5 Non X-ray background (NXB)

  A total of 498.6 ks of NXB data was accumulated using LECS dark Earth pointings. The dependence of the 0.1-10 keV count rate, ${C_{\rm T}}$, (s^-1) on time is shown in Fig. 5, where each data point is averaged over the entire FOV. The interval without data points results from successive gyro failures on BeppoSAX, which led to a 3 month observing hiatus. A gradual reduction in the counting rate by $\sim$15% over an interval of $\sim$2 years is evident. This may be modeled as ${C_{\rm T} = 0.1069 - 2.2 \ 10^{-5} \Delta T}$,where ${ \Delta T}$ is the number of days since launch. Alternatively, an exponential fit indicates a decay constant of 11.8 years. The same trend is visible in data extracted from the central 8$^\prime$ radius and background semi-annuli regions (see Sect. 6). The intensity of the primary (5.89 keV Mn K$_\alpha$ and 6.49 keV Mn K$_\beta$) radiation from the ⁵⁵Fe calibration sources decays with a much shorter half-life of 2.73 years. However, the calibration sources contain small amounts of other radioactive materials, the decay of which could contribute to the decrease in NXB intensity. Another factor may be the decreasing particle background in low-Earth orbit as the next solar maximum is approached, due to the varying height of the Earth's atmosphere. Such an effect is evident in studies of the ROSAT High Resolution Imager NXB (Snowden 1998). There is little evidence in the LECS data for "Long Term Enhancements'' or LTEs as seen by the PSPC (Snowden et al. 1995). These count rate increases are strongest below 0.28 keV and are approximately uniform over an orbit. They last between 15-30 ROSAT orbits and have peak intensities comparable to the low-energy CXB.

In the central 8$^\prime$ radius extraction region the 0.1-10 keV NXB count rate is $5.2 \ 10^{-6}$ s^-1 keV^-1 arcmin^-2. The dependence of the NXB on position within the FOV is shown in the right panel of Fig. 3. The difference in intensity between the most and least intense regions is a factor 3.1. In contrast to the standard background (left panel of Fig. 3), the NXB intensity near the center of the FOV does not show an enhancement. Within the central 5$^\prime$ radius, any such increase is <4% of the total intensity.

  

Figure 5: The dependence of the FOV averaged LECS dark Earth count rate on time. The launch date was 1996 April 30

Spectra of the LECS NXB are shown in Figs. 4 and 6. The overall level is approximately constant with energy with 3 discrete features superposed on a smooth increase $\mathrel{\hbox{\rlap{\lower.55ex \hbox {$\sim$}} \kern-.3em \raise.4ex \hbox{$\gt$}}}$4 keV. These features can be modeled as narrow Gaussian emission lines at $5.12 \pm 0.04$ keV (at 68% confidence), 8.51 $\pm$ 0.04 keV and 10.73 $\pm$ 0.07 keV. The first feature shows a clear position dependence, being stronger close to the calibration sources. It is almost certainly produced by 5.9 keV characteristic X-rays from the ⁵⁵Fe calibration sources that penetrate deeply into the detector before being absorbed. The detected energy is lower than the natural energy of the events, because X-rays absorbed deep within the detector produce, on average, less light than X-rays of the same energy absorbed close to the entrance window due to the different scintillation lengths in the driftless gas cell (Parmar et al. 1997b). The other 2 features do not exhibit an obvious position dependence within the FOV and may originate from fluorescent excitation of L-shell transitions in the tungsten window support structure (see Parmar et al. 1997b).

To illustrate the spectral changes associated with the decrease in count rate shown in Fig. 5, Fig. 6 shows the spectrum of the NXB before and after day 400. This shows that at energies $\mathrel{\hbox{\rlap{\lower.55ex \hbox {$\sim$}} \kern-.3em \raise.4ex \hbox{$<$}}}$8 keV the long term temporal evolution of the LECS NXB is not strongly energy dependent and may be simply modeled as the change in overall normalization given above. Above 8 keV, the intensity variation with time is less marked. This may indicate that the fluorescent line features exhibit less intensity variability than the rest of the spectrum.

  

Figure 6: The dependence of the rebinned LECS NXB spectrum, averaged over the entire FOV, on time. The filled circles show the spectrum obtained before day 400 and the open circles after day 400. The lower panel shows the ratio of the two spectra

The geomagnetic rigidity is a measure of the minimum momentum required by a cosmic particle to penetrate the Earth's magnetic field down to the position of the satellite. Due to the almost circular, low inclination BeppoSAX orbit, the variation in rigidity around the BeppoSAX orbit is less than is typical for low-Earth orbiting spacecraft such as ASCA which has an orbital inclination of 31$^\circ$ and experiences rigidities between 6-14 GeV c^-1. In the case of BeppoSAX, the rigidity typically varies from 10 to 16 GeV c^-1 around the orbit. In order to investigate the dependence of the NXB spectrum on rigidity, spectra were accumulated for intervals when the rigidity was $\leq$ and >13 GeV c^-1. The two spectra have almost identical shapes, with the spectrum accumulated when the rigidity was $\leq$13 GeV c^-1 having an overall normalization $5 \pm 1$% higher. This means that for all but the shortest observations temporal averaging will ensure that the dependence of the NXB on geomagnetic rigidity can be ignored.