Can the 1D viscosity profiles inferred from postglacial rebound data be affected by lateral viscosity variations in the tectosphere?