ProtoDUNE detector reconstruction stability Yongshun Zhou Aleena - - PowerPoint PPT Presentation

▶
protodune detector reconstruction stability
SMART_READER_LITE
LIVE PREVIEW

ProtoDUNE detector reconstruction stability Yongshun Zhou Aleena - - PowerPoint PPT Presentation

Jul 22, 2023 356 likes •811 views

ProtoDUNE detector reconstruction stability Yongshun Zhou Aleena Rafique 1 Data Samples Good beam run 5141-5844: (https://docs.google.com/spreadsheets/d/1O4o9_q8F-KynQltKDAfmco3e_s1eKItFkzZ4-g_Vls0/edit#gid=0) Dataset name:

slide-1
SLIDE 1

ProtoDUNE detector reconstruction stability

Yongshun Zhou Aleena Rafique

1

slide-2
SLIDE 2

Data Samples

  • Good beam run 5141-5844:
  • (https://docs.google.com/spreadsheets/d/1O4o9_q8F-KynQltKDAfmco3e_s1eKItFkzZ4-g_Vls0/edit#gid=0)
  • Dataset name: np04_full-reconstructed_v07_08_00_unified_physics
  • Cosmics run 5000-9000:
  • Runs tagged by “cosmics” during good beam runs (5000-5850)
  • Runs in the following dataset (np04_full-reconstructed_v07_08_00_05_cosmics)
  • (http://samweb.fnal.gov:8480/sam/dune/api/definitions/name/np04_full-reconstructed_v07_08_00_05_cosmics/)
  • 1 GeV/c beam+cosmics MCC11 is included in all plots
  • Dataset name: jhugon_mcc11_pd_sp_reco_sce_1.0GeV
  • The same MC is used for both good beam run and cosmics run

2

slide-3
SLIDE 3

Data analysis

  • Module scripts are taken from MicroBooNE experiment
  • (https://microboone.fnal.gov/wp-content/uploads/MICROBOONE-NOTE-1013-PUB.pdf)
  • 10 reconstructed data files are chosen from each good beam run
  • Contain about 100 data events per file
  • The module gets the target mean/RMS over all events in each run

à mean/RMS of a run for a given variable is plotted

3

slide-4
SLIDE 4

Reconstructed quantities

  • Track related:
  • Number of tracks
  • Track length
  • Track position

(begin/end, x/y/z positions)

v Comparison between pmtrack and pandoraTrack

  • Hit (linecluster hit) related:
  • Number of hits
  • Hit peak amplitude
  • Hit charge
  • Optical flash related:
  • Number of opflashes
  • Opflash peak amplitude
  • Opflash center (y/z)
  • Opflash width (y/z)

v Comparison between internal and external opflashes

  • Others:
  • Number of vertices (pmtarck/pandora)
  • Number of PFPs (pmtarck/pandora)
  • PandoraShower

4

slide-5
SLIDE 5

Notice:

  • A “sudden drop” as beam momentum

is changed from 7GeV/c to 1 Gev/c

  • Average track number per event is

slightly increasing during 1GeV/c beam runs

  • The actual pmtrack data diverges a lot

from the pmtrack MC data.

Number of tracks

Good Beam Run

5

slide-6
SLIDE 6

Number of tracks

Cosmics Run

Notice:

  • Several “Vertical” lines.
  • Caused by the wrong “cosmics” tags.

They are test runs with high trigger rate.

  • Relatively stable after around run

7500 (around 3/31/2019)

6

slide-7
SLIDE 7

Notice:

  • In real data, PandoraTrack has longer tracks

than pmtrack

  • In MC, pmtrack has longer tracks than

pandoratrack

  • Overall stable

Length distribution of a single run analysis : run 5455

Track Length

Good Beam Run

7

slide-8
SLIDE 8

Track Length

Cosmics Run

Notice:

  • Similar pattern as in plots of

number of tracks for cosmics run

8

slide-9
SLIDE 9

Track begin position

X Y Z

Good Beam Run

9

Overall Stable

slide-10
SLIDE 10

Track begin position

Cosmics Run

10

X Y Z

Overall stable except for the weird data points as corresponding to previous “vertical” lines.

slide-11
SLIDE 11

Track end position

X Y Z

Overall Stable

Good Beam Run

11

slide-12
SLIDE 12

Track end position

Cosmics Run

Overall stable except for the weird data points as corresponding to previous “vertical” lines.

12

X Y Z

slide-13
SLIDE 13

Notice:

  • A similar drop from 7Gev/c to 1Gev/c
  • A similar increasing pattern in 1GeV/c

region

  • Stable after around run 5429

(10/17/2018)

Number of hits

Good Beam Run

13

slide-14
SLIDE 14

Number of hits

Cosmics Run

Notice:

  • Similar pattern as in plots of

number of tracks for cosmics run

14

slide-15
SLIDE 15

Notice:

  • Increasing pattern before run 5429

(10/17/2018); stable after that

  • In 6GeV/c beam runs, RMS is very

unstable and big

  • V-plane data is the highest, while V-

plane MC is the lowest

Hit peak amplitude

Good Beam Run

15

slide-16
SLIDE 16

Hit peak amplitude

Cosmics Run

Notice:

  • Similar pattern as in plots of

number of tracks for cosmics run

16

slide-17
SLIDE 17

Notice:

  • Similar pattern as hit peak amplitude,

which is reasonable

  • Seems like there is a gradual increase

in hit charge overall

  • V-plane data is the highest, while V-

plane MC is the lowest

Hit charge

Good Beam Run

17

slide-18
SLIDE 18

Hit charge

Cosmics Run

Notice:

  • Similar pattern as in plots of

number of tracks for cosmics run

18

slide-19
SLIDE 19

Notice:

  • Int opflash data is unstable before run 5429 (10/17/2018), and stable after that
  • Most optical flashes have high peak amplitude

Number of Opflashes Good Beam Run

19

slide-20
SLIDE 20

Number of Opflashes Cosmics Run

Notice:

  • Around run 7350-7500, flash data is abnormal; overall stable after that
  • Around 7350-7500, telescope test runs are wrongly tagged by “cosmics”.

20

(At the end of March 2019)

slide-21
SLIDE 21

Notice:

  • Similar pattern as in plots of number of opflashes for good beam run
  • External opflashes have much larger peak amplitude (notice the 103)

Opflash peak amplitude Good Beam Run

21

slide-22
SLIDE 22

Notice:

  • Abnormal data around run 7350-7500

Opflash peak amplitude Cosmics Run

22

slide-23
SLIDE 23

Notice:

  • Similar pattern as in the previous one for good beam run

Opflash y center Good Beam Run

23

slide-24
SLIDE 24

Notice:

  • Abnormal data around run 7350-7500

Opflash y center Cosmics Run

24

slide-25
SLIDE 25

Notice:

  • Similar pattern as in the previous one for good beam run

Opflash z center Good Beam Run

25

slide-26
SLIDE 26

Notice:

  • Abnormal data around run 7350-7500

Opflash z center Cosmics Run

26

slide-27
SLIDE 27

Notice:

  • Similar pattern as in the previous one for good beam run
  • External opflashes have larger width, which is reasonable

Opflash width Good Beam Run

27

slide-28
SLIDE 28

Notice:

  • Abnormal data around run 7350-7500

Opflash width Cosmics Run

28

slide-29
SLIDE 29

Notice:

  • Similar pattern as number of tracks

and hits for good beam run

Number of vertices

Good Beam Run

29

slide-30
SLIDE 30

Number of vertices

Cosmics Run

Notice:

  • Similar pattern as in plots of

number of tracks for cosmics run

30

slide-31
SLIDE 31

Notice:

  • Similar pattern as number of tracks

and hits for good beam run

Number of PFPs

Good Beam Run

31

slide-32
SLIDE 32

Cosmics Run

Notice:

  • Similar pattern as in plots of

number of tracks for cosmics run

Number of PFPs

32

slide-33
SLIDE 33

Notice:

  • Similar pattern as number of tracks

and hits for good beam run

  • The number of showers is relatively

large.

Number of showers

Good Beam Run

33

slide-34
SLIDE 34

Cosmics Run

Notice:

  • Similar pattern as in plots of

number of tracks for cosmics run

Number of showers

34

slide-35
SLIDE 35

Notice:

  • Overall stable
  • Slight decreasing pattern in RMS

Shower length

Good Beam Run

35

slide-36
SLIDE 36

Cosmics Run

Notice:

  • Similar pattern as in plots of

number of tracks for cosmics run

Shower length

36

slide-37
SLIDE 37

Summary and discussion

Good beam run

  • Something should have happened after around run 5429, when beam

momentum is changed from 1GeV/c to others. Data of hits, internal optical flashes, vertices, PFPs and showers all share a similar improvement in stability after that.

  • In Elog, we found that after run 5387, problems with CRT emerged. Several test runs were

initiated without beam after that. Before run 5423, many software upgrades were also completed, and beam was back.

  • During 1GeV/c beam run, data of all mentioned also share a similar increasing

pattern.

  • On 10/13/2018, the purity of liquid argon was relatively low (on average 2.7ms); it gradually

changed back to 4.3ms on average on 10/17/2019. This could be a potential cause.

  • From 7GeV/c beam run to 1 GeV/c beam run, a sudden “drop” is noticed quite
  • ften, too.
  • Beam momentum change should be the main reason.
  • In Addition, before run 5213, which was before beam momentum change, high voltage of

cathode plane changed from 180kV to 130kV.

37

slide-38
SLIDE 38

Summary and discussion

Cosmics run

  • Around run 7350-7500 (3/29/2019-3/31/2019), optical flash data was

abnormal.

  • Runs around that range were telescope runs, causing optical flash data to be

unstable; they should not be tagged as “cosmics”.

  • Several “vertical” lines are noticed in most plots. That’s around run

5751 (11/01/2018), run 5850 (11/12/2019) and run 6509 (01/22/2019).

  • In Elog, it turned out that these runs should all be test runs rather than

cosmics runs; tagging is a little messed up. The “vertical” line in data was caused by CRT problem: trigger rate was too high.

38

slide-39
SLIDE 39

Thanks!

Conclusions and to do

  • The detector stability script works well for checking the stability of ProtoDUNE-SP

detector over runs

  • Some problems are detected, and potential reasons for unstable data can be

checked via Elog

  • Some tags for the data file are messed up
  • We can further check MC for with other beam momentum, and compare them

with the data

  • We can also look into other reconstruction algorithms, and check the stability for

them

39

slide-40
SLIDE 40

Appendix—Single run analysis

  • Good beam run corresponding date
  • Data from all events in run 5455 included
  • Direct plot—no calculation of mean included
  • Variables included:
  • Track position (begin y, end y, begin z, end z) Vs. Track length
  • Internal optical flash center (y center, z center) Vs. flash peak amplitude

40

slide-41
SLIDE 41

Good beam run corresponding date

Run Range Corresponding Date Corresponding Beam Momentum 5141-5213 10/10/2018-10/12/2018 7 GeV/c 5216-5426 10/12/2018-10/19/2018 1 GeV/c 5429-5452 10/19/2018-10/23/2018 2 GeV/c 5455-5460 10/23/2018-10/24/2018 1 GeV/c 5461-5757 10/24/2018-11/01/2018 No Good Beam Run 5758-5776 11/01/2018-11/04/2018 6 GeV/c 5777-5797 11/04/2018-11/07/2018 3 GeV/c 5800-5824 11/07/2018-11/09/2018 Random 5825-5833 11/09/2018-11/10/2018 0.5 GeV/c 5834-5841 11/10/2018-11/11/2018 0.3 GeV/c

41

slide-42
SLIDE 42
  • “Slope” is related with how the algorithm analyze the track position.
  • Longer tracks are more or less cosmic rays entering from the top and leaving at the bottom of the detector.
  • Not clear why pandoraTrack has data beyond 600cm (the dimension of the detector).

42

slide-43
SLIDE 43
  • “Vertical lines” are the position of the edge of APA plane (especially clear in pmtrack).
  • Some random vertical lines appear in pandoraTrack—reason not known.
  • Longer tracks “gather” more to the left in end plot, and to the right in begin plot, which makes sense.

43

slide-44
SLIDE 44
  • In Z center plot, the “vertical lines” are

around middle of each APA plane.

  • Not clear why there is no data in the

range of 0-120cm and 580-700cm

44