Speaker
Description
Non-local operator has linear divergence on lattice. RI/MOM scheme cannot eliminate the linear divergence in quasi-PDF operator, especially for the clover valence quark. Using RI/MOM scheme will undermine the credibility of our results.
We try to use the square root of Wilson loop to renormalize bare matrix element of TMD-PDF. When calculating TMD-PDF in the rest frame, we found renormalized matrix elements on different lattice spacings are separated, especially for the finer lattice. But if we change the scale from fm to a(lattice spacing), the curves of different lattice spacings are consistent, which indicates that the linear divergence has been eliminated and only the log(z/a) and log(b/a) are left. In fact, perturbation theory tells us that Wilson loop is not able to cancel out all the log divergences, even in the one-loop level.
We are trying to remove those log divergences and renormalize quasi TMD-PDF.
