Endcap e/h desicrimination: using the ESMD and EPRS

A few weeks ago I gave a short presentation on a preliminary endcap W algo.  This preliminary algo had no requirements on SMD or pre/post shower.  This blog shows some SMD and pre/post shower distributions for both the data and some single particle MC to get an idea where background might be reduced.  Single particle MC samples were used as a crude estimate of our background signal since no filtered Pythia QCD sample exists with the correct endcap geometry.  The cuts used to select W candidates can be found here, and a movie showing the cut distributions for a W+ MC sample is here.

Single particle MC samples:

    An electron, pi plus, pi minus, and pi0 sample of 25K events each were all privately generated.  Each sample was thrown with flat in eta=[1,1.5], flat in pt=[10,50], and with a zVertex distribution similar to Run 9 data.

In each figure below there are 5 figures for each distribution

Figure 1: ESMD U+V summed energy (+/- 20 strips from the centroid) vs ETOW 2x1 cluster ET

Figure 2: ESMD U+V ratio of 7 central strips to 41 strip summed energy vs ETOW 2x1 cluster ET

Figure 3: ESMD shower profile sigma: V vs. U

Figure 4: Distance from ESMD shower centroid to extrapolated track at ESMD plane: Delta Phi vs Delta Eta

Figure 5:  EPOST 2x1 summed energy vs ETOW 2x1 cluster ET

Here is a 1D projection of the charged pion post-shower distribution, which is ~constant in the plot above.