Dresden 2013 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

T: Fachverband Teilchenphysik

T 22: Gittereichtheorie

T 22.8: Talk

Monday, March 4, 2013, 18:30–18:45, WIL-C129

Mass–reweighting: Problems und Prospects — •Jacob Finkenrath, Francesco Knechtli, and Björn Leder — Bergische Universität Wuppertal, Germany

Today’s simulations in lattice quantum chromodynamic get closer and closer to the physical point by simultaneously minimizing the statistical errors. Systematic effects like isospin symmetry–breaking start becoming significant. An effective method which can incorporate these effects is the correction of the Boltzmann factor of the ensemble.

This correction enters as the reweighting factor which is the ratio of the new over the old Boltzmann factor. In the case of a mass shift Δ m the resulting determinant ratio can be estimated stochastically. Thereby the effectiveness of reweighting is regulated by the fluctuations in the range Δ m (ensemble fluc.) and by the computational cost (stochastic fluc.).

We want to discuss some problems of mass–reweighting, arising from the estimation but also from the reweighting range Δ m, and some prospects like reduction of costs by employing domain decomposition and by using correlations between parts of the action. Our result are obtained on ensembles with lattice spacing a=0.66  fm by reweighting from 440  MeV to 310  MeV pions.