The accuracy of a magnetic field experiment on a spacecraft is sometimes limited by the spacecraft field and not by the accuracy of the magnetometer itself. It has been suggested by Ness and collaborators that N?2 magnetometers on a radial boom may be employed to estimate the first N-1 multipole contributions to the spacecraft field and thereby improve the total accuracy of a magnetic field experiment. In this study the total error in the measurement of the ambient field for systems of one to four magnetometers is investigated. The optimal magnetometer locations, for which the total error is a minimum, are computed for given boom length, instrument errors, and very conservative magnetic field models characteristic for spacecraft with a only a restricted or ineffective magnetic cleanliness program. It is found that the error contribution by the magnetometer inaccuracy is increased as the number of magnetometers is increased, whereas the spacecraft field uncertainty is diminished by an appreciably larger amount. Therefore for the field models chosen and a boom length less than 10 m from the spacecraft center there is always a gain in accuracy compared with results for one magnetometer. For example, for a boom length of 6 m, instrument errors of 0.1 &ggr;, and a rather conservative model describing MVM73 the total error bounds are 2.1, 0.856, 0.822, and 0.807 &ggr; for one to four magnetometers, respectively. |