Preliminary Study of CC-Inclusive Events in the P0D using Global

ReconstructionRajarshi Das

(w/ Walter Toki)Nu-Mu Prelim. Meeting Dec 2010

CSU

Overview• Preliminary look at charged current events

originating in P0D using Global Reconstruction• Distributions shown (MC and Data):

– Pseudo-vertex (track start) distributions in P0D– Momentum and Cos θ

• Similar study with Central ECAL pseudo vertices only

• Quick look at status of related Global Reconstruction parameters

Sample and Selection for CC Study

• Sample:– Monte Carlo (v7r21) NEUT Kinematics Generator corresponding to 3.3 x 1020 Protons on Target (POT)– Data (v7r19) for Full P0D study from Run 2010a corresponding to 2.9 x 1019 Protons on Target (POT)

• Selection

BEAM DIRECTION

CUTS REASON

Require Track TPC Segment • Access to momentum measurement from TPC• Reduce high angle tracks that cause reconstruction failuresRequire Track P0D Segment

Negative ChargeMuon hypothesis

Highest Momentum

Track Start in P0D Fiducial Volume•Reject sand muons•Reject Magnet interactions

Track Start in C-ECAL Volume Only•Look for variation in distributions from change in target density

TPC 1P0D

TPC 3TPC 2

X and Y Pseudo-Vertex Distributions

As expected

MC MC

Muon Vertex Distributions

BEAM

BEAM

• Good agreement between Data and Monte Carlo in Z position of vertex• Buildup of vertices in the downstream end of the P0D• Muons originating further upstream more likely to range out (~500 MeV for P0D)

BEAM

BEAM

MC

MC

MC

MCData

•Data comparison agrees rather well after POT normalization for both Muon Momentum and Cos(θ)•Post-selection contamination (Non-CC events) very low!•Cos(θ) distribution biased towards forward going tracks due to TPC segment requirement•Post-selection muons show deficiency in < 500 MeV region (Muon range out momentum in P0D is ~500 MeV)•Complimentary study would use range out to determine momentum of muons stopping in P0D

Full P0D: P and Cos(θ)

CC-QE1 PiMulti PiOther (DIS, Eta0, K)Non-CC events

CC-QE1 PiMulti PiOther (DIS, Eta0, K)Non-CC events

Reconstructed Cos(θ)

Data

Entries: 396689 MC Entries: 43604

Data Entries: 4259

•Data and Monte Carlo remain reasonably comparable•Momentum distribution cut off below 400 MeV, but range out in ECAL is only 70 MeV!!•Possibly failed reconstruction due to short P0D Track•Lower bias towards forward going tracks (understandably)

C-ECAL Only: P and Cos θ

CC-QE1 PiMulti PiOther (DIS, Eta0, K)Non-CC events

Entries: 8381

CC-QE

1 Pi

Multi Pi

Other (DIS, Eta0, K)

Non-CC events

Full P0D: P vs. Z Vertex Position

Note location of misclassified events.

Low energy and smaller lever arms

Momentum Checks

• Sample: Muon Particle Gun– 1000 events processed with v7r21 and v8r5p3– 0 to 2 GeV flat Energy distribution– All vertices in P0D

• Cuts:– Only single Lepton track in Event– Vertex in P0D Fiducial region– Muon Track doesn’t stop in P0D

• From v7r21 to v8r5p3, we see an improvement in reconstructed momentum, Energy Loss application corrected?

• Algorithm: Should take TPC measured momentum, project track to P0D vertex, correct for energy loss in P0D, but Global Reconstruction possibly missing P0D tracks

v7r21 v8r5p3

Success/Failure RatesCategory V7r21 v8r5p3

Total in Selection 484 429Evts. w/ Single

Global Muon Track reconstructed

262 (54%) 227 (53%)

Evts. w/ Single P0D 3D Track

Reconstructed387 (80%) 424 (99%)

Evts. w/ > 1 Global Track reconstructed 132 (27%) 382 (89%)

Evts. w/ > 1 Global Muon Track

reconstructed24 (5%) 60 (14%)

Evts. Where Global picked up a P0D

Piece248 (51%) 0 (0%)

Conclusion• Provided a cross check for P0D-TPC studies preceding

this one• To first order, reasonable agreement between NEUT MC

and Data• C-ECAL only momentum distribution not fully

understood; Cut off in momentum (400 MeV) doesn’t match energy loss range out value of ECAL (70 MeV)

• Marked increase in Global Reconstruction failure at:– Matching multiple global tracks to the same muon track (5%

14%)– Picking up the Track piece in P0D (51% 0%)

• Future Work:– Continue checks on Global Reconstruction– Look at incorporating PID cuts