[EXOSAT Home] [About EXOSAT] [Archive] [Software] [Gallery] [Publications]

GSPC CALIBRATION UPDATE
(
ref. EXPRESS no. 6, p. 18)

 

Background Line Energies, Subtraction and Detector Gain

 

Further work on Cas A has quite clearly demonstrated that the two lines in the background are caused by the L alpha symbol and beta symbol transitions of Bismuth. These lines arise from the activation of the lead collimator material to Pb21O, which then decays, with a half-life of 4 days to Bismuth to give the observed L emission lines. The energies of these lines are 10.83 and 13.01 keV respectively (averaged over the alpha symbol1, alpha symbol2, beta symbol1, beta symbol2 branching ratios of 100, 10, 50 and 20).

 

Experience now shows that, except for extremely bright sources (e.g. Sco X-1), the channel number of the 13.01 keV line can be established to within one channel (gain 2.0) by fitting two gaussian-lines and a polynomial of order 3 or less to the data. The same is true for the 10.83 keV line although for some brighter sources the uncertainty in the measured channel number may increase (e.g. for a bright bulge source). The 13.01 keV line can be reliably used to lock the gain of the GSPC for any particular observation. Photomultiplier stimulation data should be used to check only the gain drift throughout an observation.

 

It has been found that the apparent energy of the 4.75 keV feature, when fitted as a line depends by up to 10% on the shape of the underlying continuum. It is now unnecessary to fit this feature which should be included directly in the response matrix, by overlapping the energies of the matrix by 70 eV at 4.78 keV.

 

The Effective Areas

 

Analysis of several long exposures on the Crab in October and November this year produced further refinement of the effective areas. A series of splined three order polynomials were fitted to the Crab data in order to force the response to give the correct fit. The increased exposure time for a burst length window of channels 89-107 of 25800 s has revealed some small deficiencies in the previously quoted values. The revised values are given in Table 1 and, are good to a systematic uncertainty of 1%.

 

An update to the CCF is in progress and will be distributed as soon as possible (details in the next EXPRESS).

 

N.E. White

Table I: GSPC Effective Area
Energy89-10489-107
1.6 .072 .223
1.9 1.040 1.591
2.2 4.644 5.363
2.5 9.722 12.757
2.8 15.611 20.190
3.1 22.336 28.424
3.4 30.464 38.072
3.7 36.348 45.553
4.0 41.842 53.136
4.3 46.024 58.534
4.6 50.498 63.463
4.9 54.328 68.758
5.2 55.477 70.439
5.7 58.087 74.514
6.2 62.543 79.298
6.7 65.880 83.188
7.2 67.803 86.548
7.7 70.007 89.545
8.2 72.351 90.589
8.7 75.331 91.819
9.2 75.859 92.550
9.7 77.415 93.926
10.2 78.601 04.427
10.7 78.435 94.526
11.2 78.255 94.225
11.7 78.016 93.531
12.2 77.676 92.455
12.7 77.194 91.014
13.2 76.471 89.154
13.7 75.592 87.040
14.2 74.466 84.686
14.7 73.192 82.249
15.2 71.346 80.093
16.2 66.658 76.419
17.2 61.906 71.541
18.2 57.218 66.080
19.2 52.723 60.579
20.2 48.484 55.412
25.2 31.338 39.613
30.2 20.485 25.894
35.2 54.906 69.404
40.2 44.888 56.741
45.2 36.476 46.107
50.2 29.604 37.421
55.2 24.082 30.440
60.2 19.779 25.002
65.2 16.311 20.618
70.2 13.614 17.209
75.2 11.431 14.449
80.2 9.633 12.176


[EXOSAT Home] [About EXOSAT] [Archive] [Software] [Gallery] [Publications]