Kaon_LT Experiment

                 Feb 13/25 PionLT/KaonLT Analysis Meeting Notes

                 ----------------------------------------------

                                (Notes by GH)

                     Today: KaonLT will be discussed first

Present

-------

Regina - Garth Huber, Ali Usman, Muhammad Junaid, Vijay Kumar, Nacer Hamdi

JLab - Dave Gaskell

Ohio - Julie Roche

CUA - Tanja Horn

Virginia - Richard Trotta

FIU - Pete Markowitz

Vijay

-----

Continuing to set up Q2=0.425 PionLT analysis

- shows diamond cut, center SHMS setting data shown

- diamond vertices selected using a Python script similar to Jacob's script

- the diamond seems a bit tight in comparison to low epsilon data, will adjust

  the vertices a bit by hand to optimize

  - also will confirm the cut is good for L1,L2,R1,R2 SHMS settings, the idea

    is to use a single diamond cut for all 5 settings, so if the diamond is

    shifted slightly for some settings, then the diamond needs to be made a bit

    smaller in the direction away from the shift

- will look at t-binning after diamond cut finalized

  - -t_min is slightly higher than Q2=0.375 so bins will require some

    adjustment

  - had 8 t-bins for Q2=0.375, but L/T-sep failed for 8th bin due to poorer

    statistics

  - considering to have 7 t-bins for Q2=0.425, with 7th bin wider

  - hopefully we will have some plots on t-binning at next meeting

- last step will be to calculate experimental normalized yields

- running SIMC simulations for Q2=0.425

Richard

-------

Q2=4.4 W=2.74 KaonLT analysis

- looked into different forms of Wfac:

  - simple version:   Wfac=1/(W^2-M^2)**2

  - more complicated: Wfac=1/(W^2-M^2)**(0.85*W^2-5.97*W+12.68)

    this more complicated version is from a few months ago, where the

    polynomial coefs came from a fit to all setting data

  - in the polynomial, the <W> value for each t-bin was used

  - GH: the purpose of Wfac is to flatten some of the variation of yield across

    an individual diamond

    - (W^2-M^2) needs to use <W> per t-bin, but the exponent really shouldn't

      need to vary that rapidly

    - suggest to compute the exponent using the W-central for each setting and

      use that instead of the polynomial

    - Richard agrees that using the full polynomial complicated the fitting,

      will look into this

  - Data/MC Ratios with simple Wfac: R=0.2-0.4 after 5 iterations

    - with complicated Wfac: R=1-2 after 2 iterations

Next step:

- wants to see what different Wfac choices look like at other Q2

  - hope to have 3 settings to compare next wek

Ali

---

- nearly done thesis writing, the plan is to get back to piDelta BSA

  systematics and other missing parts of analysis while the thesis is being

  reviewed by committee and external examiner

Nacer

-----

Continuing to set up Low Q2 KaonLT analysis

- looking closely at 4.9 GeV beam energy data, cut studies so far looked mostly

  at 3.8 GeV data

- shows plot of RFtime vs MMk

  - also computing efficiency of RFcut as a diagnostic

  - the RFcut appears to need adjustment, particularly at 4.9 GeV

  - Junaid: suggests to compare 1D plots of RFtime at both energies

    - the hcana RFtime offsets can be different between the two beam energies,

      the 4.9 GeV offset should be checked

    - also the width of the RFtime distribution can be slightly different

      between the two beam energies

  - Garth: the cut appears to be too tight for RFtime~0.2, some clear Lambdas

    are being eliminated on this side, while the cut is too loose on the other

    side RFtime~2, where lots of pion leakthrough is evident

    - for the RFtime~0.2 cut, it is clearly too tight at both 3.8, 4.9 GeV,

      while on the other side it seems only to need adjustment at 4.9 GeV

    - Nacer will review cuts and calculate new cut efficiencies, with the

      greater statistics compared to previous plots it is clear things need

      further optimization

  - Ali: when optimizing the RFcut, it's important to minize systematic

    uncertainties

    - if the RFcut is too tight, the cut efficiency will be lower, leads to a

      larger systematic uncertainty

    - if the RFcut is too loose, the pion leakthrough subtraction will be

      larger, leading also to a larger systematic

    - there is an optimal RFcut which balances the two systematics

- shows CoinTime plots for both 3.8, 4.9 GeV

  - cointime peaks line up very similarly for both energies, cut looks good

- HMS Cherenkov cuts at both energies checked as well

Next step (after adjusting RFcut):

- need to make diamond cut on low epsilon, center SHMS data

  - then confirm the cut is okay on Left, Right SHMS settings

Junaid

------

PionLT HMS optics matrix elements

- implemented 5.8, 6.1, 6.7 GeV/c matrix elements from NPS analysis

- expecting to receive new 5.6 GeV/c matrix elements soon

- HMS theta, phi offsets

  - using offsets from NPS analyis for 5.8, 6.1, 6.7 GeV/c

  - for unsaturated region <5.5 GeV/c, using phi=0

    - emailed Mark Jones re. what offsets to use, no reply yet

    - Dave: agrees this zero offset seems strange, also contacted Mark about it

      as needed also for SIDIS analysis, hopefully one of us gets a reply soon

- Heep offsets with new HMS matrix elements

  - Garth is running new Heep In-Plane offsets, job still running, should be

    done soon

    - will give global offsets using all 9 Heep settings

    - can also make offsets for subsets of Heep settings if needed

  - shows new OOP offsets compared to analysis with old matrix elements

    - SHMS: -0.05 -> -0.155 mr  a slight change for SHMS

    - HMS: +1.875 -> +1.90 mr   change is very small for HMS

  - compare to KaonLT OOP offsets determined by GH

    - SHMS: -0.11 mr   HMS: +2.51 mr

    - good news that the SHMS offsets are more similar now, not sure why the

      HMS offsets are so different

    - one possibility is that the PMY data errors are treated differently

      between Junaid and GH, Junaid is using actual errors (which is better),

      while GH used estimated errors (based on scatter of data)

      - GH errors could be off, but plots are visually similar

      - KaonLT offsets also use old ME, but the HMS difference seems to be

        stable against the optics change

    - Junaid will compute errors in the new OOP offsets, which will indicate by

      how many sigma the PionLT and KaonLT offsets differ

- setting up for Pass-2 replay

  - implemented Nathan's LH2 boiling correction

  - still implementing new ELLT calculation in report files

Next steps:

- updated HeepCoin Data/MC ratios, including hopefully also Alicia's proton

  absorption correction

- Pass-2 full replays

Garth

-----

- Richard, Nacer, GH met with Ioana and Gabi to give them an update on the

  status of the KaonLT analysis, where to find Richard's replay data and

  scripts

- Nacer is at an early stage of L/T-sep analysis, so they can learn a lot from

  each other as they progress

- they won't be able to attend our Thursday meetings until classes end in ~2

  months, can receive questions by email or other dedicated meeting in the

  meantime

Next Meeting

-------------

- Thur Feb 20 @ 15:30 Eastern/14:30 Regina

  - PionLT will go first