V Friese GSI analysis and paper status NA

V. Friese GSI analysis and paper status NA 49 collaboration meeting, Trento, April 2004 Volker Friese

Invariant mass signals y > y. CM 0 < pt < 2. 5 Ge. V Fit of rel. BW with mass resolution + linear residual background gives good description NA 49 collaboration meeting, Trento, April 2004 Volker Friese

Fit results Ebeam m 0 σm [Me. V] 72 1019. 0 ± 0. 1 1. 8 ± 0. 1 31. 1 57 1019. 1 ± 0. 2 1. 1 ± 0. 2 37, 000 26. 2 53 1018. 9 ± 0. 2 2. 1 ± 0. 3 394, 088 16, 500 12. 5 104 1018. 4 ± 0. 5 2. 5 ± 1. 3 356, 070 6, 500 9. 4 70 1018. 8 ± 0. 6 2. 6 ± 0. 9 events S SNR B/S 158 345, 543 180, 000 49. 4 80 300, 173 56, 000 40 586, 768 30 20 [AGe. V] NA 49 collaboration meeting, Trento, April 2004 Volker Friese

phi mass vs. energy measurements are sytematically below PDG value Possible reason: acceptance probability is mass dependent (inv. mass ↔ opening angle) NA 49 collaboration meeting, Trento, April 2004 Volker Friese

transverse spectra Good description by thermal distribution NA 49 collaboration meeting, Trento, April 2004 Volker Friese

Slope parameters Ebeam [AGe. V] NA 49 collaboration meeting, Trento, April 2004 Δy pt, max T [Ge. V] [Me. V] χ2 / ndf 158 1. 0 299 ± 7 1. 5 80 1. 6 251 ± 11 0. 2 40 1. 5 2. 0 245 ± 9 1. 5 30 1. 9 1. 8 237 ± 18 0. 3 20 1. 8 2. 0 203 ± 22 0. 3 Volker Friese

rapidity spectra good description by Gaussian no complete forward coverage at 158 AGe. V due to momentum cut on kaons no midrapidity coverage for 40, 30, 20 AGe. V (gap 0. 3 units) due to d. E/dx crossing K/ Total yield by summing the bins and extrapolating with gaussian (or gaussian + linear) difference → syst. error NA 49 collaboration meeting, Trento, April 2004 Volker Friese

Fit results (2) Ebeam σy dn/dy (y. CM) χ2 / ndf < > 158 1. 47 ± 0. 09 2. 40 ± 0. 06 0. 43 8. 46 ± 0. 38 ± 0. 33 80 1. 01 ± 0. 03 1. 37 ± 0. 06 2. 25 3. 36 ± 0. 10 ± 0. 06 40 0. 86 ± 0. 03 1. 18 ± 0. 08 0. 80 2. 50 ± 0. 12 ± 0. 05 30 0. 77 ± 0. 05 0. 91 ± 0. 14 0. 70 1. 73 ± 0. 20 ± 0. 02 20 0. 58 ± 0. 04 1. 23 ± 0. 24 0. 46 1. 77 ± 0. 27 ± 0. 00 [AGe. V] no full phase space yield available at other energies NA 49 collaboration meeting, Trento, April 2004 Volker Friese

yield ratios Strong increase from SPS to RHIC Situation unclear AGS / RICH : Au+Au, midrapidity ratios NA 49 collaboration meeting, Trento, April 2004 Volker Friese

Checks (1) Method to fit inv. mass spectra: Fix Γ 0 = 4. 43 Me. V, leave m 0 and σm free Reason: Γ 0 and σm hard to disentangle For 158 AGe. V: Quality of data allow a free fit Results: Γ 0 = (4. 41 ± 0. 61) Me. V, σm = (1. 81 ± 0. 26) Me. V → No deviation from natural width seen No influence on T, σy, dn/dy, < > Desirable: Show that σm is reproducable by simulation NA 49 collaboration meeting, Trento, April 2004 Volker Friese

Checks (2) Standard method: Fix m 0 and σm in pt/y bins to values obtained from the mass spectrum in the full acceptance. For 158 AGe. V: data quality allow to leave parameters free in each pt/y bin No systematic variation of m 0 or σm with y or pt seen Influence on results are within statistical errors fixed parameters free parameters T [Me. V] 299 ± 7 302 ± 9 σy 1. 47 ± 0. 09 1. 51 ± 0. 12 dn/dy (y. CM) 2. 40 ± 0. 06 2. 39 ± 0. 08 < > 8. 45 ± 0. 38 ± 0. 33 8. 86 ± 0. 61 ± 0. 34 NA 49 collaboration meeting, Trento, April 2004 Volker Friese

Phase space dependence of mass and resolution NA 49 collaboration meeting, Trento, April 2004 Volker Friese

Checks (3) When correcting the raw yields for acceptance, independence of T from y is assumed (statistics do not allow two-dimensional binning in y and pt) 158 AGe. V data allow 2 -dim binning (in 4 rapidity bins with Δy = 0. 4) Slight (if any) y dependence of T Effect on yields is well below statistical error NA 49 collaboration meeting, Trento, April 2004 Volker Friese

Check (4) For 158 AGe. V and 40 AGe. V: Seperate analysis for both field polarities Field T [Me. V] σy dn/dy < > ALL 299 ± 7 1. 47 ± 0. 09 2. 40 ± 0. 06 8. 45 ± 0. 38 STD+ 305 ± 10 1. 50 ± 0. 14 2. 37 ± 0. 09 8. 49 ± 0. 53 STD- 293 ± 9 1. 44 ± 0. 11 2. 42 ± 0. 09 8. 41 ± 0. 55 Negligible difference for 158 AGe. V, 2 stddev for 40 AGe. V (vanishes for stricter d. E/dx cut on kaons) Field T [Me. V] σy dn/dy < > ALL 248 ± 9 0. 86 ± 0. 03 1. 18 ± 0. 08 2. 50 ± 0. 12 1/4 STD+ 251 ± 13 0. 91 ± 0. 06 0. 96 ± 0. 11 2. 16 ± 0. 16 1/4 STD- 243 ± 12 0. 82 ± 0. 04 1. 44 ± 0. 13 2. 92 ± 0. 19 NA 49 collaboration meeting, Trento, April 2004 Volker Friese

To do • determine systematical errors for 20 and 30 AGe. V • proof that fit method to inv. mass spectra is correct (simulating background kaons with strong fsi) • proof that rec. efficiency is close to 100 % for 158 AGe. V (Marco? ) • blast wave fit to mt spectra • compare yields with predictions (Ur. QMD, Becattini, . . . ) • interpretation of y spectra ? NA 49 collaboration meeting, Trento, April 2004 Volker Friese