May 23rd 2008 H1/ZEUS combination meeting A M Cooper-Sarkar

1. May 23rd 2008 H1/ZEUS combination meetingA M Cooper-Sarkar Look at predictions for W/Z production at the LHC These predictions are done analytically using code of James Stirling for the NLO corrections and the W → lepton decays. • W/Z and lepton rapidity spectra • W asymmetry AW = (W+ - W-)/(W+ + W-), lepton asymmetry • Z/(W+ + W-) ratio and Z/(leptons) ratio Compare to other PDFs CTEQ, MRST

2. What has HERA data ever done for us? A little history… Note difference in scale for fractional errors Pre-HERA PDF predictions: global PDF fit without HERA data Post-HERA PDF predictions: global PDF fit with ZEUS data from HERA-I As shown at the first HERALHC workshop HERA data brings a HUGE improvement to the predictions of the W, Z cross-sections because of the improvement in knowledge of the low-x gluon: at high scale (Q2=MW2),in the central rapidity region,the W+ (and W-, Z ) are produced by low-x sea q-qbar collisons. These q-qbar are produced from the gluonby g→qqbar splittingas PDFs evolve. Hence it is the uncertainty on the low-x gluon at low-scale (Q2=Q02) which feeds into the uncertainty on these cross-sectons. (further explanation in EXTRAS)

3. HERA-II PDF projections: PDF fit with ZEUS pseudo-data projected to the end of HERA NEW !! HERA-I PDF predictions: using optimally combined H1 and ZEUS data For previous HERALHC workshops we even made a projections of how good it could get with final HERA-II data. But we were pessimistic We were not expecting the improvement in systematic error that the 2008 H1/ZEUS combination has made. The new predictions are very precise ~1.5% error in the central region

4. Compare to other PDF predictions HERA-I PDF predictions: using optimally combined H1 and ZEUS data The new predictions are very precise ~1.5% error in the central region Remember we will actually measure lepton spectra not W. Lepton +, lepton- spectra retain similar features to the W+, W- But wait.. this does NOT have model dependence

5. Model dependences Varying mc and mb (not shown) gives results well within errors, similarly for fs fc variation is just within errors Q2min variation is well within errors Q2_0 variation is the biggest effect Varying αs is just within errors Varying the parametrization is also significant

6. Variation of Q20: Q20=2 GeV2 Variation of Q20: Q20=6 GeV2 Let’s look at model dependence as a function of y: Variation of Q20 is the most significant model error in the measurable range

7. Variation of αs(MZ) αs(MZ) = 0.1156 Variation of αs(MZ) αs(MZ) = 0.1196 Variation of αs(MZ) is also significant

8. Variation of parametrization Humpy gluon solution Variation of parametrization ZEUS-Style parmetrization Variation of parametrization is also significant- mostly outside central region Overall conclusion: for the W/Z and lepton rapidity spectra there is some irreducible model dependence from choice of parametrization at Q20 and Q20 itself.

9. Now let’s look at ratios W asymmetry and Z/(W+ + W-) ratio The Z/W ratio is the most reliably predicted quantity of all, PDF Uncertainties from experimental input and from model choices almost cancel out of this ratio. The W asymmetry has a larger PDF uncertainty from experimental input, but little model uncertainty in the central region. It is related to uv-dv PDFs rather than to the gluon (further explanation in EXTRAS). Illustrated here: model uncertainty from choice of Q20 cancels out at central rapidity Q20=2 GeV2 Q20=6 GeV2 lepton asymmetry and Z/(leptons) ratio Q20=2 GeV2 For the lepton ratios the wash out of model uncertainty in the measurable region is not quite so perfect as for the W but it is still quite impressive Q20=6 GeV2

10. Now let’s look at ratios W asymmetry and Z/(W+ + W-) ratio αs(MZ) = 0.1156 The model dependence from choice of αs(MZ2) cancels out in W asymmetry at central rapidity but is visible in the Z/W ratio αs(MZ) = 0.1196 lepton asymmetry and Z/(leptons) ratio αs(MZ) = 0.1156 For the lepton ratios the wash out of model uncertainty in the measurable region is not quite so perfect but it is still quite impressive αs(MZ) = 0.1196

11. Now let’s look at ratios W asymmetry and Z/(W+ + W-) ratio Humpy param The model dependence from choice of parametrization is small but visible in these ratios in the central region ZEUS param. lepton asymmetry and Z/(leptons) ratio For the lepton ratios the wash out of model uncertainty in the measurable region is not quite so perfect - parametrization uncertainty does affect the lepton asymmetry in the central rapidity region. Humpy param ZEUS param

12. Comparsion to other PDFsCTEQ6.5 and 6.1 (MRST01/4 and CTEQ6.6 in EXTRAS)

13. Now we need some comparison to other PDFs: ratios: AW CTEQ6.5 PDF predictions Thered line is central value of HERA-I PDF HERA-I PDF predictions: using optimally combined H1 and ZEUS data HERA-I PDF predictions for W/Z and lepton rapididty spectra are in agreement with those of CTEQ6.5 in central values, but quite a bit more precise even after model uncertainty is accounted (~3% vs 5-6%). See E. Perez slides.

14. CTEQ6.1 PDF predictions. Thered line is central value of HERA-I PDF HERA-I PDF predictions: using optimally combined H1 and ZEUS data HERA-I PDF predictions for W/Z and lepton rapidity spectra are consistently higher than those of CTEQ6.1 in central values (despite using a similar zero-mass scheme)

15. Now let’s look at ratios W asymmetry and Z/(W+ + W-) ratio W asymmetry and Z/(W+ + W-) ratio lepton asymmetry and Z/(leptons) ratio lepton asymmetry and Z/(leptons) ratio The Z/W ratio and the Z/lepton ratio are predicted very consistently between different PDF providers. The W asymmetry and lepton asymmetry are not so consistent. This is due to differences in the uv-dv PDF HERA-I PDF predictions: using optimally combined H1 and ZEUS data CTEQ6.5 PDF predictions. Thered line is central value of HERA-I PDF

16. Now let’s look at ratios W asymmetry and Z/(W+ + W-) ratio W asymmetry and Z/(W+ + W-) ratio lepton asymmetry and Z/(leptons) ratio lepton asymmetry and Z/(leptons) ratio The Z/W ratio and the Z/lepton ratio are predicted very consistently between different PDF providers. The W asymmetry and lepton asymmetries are not so consistent. This is due to differences in the uv-dv PDF HERA-I PDF predictions: using optimally combined H1 and ZEUS data CTEQ6.1 PDF predictions. Thered line is central value of HERA-I PDF

17. Summary • Prediction of W/Z at LHC from HERA-I PDFs based on optimal data combination : • Very small experimental errors. • Model uncertainty from choice of parametrization at Q20 and choice of Q20 is significant for W,Z and decay lepton spectra. • Model uncertainty cancels out of W asymmetry in central rapidity range, some model uncertainty is left in lepton asymmetry • Small model uncertainty from alphas in Z/W ratio and Z/lepton ratio (about the same size as from freeing strangeness content of the sea).

18. extras

19. Where did the improvement in W/Z predictions come from? There has been a tremendous improvement in our knowledge of the low-x glue and thus of the low-x sea The uncertainty on the W/Z rapidity distributions is dominated by –- gluon PDF dominated eigenvectors Both low-x and high-x gluon It may at first sight be surprising that W/Z distns are sensitive to gluon parameters BUT our experience is based on the Tevatron where Drell-Yan processes can involve valence-valence parton interactions. At the LHC we will have dominantly sea-sea parton interactions at low-x And at Q2~MZ2 the sea is driven by the gluon- which is far less precisely determined for all x values

20. Where did the improvement in W/Z predictions come from? There has been a tremendous improvement in our knowledge of the low-x glue and thus of the low-x sea Pre-HERA sea and glue distributions Post HERA sea and glue distributions

21. Compare to published ZEUS/H1 results which also used only HERA data Now we have yet more significant improvement in our knowledge of the gluon Note in published PDFs H1 has alphas variation included in model error, ZEUS does not. Resolution of previous discrepancies, improvement in level of uncertainty

22. MRST04 CTEQ6.1 uv – dv Dominantly, at LO Aw= (u d – d u) (u d + d u) And u = d = q at small x So Aw~ (u – d) = (uv – dv) (u + d) (uv + dv + 2 q ) Actually this pretty good even quantitatively The difference in valence PDFs you see here does explain the difference in AW between MRST and CTEQ Q2=Mw2 x- range affecting W asymmetry in the measurable rapidity range

23. Now we need some comparison to other PDFs: ratios: AW Hera-I pdfs exp cteq65 cteq61 Mrst01(4) Negligible model dependence in HERA PDF. Below see AW across full kinematic range

24. But we measure leptons: comparison to other PDFs: lepton asymmetry Hera-I pdfs exp Mrst01(4) cteq61 cteq65 Small model dependence in HERA PDF from both Q2_0 and alternative parametrizations Lepton asymmetry in full kinematic range is in EXTRAS Hera-I pdfs: Q2_0 model dependence Hera-I pdfs: alternative parametrization

25. MRST01 PDF predictions HERA-I PDF predictions: using optimally combined H1 and ZEUS data

26. Now let’s look at ratios W asymmetry and Z/(W+ + W-) ratio W asymmetry and Z/(W+ + W-) ratio lepton asymmetry and Z/(leptons) ratio lepton asymmetry and Z/(leptons) ratio HERA-I PDF predictions: using optimally combined H1 and ZEUS data MRST01 PDF predictions

27. Now we need some comparison to other PDFs: ratios: Z/W Hera-I pdfs exp cteq61 cteq65 cteq66 This is a bit wider than 65 because of strangeness uncertainty Small model dependence in HERA PDF from alphas Hera-I pdfs: alphas model dependence No matter what their other discrepancies all PDFs are agreed on this ratio. Recently strangeness uncertainty has been introduced and this affects it- but it is NOT a big deal, see CTEQ66 Mrst01(4)