1. Progress Report on GEANT Study of Containerized Detectors R. Ray 7/11/03 What’s New Since Last Time?  More detailed container description in GEANT o Slightly less steel in new container o Slightly larger gaps  RPC efficiency of 95% implemented  Cross-talk implemented  Plywood absorber rather than plastic + air  Generating large samples of events on the Farm  Generated beam e samples  New event number normalizations (used LOI numbers last time)

2. General Strategy  Implement containerized detectors in GEANT.  Use MINOS event generator with a flat energy distribution  Weight interaction vertex in GEANT by number of target nucleons in various materials  Parabolic fit to multiple tracks in an event. Assume longest track is electron.  Weight final distributions by evolved energy distribution of the beam  Generate event samples with: o Interaction vertex in transverse center of a container o Interaction vertex uniformly distributed o All steel structures replaced with air

3. 2.8 cm thick Wood Floor Corner Blocks 17.5 x 8.5 x 16.0 cm 3 GEANT Implementation Detector dimensions: 5.89 x 2.39 m 2 Corrugated skin 2 mm thick Corner Post (6 mm thick steel) O.D. (m)I.D. (m) Length6.0585.898 Width2.4382.350 Height2.5912.393

4. Top View

5. RPC 11 planes per container x, y readout at each station 3 cm cells Plywood Absorber 10 full layers + 2 half layers Full layers 18.3 cm thick, 1/3 X 0

6. 50 kton Detector 4 x 10 x 50 Stack of Containers x y z 20.1 cm horizontal and 19.8 cm vertical gaps between detectors in adjacent containers. Vertical gap determined by difference between inside and outside dim. of container. Horizontal gap determined by container dimensions and cell guide spacing Cell Guides 50 kt of absorber 2.1 kt of steel 0.7 kt of wood 0.14 kt of Al.

7. Neutrino Energy Spectra Flat neutrino spectrum generated between 0.5 - 3.5 GeV for e and 0.5 - 20 GeV for  and Beam e. Weight applied at ntuple level. e CC  NC Beam e Expected Number of events after oscillation for 50 Kt detector, 5 yr run, 85% fiducial, before cuts:  e CC669   NC6878 Beam e 463

8. Vertex Distributions Center of detector

9. Vertex Distribution Vertex in plywood absorber

10. Cross Talk Implemented in GEANT Data provided by Valeri. New measurements are 4X better. (cm)

11. electron  proton   neutron e CC Event Display Using NUGEN Input

12. electron  proton   neutron e CC Event Display Using NUGEN Input

13. electron  proton   neutron  NC Event Display Using NUGEN Input

14. electron  proton   neutron  NC Event Display Using NUGEN Input

15. electron  proton   neutron  NC Event Display Using NUGEN Input

16. Event Reconstruction  Hough Transform - step through slope and intercept space  Find best straight line fit. Iterate.  Assign hits within  15 cm of fit to track.  Final parabolic fit to track.  Eliminate all but the first few hits on track from further consideration  Repeat to look for additional tracks.

17. e charged current

21.  Neutral Current

24.  Neutral Current that Passes Cuts

25. Basic Reconstruction cuts to reduce ntuple size:    1  reconstructed track in each view   2 of longest track in each view  100  Total number of hits  150 Analysis cuts at ntuple level  30  Total Hits  100   1.6 hits/plane ave for longest track in each view  Width of longest track (RMS) in each view  8.5 cm 2  Fraction of hits on longest track/total hits  0.575, in each view  12  number of hit planes  25 in each view  20 - 50 hits on longest track in each view CUTS Calculate Figure of Merit = signal events/sqrt( e beam +  NC) Number of events normalized to flux for 50 kton detector x 5 years of data x 85% fiducial volume.

26. e CC  NC Distributions are weighted, successive Beam e

27. e CC  NC Cut distributions are weighted, successive Beam e

28. e CC  NC Cut distributions are weighted, successive Beam e

29. Cut distributions are weighted, successive e CC  NC Beam e

30. Cut distributions are weighted, successive e CC  NC Beam e

31. e CC  NC Cut distributions are weighted, successive Beam e

32. e CC  NC Beam e Energy Distribution of Surviving Events

33. Total Hit distribution for e CC Vertex in center of container Uniform vertex distribution No Steel Distributions normalized to total number of hits Steel reduces number of hits

34. e CC Container Center  NC Container Center Beam e Container Center e CC Uniform Vertex  NC Uniform vertex Beam e Uniform Vertex Total Events (weighted)50,53099,8425,571202,150397,22421,196 Reconstruction Cuts40,97128,4631,817160,440110,6487,238 Total Hits between 30 - 100 (25 - 100) 33,9569,4191,149138,14145,6484,783 Ave hits/plane  1.6 in each view 28,9005,5981,057110,58823,7384,225 Track width < 8.5 cm 2 in each view 25,4102,10088096,0288,4503,382 Frac of hits on track > 0.575 in each view 21,0021,05262178,9574,4762,321 Number of hit planes bet 12 - 25 in each view 19,00678254566,6602,6281,972 Hits on longest track bet 20 - 50 in each view 18,64872441365,6062,4651,529 Efficiency0.3690.0070.0740.3240.0060.072 Number of Events (50 kt, 5 yr)246.848.134.2216.741.333.3 FOM27.225.1 Results

35. e CC Vertex in Absorber  NC Vertex in Absorber Beam e Vertex in Absorber e CC Uniform Vertex  NC Uniform vertex Beam e Uniform Vertex Total Events (weighted)18,153349,55718,652202,150397,22421,196 Reconstruction Cuts145,735101,0396,575160,440110,6487,238 Total Hits between 30 - 100 (25 - 100) 126,23541,9174,333138,14145,6484,783 Ave hits/plane  1.6 in each view 101,20721,7883,837110,58823,7384,225 Track width < 8.5 cm 2 in each view 88,0217,8243,09396,0288,4503,382 Frac of hits on track > 0.575 in each view 72,4044,1652,12678,9574,4762,321 Number of hit planes bet 12 - 25 in each view 61,4102,4371,80966,6602,6281,972 Hits on longest track bet 20 - 50 in each view 60,4442,2791,40265,6062,4651,529 Efficiency0.3330.0060.660.3240.0060.072 Number of Events (50 kt, 5 yr)222.741.330.5216.741.333.3 FOM26.325.1 Results

36. Last Time e CC Container Center  NC Container Center Beam e Container Center e CC Uniform Vertex  NC Uniform vertex Beam e Uniform Vertex Efficiency0.3690.00090.3260.002 Number of Events (50 kt, 5 yr)270.312.625238.628.525 FOM44.132.6 Comparison with Previous Results This Time e CC Container Center  NC Container Center Beam e Container Center e CC Uniform Vertex  NC Uniform vertex Beam e Uniform Vertex Efficiency0.3690.0070.0740.3240.0060.072 Number of Events (50 kt, 5 yr)246.848.134.2216.741.333.3 FOM27.225.1  Signal efficiencies very similar  Backgrounds larger  FOM smaller  Difference between center of container and Uniform vertex smaller

37. Summary  New more realistic simulation implemented with better container description. Inefficiency and cross-talk effects now included.  Cross-talk likely smaller than currently implemented. Smaller gaps between containers possible (J. Cooper)  e efficiency similar to last time. Background larger.  Electron efficiency and background rejection attainable with containerized detector appear reasonable, but you do pay a price in electron efficiency. Still to do: o Simulate  CC o Add noise? o Explicitly implement a monolithic detector in GEANT for comparison o Categorize  NC events that pass cuts and e CC events that fail cuts. o Get some other people to look at GEANT output (Leslie).