Subject:
DoAr 21 ISIS parameter limits
From:
Marc Gagne <mgagne@wcupa.edu>
Date:
Fri, 08 Aug 2008 18:12:57 -0400
To:
David Cohen <dcohen1@swarthmore.edu>

David,

Please keep this file for your records: it details my ISIS CHISQR fits to
the combined (but not coadded) +/-1 HEG/MEG spectra of DoAr 21, optimally
rebinned to 20 counts per bin, with the exception of the He-like Si, Mg, Ne
density fits which used CSTAT on the unbinned combined spectra in only
narrow wavebands around the He-like complexes.

For reasons we could discuss ad nauseum, the CHISQR fits on binned data
converge on finite limits, whereas the CSTAT fits on unbinned data usually
do not.

Summarizing, the parameter, best-fit value, and 1-sigma lower and upper
confidence limits are as follows. Note that as better fits emerge through
the fitting of other parameters, the best-fit values also wander around,
sometimes, outside the prior formal limits.

nH      0.9831352   0.963041    1.0006 (1e+22)
norm1   0.001489894 0.00135286  0.00172229
norm2   0.01380541  0.0133497   0.0139204
temp1   1.38829e+07 1.32374e+07 1.54275e+07
temp2   6.4309e+07  6.26475e+07 6.9524e+07
density 1.31567e+08 1e+06       1.15973e+13 (ul)
Fe      0.3493926   0.310759    0.396759
Si      0.7022756   0.64686     0.760835
vturb   130.9176    0           200.654 (ul)

Let me know when you've got the time resolved data ready to go. Have a good
weekend. 

Marc



doar21_he_modifier_chisqr_1sigma_limits.out
	
Content-Type:
	application/octet-stream
Content-Encoding:
	base64