Abstract: We consider cumulant moments (cumulants) of the thrust distribution using predictions of the full spectrum for thrust including O(alpha(3)(s)) fixed order results, resummation of singular (NLL)-L-3 logarithmic contributions, and a class of leading power corrections in a renormalon-free scheme. From a global fit to the first thrust moment we extract the strong coupling and the leading power correction matrix element Omega(1). We obtain alpha(s)(m(Z)) = 0.1140 +/- (0.0004)(exp) +/- (0.0013)(hadr) +/- (0.0007)(pert), where the 1-sigma uncertainties are experimental, from hadronization (related to Omega(1)) and perturbative, respectively, and Omega(1) = 0.377 +/- (0.044)(exp) +/- (0.039)(pert) GeV. The nth thrust cumulants for n >= 2 are completely insensitive to Omega(1), and therefore a good instrument for extracting information on higher order power corrections, Omega'(n)/Q(n), from moment data. We find ((Omega) over tilde '2)(1/2) = 0.74 +/- (0.11)(exp) +/- (0.09)(pert) GeV.

Abstract: We determine the (MS) over bar charm quark mass from a charmonium QCD sum rules analysis. On the theoretical side we use input from perturbation theory at O (alpha(3)(s)). Improvements with respect to previous O (alpha(3)(s)) analyses include (1) an account of all available e(+)e(-) hadronic cross section data and (2) a thorough analysis of perturbative uncertainties. Using a data clustering method to combine hadronic cross section data sets from di ff erent measurements we demonstrate that using all available experimental data up to c. m. energies of 10 : 538 GeV allows for determinations of experimental moments and their correlations with small errors and that there is no need to rely on theoretical input above the charmonium resonances. We also show that good convergence properties of the perturbative series for the theoretical sum rule moments need to be considered with some care when extracting the charm mass and demonstrate how to set up a suitable set of scale variations to obtain a proper estimate of the perturbative uncertainty. As the fi nal outcome of our analysis we obtain (m(c)) over bar((m(c)) over bar) = 1 : 282 +/- (0.009)(stat) +/- (0.009)(syst) +/- (0.019)(pert) +/- (0.010)(alpha s) +/- (0.002)(< GG >) GeV. The perturbative error is an order of magnitude larger than the one obtained in previous O (alpha(3)(s)) sum rule analyses.

Abstract: We introduce an operator depending on the "transverse velocity'' r that describes the effect of hadron masses on the leading 1/Q power correction to event-shape observables. Here, Q is the scale of the hard collision. This work builds on earlier studies of mass effects by Salam and Wicke [J. High Energy Phys. 05 (2001) 061] and of operators by Lee and Sterman [Phys. Rev. D 75, 014022 (2007)]. Despite the fact that different event shapes have different hadron mass dependence, we provide a simple method to identify universality classes of event shapes whose power corrections depend on a common nonperturbative parameter. We also develop an operator basis to show that at a fixed value of Q, the power corrections for many classic observables can be determined by two independent nonperturbative matrix elements at the 10% level. We compute the anomalous dimension of the transverse velocity operator, which is multiplicative in r and causes the power correction to exhibit nontrivial dependence on Q. The existence of universality classes and the relevance of anomalous dimensions are reproduced by the hadronization models in Pythia 8 and Herwig++, though the two programs differ in the values of their low-energy matrix elements.