Test of Corrected EEMC geometry: Pion Reconstruction

After the bug in the EEMC geometry was fixed a simple test would be to compare the neutral pion invariant mass spectrum before and after. 

The simple MC events have 2 pions in each event

• flat in eta [1,2]
• flat in phi [20,40]  ie. center of sector 2
• flat in pT: two samples of 5000 events each -> [5,7]  [7,9]

Simulation using full STAR

    in Kumac use line "detp geom y2006g; gexec $STAR_LIB/geometry.so" in DEV and SL09c to get new and old geometries.

Figure 1: Invariant mass distributions for new and old geometries (left figure -> 5<pT<7  and right figure -> 7<pT<9)

Summary:  So far it looks like the new geometry doesn't contribute to the low mass background like what we see in the data.  This may not be to surprising since with the Data Driven simulations the dipion samples also didn't generate the same low mass background as the data,  but once we used the Data Driven Maker with PYTHIA the low mass background become more apparent.

The pT dependence of the invariant mass for the the new and old geometry:

For the figures below I summed the 2 pT samples above to give a total sample of 10K events flat in pT [5,9].  The plot titled "Centroid of Inv Mass vs pT" is the centroid of a fit to a few pT slices in the 2D plot above it over a mass range of [0.1,0.18].

Figure 2: Invariant Mass vs. Reconstructed pT of the pion for the New Geometry (individual fits can be found here)

Figure 3: Invariant Mass vs. Reconstructed pT of the pion for the Old Geometry (individual fits found here)