Crab Nebula 240 channels spectrum

We have fitted LS1 and LS3 on-axis Crab Nebula spectra with a single power-law, to compare the results with those obtained with the 1s ratemeters. Also, because the very low signal above $\sim$400 keV reduces the energy band in which we can be confident of not being influenced by background subtraction effects to $\sim$20-$\sim$400 keV, we cannot constrain significatively the parameters of a broken power-law model.

By comparing the data with the expected spectrum obtained by convolving the canonical Crab spectrum with detector response matrix, we find a discrepancy in the range $\sim$40-$\sim$70 keV, in terms of a 'depression' in the measured spectrum. This effect is seen both in LS1 and LS3 Crab spectra, and is also generally seen in sufficient bright GRBs when fitted with a continuous law. The reason for this is still under investigation (see section 6.2) and therefore we exclude channels in this energy range when performing fits on GRB sources. If we fit GRBM Crab spectra with a single power-law fixing the photon index to the X-ray canonical value of 2.1, we obtain a normalization of 9.4$\pm$0.5 and 9.3$\pm$0.6 photons cm^-2 s^-1 keV^-1 for LS1 and LS3 respectively. This normalization values are fully consistent with the canonical value of 9.7 photons cm^-2 s^-1 keV^-1. This indicates a good normalization of the response matrices.
The power-law fits with both parameters free to vary give photon indices of 2.25$\pm$0.09 and 2.27$\pm$0.11 for LS1 and LS3 respectively, fully consistent with the 1s ratemeters values and other hard X-rays measurements. The $\c2r$ of these fits are of 1.12 and 1.15 for 81 d.o.f. Extending the energy range of the fits to 600 keV we find higher ($\sim$1.4) values of $\c2r$ despite the increased number of d.o.f. (188). This indicates that in the region were the source contribution is negligible, the BKG subtracted spectrum is affected by distortion due to background variations effects that are not well interpolated with the technique described in section 4.4. Thus, Crab Nebula measurements tell us that LS1 and LS3 240 channels around axis response matrices are satisfactory in the region 20-400 keV (excluding the 40-70 keV region), and don't give significative information above 400 keV. Nevertheless, the fits of the spectra of bright GRB (e.g. GRB980329, see Fig. 4.9), show us that the spectral response description of the two detectors is good also above this energy.

The LS1 Crab Nebula spectrum fitted with a simple power-law is shown in Fig. 4.11.1. The fit obtained excluding the 40-70 keV region is oplotted to the data, which have been grouped in order to obtain good statistics in each channel. The 40-70 keV depression in the data with respect to the model is apparent.

  

Figure 4.10: LS1 photon spectrum of the Crab fitted with a single power-law