Nov 13-14/25 PionLT/KaonLT Analysis Meeting Notes ------------------------------------------------- (Notes by RLT & GH) Today: KaonLT will be discussed first Please remember to post your slides at: https://redmine.jlab.org/projects/kltexp/wiki/Kaon_LT_Meetings Thursday: Present ---------------- Regina - Nathan Heinrich, Nacer Hamdi, Alicia Postuma, Nermin Sadoun Virginia - Richard Trotta CUA - Tanja Horn, Chi Kin Tam JLab - Dave Gaskell Richard ------- KaonLT Q2=3.0, W=2.32 LT-sep - Using the same functional forms as Q2=4.4, W=32.74 - Ratios look fairly flat with little structure. There are 2-3 bins that dip and need investigation - The distributions look okay overall, but more iterations will be needed (currently only 3). The delta distributions, in particular, are a bit off. - New outline for analysis going forward - Finish and refine Q2=3.0/W=2.32, Q2=4.4/W=2.74,Q2=5.5/W=2.95 (scaling settings) - Then work on the fit algorithm paper and studies (fitting Fpi1/2, low Q2 Pion-KaonLT, etc. data) - Chi Kin to finish up Q2=3.0/W=3.14 - Then finish cross section systematics checks on these settings - Once Sameer is finished up with cointime blocking correction, a final full replay will be done - With the full replay finished, Richard will check and finalize systematics Chi Kin ------- KaonLT Q2=3.0, W=3.14 analysis - Rebinning in t and adjusting t-range - Showed average t per bin vs t bin center - Same functional form as Richard - Initial parameters determined from distributions - 0th iteration lookd reasonable with high > low eps in unseparated cross sections - Shape a bit odd for sigT and just one iteration dramatically changes the cross sections and ratios look much worse - Unphysical parameters (local minima likely) result in a negative total cross section - Will be using Richard's fit algorithm for a better set of initial parameters - Nathan: given the plots, how can the total cross section be negative? - Likely a bug in the code because the 0th iteration overall looks very reasonable (unspe/ratios/sep/kin/etc). - *NB* GH: it's important to plot cross sections over the full kinematic range of the data, not just the range spanned by the nominal t-values of the bins. Maybe the parameterization is doing something weird on the edge of the distribution? - Discussion if Kin should shift data and simc MM to lambda mass of 1.115 or just shift MM data to simc. - Tanja: simc is "truth" so data should shift to it Alicia ------ pi+n BSA PLB paper - Has been accepted, there are a few, brief comments from one referee - There's some confusion about exact coverage and correlation of kinematics for different settings - Garth suggested adding a t-phi plot but space is limited. - Tanja suggested a reference to previous LT papers - what Garth actually meant was to put the plot in the supplemental material, not the paper - Referee unsure why bin widths are indicated by standard deviation instead of just edge of each bin - Alicia and Garth have showed to our group the different methods so just explaining to referee is probably fine - Radiation corrections are still not discussed in text - Alicia is expanding some of the text - Fix Q2/x but correlated with t - Leave as it - Proper citations - PLB Website issues - Figure 7: GK model has shows such strong dependence on Q2 - Alicia: scaling of denominator to show dependence, but is that worth doing? - Tanja: Probably fine as is, not a theory paper - Various minor edits to wording, etc. - No comments, looks good - expect to submit final version for publication in a few days Nacer ----- KaonLT Q2=0.5 LT-sep - Verbal: Trying to improve ratios with continued adjustments Nathan ------ PionLT finishing up CoinLumi studies - Showed coin yield vs current - Using Dave's previous suggestion, fit to constant - No chi2 yet from Nathan, Dave says that should be first step. - If good chi2 then statistically consistent with a constant so no systematic correction. - If bad, then need to think about systematic. Possibly, amplify error bar until good chi2 achieved then quantify that amplification. - Friday update: did a combined fit using error-weighted average, get flat vs rate within ~1 sigma - Note added after meeting: Nathan finds that adding 0.72% to the error bars results in a reduced Chi-square of 1. Dave and Garth agree the adjusted plots look good - Coinblock corr vs coin rate - A bit of a spread, but trends per setting are consistent - Two settings (Nathan's categorization, exclusive lumi 3 vs 4/5) should be the same but aren't. Someone may have adjusted the electronics, but no log entry on this. - Garth has no recollection of any such electronics changes during the run - One other point is off trend, but error bars are large so likely low statistics run. Friday: Present --------------- Regina - Garth Huber, Nathan Heinrich, Alicia Postuma, Muhammad Junaid, Vijay Kumar, Nacer Hamdi CSULA - Konrad Aniol FIU - Pete Markowitz Virginia - Richard Trotta CUA - Tanja Horn, Chi Kin Tam, Sameer Jain JMU - Gabriel Niculescu York - Stephen Kay Ohio - Julie Roche Vijay ----- New Low Q2 PionLT results - CoinTime blocking and pion absorption corrections now included for both Q2 - pion absoprtion correction: 2.02%+/-0.045% (see June 5/25 slides), correction is applied by dividing by 0.979 - Q2=0.375 results - mid epsilon Data/MC ratios: 0.9-1.2 with some scatter - Rosenbluth fit of sig_uns has smaller chi-square than before - 5-10 iterations done after corrections applied - Q2=0.425 results - mid epsilon Data/MC ratios: some larger ratios at edges of phi-dist, 0 and 360 deg - TT trends to 0 at -t=0.01, LT is constant (not zero) at -t=0.01 - sigL vs t is less smooth than at Q2=0.375, big jump for lowest t-bin - next steps: - working on systematic uncertainties - preparing kinematics and acceptance plots for next week - will send tar files of results to spokespersons Sameer ------ KaonLT CoinTime blocking at 6.2, 8.2, 10.6 GeV beam energies - still need ROOT files for Q2=4.4 (10.6 GeV) - no error bars yet - blocking corr @ Q2=3.0, W=3.14 - get 3 parallel bands vs Rate with values of 0.92, 0.93, 0.94 - *NB* Garth: this seems weird, is it a round off issue when calculating the ratio? - blocking corr @ Q2=5.5, W=3.02 - high epsilon shows a band at 0.78, with rest at 0.85 - *NB* Nathan: concerned that the timing window is not set correctly for these - maybe standard.database is not pointing to the correct file for these runs? - low epsilon shows some corrections of 0.4 at low rate - *NB* Nathan: this needs more investigation, compare the shape of timing plots for good and bad runs - if shapes are similar, the correction can't be very different - if shapes are different, then need to confirm the applied cuts are correct - *NB* Garth: please show sufficient information next time that we can help you determine what might be wrong Junaid ------ PionLT Q2=3.85, W=2.62 LT-sep - working on pion absorption correction following Alicia's instructions - turned on NGC in Geant4 code - set Aerogel tray to n=1.011 thickness - ran 100k events - applied default cuts should be okay for this configuration - resulting correction is similar to Vijay's but there are more secondary particles produced due to the higher momentum - *NB* Garth: please give more info next time on how the secondaries are handled in your Geant4 analysis - all other corrections are applied to the data Next Week Meetings ------------------ - Thurs: Nov 20 @ 16:00 Eastern/15:00 Regina - PionLT will go first - Fri: Nov 21 @ 11:00 Eastern/10:00 Regina - we will continue where we left off