Penguin Loops for Nonleptonic B-Decays in the Standard Model: Is there a Penguin Puzzle?

Abstract

We compute standard model penguin amplitudes in nonleptonic B-decays to light charmless mesons using tree amplitude data to fix hadronic parameters. The leading calculation is carried out for the alphas(mb) penguin contributions from charm quark, up quark, and magnetic penguin loops in the NDR and HV renormalization schemes. Power suppressed penguins that are proportional to the chiral condensate are also computed using a new factorization formula for these terms, which is derived working to all orders in alphas(sqrt{mbΛ}). We demonstrate using SCET1 that this formula exhibits only small perturbative phases and does not have endpoint singularities. Due to our use of data to fix hadronic parameters we obtain significantly more accurate predictions for the short-distance standard model penguin amplitudes than have been found in the past. Analyzing data in B-> pi pi, B->K pi, and B->rho rho for the penguin amplitudes we find that standard model short-distance imaginary parts are an order of magnitude smaller than current measurements, while real parts are up to a factor of two smaller with the correct sign. This difference is most likely a consequence of long-distance charm contributions or new physics. Constraints on the type of new physics that could help explain the data are derived, and used to show that current data favors sizeable long-distance strong phases.