There is a problem with the Fe 4-2 transition strength
All the cooling flows whose spectra we have analyzed show negative
residuals in the range 1.4 - 1.5 keV. This is best illustrated by
the spectrum of the center of the Centaurus cluster. In the figure
below the 1.4 - 1.5 keV region has been ignored in the fit. This
problem has been identified by Liedahl et al. as an incorrect ratio
between the Fe 3-2 and 4-2 transition strengths. The correction for
this is known but has not yet been implemented in any commonly available
plasma emission codes. Until this is done we have ignored the relevant
channels in all spectral fits.