I'll start with the data Ben has from runs 19132074, 19132075, and 19132076 (top to bottom) before (left) and after (right) he applies the correction....

(<sDCA> vs. time [sec])

Run 076 clearly has some really large event not seen in the DAQ data. Ben has shown that it persists at all phi...

(sDCA (color axis) vs. phi and event in file)

and on both the east and west sides...

(sDCA vs event in file for east and west)

...so it appears to affect the entire TPC!

-Gene