Resources needed for rapid analysis of W events from 2009 pp 500 GeV run
(relay only RCF resources)

Updated 16 April 2009 (UTC)

Primary goal is publication of W xsection. Approaching conferences do not define deadlines for intermediate results.
We will work first on sufficient quality calibration.
If final TPC tracking accuracy is not adequate we will not publish separately W+ vs. W- .

1. real events data sets , obtained ~1TB (mit+iucf+data05+06+07)
   • 1.1M L2W tagged events saved in W-stream (muDst=400GB)
   • 1.5M VPD minB events from Fill F10471 (muDst=500GB)
2. MC events Pythia 500 GeV, geometry 2009, needs ~200GB (uky)
   • 1M filtered QCD Pythia events partPT>10 GeV/c, LT=10/pb,filter: stable particle 'cluster' ET>20 GeV, 5e8 Pythia events thrown
   • assorted rare Pythia events involving Ws, Zs, W-jets, etc, above 100/pb apiece
3. TPC calibration needs : low lumi data set at 200 GeV 3M events (to be taken!) , 200GB disk for muDst (data07), 1-2 CPU years
   • status tables (time) , per longer runs
   • TPC: variety of corrections based on low lumi run (to be taken at pp 200 GeV) and sampled pp 500 data
   1. twist, inner/outer alignment done by STAR core team - 1 week of work according to Gene
   2. Space Charge corrections by Joe: muDst=200GB, 2.5 CPU month per pass, 3-8x iterations
   • TPC: evaluation of achieved pT resolution - 1 week
   • determine beam line constraint
   • Vertex finder QA - 1 week
4. EMC calibrations: ~week of CPU , few weeks of FTE per detector
   • QA of runs
   • status tables (BTOW,ETOW,BSMD) , pedestals (BTOW,ETOW), per fill
   • BTOW: relative gains (per towers) from slopes from minB eve, absolute=ideal
   • ETOW: as above
   • BSMD: relative gains (per strips) from slopes from minB eve , absolute using 2008 pp calibration (done)
   • polarization related: STAR bXing, fill pattern, CNI : fill pol(time)
   • BPRS, TOF - not used unless needed
5. CPU needs for data analysis: 2 CPU years (TPC calibration effort not included)
   • full BFC production of L2W events: (28sec/eve), 250 CPU days per pass (assume 2nd iteration may be needed)
   • real muDst analysis: 13 Hz, 1 CPU day/pass, x 20 iterations
   • MC muDst analysis: 13 Hz, 1 CPU day/pass, x 20 iterations

Approximate schedule, milestones, an estimate

1. calibrations ~1-2 months
2. BFC 1st production ~1 week
3. tuning of W-cuts ~2 weeks, first W peak in ET spectrum
4. comparison to M-C, efficiency : ~1 month
5. (if needed) BFC 2nd production : ~1 week
6. final analysis , preliminary cross section released : ~1 month
7. writing paper, cleanup of analysis, draft ready for spin PWG : ~3 months
8. comments from spin PWG, analysis : ~2 months, draft ready for GPC
9. working w/ GPC : ~3 months, paper sent to publisher