May 28-29/26 PionLT/KaonLT Analysis Meeting Notes ------------------------------------------------- (Notes by GH) Today: PionLT will be discussed first Please remember to post your slides at: https://redmine.jlab.org/projects/kltexp/wiki/Kaon_LT_Meetings Thursday: Present ----------------- Regina - Garth Huber, Nathan Heinrich, Alicia Postuma, Dex Yadlowski CUA - Tanja Horn, Chi Kin Tam FIU - Pete Markowitz Virginia - Richard Trotta Glasgow - Kathleen Ramage, Rachel Montgomery Ohio - Julie Roche Nathan ------- PionLT LT-sep - Q2=5.0 W=2.95 - working on setting up LT-sep code, need to modify for 3rd epsilon setting - Q2=6.0 W=3.19 - t-phi plots and raw pi+ yields per SHMS-epsilon setting - observed a weird variation of yield between Left, Right, Center settings - assumed that some runs were deleted from volatile storage, so decided to replay data again, this is half-done - will rewrite batch submission scripts for new batch system syntax Richard ------- KaonLT LT-sep - last week showed Q2=4.4, now on to Q2=3.0 W=2.32 - low epsilon data look very clean - discussing SIMC background subtraction method w/ Chi Kin - not implemented yet - in ~2 weeks expect to implement method: - first apply pi-subtraction - then compare to SIMC MM radiative tail - do empirical fit for what remains - shows a variety of Q2=3.0 W=2.32 plots using old bkd subtraction method - Data/MC ratios - kinematic plots for Data and MC - separated cross sections: sigL ~ 3X sigT (!) - looked at Rosenbluth fit plots, high epsilon > low epsilon, but hard to tell if consistent w/ sigL = 3X sigT - everything will have to be revisited after revised bkd subtraction Dex --- KaonLT u-channel pi0 and DVCS analysis - goal of study: estimate systematic uncertainty of potential u-channel DVCS events - starting with Q2=3.0 W=3.14 center low epsilon setting - showed sample fit of pi0 and DVCS MC to MM data - expt peak is better fit with sum of 2 MCs than single MC alone - the fit is sensitive to what MM-shift is assumed for the data - the question is how stable/unstable is the DVCS yield when the data MM-offset is changed? - shift MM data to the right in steps from 0.0 to 0.0055 and refit MCs to MM data - will eventually look at Left, Right, Center settings at both epsilon - also will look at pi+n, K+Lambda, p+omega data to see if get same MM-offsets everywhere Alicia ------ u-channel DVCS future prospects - presents ideas for a potential LOI in 2027 - summary of questions raised: - Julie: the Hall A VCS experiment was at much lower energy, the proton went backward. At higher energy, expect that the proton goes forward due to large Lorentz boost - will send a copy of her thesis (in French, on DVCS @ Mainz) - *NB* Dex' MM study really depends on understanding the reconstruction resolution and radiative tail, as well as the offset effects - Tanja: putting the NPS at backward angle is non-trivial, as there is no stand - interested in having Alicia present her ideas at a future NPS Collaboration meeting Friday: Present --------------- Regina - Garth Huber, Alicia Postuma, Nathan Heinrich, Dex Yadlowski, Vijay Kumar, Nermin Sadoun, Nacer Hamdi JLab - Dave Gaskell JMU - Gabriel Niculescu Glasgow - Kathleen Ramage CUA - Chi Kin Tam, Tanja Horn Virginia - Richard Trotta FIU - Pete Markowitz CSULA - Konrad Aniol Kathleen -------- PionLT LD2 yields for 7.9 GeV lumi scan - gets some weird results: HMS anti-boiling for NoTrack & Track yields - some SHMS NoTrack & Track outliers - will recheck current and PID cuts - *NB* Nathan: the EDTM is unreliable for 2-Arm Singles runs, please use estimated CPULT+ELLT instead - also, unless the PID is very wrong, it should not introduce a significant rate dependence here Nermin ------ PionLT LD2 yields for 6.4, 9.1 #1, 9.2 #2 lumi scans - tried to combine the settings together, starting first w/ Carbon runs - combined Tracked Yield slope is flat within errors: (-1.2 +/- 2.3)%/100uA - there are a few outliers at higher current (~80uA) - then combined Scaler, NoTrack, Track relative yields for LD2 - scaler yields more stable, used slope to get boiling correction HMS scaler slope: (-5.46 +/- 0.27)%/100uA - 3 data sets agree reasonably well SHMS: huge differences between the sets - *NB* Nathan: please make plots of both HMS, SHMS scaler vs 3/4-Rate - maybe the ELLT is over-corrected - could try plotting w/o ELLT correction to see - since we don't use ELLT for the physics dta, we don't need to understand ELLT over a wide rate range, need first to see if it's the problem here - *NB* Nathan: determine the systematic uncertainty from the variation in slope for the different settings - Garth: can't just combine settings which have "wrong" results. First need to understand what's going on, and see what is reliable, before combining them - Nermin & Kathleen have started meeting, will meet every 2 weeks Nacer ----- KaonLT Q2=0.5 LT-sep - working on systematics studies - did some tests on a few runs before running scripts on everything - applying -3%, -5% reduction in acceptance cuts - seeing almost no variation in data yields - Garth: a bit surprised the variation is so small - apply MM cut variations of +/- 3% - also will vary pion background subtraction under Lambda, Sigma0 pieaks by varying the normalization of the pi+n peak to the K+ data - method discussed in Regina group meeting is to vary the MM window used for the pi+ normalization Vijay ----- PionLT Low Q2 LT-sep - finishing up changes to draft PRL Chi Kin ------- KaonLT kinematic offsets - OOP offsets using PMY method - did different fits, iteratively excluding a point - 5 energy fit: phi-HMS=+2.949mr phi-SHMS=-0.828mr - since only the 3 highest energies are used in his physics analysis, will try doing a fit with just those - checked if beam angle could be an issue, but it is not - in-plane offsets: investigating dE_beam - plotted dEbeam vs Ebeam for 5 HeepCheck settings, does not lie on a trend - Garth: each beam energy is an independent linac setting, and the Arc Energy measurements are independent, each Ebeam could vary anywhere within the stated uncertainty in the Arc Energy measurement, there does not need to be any trend within this error band - the uncertainty can be looked up in the hclog energy for the measurement, - link easily accessible from the KaonLT status page: https://hallcweb.jlab.org/wiki/index.php/Kaon_LT_Status_Page - HMS saturation correction - found old post from Dave about a saturation issue, which was due to a - significant current offset in the HMS-Q3 power supply. This supply was replaced in 2021, and is not an issue for the PionLT data - Alicia: this correction is in her version of hcana, not sure if it was pushed to the master version or not - *NB* will send a link to the GitHub code to Chi Kin - applying this correction will improve the MM resolution - Gabriel: this correction is also in his analysis - presented a plot of HMS saturation C0 vs HMS P-central, showing a positive correlation - will implement Alicia's correction and recheck Next Meetings ------------- PS: Gabriel and Ioana have been patiently waiting to do their presentation and discussion. Since we still didn't get to it this week, Gabriel will go first next Thursday - Thurs: June 4 @ 13:00 Eastern/11:00 Regina - Gabriel first, then KaonLT will follow - Fri: June 5 @ 12:00 Eastern/10:00 Regina - we will continue where we left off