Run 20 AuAu9.2 EbyE T0 calibration

I looked at 4 runs spread across the three trigger setups (production_9p2GeV_2020, production_9p2GeV_2020b, and production_9p2GeV_2020c):

• st_physics_21038010_raw_3000003.daq
• st_physics_21041004_raw_1000002.daq
• st_physics_21080013_raw_1000002.daq (b)
• st_physics_21245006_raw_2500001.daq (c)

I found the EPD to be the best correlation. The VPD and BBC showed multiple bands, so they would additionally require more investigation than the straightforward EPD. Here is TPC weighted average time vs. EPD average time, with essentially no cuts (left), and then with requiring at least 5 hits in TPC inner/outer east/west (all 4) and non-zero EPD time (right):

ebyeT0tree->Draw("(tpctow*tpcnow+tpctiw*tpcniw+tpctie*tpcnie+tpctoe*tpcnoe)/
(tpcnow+tpcniw+tpcnoe+tpcnie):(epdw+epde)/ 2",
"tpcnow>5&&tpcnoe>5&&tpcnie>5&&tpcniw>5&&epdw>0&&epde>0","prof");

Next we take the profile plot and fit, as shown on the left below to arrive at the calibration numbers to use in the database:

• constant = [Mean y] - [p0] = -0.119 - (-0.09163) = -0.02737
• slope = - [p1] = 1.637e-5 

The right plot is the same data but separated by run, where we see that all 4 runs (different colors) show the same calibration validity across the entirety of Run 20. The magenta data is the same run (21038010) as the red, but after the calibration is applied. I did not investigate the oddities at the far right end, but they are in a region of low statistics, so a larger statistical sample would be needed to study further (and the oddities were not present in the other runs).



-Gene