Computing the Adler function from the vacuum polarization function

We use a lattice determination of the hadronic vacuum polarization tensor to study the associated Ward identities and compute the Adler function. The vacuum polarization tensor is computed from a combination of point-split and local vector currents, using two flavours of O($a$)-improved Wilson fermions. Partially twisted boundary conditions are employed to obtain a fine momentum resolution. The modifications of the Ward identities by lattice artifacts and by the use of twisted boundary conditions are monitored. We determine the Adler function from the derivative of the vacuum polarization function over a large region of momentum transfer $q^2$. As a first account of systematic effects, a continuum limit scaling analysis is performed in the large $q^2$ regime.