Time dependence of dome flats

15 Aug 2014 GaryB

Large (>1%) and structured changes appeared in dome flats between Y1 (20111117) and Y2 (20140807) as noted by Robert Gruendl. What are they from?

I have taken g-band dome flats (mostly just single exposures) from each month from Feb 2014 to Aug 2014 and plotted their ratio to the combined g-band dome flat from 7 Aug 2014 engineering night. The bright (~1%) ring in the center exhibits much structure and there are also sharp features at the very E-W tips of the array.

It is seen that the bright ring represents a change that occurred between 0218 and 0315. This is consistent with its origin being scattered light from the inner FCM and shutter surfaces that disappeared when these were painted black in the 14 March 2014 engineering run.

Another bright and structured feature is visible in the central CCDs in the March and April dome flats. This went away (or more accurately, approached the state of the Aug 2014 flats) sometime between 15 April and 16 May 2014. I do not know what this could be.

The attached files are labeled with the date of the numerator flat; 20140807 is always the denominator.