Searching for Signals....

11-21-02: A broad search for signals in the full data set, focusing on the effects of lowered cluster seed energies and the improvements to the clusterizer, was performed over the last two weeks; a collection of "things I've found" follows. All of the images show three histograms using identical analysis; the top using seed energies (0.35, 0.15), the middle using seed energies (0.15,0.05), and the bottom using seed energies (0.15,0.05) and the clusterizer alterations I developed during the last year.

Overall, I'm enthusiastic about the results at this point.... The combination of lowered seed energies with cluster splitoff re-combination seems to produce improved signals, especially at high multiplicity. I'm working on further improvements, additions, and balances to the cuts, currently set conservatively at:

• One and only one BSD Pixel
• lgd_cluster_cleanup is satisfied (mostly fiducial radial cuts)
• Total energy in the LGD is between 4.8 and 6.2 GeV
• Total energy in any photon is greater than 150 MeV
• No more than 2 "in-time" hits in the CPV
• Bottom Histograms: Clusters combined when closer than 6 cm

Sample of Results:

• 2pi0 -> 4g: Peaks at ~950 MeV and ~1250 MeV
• 1pi0 1eta -> 4g: Peaks at ~725 MeV and ~1000 MeV
• 2pi0 g -> 5g: Peak at ~550 MeV
  omega pi0 -> 2pi0 g -> 5g: Peaks at ~1000 MeV and ~1250 MeV
• pi0 eta g -> 5g: Peak at ~1150 MeV
  omega X -> pi0 eta g -> 5g: Peaks at ~1000 MeV and ~1200 MeV
• 2pi0 2g -> 6g: Peaks at ~600 MeV and ~900 MeV
  2pi0 eta -> 6g: Peak at ~1000 MeV
  3pi0 -> 6g: Peak at ~570
• All 7g events
  2pi0 3g -> 7g

04-19-03: As requested during the last conference call, I've been examining the minimum photon energy cut. To mimic that cut I saved the lowest photon energy as a variable.... Note that this analysis uses prune_cluster_singles, cluster_cleanup, and splitoff_cluster_reclaim.

• Lowest Photon Energy vs Number of Photons in Event
  One can see that cutting at 50 MeV will have very little effect on any channel, and cutting at 150 MeV (the first cluster seed energy) will reduce the high-multiplicity sample by a significant amount.
• Number of Photons in Events w/ Pi0; 6-gamma mass w/ 3pi0
  Solid iE = 10 MeV; Dashed iE=70; Dotted iE=100; Dot-Dashed iE=130
  Note that in raising the photon threshold to 130 MeV we lose 50% of the 6-gamma events with at least one pi0, but only lose about 20% of the eta events in 3pi0 -> 6g....
• Various photon energies
  Bachelor photons look a lot like any other....
• (3pi0)g -> (6g)g -> 7g: 6g mass
  Cutting as high as 130 MeV doesn't seem to greatly enhance or degrade the eta peak....

Examining the 3pi0 -> 6g eta peak, specifically the number of events in that peak and a ratio to those not in the peak:

We do see some modest gain in signal-to-noise in the eta -> 3 pi0 -> 6g spectrum when increasing the individual photon energy threshold, but only at the loss of statistics. Since I don't really have any nice 5g peaks to use right now, I hesitate to set a number as "most-efficient"..... At this point, I tend to say that we should keep this cut low (as the UConn group has) and maintain statistics. The high number of low-energy photons may also be an argument for re-examining the cluster seed energies.