NFI Cross-calibration

One of the peculiarities of the Beppo-SAX payload is the good broad band (0.1-200 keV) sensitivity, which allows unique studies of cosmic sources. This is achieved thanks to 4 different instruments. Furtheremore, two of these instruments are actually constitued by indipendent units (MECS, three units, and PDS four units). It is therefore clear that to make full use of the BeppoSAX capabilities a good instrument cross-calibration is vital. Despite the rather good level of the calibration achieved so far, substantial improvements in the cross-calibration field are to be expected in the next future.

The cross-calibration issue involves different areas:

  • Flux cross-calibration
  • Energy cross-calibration
  • Position reconstruction cross-calibration (LECS and MECS only)
  • Flux cross-calibration

    The source used for flux cross-calibration is the Crab nebula The table give the source exposure time, the energy range, the dead-time corrected source count rate and the best fit photon spectral index of the last BeppoSAX observation (April 11, 1997).
    ksec counts s-1 Gamma NH (10**21 cm**-2)
    LECS 2.0 0.12-9 keV 208 2.0763.28
    MECS1 19.1 1.65-11 keV108 2.0832.87
    MECS2 18.6 1.65-11 keV131 2.1042.99
    MECS3 18.5 1.65-10 keV 134 2.0983.03
    MECS 1.8-10.5 keV 362 2.0732.68
    HPGSPC 8.1 7-60 keV 192 2.0503.0 (fixed)
    PDS 9.6 13-200 keV 176 2.113...

    LECS and MECS spectra have been extracted from 8 arcmin radius regions. They are not background subtracted. the PDS spectrum is the sum of the spectra of the four PDS crystals. HPGSPC and PDS spectra are background subtracted.

    The figures below show the 6 NFI spectra (LECS, three MECS units, HPGSPC and the sum of the four PDS crystals) of the Crab nebula normalized by a simple absorbed (NH=3.3 × 10**21 cm-2) power-law spectrum with a spectral photon index equal to 2. The response matrixes are the one available on-line from September 1, 1997. The upper panel refers to the first issue of software and calibrations (January 31, 1997), while the lower panel to the current version (Rev 1.0).

    The figure and the table show that the slope of the Crab in the 2-10 keV energy range in the MECS units and in the LECS and in the MECS units and the PDS in the 13-200 keV agree within 0.5%, while it is flatter in the HPGSPC 7-60 keV range by 2.5 %.

    The Crab emission is made up of two distinct components: the pulsar emission and the nebula emission. It is known from phase resolved spectroscopy that the pulsar emission is harder than the nebula emission. An analysis is ongoing to disentangle the different components contributions. It will help in clarifying also the cross-calibration issue on the whole Beppo-SAX broadband.

    The figure also shows that there are small differences in the relative normalization. Because of the large energy range covered by the instruments, small differences in slope can reflect themselves in huge differences in the 1 keV normalizations (that can however be read from the figure if one extrapolates the spectra); for such a reason, we prefer to comment on the differences in the normalization in the energy ranges common to the various instruments. We note that:

  • the normalization of the Crab spectra in the three MECS units is consistent with 1 within 3%;
  • the LECS spectrum between 2 and 9 keV has a normalization about 7-8 % lower than that of the MECS spectra; it must be noted however that Crab normalization is not expected to be the same as the MECS due to strongback obscuration of the extended emission.
  • the HPGSPC spectrum has a normalization between 7 and 10 keV 2-3 % higher than that the MECS spectrum
  • the PDS spectrum between 13 and 60 keV has a normalization about 15 % lower than that of the HPGSPC in the same band.

    The above number are reported in the figure, where the arrows indicates the ovrelapping energy bands on which they were estimated. They give the limits of the present flux cross-calibration of the instruments. The reasons for these differences are currently under investigation. Any significant improvement in the flux cross-calibration will be promptly communicated.

    Energy cross-calibration

    LECS and MECS energy cross-calibration can be tested using the spectrum of the superova remnant CAS A, which is reach in emission lines. The table gives the position of 5 lines in the LECS and in the three MECS, computed using the redistribution matrices avalable on 1-Jan-1997, and compares them with the line positions found by the ASCA SIS (Holt et al., 1994) and the lab energy of the atomic lines.

    Ne Lya-Fe-L1.02-1.081.02-1.091.06
    Mg He_a-Lya1.35-1.481.35-1.491.40
    Si He_alpha1.851.8541.881.871.87 1.86
    Si He_beta2.202.2002.282.212.21 2.20
    S He_alpha2.452.4442.452.462.42 2.45
    Ar He_alpha3.123.1203.063.103.07 3.08
    Ca He_alpha3.873.8693.863.923.90 3.91
    Fe He_alpha6.606.5866.586.736.65 6.68
    ??Fe He_beta?? 7.877.87 7.85

    Position reconstruction

    At the moment the statistics on LECS and MECS position reconstruction is limited. This study is under way and here we present only preliminary results concerning source positions in the central 10 arcmin of the field.

    The analysis of two bright sources (MECS count rate, three units, higher than 10 counts s-1), two medium bright sources (MECS count rate between 1 and 5 counts s-1) and two faint sources (MECS count rate between 0.1 and 1 counts s-1), and of 14 short observations of LMC X-3 (a raster scan performed in the Science Verification Phase) revealed that:

  • the scatter in the reconstructed sky position in the three MECS units is of the order of 10-20 arcsec, depending on the source brightness;
  • the scatter between the LECS and the MECS units is the order of 30 arcsec;
  • the absolute position reconstruction is for both instruments within 30 arcsec.

    Very preliminary results on sources outside the central 10 arcmin of the LECS and MECS FOV show that the absolute position reconstruction is less accurate (of the order of 1-1.5 arcmin). The reason of this imperfect calibration is under investigation.

    We are confident an absolute position reconstruction significantly better than 1 arcmin can be achieved over the entire LECS and MECS FOV.

    Please send questions/problem reports to: