Near Detector Coil Current and Gains

We have two known bad periods in the near detector gains over time. The first is from Nov. 7, 2004 to Jan. 8 2005, where the numbers in the database are simply repetitions of the measurement from Nov. 4. The second is from Feb. 12 to Mar. 15, 2011, where gains were low due to a high voltage issue.

After we remove those two periods, we have the following plot, where periods in which the detector was shutdown are shown as gray circles:

Removing all gains periods that contain any time longer than 60 seconds with the coil current isOK flag set to false removes 312 gain periods; we call this the coil current mask. However, most of these appear to be during shutdown periods, as indicated by the near absence of gray circles after the coil current mask is applied.

We can also look at the effect of removing these periods on the temperature corrected plot and fit. Without the coil current mask, the infamous bump straddling 2007 and 2008 is clearly visible, as well as several outlying points.

Applying the mask yields

The bump remains, though it is less populated now. Also, the mask appears to remove all but one of the outliers.