Kaon_LT Experiment

                 Apr 17/25 PionLT/KaonLT Analysis Meeting Notes

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

                                (Notes by GH)

                     Today: PionLT will be discussed first

Present

-------

Regina - Garth Huber, Nacer Hamdi, Alicia Postuma, Muhammad Junaid,

   Nathan Heinrich, Vijay Kumar, Nacer Hamdi

JLab - Dave Gaskell

FIU - Pete Markowitz

Virginia - Richard Trotta

CUA - Tanja Horn

Junaid

------

Q2=3.85, W=2.62 t-binning

- went from 7 to 5 t-bins, covering range 0.16<-t<0.50

- roughly 5k events/bin, up from ~3.8k before

  - lowest number of counts in a bin is 4315

- will do 15 phi-bins

- scripts to calculate yields for each t-phi bin in progress

- will give a summary of software framework and Qeff error calculation in a

  future meeting

Nathan

------

PionLT Luminosity Scans

- Dave has looked briefly the CoinTime plots, needs to consider them in more

  detail

  - arranged a meeting to discuss them with Nathan: Tue Apr 22 @ 15:30 EDT

- Tanja reports that NPS found that a small non-linear BCM offset was needed to

  get their Carbon scans to work out

  - done by Avnish NPS RG1A scaler lumi analysis (from the NPS elog)

  - corrected_current=H_BCM4A_Current + corr*sqrt(BCM_rate)

    where corr is approximately -1.5nA/sqrt(Hz)

  - applied a rate dependent correcton to Carbon Lumi data and then determined

    current offset, which is about -200nA, compared to Nathan's ~35nA offset

Nacer

-----

KaonLT Q2=0.5 LT-sep t-phi binning

- solved some problems w/ error propagation in data binning script

- wrote SIMC binning scripts, they run well and quickly

- Data/MC ratio scripts, kinematics comparison plot scripts in progress

- had a very useful meeting with Richard yesterday on his LT-sep framework and

  SIMC event weighting

  - will adapt codes to K+ analysis

- will present some slides on Yield calculation and Error propagation in a

  future meeting

Vijay

-----

Q2=0.50 MM offsets

- showed plots for low epsilon: center, Left1, Left2

  - resulting offset about 6 MeV gives good agreement between data and MC

  - Nacer comments on how well SIMC reproduces the pi+n radiative tail

  - simply matching the highest bin of data and MC, works well for these

    settings, but in general this method is sensitive to statistical

    fluctuations and a peak fit is typically used

Richard

-------

KaonLT Q2=3.0, W=2.32 LT-sep, 4 t-bins

- functional forms from Q2=4.4, 5.5 seem okay

  - Q2-dependence was removed for now to simplify process, only t-dependence,

    will need to restore Q2-dependence later

  - Data/MC ratios closer to 1 now for high epsilon, but >2 or higher for low

    epsilon

- kinematic comparison plots overlaying Data and MC look good

  - some very interesting differences in SHMS-delta between SHMS Center and

    Left settings, but both are well described by SIMC

- preliminary L/T/LT/TT cross sections

  - L,T have sensible t-dependence, sigL>sigT to middle of t-range

  - main surprise is that LT,TT are very large, and LT doesn't go to zero as t

    goes to -t_min

  - we should consider some systematic studies to do if this holds up

Planned next steps:

- finish Q2=3.0, W=2.32

- parameterize Q2=2.115, W=2.65

- re-parameterize Q2=3.0, W=3.14 with new functional form

- rull SIMC runs with new functions and parameters

- full data replay

- systematic studies

Alicia

------

First look at Q2=3.0, W=3.14 u-channel data

- ran SIMC with omega, rho, phi, eta, eta-prime, Xphasespace generators

- modified Ali's piDelta shape study code for u-channel shape study

- for data, start with aggressive proton cuts to study PID and MM shapes

  - decided on a CoinTime +/-2.25ns cut for proton selection

- RF vs Aerogel: applied a HGC<1.5 npe cut to select protons

  - preliminary RF<1.7, Aero-NPE<3 cuts reduce pi+ leakthrough a lot, some

    remains due to HGC hole

- RF vs CoinTime: if tight Cherenkov cuts, things are cleaned up a lot

  - observe that the proton reals peak is not quite at zero, some shift is

    planned to center peak at zero

- MM plot with progressive PID cuts applied

  - pi+n peak shrinks a lot, much easier to see eta, eta-prime peaks after cuts

    while they were not evident before cuts

  - Q2=3.0, W=2.32 center MM plot has much less obvious peaks

    - *NB* Vijay suggests to tighten the CT cut to see if that helps

- Q2=3.0, W=3.14 shape study

  - sum of Xphasespace and rho MCs do not reproduce shape of continuum

    distribution at higher MM, unlike Bill's thesis

  - plot of u vs MM indicates that omega events don't extend below -u<0.6,

    placing a cut to exclude this region removes a lot of the high MM events

    - Ali says he does something similar in his piDelta BSA analysis

  - omega peak doesn't quite line up with MC, will need to adjust MM peak by a

    few MeV

- Lots of discussion about the PID and high MM region

  - can place a but on HGC hole to reduce pi+ leakthrough

  - *NB* if so, then should apply same cut to SIMC

  - Richard says it's straightforward to implement this in recon_hcana

  - Nacer: suggests to look at events only passing through HGC hole to better

    understand the leakthrough

  - Aerogel: applying NPE<3 cut, can look at higher MM region with varying PID cuts

  - Garth: maybe the extra events at MM>1 GeV are due to proton SIDIS?  Hem

    reports seeing lots of u-channel proton events in pi-SIDIS data

    - Dave: there is no model for this in SIMC yet, would have to modify

      pi-SIDIS model accordingly

  - *NB* Xphasespace model was written by Pawel Ambrozewicz for 4 GeV data, maybe

    there is some limit in the code?  Alicia will look more carefully at the

    code

Ali

---

piDelta BSA

- new (finer) t-binning finalized

- to be efficient, will do some systematic studies in parallel with rebinned

  shape study

- planned systematic studies:

  - MM shape differences between data and SIMC

  - new idea after discussion with Garth: tweak bkd fit by ~1 sigma (or

    whatever seems appropriate based on reasonable fit to data) and extract

    different asymmetry

  - wrote new script to do these studies, some debugging underway

- plans to show some fitting plots next week

Next Meeting

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

- Thur Apr 24 @ 15:30 Eastern/13:30 Regina

  - KaonLT will go first

  ** THIS TIME WILL BE USED AT LEAST UNTIL THE END OF AUGUST **