dQ/dx-d/Edx Calibration

Contact Persons:

  • Varuna Meddage, Tingjun Yang (Framework, tools)
  • MCC8
    • Varuna Meddage
    • Davide S. Porzio, Yun-Tse Tsai (Run 3 data)
  • MCC9
    • Maya Wospakrik, Wanwei Wu


We use through-going and stopping muons to determine the relative response correction factors - XYZ (T) of the LArTPC detector and want the detector response to the charge collection to be uniform over space and time.

The detector response is not uniform because:

  • Dead wire regions and misconfigured ASICs
  • Space charge effects
  • Electron attenuation and diffusion
  • Recombination
  • Temporal variations

By unifying the detector response, we mean to calculate the correction factors for taking care of those effects.


Prerequisites (in the following order)

  1. Updated bad channel database (refer: Calibration Database)
  2. Updated ASIC misconfiguration database (refer: Calibration Database)
  3. Good run list (refer: Good Run Selection)
  4. Stage Reco 1 sample (from Ext. BNB)
  5. Stage Reco 2 (Calorimetry reconstruction on the Reco 1 sample) sample

Step by Step

1. YZ Correction for dQ/dx values:
  • Aim to remove the effects from the misconfigured ASICs, space charges, and transverse diffusion.
  • see example plot below (from docdb: 19039)



2. X Correction for dQ/dx values:
  • Deal with electron attenuation, longitudinal diffusion, space charge effects, and recombination.
  • see example plot below (from docdb: 19039)



One can calculate the XYZ correction using: XYZ_correction = YZ_correction * Drift direction (X) correction.

3. T Correction for dQ/dx (in addition to XYZ correction):
  • Study the variation in global dQ/dx median over time to derive the timing corrections.
  • see example plot below (from docdb: 19039)


One can obtain the total correction for dQ/dx now: Total correction = Time correction * XYZ correction.

4. dE/dx calibrations

  • Once the data sets are corrected by the total correction (dQ/dx), then we try to convert all dQ/dx values to dE/dx values.
  • See the formula and plot below (from docdb: 9248). Here the calibration constant was chose as 4.400 e-3.


  • dE/dx distribution is then fitted with a Landau convoluted Gaussian and extracted the MPV (most probable value) for dE/dx.
  • Ideally MPV value should be peaked around 1.7 MeV/cm (MIP energy for a 5 GeV Muon).


Post Steps

  • Apply these corrections and calibration constant to data and see whether we can use dE/dx information for particle identification.
    • See example for proton identification below (from docdb: 9248)




  1. Detector Calibration Public Note: docDB # 15584
  2. Detector Calibration Internal Note: docDB # 14754
  3. Detector Calibration of Run 3 Data Internal Note: docdb # 18333

Documents for methods:

  1. Update on Calibrating the detector using cosmic Muons: docDB # 9248

MCC9 Retreat at CSU (March 18-22, 2019)