Reply to comment by Rae et al. on “Formation of substorm Pi2: A coherent response to auroral streamers and currents”

[1] The initiation of Pi2 pulsations is often associated with substorm onset, and their generation mechanism has been extensively discussed. Nishimura et al. [2012, hereinafter referred to as N12] performed a multi-event study by using the all-sky imager network and ground magnetometers of the Time History of Events and Macroscale Interactions during Substorms (THEMIS) mission [Angelopoulos, 2008; Mende et al., 2008] and showed that expansion-phase auroral intensifications near the poleward edge of the auroral bulge occur quasiperiodically with a one-to-one correspondence to Pi2 pulses. Those auroral intensifications often propagate equatorward as longitudinally narrow auroral forms, which are called auroral streamers. N12 found that auroral zone Pi2 near the streamer meridian is antiphase with midlatitude-equatorial Pi2, and their amplitude peaks occur near peak intensities of the streamers. This phase relation can be explained by a current wedge that quasiperiodically intensifies in association with streamers, which mark the upward field-aligned current portion of the current wedge. [2] The streamer-Pi2 correlation shown by N12 indicates that substorm Pi2s are driven by multiple plasma sheet flow bursts, each driving a Pi2 pulse. In situ measurements in the magnetotail have shown existence of magnetic field fluctuations (called multiple activations) [Sergeev et al., 1996] and quasiperiodic flows [Kepko and Kivelson, 1999] in the Pi2 frequency range, and such flows are indeed coupled with the ionosphere as evidenced by correlated auroral intensifications near the footprint of the spacecraft [Keiling et al., 2008; Panov et al., 2013]. Their keograms that include all usable data points do not show quasiperiodic auroral intensifications in the Pi2 frequency range. Thus, the conclusions of N12 could be affected by the data selection. Their keograms and magnetograms do not show a clear phase relationship between maximum auroral brightness and ground magnetic field changes. Furthermore, N12 only showed some available magnetometer stations, and the coherence found by N12 between magnetic field pulsations along a meridian does not persist when data from all stations are included. The coherence shown by N12 is due to the similarity of the magnetic bay signature rather than to Pi2. [7] The first two of the above claims result from R13 not considering one of the most important, and fundamental properties of streamers, that being they are localized in both latitude and longitude. Expansion-phase auroral streamers often form in a somewhat localized region that is typically the eastern edge of an auroral surge [Nakamura et al., 1993]. Since an auroral surge is a bright auroral form that evolves on a different time scale from individual streamers, one has to distinguish surge and streamer regions for investigating correlations with Pi2. Since streamers dynamically evolve and generally do not stay at a single meridian, N12 constructed keograms using wide longitude ranges to allow tracing streamers that have any orientation or propagation direction. The keograms from four stations are shown in Figure 2 of N12. The keogram from the FSMI imager data covers the streamer meridians and identified quasiperiodic auroral intensifications as indicated by the red arrows (Figure 2c). R13 successfully reproduced the quasiperiodic activity in their Figure 1a and gave us confidence that the quasiperiodic auroral intensifications do exist in this event. Their Figure 2 (left) demonstrates that the quasiperiodic intensifications can be seen as brightness increases starting from the vertical lines. [8] N12 removed the meridians that are contaminated by moonlight, which is the eastern portion of each imager field of view (FOV). R13 created their Figure 1f by including imager data poleward of the moonlight contamination FSMI imager FOV. This region contains the surge that is also covered by eastern portion of the FSIM imager FOV. The FSIM imager keogram (Figure 2b of N12) mainly detected the surge whose intensity slowly increased and decreased in time without quasiperiodic auroral intensifications. The intensity evolution in this keogram is similar to that of Figure 1f of R13, which contains the moonlight-contaminated meridians of FSMI. Thus, this surge brightness dominated their keogram and masked the quasiperiodic streamer intensifications. The surge region should be separated from the streamer meridians in order to identify auroral structure that is correlated with Pi2. The reanalysis by R13 did not distinguish different types of auroral forms, and therefore, the streamer activity is masked by the bright and slowly varying surge intensity. [9] R13 also mentioned that their keogram did not find a single auroral form that extended latitudinally at a time. That notion could arise because they chose to consider an extended latitude range. The streamers were limited to a degree or two in magnetic latitude, and the keogram is expected to track other auroral forms or light contaminations equatorward and poleward of the streamers. It is thus not appropriate to identify streamers by only examining a wide latitude range as in their Figure 1b. Latitudinally and longitudinally localized auroral intensifications do indeed exist, as is clearly seen in their Figure 1a. [10] N12 selected several magnetometer stations that represent magnetic field variations seen from high to equatorial latitudes. Figure 1 shows the same set of stations as used in Figure 4 of R13 but with all three components. The repetitive magnetic field increases and decreases can be seen at all stations. Pi2 pulses in the horizontal (H) component start to fall (auroral zone) or rise (mid-to-low latitude) at the vertical lines and reach the maximum amplitudes between the vertical lines, in a similar way as the auroral intensity does. [11] Note that the Pi2 pulses at FSMI to ATHA show a mixture of the high- and middle-latitude types of behavior, reflecting these stations' location at intermediate latitudes. Although R13 only considered the horizontal component, the vertical (Z) component has comparable amplitudes of oscillations and shows coherent pulses at these stations, while the Z-component oscillations at the lower latitude stations are much smaller. This suggests that these stations detected oscillations of the westward electrojet that created the H-component oscillations at YKC. The D-component oscillations remained small and are coherent at MSTK to CCNV, indicating that this meridian is close to the center of the current wedge and that the current wedge system essentially explains all components of oscillations. [12] One point that we have not considered is the motion of the current wedge. Since these stations (FSMI to ATHA) are located near the streamer longitude and near the equatorward boundary of the auroral oval, their distances to the streamers significantly vary during the course of the streamer propagation. The associated motion of the current wedge affects the magnetometer records for those stations more significantly than at lower latitude stations. In addition, the simple current wedge model does not include eastward electrojets, which lie equatorward of the westward electrojet and whose behavior during streamers is not well known. Those two effects should be considered for interpreting stations near the equatorward boundary of the oval. [13] Since N12 referred to the unfiltered magnetometer records for identifying streamer-Pi2 correlation, the last comment by R13 does not affect the coherence result of N12. Individual Pi2 pulses can be clearly distinguished from the slowly varying bays and are not filtering artifacts. As discussed above, the surge activity in the FSIM imager FOV lasted longer than individual streamers, and thus could be related to a large-scale current wedge that creates the more persistent positive and negative bays. [14] In summary, the issues raised by R13 do not affect the conclusions given by N12. Considering that streamers form in limited latitude and longitude ranges, the streamer regions should be highlighted without being mixed with other bright, slowly varying auroral forms such as an auroral surge. As confirmed by R13, streamer intensifications occur quasiperiodically and are correlated with Pi2 pulsations over a wide latitude range. The conclusions of N12 are based on a multi-event study and all events show similar correlations between quasiperiodic streamers and Pi2. However, the reanalysis by R13 was limited to only one of the events, although other cases in N12 show streamers extending over a wider area of the sky with fewer effects of surges than in the first event. We believe that R13 could have found answers to their questions by investigating more than just the one event. We again emphasize that Pi2 pulsations are correlated with spatially limited, quasiperiodic auroral streamers, which are likely the ionospheric signature of quasiperiodic fast flows in the plasma sheet during the substorm expansion phase. [15] This work was supported by NASA contracts NNX09AI06G and NAS5-02099, NSF grants AGS-1101903 and AGS-1004736, and CSA contract 9F007-046101. [16] Robert Lysak thanks the reviewers for their assistance in evaluating this paper.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.