Just an idea I'm putting here. The main noise components of a master dark are fixed-pattern noise (FPN), dark current, and clock-induced charge (CIC). Dark current is temperature-dependent, so a synthesized master dark created from data taken at temperature T1 will not be the same as one created from data taken at temperature T2. In practice, if the temperature of the CGI actually changed substantially in flight, from T1 to T2, more frames could be taken at T2 to make another synthesized master dark. However, if time were short, a back-up plan could be to use a synthesized master dark created from noise maps created from T1 data with an adjustment to the dark current from T1 so that it is appropriate for T2. There are models for the temperature dependence of dark current that could be employed.
Just an idea I'm putting here. The main noise components of a master dark are fixed-pattern noise (FPN), dark current, and clock-induced charge (CIC). Dark current is temperature-dependent, so a synthesized master dark created from data taken at temperature T1 will not be the same as one created from data taken at temperature T2. In practice, if the temperature of the CGI actually changed substantially in flight, from T1 to T2, more frames could be taken at T2 to make another synthesized master dark. However, if time were short, a back-up plan could be to use a synthesized master dark created from noise maps created from T1 data with an adjustment to the dark current from T1 so that it is appropriate for T2. There are models for the temperature dependence of dark current that could be employed.