Kaon_LT Experiment

                 Mar 20/25 PionLT/KaonLT Analysis Meeting Notes

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

                                (Notes by GH)

                     Today: PionLT will be discussed first

Present

-------

Regina - Garth Huber, Nathan Heinrich, Ali Usman, Alicia Postuma,

   Muhammad Junaid, Vijay Kumar, Nacer Hamdi

JLab - Dave Gaskell

Virginia - Richard Trotta

CUA - Tanja Horn

FIU - Pete Markowitz

Nathan

------

PionLT Coin Lumi scan update

- LH2 boiling correction for singles is applied throughout, 2.8%/100uA

- randoms subtraced and pion PID are applied

- two studies presented:

Pass 10: boiling and tracking (both arms), EDTM TLT, BCM correction all applied

- normalized yield vs Beam current shows a decrease of ~6% @ 80uA with lots of

  scatter

  - best set (least decrease) is SIDIS setting 2

- same data vs PS5 rate gives a bimodal distribution, one lobe with steeper

  slope than the other

- vs SHMS rate also bimodal dist

- vs HMS rate, data merge closer to one dist, but much scatter

Pass 12: same as 10 plus DMack hodo corr (both arms) - beam off periods are

  removed when calculating hodo correction

  - Dave: concerned that applying DMack hodo correction is double counting

  - Nathan: maybe it's actually an inefficieincy in the hodo at high rate??

- normalized yield vs Beam current shows scatter but less than Pass 10

- vs SHMS rate gives fairly flat distribution but still lots of scatter

  - Nathan interprets to mean that the SHMS rate dependence is correctly

    modeled when applying DMack correction

- vs HMS rate, strong correlation of 3.4%/100kHz but much less scatter, with

  the data extending up to ~180kHz, i.e. ~5% correction, which would be

  presumably different at high and low epsilon

Dave: the data seem convincing that there is an unresolved rate dependence,

  still concerned applying DMack correction and EDTM is double counting

  - Nathan will make a version with CPULT and DMack hodo correction instead

  - Dave expects there will be a rate dependence, but is interested to see if

    the scatter in the data is less

Garth: baffled that the strong correlation is with HMS rate, which is not

  large, and not with SHMS rate which is 3x larger

  - would like a better explanation to justify such a large correction to the data

Dave: in the 6-GeV era we had to determine a coincidence blocking correction from

  CoinTime data, maybe something similar is going on here?

  - we discuss whether coincidence blocking is still possible with the 12-GeV

    DAQ

  - Nathan will give Dave some run numbers where the HMS rate correction is

    fairly large to look for evidence of coincidence blocking in the CoinTime

    histo

Junaid

------

Setting up for Q2=3.8 physics analysis

- ran SIMC for these settings

- calculating MM offsets to be applied to data to line up with SIMC

- then will finalize MM cut and Diamond cuts

Vijay

-----

Diamond cut for Q2=0.425 PionLT

- checked SIMC diamond cut compared to data and found a shift

  - it turned out the kinematic offsets weren't applied to MC

  - when offsets appied, shift in SIMC diamond is much less

  - will adjust diamond in one corner which is not filled by SIMC

  - will shift data MM peaks to overlap SIMC

Nacer

-----

KaonLT Q2=0.50 physics analysis

- all cuts applied to data but not random subtracted for purposes of

  determining t-binning and diamond cuts

- Lambda: chose 7 bins, ~15k events/bin at both high and low epsilon

  - made plot of SIMC t-ti to check t-migration

  - ti has no radiation or multiple scattering, while t is simulated detected

    data

  - fits t-ti with a Gaussian, sigma ~0.002 for both high and low epsilon

    however there is a tail (due to radiation) which is about 0.02 long

  - the narrowest t-bin is 0.007 wide

  - Dave: theere will be a small amount of t-bin migration but this should be

    acceptable

  - Vijay notes that one bin limit looks wrong, there is a gap between bins,

    Nacer will investigate

- t-phi plots at high and low epsilon, with all SHMS settings superposed

  - Garth: the highest t-bin extends to 0.166 and the t-phi plot looks very

    sparsely populated there at 6 and 12 o-clock

    - Nacer will make binned plots

    - also no diamond cut was applied so actual t-coverage will be smaller

    - most likely the t-range of the data needs to be narrowed, such as by

      dropping the highest t-bin

  - Garth also suggests that Vijay and Junaid make similar plots when

    finalizing their t-binning

Richard

-------

KaonLT high Q2 analysis

- summarizes meeting on Wed with Tanja, Pete, Garth

  - if possible, we should aim to have an xB=0.40 scan in our first paper

- summary of xB=0.40 status:

  Q2=3.0, W=2.32 hasn't been looked at in ~1 year.  There have been lots of

    improvements since then, will look at again

   - a question to be investigated is whether this point can use the same

     functional form as W=2.72 points, given that W=2.32

  Q2=4.4, 5.5: nearly done, will try to finish them up soon

  - likely only 1 t-bin for Q2=5.5 due to statistics

- status of other points:

  Q2=2.115, W=2.95 was looked at last summer, seemed OK.  Will look at briefly

  again but not expecting problems

  Q2=3.0, W=3.1 nearly finalized.  T,TT forms different than Q2=4.4,5.5

- updates on Q2=4.4 L/T-sep

  - removed Q2-dependence to give very simple parameterizations

    - this seems to work very nicely for the lowest t-bin, -t=0.470 with ratios

      fairly flat and close to 1

    - ratios increase for higher t-bins, but still seem fairly flat

  - comparison of data and MC for kinematic and acceptance plots

    - integrated over t-bin, not renormalized to data

    - agreement looks pretty good, which is encouraging

Alicia

------

Proton absorption Geant4 model

- earlier, some anomalous absorption was reported for electrons

  - this was traced to a problem with S2Y NPE and Energy step action, data were

    being reset for each step through detector instead of being accumulated

  - with this fixed, things look more reasonable now

- detection efficiency vs momentum for all particle types

  - very low efficiency for P<0.4 GeV/c

  - no protons for P<0.1 GeV/c reach the hodoscope at all

- plot of absorption vs momentum

  - a significant momentum dependence is observed which is unexpected

  - electrons have a flat momentum dependence for P>3 GeV/c

  - pions ~1 GeV/C have very large absorption

    - can check the MM but expecting this to be in the resonance region, where

      the proton absorption cross section is very large due to Delta resonance

      creation

  - will plan to run larger event samples for next week to remove some possible

   statistical scatter (10k vs 2k events)

- Ali: good to look at where the particles stop in the detectors for different

  momenta

  - yes, this is in progres

- Garth: the simulation covers the full focal plane, using actual (x,x',y,y')

  focal plane values from data

  - maybe some of what is seen is due to edge effect issues.  One way to study

    this might be to use particle gun covering only the edge of the focal plane

  - if the gun sees a smaller momentum dependence, this would indicate the

    observed dependence here is due to edge effects

- spreadsheet of DEMP data for Simonetta Liutti

  - version 1 of spreadsheet is completed

Next Meeting

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

- Thur Mar 27 @ 16:30 Eastern/14:30 Regina

  - KaonLT will go first

- NOTE: meetings start 16:30 Eastern/14:30 Regina from March 13-April 10

  - after that, we can return to 15:30 Eastern for the convenience of our USA

    colleagues