Minutes y05m09d09

Present: David, Aya, Sean, Kotoyo (UW-Madison)
         Keiichi, Carsten, Kael, Shigeru (Chiba)

1 Status Report

   a. MC data production (Keiichi, Aya)

     Started on both Madison and Chiba side.
     Takes long CPU time to produce EHE data above 10 EeV.
     Keiichi reported that it took an 1 hr/event for 10 EeV,
     4 hrs/event for 100 EeV (EeV=1000 PeV).

     The hit-constructor/PSI are the prime suspects
     to slower the computing time. Keiichi will investiate
     to pin down a location of the code to cause this problem.

  b. Weighting module, a first MC data look (Aya)

     Aya uploaded the weighing module called "weighting-module"
     to the sandbox. It handles events generated by both Juliet
     and Anis generators. The pre-calculated Juliet matrix data
     files are sucessfully read out to calculate the intensity
     of muons/taus at the IceCube depth, which is one of the key
     elements among the MC weights.

     Aya also devloped the nice checker module to make a TTree
     to run a quick analysis. Some initial plots are found in
      here

     They generally look good except

     (1) a strange split of energy population in the brightest cascade.
     Shigeru pointed put that this may be due to the JULIeT's trick
     to interpolate differential cross section data into lower energy regime
     below 100 TeV. Needs to check its effects in the analysis ends.

     (2) Azimuth values in the MC weight dict are all negative.
     Keiichi said it might be due to double-call the convert function
     from the JULIeT definition. But as far as I (Shigeru) see her code,
     looks OK. Needs to check.

     (3) The spectrum of MC primary particles has a higher cutoff
     at 1 PeV, which should have been 100 TeV.

     (4) Both primary differential flux AND Juliet's propagating neutrino flux
     exhibits a bump structure at around 2-3 PeV. This is hard to understand
     because they are independent variables each other.

     (5) A strange behavior in number of total PEs as a function of number
     of OMs with hit.


  c. Cascade reco module (Sean)

   Sean developed the waveform-based reco module "wf-llh". Currently using
   the Pendal's function as a PDF, but it produced encouraging results.
   See his webpage
   for details. A simple noise cleaning like removing DOMs with only a single!
   ATWD bin would work. A next move is to use Photonics-based PDF, which is
   expected to come soon.


2. Near Future plan about the reco (everybody) - after the London!

   Would need to update PSInterfaceLite to implement the photonics PDF.
   Kotoyo will be in charge of this.

   It is necessary to have a fast and relatively simple "first guess" reco
   program of muon track before applying the cascade reco module to EHE events. 
   David nicely suggested that "line-fit" project should work
   good enough for high energy muons. It relies only on "TDC and ADC" info,
   which is very good.


3. The Oxford/London meeting.

   Shigeru encouraged everybody to give a short talk in the Oxford
   to introduce what you are doing to the collaboration. 
   Send the short abstract to Shigeru by next week.