Outlined here is the process of selecting the status table to apply for 2005 simulation data.

Obviously, we want a status table that is most representative of the EMC status over the entire run period.

First of all, we take a look at the SMD status tables acceptance over time (Eta, Phi) and see a number of valleys. Separating the tables into groups by the average acceptance:

• SMD table index #13-15, acceptance ~65% (27 runs, 6131092-6134011)
• SMD table index #45, acceptance ~70% (20 runs, 6157050-6158086)
• SMD table index #46-55, acceptance ~80% (Eta), ~85% (Phi) (121 runs, 6160039-6168044)
• SMD table index #56-58, acceptance ~75% (83 runs, 6168068-6170018)
• SMD table index #59-62, acceptance ~60% (36 runs, 6170031-6172093)
• SMD table index #63-64, acceptance ~40% (30 runs, 6174010-6174072)
• The rest have average acceptance of ~83% for Eta and ~90% for Phi.

BTOW acceptance is a lot more uniform than BSMD acceptance. However, there are a number of tables that have noticeably different acceptance.

• BTOW table index #1-8, 82-83, acceptance ~90%
• BTOW tables index #49,50,52,55 have slightly lower acceptance
• The rest have practically uniform acceptance at ~95%

The list of runs with their corresponding BTOW and BSMD tables.

The runs with lower acceptance can either be removed or treated separately and be given separate correction factor based on their acceptance.

Out of the above-listed groups, I will most likely remove runs with the following tables: 13-15, 45, 59-64, for a total of 113 runs or 15.3% of the number of runs. The next step before doing this is to check the number of events, and even the number of electrons that would be removed by removing these runs.

Looking at the above groups by number of events (by trigger, after cuts):

• Group 1: SMD table index #13-15 - MB: 43.5K, HT1: 139.3K, HT2: 109.3K
• Group 2: SMD table index #45 - MB: 38.9K, HT1: 89.3K, HT2: 101.5K
• Group 3: SMD table index #46-55 - MB: 280.1K, HT1:713.9K, HT2: 710.6K
• Group 4: SMD table index #56-58 - MB: 105.4K, 314.1K, 271K
• Group 5: SMD table index #59-62 - MB: 53.5K, 147.8K, 146.5K
• Group 6: SMD table index #63-64 - MB: 43.1K, 112.0K, 117.9K
• Total number of events (including the above runs) - MB: 960.6K, HT1: 3,085.5K, 2,625.2K

So groups 1, 2, 5 and 6 comprise of 18.6% (MB), 15.8% (HT1) and 18.1% (HT2) of events.

Looking at groups by number of inclusive electrons:

• Group 1: SMD table index #13-15 - HT1: 207, HT2: 171
• Group 2: SMD table index #45 - HT1: 139, HT2: 205
• Group 3: SMD table index #46-55 - HT1: 1,142, HT2: 1,334
• Group 4: SMD table index #56-58 - HT1: 538, HT2: 573
• Group 5: SMD table index #59-62 - HT1: 312, HT2: 343
• Group 6: SMD table index #63-64 - HT1: 147, HT2: 186
• Total number of events (including the above runs) - HT1: 7,207, HT2: 7,162

So groups 1, 2, 5 and 6 comprise of 11.2% and 12.6% of HT1 and HT2 electrons, respectively.

Considering the above, runs included within groups 1, 2, 5 and 6 will be removed from analysis to reduce the need to apply large corrections to the electron reconstruction efficiency. Runs in groups 3 and 4 will be kept to preserve the statistics, but will most likely need some sort of correction to bring their efficiency to the average efficiency for the rest of the run.

Runs removed: 6131092-6134011, 6157050-6158086, 6170031-6172093, 6174010-6174072

For the rest of the run, the SMD acceptance is fairly flat. The differences in status is probably due to single strips changing status over time, instead of large scale (module or even crate) failures. A single status table with acceptance closest to the average acceptance over runs will be used as the table in analyzing simulation data and calculating reconstruction efficiencies. To find which table is best, we'll first average the acceptance over runs, after removing runs that use the tables outlined above.

Removing runs with lower acceptance tables, the average (over run) acceptance is 83.1% and 90.1% for BSMD Eta and Phi plane, respectively. While the bad status area of Eta and Phi planes are not independent, for the purpose of selecting a status table to represent the average status over the run, I'm combining the acceptance (by multiplication) for an average of 74.9% overall SMD acceptance (in practice it's likely higher). I then find the status table with the closest SMD acceptance, which gives me SMD table #9. If done separately, the closest to average tables for Eta and Phi are #35 and #41, respectively. Since I already picked tables that have similar acceptance, it really shouldn't make a big difference which of the table is used.

Table 9: Run 6122010 - 6122018, time stamp: 20050502,002411 - 20050502,065708

Table #16 is excluded because while the overall SMD acceptance is in line with the run average, the individual Eta and Phi acceptance is actually noticeably lower and higher, respectively than the average over the runs.

Table 16: Run 6134024 - 6134060, time stamp: 20050514,085956 - 20050514,221550