2KM Group Wiki

To do list for 2km-geant4 code

NEW ITEMS:

• 2KM code to generate vectors (based on Hayato's own code for 280m & SK ) : DO NOT USE TRandom3(0) for the objects. Explicitly set the random seed yourself.

  • Since each file has a unique number (from 1 to whatever) use that number. Otherwise the beam profile is seriously messed up. The first 0.9 years of numu ntuples have this problem.

• MuonMinusCapture was removed from the GEANT4 simulator code, because of a possible bug. Will need to be turned back on next time.

• polfit5 : understand why narrow 10 ns timing cut fails at 2KM.

OLD ITEMS :

• Look more carefully at the NC events
• Are there "fake low energy events" in the SK vectors which will mess up f/n ratio?(Can we apply the same explicit cuts @SK?)
• numu spectrum
• OD size

• understand the PID disc between MC & data in 1KT (related to PMT shapes ?)

• Fix the reconstruction code so we can use full fiducial volume - DONE!!!

• For later on (definitely not for the CDR/proposal) :
  • fix the digitizer :

    1. constant time offset +940 ns added to individual WC hits (works if events are generated inside the water tank. What if they come from outside ? the G4 clock starts earlier so all the WC hit times will be artificially delayed) + only the time of the first hit is used for each PMT. STILL a problem but trigger T0 is now taken into account

    2. all Cherenkov photons ever produced are added. Need an integration window (1300ns @ SK?) DONE

    3. Decay electron finder
    4. add a real cathode in the PMTs and make it the sensitive detector
• Use different QE for 8" tubes.
• Time jitter : at the moment we use 2 different functions based on a scaling argument. Make sure that it is valid for 8 inch tubes.
  • Cutoff : I prevent the jitter from being lower than 0.58 ns (based on 1KT tabulated data in /skam). This is reached for ~30 p.e. with 8" tubes using the present

function. TO BE CHECKED

Here is info on the number of tubes.  If we can use the smaller # it would be cheaper.  We should test this sometime in future.

Both Kajita-san and Maxim agree that the PMT ring is at 8.5 meters and
the "tank" is at least 9.3 meters away.  The geometry agrees with
Kajita-san's PPT file.

The tube numbers are as follows:
           endcap*2 + rings*tubes/ring = total
Maxim:    688*2    + 42*102           = 5660
Kajita:   696*2    + 42*93            = 5298

Both allow a 3x4 tube grid 3*31 = 93
                           3*34 = 102

Kajita-san:  uniform 28 cm tube to tube spacing.
Maxim:       28 cm ring-to-ring 26.2 cm tube-to-tube in phi

So the difference is due to the 3x4 PMT grid being different. 

Here is how many new events we should get...

SK : 50,000 nue events
     500,000 numu events . That's what we have now with 03b.
  With the web page we have about half that number of useful numus.

Correction, the previous numbers were not quite right : 

volume              |    numus/1year

56t                    ~150,000

100t                   ~268,000

56t+1m = 185t          ~496,000

100t+1m = 275t         ~736,000

The final reduction rate in the present nue appearance analysis
is ~1.2e-3 (counting numu CC & NC together) (= #final/#interactions).

56t     -->      180  useful numus/ 1year
100t    -->      320   '  '  '  '  '  '

We presently have 1/3 year (counting the 'good' events only).

We need 10 years for numus. For nues we already have 14.3 years in FV56t. That's less of a problem.