As can be seen in the first plot, the agreement between the measurement and the simulation of the Bz component of the magnetic field along the magnetic axis is excellent.
The above simulation has been used to get the radial component of the magnetic field Br as a function of z for different values of the distance r from the magnetic axis:
Using the simulated field map it is possible to calculate the drift lines of electrons in a TPC located in this magnet by neglecting the E and E × B contributions in the Langevin formula (for more information on this, look here) which is a good first order approximation for the drift region in the case of high magnetic fields. The drift lines are indicated in the following plot together with the contour of the new TPC prototype which is currently being built at DESY (indicated by the red dashed lines).
This picture can be transformed into the radial displacement r - r0 from the start of the drift line.
As can be clearly seen, one would get displacements of the order of a couple of millimeters up to one 1 cm if the reconstruction does not correct for these field inhomogeneities.
Further measurements to verify this simulation are foreseen as soon as the magnet can be operated again after the HERA shutdown.