6.1.1 Position and Source Size
The measured signal position was found to be clearly coincident with the position of the bi-nary system. However, in the vicinity of the system position exists a known variable hard X-ray source, 1RXP J130159.6−635806, found by the X-ray satelliteRXTE, which is located 90 south-east of the pulsar and could be an emitter or VHEγ-rays. Given the small error on the derived position of PSR B1259−63 (≈3000in each direction), a possible source confusion can be firmly excluded.
The point source character of the detected γ-ray emission allowed to set a limit for the source extension (see Sec. 5.2). This extension,σs, was found to be< 3300 at 95% confidence level which corresponds to 0.24 pc=7.4×1017cm at an assumed distance of 1.5 kpc. The information contained in the time scale τvar of the flux variability of the order of days further limits the size of theγ-ray production region tol=cτvar≈ 1016cm.
6.1.2 Spectral Energy Distribution
Figure 6.1 shows the spectral energy distribution E2dN/dE = EFE of the X-ray and γ-ray emission of PSR B1259−63. The black symbols result from data taken near the 2004 periastron passage while the grey symbols are from archival data of the 1994 periastron (see Sec. 2.3.2).
The full points represent the VHE γ-ray spectrum obtained in this work. The open circles and triangles are from RXTEandINTEGRALsoft and hard X-ray measurements [Shaw et al., 2004], respectively, obtained in a rather short observation period between τ+ 14.1 and τ+ 17.5 days (see Fig. 6.3). The total energy contained in the X-ray emission, and thus the radiated luminosity, is roughly one order of magnitude higher than that of the measured VHE γ-ray emission. Additionally, the spectrum of X-rays follows a power law with photon indexΓX≈ 1.7, while the spectrum of VHEγ-rays is softer withΓγ ≈2.7.
The mean energy flux per energy decade contained in the VHEγ-ray emission of EFE ≈ 3×10−12erg cm−2s−1 (c.f. Fig. 6.1) represents aγ-ray luminosity of Lγ ≈ 8×1032erg s−1 for an assumed distance of the binary system of 1.5 kpc. This corresponds to less than 0.1% of the pulsar spin-down luminosity so that the pulsar wind emerging from PSR B1259−63 should be energetic enough to provide the main source of energy for the observed emission. Assuming that the γ-ray emission is powered by the pulsar alone, the efficiency of converting the spin-down luminosity into γ-rays is much greater than in the case of the Crab Nebula. The ratio of spin-down luminosities of both pulsars is L1259/LCrab ≈ 2×10−3 while the ratio of the squared distance from Earth is d21259/dCrab2 ≈ 0.56. The average flux level of the VHE γ-ray emission from PSR B1259−63 corresponds to≈ 4% of that of Crab and thus theγ-ray production in the system of PSR B1259−63 is one order of magnitude more efficient than in the Crab Nebula.
6.1.3 Flux Variability
The VHE γ-ray emission was found to be strongly variable with the fastest variation on a timescale of a few days close to the periastron passage of the pulsar, suggesting a radiation mechanism which is related to the interaction of the pulsar and its companion.
Figure 6.2 illustrates the VHEγ-ray flux variation with respect to the pulsar orbit and the line of sight. The light curve exhibits a characteristic shape with two high flux states before and after and an indication for a minimum at periastron, followed by a slow decrease untilτ∼+75 days where the observed excess is no longer significant. Note that the orientation of the disk with respect to the orbital plane is not precisely known and no safe conclusions can be drawn about the observed time difference between the time of the post-periastron peak emission and the time of the second disk crossing assuming a disk orientation with ωdisk = 90◦ as indicated in the figure.
The variability pattern is quite similar to that of the transient unpulsed radio emission mea-sured contemporaneously in radio observations [Johnston et al., 2005] as can be seen in Fig. 6.3.
In fact, the observed flux rise post-periastron seems strikingly correlated between the two wave-lengths on timescales of days. Additionally, the “double-bump” feature is similarly observed in the light curve of the non-thermal X-ray emission seen near periastron, however not coeval (c.f. Fig. 2.12). Unfortunately, the 2004 X-ray data provides an insufficient time coverage of only three consecutive days, indicated in Fig. 6.3, for which no significant variability could be detected.
periastron
Figure 6.2: Sketch of the orbit of PSR B1259−63 with respect to the line of sight [adapted from Johnston et al., 1999, see also Fig. 2.15]. The colour gradient bars along the orbit indicate the periods of H.E.S.S.
observations and show the integral VHEγ-ray flux using a smoothed light curve based on the data points from the daily light curve.
F(1.4 GHz) [mJy] 20
40
60−20 τ 20 40 60 τ
80 100
−1−2−11F(>380 GeV) [10 cm s ]
Time [days relative to periastron ]
0
53060 53080 53100 53120 53140 53160 53180
−0.5
February March April May June
pulsar radio eclipse
INTEGRALRXTE
Figure 6.3: VHEγ-ray and radio light curves of PSR B1259−63 around its periastron passage (epoch τ=0, dotted vertical line). Upper panel, open points: Flux density of the transient unpulsed radio emis-sion at 1.4GHz [Johnston et al., 2005]. The pulsed radio emission was eclipsed in the time interval between∼ τ−18days and∼ τ+15days, indicated by the two dashed vertical lines. Lower panel, full points: Daily integral flux above 380 GeV as measured by H.E.S.S.