As a test for flat field telluric line removal we tried a 270K, 260K and 250K on hi-med 1640cm-1, Sirius. That question will be explored later, but what is immediately apparent is that there is a bug in the treatment of the flux calibration. exes_makeflat.pro is supposed to normalize by a planck function using the header BB_TEMP parameter. See below that the exact same data files and reduction give very different values of the continuum level in ergs s-1 cm-1 (cm-1)-1.
The flat fields used were 40001 (blue),40002(yellow),40010(orange),40011(red),40013(magenta).
Another small issue that we intend to investigate is the changing T of the telescope mirror in flight. The decline in sky continuum level was apparent in the sky frame of these flat fields. Also a few narrow looking water emission lines were present against the blackbody and declined slowly with time in successive flat fields.
UPDATE: Nov 7, 2016
Based on the blog post "Flux versus temperature ..." from Sunday, Nov 06, 2016, we know that the effective temperatures of the blackbody with the ZnSe are different than the setpoint. From the plot of flux versus temp for 1640cm-1, we estimate that T=250K --> 275K, T=260K --> 278 K, T=270K --> 282K. We sub these values into the header parameter BB_TEMP and reduce as before -- and the resulting fluxes calculated from flat fields is much more consistent. Same color scheme as in the previous plot --red 270K, blue 260K, magenta 250K.
I can get to the same rescaling by just multiplying the data by the ratio of fluxes for the actual temperature / assumed temperature, so no re-reduction is actually needed.
The units here are erg/s/cm2/cm-1/SR, according to the pipeline headers.
