LEA_logo Laboratory for Experimental Astrophysics

Konus-Wind Magnetar Data Analysis

The main steps that are performed during a standard data analysis procedure are the following.

  • 1

    Source Identification

    We start our analysis with the identification of the burst source.

    We use precise localization derivied by single instrument with imaging capabilities (e.g. Swift-BAT, INTEGRAL IBIS/ISGRI, etc.) and autonomous burst localizations obtained by Fermi (GBM).

    When none of these are available, for events that were observed by at least one other mission the localization is calculated to some extent using InterPlanetary Network (IPN) triangulation.

    For the nonlocalized events detected solely by KW we constrain the source location considering the information we have from KW ecliptic latitude range, ongoing source activity and Earth-blocking.

  • 2

    Temporal Analysis

    For the temporal analysis, we used time histories in G1 + G2 energy band (~ 20 − 320 keV at present), with a time resolution of 2 and 16 ms. The data starts at T0 − 0.512 s; where 2-ms resolution light curves are available up to T0 + 0.512 s, 16-ms – up to T0 + 33.280 s.

    The total burst duration T100, and the T90 and T50 durations (the time intervals that contain 5% to 95% and 25% to 75% ranges of the total burst count fluence, respectively), were calculated using a concatenation of waiting-mode and triggered-mode light curves with a method similar to that developed for BATSE.

    The burst’s start Ti and end Tf times are determined by searching an excess above background at the 5σ level on timescales from the best available lightcurve resolution up to 100 s. The background is approximated by a constant, using, typically, the interval from T0 − 800 s to T0 − 20 s.

  • 3

    Spectral Analysis

    We performed the spectral analysis in XSPEC using two spectral models, which have been shown to be the best fits to the broadband spectra of magnetar bursts.

    The first one is a sum of two bbodyrad (2BB) XSPEC models with the normalization proportional to the surface area.

    The second model is a power law with an exponential cutoff (CPL), parametrized as \(E_p\): \(f(E) \propto E^{\alpha} \exp(−(2 + \alpha)E/E_p)\) , where \(\alpha\) is the power-law photon index and \(E_p\) is the peak energy in the \(\nu F_\nu\) spectrum.

    We also tried to fit the spectra to an optically thin thermal bremsstrahlung (OTTB; CPL with \(\alpha=-1)\) and found that this model may be rejected on statistical grounds. We provide the OTTB model parameters only when both CPL and 2BB models fail to adequately fit the burst spectrum.

    The total energy fluence and the peak energy flux were derived using the energy flux of the each spectral model in the 20 − 500 keV band. Peak flux was calculated on 16 ms timescale.