Updated on Electron Energy Reconstruction Aaron Higuera University - - PowerPoint PPT Presentation

Aaron Higuera

University of Houston

Updated on Electron Energy Reconstruction

2

Energy Reconstruction

Ecalo =

i=N hits

X

i=1

✏i(X, Y Z)dQiWion calorimetry factor · Recombination factor

• epsilon_{i} = correction factor X(life time) and

YZ(wire response, etc.) run 5809

• dQ_{i} = hit charge
• W_{ion} = 23.6e-6, from Argoneut
• calorimetry factor = 5.58e-3 run 5809
• Recombination factor = 0.63, from FERMILAB-

PUB-15-458-ND

200 400 600 800 1000 1200 1400 1600 1800 2000 Shower Energy [MeV] 200 400 600 800 1000 1200 1400 1600

MC Data

Area Normalized

3

Energy Reconstruction

200 400 600 800 1000 1200 1400 1600 1800 2000 Shower Energy [MeV] 200 400 600 800 1000 1200 1400 1600

MC Data

Area Normalized

Two electrons coming in the same beam spill This is a common feature at all momenta BTW beaminfo says there is just one

4

Energy Reconstruction

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Shower Energy/Beamline Energy 500 1000 1500 2000 2500 3000

MC Data

Area Normalized

1000 − 800 − 600 − 400 − 200 − 200 400 600 800 1000 Beamline Energy - Shower Energy 500 1000 1500 2000 2500 3000

MC Data

Area Normalized

There is a bias between the shower energy and the beamline ~150 MeV Where is this coming from?

5

Energy Reconstruction

There is a bias between the shower energy and the beamline ~150 MeV Where is this coming from? Look at true energy deposited using sim::SimChannel

200 400 600 800 1000 1200 1400 1600 1800 2000 True Energy Deposited [MeV] 500 1000 1500 2000 2500 3000

h_Eres2 Entries 26890 Mean 17.23 Std Dev 9.373

30 − 20 − 10 − 10 20 30 40 50 60 Beamline Energy - True Energy Deposited 1000 2000 3000 4000 5000

h_Eres2 Entries 26890 Mean 17.23 Std Dev 9.373

Energy loss upstream of the TPC is ~18 MeV Is this consistent with beam experts information?

6

Energy Reconstruction

There is a bias between the shower energy and the beamline ~150 MeV Energy loss upstream of the TPC is ~18 MeV Is the calorimetry reconstruction introducing a bias? Using true hits(charge) calculated the energy using our calorimetry method

200 400 600 800 1000 1200 1400 1600 1800 2000 True Energy Deposited (by Calo) [MeV] 500 1000 1500 2000 2500 200 400 600 800 1000 1200 1400 1600 1800 2000 True Energy Deposited [MeV] 500 1000 1500 2000 2500 3000

true deposited energy hit charge—> energy using our calorimetry method does not introduce a bias

7

Energy Reconstruction

There is a bias between the shower energy and the beamline ~150 MeV Energy loss upstream of the TPC is ~18 MeV Is the calorimetry reconstruction introducing a bias? NO Look at shower completeness

h_compl Entries 23700 Mean 0.8802 Std Dev 0.1117

0.2 0.4 0.6 0.8 1 1.2 1.4 completeness 200 400 600 800 1000 1200 1400

h_compl Entries 23700 Mean 0.8802 Std Dev 0.1117

compl = P

i reco pandora hiticharge

P

i MC particle hiticharge

8

Energy Reconstruction

There is a bias between the shower energy and the beamline ~150 MeV Energy loss upstream of the TPC is ~18 MeV Is the calorimetry reconstruction introducing a bias? NO Look at shower completeness

50 100 150 200 250 300 350 400 450 500 wire 3600 3800 4000 4200 4400 4600 4800 5000 time 50 100 150 200 250 300 350 400 450 500 wire 3800 4000 4200 4400 4600 4800 5000 time

true hits Pandora shower hits

h_compl Entries 23700 Mean 0.8802 Std Dev 0.1117

0.2 0.4 0.6 0.8 1 1.2 1.4 completeness 200 400 600 800 1000 1200 1400

h_compl Entries 23700 Mean 0.8802 Std Dev 0.1117

9

Energy Reconstruction

50 100 150 200 250 300 350 400 450 500 wire 3600 3800 4000 4200 4400 4600 4800 5000 time

Can we recover some of the missing hits? Project the shower direction into a the XZ plane (collection) Look for hits within a 2D cone given then shower length and 30 degrees Calculated completeness again New completeness does not look better it seems that we have some cosmic contamination from intersecting cosmic with the cone… more work need to be done, what about a 3D cone? Look also at purity

0.2 0.4 0.6 0.8 1 1.2 1.4 completeness 200 400 600 800 1000 1200 1400

shower shower + cone

10

Energy Reconstruction

Recombination factor: Given our E-field recombination factor is approximately(box mode) ~0.7 However a recombination value of 0.63 gives better results?

200 400 600 800 1000 1200 1400 1600 1800 2000 True Energy Deposited (by Calo) [MeV] 500 1000 1500 2000 2500 3000 3500

True energy deposited True hits w/ calorimetry & R = 0.63 True hits w/ calorimetry & R = 0.7

11

Energy Reconstruction

200 400 600 800 1000 1200 1400 1600 1800 2000 True Energy Deposited [MeV] 500 1000 1500 2000 2500 200 − 150 − 100 − 50 − 50 100 150 200 True Energy Deposited - True Energy Deposited (by Calo) 200 400 600 800 1000

using calo Fitting multiple gaussians to a long pulse is just an approximation An alternative would be use recob::Wire signals and sum up the ADC values on each wire based on the shower hit peak time

12

Summary

• We understand (sort of) where the bias is coming from in the energy

reconstruction

• Biggest contribution comes from missing hits in the shower
• Upstream energy loss according to simulation is ~18 MeV

according to beam experts is ~50 MeV

• Recombination factor still an open question

13

The End