The SNR RX J1713.7-3946 is reminiscent of SN1006 both in the synchrotron X-ray emission from the shell far from the centre of the remnant and also in the TeV gamma-ray emission from an extended region coincident with that of the non-thermal X-rays. This suggests that the particles responsible for the emission of the high energy photons are accelerated in shocks.
There are several possible emission processes of TeV gamma-rays: the emission induced by accelerated protons (by the decay process) and by electrons - through bremsstrahlung and/or the Inverse Compton (IC) process. The expected integral flux of gamma-rays above our threshold energy of TeV by the decay process is estimated to be photons cm-2 s-1 (Drury et al. 1994, Naito & Takahara 1994), where we assume the distance and the upper limit for the number density in the ambient space of the remnant as 6 kpc and 0.28 atoms/cm3, respectively (Slane et al. 1999). This flux value is too low to explain our observed flux, even taking into account the large uncertainties in the estimates of the distance and the ambient matter density of the remnant (Slane et al. 1999, Tomida 1999). However, there remains the possibility of some contribution of the decay process if the remnant is interacting with a molecular cloud located near the NW rim (Slane et al. 1999). The relative contribution in emissivity of the bremsstrahlung process compared to decay process is estimated as %, assuming the flux ratio of electrons to protons is and that both have power law spectra with the index of 2.4 (Gaisser 1990), indicating this process is also unlikely to dominate. Therefore, the most likely process for TeV gamma-ray emission seems to be the IC process.
Under this assumption, the magnetic field strength in the supernova remnant can be deduced from the relation between the IC luminosity and synchrotron luminosity , where and are the energy densities of the magnetic field and the target photon field, respectively. and in the above formula must be due to electrons in the same energy range. The value of which should be compared with our TeV gamma-ray data is estimated from the ASCA result to be /, extrapolating the synchrotron spectrum with the same power law out of the energy range of 0.5-10 keV covered by ASCA (Tomida 1999). Here erg cm-2 s-1 is the X-ray luminosity in the 0.5-10 keV energy band observed by ASCA from the NW rim of the remnant and the power law index of -1.44 is the mean value for index of X-rays in the same energy range (Tomida 1999). keV is a typical synchrotron photon energy emitted by electrons which emit 1.8 TeV photons (the threshold energy of our observation) by the IC process when we assume the target photons to be from the CMBR. The value of is calculated to be erg cm-2 s-1 from our result for the number of photons of TeV gamma-rays, and using the fact that the spectra of synchrotron photons and IC photons follow the same power law when the electrons have a power law spectrum. Thus inserting , , and erg cm-3 of the energy density for the CMBR into the above relation, we can solve for the magnetic field strength B. Finally, the magnetic field at the NW rim is estimated to be G. The extrapolation used to estimate is reasonable, because is estimated to be 0.15 keV; this is not so different from the minimum energy of the ASCA band (0.5 keV).
The electrons responsible for the synchrotron and IC photon emissions are likely to have been accelerated by the shocks in the remnant as discussed above. If the maximum electron energy is limited by synchrotron losses, this maximum energy can be estimated by equating the cooling time due to synchrotron losses with the time scale of acceleration by the first Fermi process in a strong shock as TeV, where is the shock velocity (Yoshida & Yanagita 1997). On the other hand, equating the acceleration time with the age of the remnant, the maximum energy can be expressed TeV. In either case, whether it is synchrotron losses or the age of the remnant that limits the maximum electron energy (Reynolds & Keohane 1999), electrons should exist with energies high enough to emit the observed synchrotron X-rays and TeV gamma-rays by the IC process.
It is notable that both RX J1713.7-3946 and SN1006 have relatively low magnetic field strengths and low matter densities in their ambient space. These common features may have arisen if the magnetic field was `frozen in' to the matter without amplification other than by compression by shocks and may be the reason why electrons are accelerated to such high energies. These facts may also explain the radio quietness (Green 1998) and the weak emissivity of decay gamma-rays of the remnants. For SN1006, the low matter density in the ambient space might result from the remnant being located far off the galactic plane and the supernova being of type Ia. For RX J1713.7-3946, the low matter density may be caused by material having been swept out by the stellar wind of the supernova progenitor (Slane et al. 1999). The low magnetic field and the low matter density in the ambient space of SN1006 and RX J1713.7-3946 may explain why TeV gamma-rays have been detected so far only for these two remnants.
In conclusion, we have found evidence for TeV gamma-ray emission from RX J1713.7-3946 at the level of 5.6 sigma. If confirmed (à la Weekes 1999), this would be the second case after SN1006 to show directly that particles are accelerated up to energies of TeV in the shell type SNR.
© European Southern Observatory (ESO) 2000
Online publication: January 31, 2000