BTOW Longitudinal Shower Leakage

Motivation:  Determine how much energy in a EM shower "leaks" out the back of the Barrel Towers.  Find a correction, if any, necessary to determine electron ET for high ET electrons from W decay.

Simulation Used:

Single thrown electrons flat in eta=[-1,1] and phi=[0,2pi] for a few fixed energies (10, 20, 30, 40, 50, 60, 100).  There are 25k thrown events per energy bin.  The geometry setup creates only the magnet, barrel and endcap (no material in front of the barrel)

detp geom y2009a

make geometry/cavegeo
make geometry/calbgeo
make geometry/ecalgeo
make geometry/magpgeo

Determining the longitudinal shower leakage

The sampling fraction is computed as the sum of energy deposited (from geant) in all barrel towers divided by the thrown energy.  If there is a substantial amount of longitudinal leakage one would expect to see a decrease in the sampling fraction at higher energies.  In the attched PDF file you can see the sampling fraction vs. energy for the 20 eta bins on page 2.  The samping fraction appears to be a constant for energies up to 60 GeV at mid-rapidity, however as expected at 100 GeV sampling fraction decreases.  Near the edges of the barrel (|eta|~1) the sampling fraction is increasing with energy (in the lower energy bins) so it may be harder to determine if a there is a decrease due to longitudinal leakage, but at 100 GeV the sampling fraction has not decreased like in the bins at mid-rapidity.  The sampling fraction for each energy, eta bin is obtained by fitting a gaussian to the sampling fraction distribution (individual plots on pages 3-8).