During 15 days in a relative quiet geomagnetic time around the M8.6 Sumatra earthquake on 28 March 2005, ionospheric perturbations occurred before this event on March 20, 22, 23, and 28，2005 in many parameters such as the electric field and plasma parameters in the equatorial area. Since it was a quiet magnetic time, those four days referred above may be related to the preparation for this strong earthquake. The electron density showed two types of anomalies, one being a monotone increase in the single peak values with amplitudes exceeding 1σ, such as on March 20 and 28 in 2005; the other one changing the normal single peak to double crests and a trough in the equatorial area, which occurred on March 22 and 23, 2005. Both the anomalies are distributed over a very large area of 100°longitude and 30°latitude, presenting a long duration and global feature. The study of VLF electric field spectrums demonstrated that those anomalies were not presented in the geomagnetic conjugate feature. Electric field perturbations were concentrated in a relative smaller area than those of Ne, occurring only in the nearest orbit to the epicentre with only monotone increasing Ne anomalies. It is thought that it is difficult for an electric field generated close to the seismic preparation area in the equatorial area to penetrate into the ionosphere directly, but it may occur through the coupling mechanism of lithosphere-atmosphere-ionosphere. Combined with the analysis of ion composition and varying shape, this shows that prior to this earthquake the variation in oxygen density reversed to that of hydrogen and helium density, similarly to other ionospheric precursors before other earthquakes. It is concluded that conjoint analysis among different parameters should be applied in order to distinguish the precursors related to the earthquakes.
