In a suggested radio propagation experiment using a deep space antenna, accurate calibration of the propagation delay through the Earth's atmosphere is essential. One or two microwave radiometers can be used for this purpose. Differences in precise locations of the radiometer(s) and antenna to be calibrated leave a residual wet path delay value. We computed the Allan Standard Deviation (ASD) of this residual, as well as the one resulting from different pointing positions in the plane of the sky, by simulations. Pointing offsets, e.g., to avoid solar radiation into the radiometer beam, lead in general to an increased ASD. However, for many observation geometries a deliberate pointing offset can compensate for the location differences. In the case studied we found a reduction of the ASD with up to 45% compared to the ASD obtained for a zero pointing offset. The size of the calculated ASD depends strongly on the model parameters used, e.g., the turbulence strength parameter Cn 2, which has a significant natural variation over a year.