Influence of epoch time selection on the results of superposed epoch analysis using ACE and MPA data

The influence of the reference time selection when conducting a superposed epoch analysis is examined for intense geomagnetic storms at solar maximum. The events were selected according to the minimum pressure-corrected Dst, Dst*, being less than -100 nT. Solar wind data from ACE are used, along with near-Earth data from the magnetospheric plasma analyzer (MPA) instruments on the Los Alamos National Laboratory-operated geosynchronous spacecraft. Numerous choices for the zero epoch time are used, ranging from the storm sudden commencement (SSC), the peak of the ring current enhancement (minimum Dst* slope), to the time of the storm peak (minimum Dst* value). When doing superposed epoch analysis (SEA), the choice of the time stamp can be very important; for different choices, different storm characteristics are evident in the averaged data. In the superposed ACE data we find that when using the SSC as a time reference, the SSC-related jump in solar wind parameters is very well defined, but near the storm peak, B_z does not apparently follow the well-known criteria for intense storms (B_z ≤ -10 nT for more than 3 h), even though this criterion is met by most of the individual storms selected for this study. When the zero epoch time is chosen near the storm peak, the jump in solar wind parameters is less distinct (and eventually lost), but the criterion for B_z is met. Regarding the MPA data, there are certain parameters that require the choice of a specific epoch time in order to produce a systematic behavior in the SEA analysis and others that are less sensitive to this choice of epoch time, since they appear to be less distinct in their temporal and spatial location. For instance, the nightside and morningside hot-ion density and temperature are main phase traits, and a zero epoch time near the peak of the ring current enhancement is required to make these features distinct.