GLAST Calorimeter Software Weekly Report
Week of  9 August 2004

NRL

Sasha:  Produced step-by-step plan for new single CDE recon:

• For use in CalRespSvc
• Energy and position reconstruction (using geom mean for energy)
• Improved performance
• Deals with "mixed" events where one end of the CDE comes from one size diode and the other end from the other.

Working on similar plan for changes that need to be made to CalibGenCal:

• Supports current understanding of FLE crosstalk contribution to IntLin
• Muon calibration algorithm changes to support geometric mean energy estimator
• Supports new requirement for FLE thold calibration (currently looking at simplified technique)

Zach:  Continued work on CalRespSvc infrastructure and interface.  Will get a "toy" version (responds predictably but with fake numbers) out by end of week for use by digi and recon coders.  Real version shortly thereafter.

Andrey: 

For the past week I was looking at the tapering curves (Sqrt(P*M) and (P+M)/2 vs longitudinal position) for the crystals in the first 2 layers of the EM on high-energy muons.

The conclusion we came up with Sasha is that the behavior of the curves varies from crystal to crystal (i.e. they have different shapes).  Particular shape of the curve can possibly be explained by what kind of light component is dominating in the parts of crystal - exponential or linear.  Although tapering curves look quite different from crystal to crystal, the variations (difference between max an min) are of the order of 2% or less.  Most curves can be fit pretty good with the quadratic polynomial, if we throw always the most left and most right bins along the crystal length (I am using a total of 10 bins).

The geom. average of the ADC signals from 2 ends, seems to have a smaller difference between min and max, sometimes it is smaller by a factor of 1.5 when compared to the arithm. average.  Also the muon peak itself is narrower for the geom. avg. than for the arithm. avg.

Right now I am collecting data (from the top EM CAL layer) to have a more detailed pictured of how does the P/M (M/P) ratio vs transverse coord. vary with the longitudinal position, since I just have data for distances 145 mm and closer to the end.  Sasha was interested in how P/M behaves when we move further apart from the photodiode.

Mark:  Analyzed test data from 1620 flight CDEs to confirm that geometric mean of ends gave an energy estimator that was effectively independent of position.  Similar to Andrey's study but with more xtals in less detail.  Will post a report of results shortly but agree with what Andrey has found on EM

Continued documentation work.

LLR

Pol:  I worked on high energy gamma reconstruction, both for energy and direction.

CENBG

Benoit: I worked on the quenching note, and included my findings concerning the presence of a slow component with a decay time of about 10 microseconds in the muon pulse, with a weight of 25%. Sasha mentioned to me that similar results were obtained by the BELLE collaboration.  The antiquenching could come about from the strong reduction or absence of this component for heavy ions as holds true at low energy, combined with the filtering features of our electronics.  The note should be distributed soon.

In the I&T area, I looked at the improvement in the position resolution using the information from the bottom trays instead of that from the top trays. The improvement is about a factor of 2 for 4 GeV muons.

GAM no report

SLAC  no report

UCSC  no report