Abstract. For two events observed by the CLUSTER spacecraft, the field line distribution of mass density ρ was inferred from Alfvén wave harmonic frequencies and compared to the electron density ne from plasma wave data and the oxygen density nO+ from the ion composition experiment. In one case, the average ion mass M≈ρ/ne was about 5 amu (28 October 2002), while in the other it was about 3 amu (10 September 2002). Both events occurred when the CLUSTER 1 (C1) spacecraft was in the plasmatrough. Nevertheless, the electron density ne was significantly lower for the first event (ne=8 cm−3) than for the second event (ne=22 cm−3), and this seems to be the main difference leading to a different value of M. For the first event (28 October 2002), we were able to measure the Alfvén wave frequencies for eight harmonics with unprecedented precision, so that the error in the inferred mass density is probably dominated by factors other than the uncertainty in frequency (e.g., magnetic field model and theoretical wave equation). This field line distribution (at L=4.8) was very flat for magnetic latitude |MLAT|≲20° but very steeply increasing with respect to |MLAT| for |MLAT|≳40°. The total variation in ρ was about four orders of magnitude, with values at large |MLAT| roughly consistent with ionospheric values. For the second event (10 September 2002), there was a small local maximum in mass density near the magnetic equator. The inferred mass density decreases to a minimum 23% lower than the equatorial value at |MLAT|=15.5°, and then steeply increases as one moves along the field line toward the ionosphere. For this event we were also able to examine the spatial dependence of the electron density using measurements of ne from all four CLUSTER spacecraft. Our analysis indicates that the density varies with L at L~5 roughly like L−4, and that ne is also locally peaked at the magnetic equator, but with a smaller peak. The value of ne reaches a density minimum about 6% lower than the equatorial value at |MLAT|=12.5°, and then increases steeply at larger values of |MLAT|. This is to our knowledge the first evidence for a local peak in bulk electron density at the magnetic equator. Our results show that magnetoseismology can be a useful technique to determine the field line distribution of the mass density for CLUSTER at perigee and that the distribution of electron density can also be inferred from measurements by multiple spacecraft.
