B.2 EXPLORATIONS WITH THE STANDARD DATA PRODUCTS

Our initial approach on this question was that we might actually find usable data from OSSE at energies greater than 10 MeV. Our other concern was that if this data were unusable, were there any 'false detections' which might mislead other researchers with less experience with the instrument and dataset.

The standard OSSE data analysis products utilizes what is called the low and medium energy ranges. These ranges cover from  50 keV to  10 MeV. To include the high range channels in the analysis, some non-standard processing must be performed. In the standard products, the detector response matrix is generated for energies up to 10 MeV.

Initially, we utilized some of the high-level data products currently available in the OSSE archive as the starting point for this analysis. In the standard processing the background subtraction is performed and the resulting 2-minute spectra are summed into orbital or daily total spectra before the detector response matrix is generated. We examined the counts in the background-subtracted datasets accumulated on daily timescales. This data still had the high-range detector channels (channel numbers 528-544) in the dataset. Summing these channels over the length of the observation, we performed a selection on observations with at least two channels of positive counts at the 2-sigma level. This selection yielded 13 `detections' out of the 197 observations checked. Among these `detections' were the Galactic Center, 3C 279, LMC X-3, IC 4329A, etc. As a sanity check, we also looked for `sources' at a negative 2-sigma count level. This search yielded 15 `detections' with 3C 279, Cas A, and Cygnus X-3 among the group. Our conclusion was that we were selecting on random noise. We decided to refocus the exploration on sources that we expected to see above 10 MeV.

We decided to utilize observations of the Crab pulsar for three reasons: 1) the source is bright (though not the brightest) in the EGRET energy range; 2) there is a lot of publicly available OSSE data on this target; and 3) the Crab is something of a `standard candle' in X-ray astronomy. If the Crab is undetectable by OSSE above 10 MeV, it bodes ill for many sources, particularly blazars, which EGRET readily detects.

First we needed to generate an instrument response matrix which included the high range detector channels. In the IGORE data analysis environment, the FFIT_PREP routine generates the detector response matrix and stores it with the count spectra summed by detector.

status=ffit_prep(file='igr:[dat]nom_cfg.fp', $
       fileroot='crabpulsar', sdbfile='CRABPULSAR_1_GSM.SDB',$
       /highrange,mtrx_nphot=550,mtrx_prange=[0.05,250.0])

We force the high-range response matrix generation with the /HIGHRANGE keyword and expanding the energy range and number of photon energy channels.

Using the response matrix thus generated, we use the standard fitting tools to fit a power law to the spectrum. In this particular case, the detectors only go up to about 150 MeV. We also consulted Nolan et al. (1993) for the Crab flux in the 30 MeV - 10 GeV range and included it for comparison. The results are presented in Figure B.1 below.

  
Figure B.1: Using OSSE High-Range data through the standard data products. The Crab clearly does not appear above the error bars at energies above 10MeV.