Pulsars were discoveredat radio frequencies and this is still the way we learn most about theirpopulation and properties. This is the case despite the fact that the energyemitted in the radio regime is only a tiny fraction (typically one millionth or so) of the total loss in rotational energy. However,recent advances in high energy astronomy, in particular with the adventof new X-ray satellites like ROSAT or, more recently,XMM orChandra, have opened up new largewindows to study these exciting objects. In these parts of the electromagneticspectrum, the energy output is typically a few percent of the spin-down luminosity.
Moving in the electromagneticspectrum from radio to shorter wavelengths (i.e. higher frequencies) wefirst pass infrared emission, then the short range of the visible spectrum,then ultraviolet radiation before we reach X-rays and finally gamma rays, Figure 1.
 |
Figure 1. The electromagnetic spectrum. |
While the large majority of pulsars have only been detected in the radioregime, an increasing number are now also detected at optical frequencies,and in particular at X-rays. The Crab pulsar is again something special,as it is detectable across the whole electromagnetic spectrum.The origin of the emission from pulsarsin other parts of the electromagnetic spectrum usually has a differentorigin than in the radio regime, as we will now discuss.
 |
Figure 2. This sequence of optical images shows thevarious phases of the Crab pulsar rotation as observed byA.Golden and co-workers. The optical pulses of the Crab pulsar can becompared to the reference star to the west (the star on the right, see also the optical image shown in Part 1 of these notes). |
Only a handful of pulsarsare detected in the optical regime. For most of these sources still, oneonly observes a star-like point source at the position of the pulsar. Thatmeans that as long as individual optical pulses cannot be observed, the objectcould still be an ordinary star which happens to be located along the lineof sight to the pulsar. One way to distinguish between such a case anda genuine pulsar detection is to study the spectrum and the properties of theemission. If the optical emission happens to be polarized, it can be fairlysafely assumed that the emission is originating from the pulsar magnetospherein the form of optical synchrotron emission. If the emission appears tobe blackbody radiation, the likely cause is a positional coincidencewith an ordinary star. However, there is also a chance that one has detectedthe hot surface of the neutron star. But since the temperature of the surfaceis of the order of a few million Kelvin, the peak of this radiation shouldbe expected in the ultraviolet or X-ray regime.
The optical emission ofthe Crab pulsar had been observed as a strange star inthe centre of the Crab nebula long before the pulsar was identified assuch. Indeed, the Crab pulsar is the prime example of a radio pulsar alsoemitting pulses in the optical regime, Figure 2.
 An animation of the optical Crab pulsar observations(made by A. Golden). |
About 40 pulsars have been detected at X-ray frequencies. For some pulsars we apparently see magnetospheric emission, i.e. the X-rays are created by a non-thermal emission processat some distance above the surface. For other pulsars however, we seemto detect the hot surfaces, in particular the polar cap regions aroundthe magnetic poles of the pulsars. These areas are hotter than the restof the surface since the very same process that creates the plasma forthe radio emission also bombards the polar cap with particles, heatingit up. Such observations can be used to determine the size of theneutron star. The results are consistent with the general size estimatesas discussed in Part 1. Furthermore, X-ray spectroscopy reveals the compositionof the neutron star atmosphere which is only about a centimetre thick!
 |
Figure 3. Two X-ray images of the Crab pulsar in its off (left) and on (right)phases. |
Again, the Crab pulsaris a prominent source in X-rays. In the ROSAT observation of Figure 3one can see the pulsar as a strong pulsed point source in the centreof the nebula. In the left image the beam is directed away from us andthe pulsar is seemingly off (faint). In the right image the pulsar seems to beon, as the pulsar beam points towards us.
Until recently, X-rayastronomers used information from the radio regime to detect and studypulsars. Times are changing, however, as the new X-ray satellite telescopeslike XMM and CHANDRA are sensitive enough to detect pulses from objectswhich had not been identified as pulsars before. For the first time, radioastronomers can follow up X-ray discoveries rather than the other way round. Usually, it still needs thepower of pulsar timing in the radio regime to study the rotation propertiesof these new pulsars in detail. However, as we will discuss in some detailbelow, there is an interesting new class of pulsars which are only visibleat X-ray frequencies.
At the moment, only sevenradio pulsars are known to exhibit emission at gamma-ray frequencies. The currentsample is small because the sensitivity of the previously available gamma-raysatellites has been rather limited. Figure 4 shows six of theseven pulsars, plus a source called Geminga, all observed with theEGRET gamma-ray satellite.Geminga is, next to the Crab pulsar, the second strongest gamma-ray pulsarand also visible at optical and X-ray frequencies. However, many attemptsto detect Geminga in the radio regime have failed. Reports about detectionsat a low frequency of 102 MHz exist, but have not yet been independently confirmed.
 |
Figure 4. Gamma-ray pulsars. In each case the emission over a full pulse periodis shown at wavelengths from gamma-ray to radio. Where no pulse has been detected ahorizontal line is shown. Three famous sources, the Crab and Vela pulsars and Geminga, are labelled. |
As is obvious from Figure 4, the Crab pulsar is the only source for which the pulses arrivesimultaneously across the whole electromagnetic spectrum. Current theoriesindeed suggest that the mechanism for the high-energy emission as wellas the location in the magnetosphere where it is created, differs from theradio emission, explaining the lack of coincidence of pulses for the majority of the sources.It seems that the high-energy emission is created in so called "outer gaps" close tothe light cylinder. Moreover, there is an obvious common property amongthe pulsars detected at gamma-ray frequencies: they all show the largestrates of loss of rotational energy among the known pulsars, indicatingthat this is the main parameter driving this high energy emission.
There are a large numberof still unidentified gamma-ray point sources and it seems likely thata significant fraction of those are in fact pulsars. The current satellitedata are however not sensitive enough to search for gamma-ray pulses, andso confirmation has to be found in the radio regime. Indeed, a number ofyoung pulsars discovered in the Parkes Multibeam Survey that exhibit largeloss rates of rotational energy seem to be associated with previously unidentifiedpoint sources. In the near future, more sensitive gamma-ray satellitesare planned to be launched. With satellites such asGLAST manymore sources should be able to be identified as pulsars, allowing a much more profound study of their properties.
 |
Figure 5. An artists impressionof a magnetar. |
In recent years, anew class of rotating neutron stars has been identified which is only visible at highenergies. These objects are called "Soft Gamma-Ray Repeaters" (SGRs) and"Anomalous X-Ray Pulsars" (AXPs) and have been identified as rotating neutronstars, emitting gamma-rays or X-rays, respectively. The rotational periodsare rather long and are all found within a narrow window of 5 to 10 seconds.But, in contrast to radio pulsars, these objects do not obtain their radiatedenergies from the rotational energy. Rather they find a mechanismto convert magnetic field energy into radiation. It turns out that allthese objects have very high magnetic fields that are even larger than thosefound for radio pulsars. The field strengths are of the order of 10Gauss (10 Tesla) - see the P-Pdot diagram in Part 1, where we had included these sources.
In the case of SGRs, whichare also sometimes called "Magnetars"due to their high magnetic field strengths, Figure 5. It is currently believed that the emissionoriginates from giant starquakes which are caused by a cracking of theneutron star crust due to the strong magnetic field.
 An animation of a magnetar (fromNASA). |
There has been no confirmeddetection of radio emission from SGRs or AXPs. However, the proximity ofthese sources to radio pulsars in the P-Pdot diagram suggests that theymay emit radio emission after all. However, for long period pulsars the radiopulsar beam narrows and it may be possible that in these cases it has become so narrowthat it is likely to miss the Earth altogether. Future observationswill hopefully be able to answer this question.