160 likes | 268 Views
Finalizing MLLA comparisons. Andrey Korytov Alexei Safonov. Analytical results are infrared stable cut-off scale parameter Q eff can be pushed down to QCD ~250 MeV . Soft partons are accounted for!.
E N D
Finalizing MLLA comparisons Andrey Korytov Alexei Safonov
Analytical results are infrared stable cut-off scale parameter Qeff can be pushed down to QCD~250 MeV. Soft partons are accounted for! Hadronization occurs locally at the last moment hadrons “remember” features of parton distributions. Nhadrons/Npartons=KLPHD Hadron distributions are related to Parton ones! (Bassetto et al.MLLA (Modified Leading Log Approximation) + LPHD(Ya. Azimov et al., 1985) (Local Parton-Hadron Duality) Perturbative dominance scenario. Result =Perturbative Model, which potentially may coherently describe jet fragmentation! Two parameters only - Qeff and KLPHD
MLLA spectrum: • Parton level multiplicity: • Hadron momentum distribution: • Parameters: KLPHD, r, Qeff(Qeff=240±40 MeV from momentum distribution fits)
Does MLLA have a chance? CDF Preliminary (each distribution fitted separately) • Reasonable qualitative agreement • Quantitative match is not perfect (excess and not exact shape)
Does MLLA have a chance? • Q is not a “universal” constant (systematic errors are correlated!). Qeff=240±40 MeV
Does MLLA have a chance? CDF Preliminary • K is not constant for fixed energy
Does MLLA have a chance? • Obviously, we cannot talk about formal agreement between the data and theory. Do we expect one? • Simple hadronization assumption (in fact, absence of it) may be too naïve. • If the problem is in the hadronization stage, it is still possible that properties of the hadrons are mostly determined by parton distributions (perturbative dominance scenario). Qualitative agreement supports this.
MLLA - “mostly correct” model Even though MLLA is not a precise model, it describes data fairly well. • Let’s consider MLLA as a “mostly correct model” with allowed minor deviations from data. • Then MLLA parameters still can be extracted and will make sense.
MLLA - “mostly correct” model • K is not the same as KLPHD. We need to fit K vs fraction of gluon jets in the sample. • If MLLA - almost true, then Qeff is “almost universal”. • It is better to refit 9 distributions with 10 parameters (9 values for K(Ejet)+Qeff). • Result will be more consistent if we want to do something else further with these Ks
Parameter KLPHD CDF Preliminary
Fit procedure: • Chi-square: • Correlation matrix: • Coefficients can not be precisely defined and have to be varied within reasonable range.
Fit for KLPHD Separate fit Combined fit
Peak Position Systematic errors are correlated!
Peak Position • Same fit procedure. • 2 smaller cones added • Error comes from fit + comparison of fits for all 5 cones(Q=240-250 MeV) and for 3 larger cones only(Q=250-290 MeV).
Conclusion • MLLA is not perfect. Too naïve LPHD assumptions may be responsible • If MLLA is “mostly correct”, MLLA parameters can be extracted. • Qeff=240±40MeV. (250±40 MeV from peak position) • Indirectly measured KLPHD=0.75±0.06, ratio of multiplicities in gluon and quark jets r=1.8±0.4. • Agreement with multiplicity comparison to MLLA KLPHD=0.69±0.3±0.5, r=1.7±0.3