Back to the LGD

The next hypothesis to consider is that missing photons are hiding in the LGD. Maybe we fuse two photons into a single cluster, or maybe we are not seeing them at all due to inefficiency in the cluster recognition. To examine this, we again compared the simulation with the reconstruction in the LGD. From simulation, we use low angle photons (LAP) that have polar angle $\theta < 25^0$. Fig. 7 (right) shows the energy distribution of clusters (black) and low angle photons (red) from the $b_1$ reactions (mostly $b_1 \rightarrow 5\gamma$) reconstructed as $3\gamma$. Green and blue histograms show the LAP contribution when $N_{LAP}=4$ and $N_{LAP}=5$. The right plot on Fig. 7 shows the LAP multiplicity for the given cuts. We see a good agreement between photons and clusters except for very high and very low energy photons. At high energy the disagreement is probably because we fuse some soft photons to a high energy cluster. In the low energy region there is a clear threshold effect. If we can teach the clusterizer to search for soft photons in the LGD we should be able to suppress the $b_1$ background.

Figure 7: The energy of clusters and low angle photons (left) and LAP multiplicity (right) in the case of $3\gamma$ events from $b_1$ reactions.

Fig. 8 shows that this will not be an easy task. We compare cluster (black) and LAP (colored) energy for different cluster multiplicities from the same $b_1$ reactions and general requirements $N_{pixel}=1$, $E_{TOT}>4.0$ GeV (no $\eta \gamma$ selection). The red histograms represent the total LAP contribution for the given number of clusters. The other colors represent LAP's ordered by their energy. Again, we have agreement between clusters and mid-range energy photons; however we can have more then 3 photons even when we see just 2 clusters. In the case of 2 clusters, merging of clusters is particularly visible for high cluster energies. The generator produces essentially no photons above 4 GeV, while the reconstructed cluster energy distribution extends above 4. In the case of 4 and 5 clusters, a peak in the cluster energy distribution in the range 300-500 MeV comes from charged particles from the $b_1 \rightarrow \omega(\pi^+\pi^-\pi^0)\pi^0$ channel.

Figure 8: Comparison between cluster and LAP energy for different cluster multiplicities from $b_1$ reactions.