Curved sky lines and edge effects in kr*_sky.fits files

[rremigio]

I am trying to reduce some sky exposures at red wavelengths (RCWAVE = 8900 A, RM2 grating) and have run into two issues when trying to reduce the frames.

One issue is that between the creation of the master object and the master sky, the master sky has many sky lines that are curved, where the effect is more pronounced as I move towards the red and blue ends of the slices

The second issue is that at the very red and blue ends of the slices, there is either a sharp cutoff to zero counts, or a sharp increase to tens of thousands of counts. I've narrowed it down to a possible flat fielding issue since this doesn't show up in the .intd files (top left) but is present in the .intf files (top right). The stacked twilight flats look relatively normal (bottom left) but the master twilight flat (bottom right) has the issue.

Is there currently some way to address this issue I am running into, and is it wavelength-dependent? If it helps to know, I have not run into this issue at lower wavelengths (e.g. RCWAVE = 6950 A, RH2 grating).

[rremigio]

Update: I tried re-reducing the data with dome flats. This alleviated most of the sharp dropoffs at the red end of the slice, but you can still see the unusually high flux regions on the blue end.

In the next two figures, the master object of the sky exposure looks fine (left), but again I run into very bright regions on the blue end of the slice.

Here is the dome flat image used in this new run