m c = eH /m c, then the Schwinger’s limit
of nanoseconds (since close spatially) may have quite dif- w w e w e
can be written as ~ω >> m c2. The next criterium is
ferentpolarization(Fig.1,4).Moreover,polarizationmay w e
condition ω > ω, where ω is wave frequency. It means
changerapidlyinsideasinglenarrowcomponent,fromcir- w
that the motion of charged particles under the action of
cular to linear and opposite, or/and circular polarization
EM wave becomes relativistic, such wave accelerates par-
1 ticles nearly up to the speed of light over a single wave
In case of extent pulses, flux density integrated over the
whole pulse duration, rather it’s peak value, is a measure of cycle.
pulse strength. Observers name it “energy” and use the units From the observed peak flux density SJy we can esti-
of strange but convenientin practice dimension [ Jy· µ s]. mate ωw at the distance lcm from the region of GP emis-
70 V.Soglasnov: Amazing properties of giant pulses and the natureof pulsar’s radio emission
Fig.4. Example of extremely bright double giant pulse
Fig.1. Example of rare occurred multi component giant
from the Crab pulsar, S = 4.2 and 7.3 MJy, de-
pulse from MSP B1937+21,f=2.2 GHz, as it seen in two peak
tected at 2.2 GHz, displayed in total intensity (blue), lin-
polarizationchannels. The first component has nearly to-
ear (red) and circular(magenta) polarization. Both com-
talcircularpolarizationofonesign(LCP).Aleadingedge
ponents have almost 100% polarization, the first circular
of the second component also 100% circularly polarized,
and the second linear.
but ofopposite sign(RCP).Polarizationchangesthe sign
at the middle of second component.
where L is the distance from the Earth to the pulsar
kpc
in kiloparsecs,∆ν is the frequency band. This estimate is
obtained without any arbitrary adoption such as dimen-
sion of GP source, beam pattern of the emission etc., it
based only on the inverse square law, which is valid at
least up to the boundary of wave zone. Even strongest
Fig.2. Example of extent giant pulse from the Crab pul- “MegaJansky” Crab pulses detected at 2.3 and 1.4 GHz
sar,detectedat2.2GHz,∆ν =16MHz,displayedintotal are far below the Schwinger’s limit, however, they satisfy
intensity (blue), linear(red) andcircular(magenta)polar- to the condition ω > ω, which is valid up to the dis-
w
ization. tance 1010−1011cm from the emitter, or ≃ 100 radii of
the light cylinder. Since inside magnetosphere pulses are
“really giant”: if suppose that they are emitted near the
sion as
star’ssurface,theratioω /ω ≥100evenatlightcylinder,
L w
ωw ≈4.9·1012 kpcpSJy∆ν, 103 −106 near the emitter (at 104 −106cm). This ratio
l
cm may be much more for giant pulses at low frequency 23
MHz: if adopt the true (non-scattered) pulse duration of
order∼1µs, ω /ω ∼104atlightcylinderand∼106−107
w
near the emitter.
Under condition ω /ω > 1 the interaction between
w
EM wave and plasma particles becomes very specific.
The wave accelerates particles up to the Lorenz factor
γ ∼ ω /ω. Particles emit secondary waves at different
w
frequencies and in different directions than the incident
wave. Thus, in the simplest case of circularly polarized
wave, trajectory of particles is a circle of radius λ/2π (λ
is wavelength), the plane of the circle coincides with the
wavefront.Particlesemitatfrequencyω ∼ωγ2 ∼ω2/ω
em w
under right angle to the direction of the incident wave
propagation within a small angle ∼γ−1 ∼ω/ω .
w
The case of linearly polarized wave is more compli-
cate. Particles move along trajectory of eight-like shape
Fig.3.Example ofextremelybrightlinearlypolarizedgi- inthe planewhichisnormaltothe wavefront,thetrajec-
ant pulse from the Crab pulsar,Speak =5.4MJy, detected tory crosses itself at the center under angle 51◦.1,γ-factor
at 2.2 GHz, ∆ν = 16MHz, displayed in total intensity changes along trajectory from 1.03ω /ω in the middle to
w
(blue), linear (red) and circular(magenta)polarization. 0.36ω /ω at the edges of trajectory.
w
V.Soglasnov: Amazing properties of giant pulses and the natureof pulsar’s radio emission 71
Someevidenceoftheexistenceofsuch“perpendicular” creation develops very fast in a small volume, begetting
emission is a dramatic transformation of the Crab pul- rapidly rising volume charge, in turn, it generates strong
sar profile at high frequencies (Moffet & Hankins 1996), EMemission(“secondary”giantpulse),whichaccelerates
where two strong wide components HFC1 and HFC2 ap- particles, particles emit high energy synchrotron quants
pear,theyarespacedby51◦−54◦ inlongitude,the longi- which produce e+e− pairs, and so on.
tude of HFC2 is −90◦ relative the main pulse. They may
be interpreted as counterparts of giant pulses emitted at
3.3. What will happen afterwards
lower frequencies, as the sum of large number of short
durationbursts(“secondarygiantpulses”),in accordance The rate of pair production drops rapidly when the wave
withtheresultsobtainedby Sovakovska et al. (2006).Fi- and particles move away from the star’s surface both be-
nal confirmationmay be obtained from polarizationmea- causeoftheemissiondilutionanddecreasingthemagnetic
surementofsinglepulseHFCemissionwithhightimeres- field strength. However, as before, the wave is capable to
olution. accelerate particles over the whole path inside magneto-
In presence the magnetic field of strength H, in case sphere upto light cylinder.It leads to the developmentof
of wave propagation along field lines (the case important strong plasma instabilities, as a result, a “normal”pulsed
for pulsars), the particles accelerated by the wave are or- radioemission is generating by usual plasma mechanism.
biting with frequency ω ω /ω, where ω ≡ eH/m c = Pulsars with long period have a large light cylinder
H w H e
1.78 × 107 H. They emit synchrotron radiation at fre- radius,theenergyofEMwave(“giantpulse”)propagating
quency ω ∼ω ω2/ω2. Near star’s surface H ∼1012Gs over such long path, converts entirely to the ”normal”
em H w
(Crab)and∼108Gs (MSP B1937+21),since synchrotron pulsed emission. It is the reason why giant pulses are not
quantshaveenergy∼1012 and∼1016eVcorrespondingly, observed in the majority of long period pulsars.
whichisfarabovethresholde+e− paircreationbyquants
Acknowledgements. The author thanks all participants of ob-
inmagneticfieldsnearsurfaceofthesepulsars.Thismech-
servations.EspecialthanksTimHankins,JoannaRankin,Nor-
anism is much more effective than traditional curvature
bertBartel,YuriiKoalevforexclusivelyusefuldiscussionsand
radiation,because synchrotronquants are emitting under
financialsupportforvisitingobservatoriesandproviding1-TB
right angle to the field lines, while curvature quants are
Big LaCie Disks for storing the observational data. This work
emitting along the tangent field line. Thus, giant pulse
is supported by the Russian Foundation for Basic Research
may work as effective particles’ accelerator and is able to (project number 04-02-16384) and the Presidium of the Rus-
induce cascade pair creation. Moreover, under some con- sian Academy of Sciences, the project “Origin and evolution
ditionsgiantpulsebecomesself-generating,seebelow,sec- of stars and galaxies”. Wegratefully acknowledge thesupport
tion 3.2. by theWE-Heraeus foundation.
3.2. Where and how giant pulses are emitting References
All statedaboveareobservationalresultsand their direct Hankins, T. H., Kern, J. S., Weatherall, J. C., & Eilek, J. A.
2003, Nature, 422, 141
consequences. Now we concern briefly more speculative
Eilek J.A.&HankinsT.H.,2006, Theseproceedings,”Radio
question on the origin of giant pulses.
Emission Physics in the Crab Pulsar”
In the frame of standard model of pulsar’s magneto-
V.I. Kondratiev, M.V. Popov, V.A. Soglasnov, Y.Y. Kovalev,
sphere it seems rather unlikely that giant pulses originate
N. Bartel, & F. Ghigo, 2006, These proceedings, ”De-
from the region located far from the star’s surface, where
tailed study of giant pulses from the millisecond pulsar
the energyof radiationper unit volume,correspondingto B1937+21”
megajanskypeakflux,exceedsbymanyordersthedensity A. Slovakovska, A. Jessner, G. Kanbach, & B. Klein, 2006,
ofplasmaenergy.Bythisreasongiantpulsessurelycannot These proceedings, ”Comparison of giant pulses in young
arise near the light cylinder. The only place where suffi- and millisecond pulsars”
cientnumberofparticleshavesufficientenergytoproduce M. Popov & B. Stappers2006, A&A,in preparation
giant pulse is a polar cap near star’s surface. Perhaps, a Kovalev,Y.Y.,Ghigo,F.,Kondratiev,V.I.,etal.2005, ”Ob-
serving at the GBT with VLBA+Mark5A and VLBA+S2
causeofgiantpulsesisthegapdischarge,theyaredirectly
backends”, GBT Commissioning Memo 236; available
observed effect of the discharge.
at http://wiki.gb.nrao.edu/pub/Knowledge/GBTMemos/
If so, we can suppose the follow scenario. The first
GBT Mark5A S2.pdf
giant pule may be born when the gap potential reaches
Popov,M.V.,Soglasnov,V.A.,Kondratiev,V.I.,etal.2006b,
somecriticalvalue,strongcurrentsbegintorunacrossthe ARep,50, 55, transl. from: AZh,2006, 83, 62
polar cap generating strong EM emission. As described Moffet, D.A., & Hankins, T. H. 1996, ApJ 468, 779
above (in the section 3.1), interaction between strong Soglasnov, V.A., Popov,M. V.,Bartel, et al. 2004, ApJ,616,
EM wave (“giant pulse”) and chargedparticles leads in a 439
strong magnetic field to the e+e− pair creation. Because
thismechanismisveryeffective,immediatelycascadepair