Add option for linear time dependence to photom step

[stscijgbot-jp]

Issue JP-3887 was created on JIRA by Karl Gordon:

For multiple NIR instruments, we are seeing a slow throughput decline of around 0.2%/year. This has been been reported in the JWST AbsFlux Expert Team and other venues. Such a decline is expected given the slow degradation to the JWST primary due to space weathering. It would be good to proactively add the option to the photom step to expand the current allowed function from from exponential to be exponential, linear, or exponential+linear. The exponential covers the current behavior seen for MIRI, linear covers the current behavior seen for the NIR instruments, and exponential+linear will cover the expected behavior for MIRI as it should see the same primary mirror degradation as the NIR instruments in addition the current exponential.

As this affects all instruments and modes, I have left the choice at none as there is no option for all. Picked 1.17.2 for the affected pipeline version to pick one. Really applies to many version.

[stscijgbot-jp]

Comment by David Law on JIRA:

Thanks Karl Gordon ; will put this on a JP agenda to start discussing.