Calo CCWG Weekly Meeting
11.19.2025
-------------------------

Extending Forward Lambda Reconstruction
Baptiste Fraisse (CUA)
---------------------------------------

- Meson stucture function group
  - preparing for kaon structure function
    measurements @ the EIC
- Primary method
  - Semi-Inclusive Sullivan process e+p -> e+L0(+K+)
  - K structure function inferred from L0 momenta
  - 3 measurement options
    a. non-decayed L0 (~m decay length)
    b. measure charged decay (2 particles, B-field dependence)
         L0 -> p+pi-
    c. measure neutral decay (3 particles, no B-field dependence)
         L0 -> n+2gamma
- Lambda reco algorithm
  - IDOLA algorithm --> implemented for ZDC
  - measures n+2gamma
  - realistic L0 distribution --> 0% reconstructed @ 5x41 GeV,
    0.4% reconstructed @ 18x275 GeV
  - Primarily due to small geometric acceptance
- Improvements to algorithm
  - Extend inputs to encompass multiple calos (ZDC, B0 ECal, FEMC,
    LFHCAL + Insert)
    - Looking for triple coincidence in any calo + across calos
  - leads to 20% reconstructed @ 5x41 GeV, 5% reconstructed
    @ 18x275 GeV
    - latter performance due to L0 being highly boosted -->
      decays in FF
- Need calibration constants for neutrons gamma in B0 ECal, FHCal's
  - numbers in IDOLA paper are for ZDC only
  - 3 questions:
    a. Are numbers known for other calorimeters?
    b. Can we reuse ZDC numbers?
    c. Is machinery available to calculate these numbers?

Oleg: on question (b), no
  - these calorimeters have very difference chemical compositions
  - no general formula for these numbers exist
Derek: on question (c), yes
  - Jan has been working this study for BHCal

BHCal + BECal Calibration
Jan Vanek (UNH)
--------------------------

- System is complex --> so machinery not ready
  - biggest reason is the magnet --> impossible to measure
    energy deposited in magnet
  - due to neutron decay length, many start showering in
    magnet --> no energy deposit in BECal
- Energy deposition in BECal/BHCal
  - depends strongly on energy
    - 1 GeV/ --> neutron deposits most energy in BECal OR BHCal
    - 2 GeV/c --> more even distribution of energy between BE/HCal
    - 5 GeV/c --> showers starting in BECal usually reach BHCal
  - in all cases, can observe cases where shower either starts
    in magnet OR loses energy in magnet
- Investigating manual clustering
  - i.e. clustering outside of EICrecon
    - felt clustering in EICrecon is a blackbox
    - so trying out clustering scheme by hand
- Looking towards a couple manual (non-ML) calibration schemes
  - 1st is just to determine a single constant to adjust
    the sum Ebemc + Ebhcal
  - 2nd is Sebouh's technique

- Derek: can extract eta/phi indices from Cell ID
    - Will add snippet on how-to if there is none
- Derek: are the code to do both methods in a state to be shared?
  - Jan: not really --> very much WIP. But can do some
    clean-up and commit to a repo

Discussion
----------

- Oleg: to Baptiste, since you only have B0 ECal could only
  consider gammas for those detectors
  - Baptiste: True
- Derek: how does the implementation look for these improvements?
  - Baptiste: similar to original analysis --> follow
    chain back to see if particle is a neutron/gamma and if
    it came from a lambda, then look for triple coincidence 
  - Derek: and the reconstruction? There's 1st the neutral reco 
    in the ZDC, and then the iterative reconstruction of the 
    lambda (IDOLA)
  - Baptiste: vertex reco is one thing, we were just looking 
    for triple coincidences --> will be helpful for validating 
    existing algorithm