However, since the spatial and temporal resolutions of the satellite observation data are limited, it is necessary to analyze data from ground-based imager and magnetometer networks to study the spatiotemporal development of these auroral phenomena in more detail. The auroral surges occurred intermittently at intervals of about 15 min just after the expansion onsets of a substorm and propagated eastward.

We investigate the characteristics of the eastward-expanding auroral surges hereafter referred to as EEASs in detail and compare them with those of omega bands and other auroral phenomena in the post-midnight sector reported by previous studies. The goal of this campaign was to reconstruct the three-dimensional 3D distribution of the auroral emission and horizontal distribution of the energy spectra of precipitating electrons from the multiple auroral images using the tomographic inversion technique Aso et al.

In this paper, however, we focus on the characteristics of three auroral surges observed during the substorm on March 8—9, The results of the tomography analysis will be presented in another paper. Table 1 lists the geographical and the altitude-adjustment corrected geomagnetic AACGM coordinates of the observation stations used in this paper.

The AWIs at TRO measured the emissions at , , and nm using wider bandpass filters about nm at 1-s intervals and white light at 0. The sampling rate of the IMAGE magnetometers is 10 s; thus, we present all the optical data at s intervals to correspond to the magnetometer data.

Location of all-sky imager and magnetometer stations used in this paper. Figure 2 shows a north-south keogram i. Variations resembling relatively sharp negative bays with small amplitudes 30—50 nT can be seen at NOR at , , and UT. These negative and positive bays are consistent with a DP-1 or substorm current wedge current system, which typically occurs during the substorm expansion phase see Kepko et al. The magnetic variations associated with the DP-1 current system were also confirmed in the Y component at mid-latitude not shown here.

Palin et al. Three vertical dashed lines indicate the times when the fronts of the eastward-expanding auroral surges which will be shown below reached the zenith at TRO. In the keogram, each EEAS is characterized as an equatorward movement of a discrete arc and subsequent rapid poleward expansion of the diffuse auroral region. These times coincide with the occurrence of Pi 2 pulsations.

Small but similar optical and magnetic variations can also be identified at UT. North-south keogram from TRO and north-south X component of the magnetic field from the magnetometer network. Top panel shows a north-south keogram obtained from the AWI imager data for the period from to UT on March 8—9, In the bottom panel , the upper eight lines show the magnetic variations detected by the IMAGE magnetometer chain. Three vertical dashed lines indicate the time when the front of the eastward-expanding auroral surges reached the zenith at TRO.

Both solar wind and geomagnetic conditions were moderate during this interval. These color images were composed from the , , and nm images taken by the AWIs. We first focus on the second event Fig. Then, the poleward edge of the diffuse aurora gradually moved equatorward — UT. An auroral streamer marked by A. Almost simultaneously with the auroral streamer, the front of an auroral surge i. In its poleward region, the auroral streamer formed spirals and branched at several places e.

North-south-aligned auroras emerged behind the vortex structure NS in the figure at and UT and finally connected the branched streamer with the EEAS. From these all-sky images, we could not determine whether N-S-aligned auroras emerged from any of the auroral streamers or the auroral surge. The dynamic behavior of the EEAS is very similar for all three events, which occurred intermittently at intervals of about 15 min.

Thus, the EEAS can be regarded as a repeatedly occurring phenomenon during the substorm. Sequence of all-sky images for three eastward-expanding auroral surges observed at TRO. These color images were composed from the , , and nm images taken by the all-sky Watec imagers AWIs.

Magnetic north and east correspond to the top and right , respectively. Thus, we concluded that this multiple-onset substorm was small-scale and did not expand to the premidnight sector. On the other hand, the imager at LYR barely detected the EEASs at the southern edge of the FOV and did not provide additional information on the auroral activity in the poleward region. Systematic temporal variations can be clearly seen in all components of the magnetic field at times when the EEASs were observed.

These variations are marked by dashed-dotted lines. The magnetic variation at TRO shows negative-positive excursions in the X component, positive-negative-positive excursions in the east-west Y component, and positive-negative excursions in the vertical Z component for all the EEASs. The periods of the magnetic variations shown by the dashed-dotted lines are about 4—6 min.

The peak-to-peak amplitudes of the magnetic variations are between 10 and nT. Vertical dashed-dotted lines show the start and end times of the magnetic variations associated with the EEASs, determined by visual inspection. The dashed lines are the same as those in Fig. Figure 5 shows the ionospheric equivalent current associated with the EEASs, plotted with the auroral images from TRO mapped onto a horizontal plane at a km altitude.

The magnetic perturbations associated with the EEASs have been extracted by subtracting a linear trend from the raw data within each interval of the EEAS, as shown in Fig. It should be noted that if we do not remove the trend, the ionospheric current is always roughly westward because the X component is always negative during these intervals Fig. Again, we explain the temporal variation of the ionospheric current using the second event Fig.

Before the arrival of the EEAS, the ionospheric current was roughly westward along the auroral arc and showed a CW rotation centered at a higher latitude than the visible aurora UT. When the size of the vortex became smaller at UT, the CCW rotation of the ionospheric current system is evident around the vortex. This temporal variation in the ionospheric current is very similar for the other two EEAS events. It is well established that if the ionospheric conductivity is horizontally uniform, the rotational divergence-free ionospheric current that is observable on the ground is the Hall current, which flows CW CCW along the equipotential lines around the positive negative potential at high latitude in the Northern Hemisphere.

Although the ionospheric current pattern is deformed because of non-uniform ionospheric conductivity, it is possible to deduce that the CW and CCW rotations in the ionospheric equivalent current correspond to a positive and negative ionospheric electrical potential, respectively, which are primarily caused by downward and upward field-aligned currents FACs c.

Ionospheric equivalent current system associated with EEASs. The speed of the eastward expansion of the surges was estimated from the temporal variation of the auroral images mapped onto a horizontal plane at the km altitude.

The assumed altitude of km is reasonable, because we applied tomography analysis to the nm images taken by the three EMCCD imagers and confirmed that the maximal peaks of the emission were located around to km altitude not shown here. Furthermore, we calculated the footprints of the EEASs on the equatorial plane of the magnetosphere using the Tsyganenko 96 model. The large-scale vortex-like structure of the EEASs at the poleward edge of the diffuse aurora appears to be similar to the undulations of omega bands.

In addition to the spatial structure, both have some similar properties, such as 1 occurrence in the post-midnight sector, 2 eastward propagation, 3 concurrence with magnetic pulsations with amplitudes of 10 to several nT, 4 periodic recurrence with intervals of about 5—40 min, 5 horizontal scale sizes of several hundred kilometers, and 6 existence of a pulsating aurora in the diffuse aurora Opgenoorth et al.

As for item 3 , omega bands are in general accompanied by Ps 6 magnetic pulsations with an amplitude of 10 to more than nT Saito , which are more dominant in the Y and Z components. However, we note that our EEASs show strong pulsations in the X component, which would be somewhat unusual for omega bands.

Item 4 is generally accepted for omega bands because Ps 6 pulsations have periods of 5—40 min. The occurrence interval of about 15 min for the EEASs is equal to the typical average period of omega bands. The typical longitudinal as well as latitudinal extent of omega bands is about — km, which is comparable to that of the EEASs item 5. Regarding item 6 , the pulsating aurora is commonly observed at the center of the torches, i. On the other hand, there are some differences in the properties between the EEASs and the omega bands as follows: a It is deduced from the simultaneous observations of Pi 2 pulsations and the magnetic bay variations associated with the DP-1 current system that each EEAS was observed just after each onset of a multiple-onset substorm, whereas omega bands are usually observed during the substorm recovery phase.

As for item d , omega bands have a systematic and relatively stable ionospheric equivalent current system, which is caused by the large-scale wavy structure along the east-west direction. The ionospheric current shows a CCW rotation around the torch, corresponding to the upward FAC and negative potential, and a CW rotation around the dark region i.

On the other hand, the ionospheric equivalent current of the EEASs indicates a quite dynamic behavior. It is expected from the CCW rotation that the upward FAC and negative potential existed inside the auroral vortex, although they depend on the distribution of the ionospheric conductivity. However, for the case of the third event at UT Fig. The difference in the ionospheric equivalent current of the EEAS from the omega bands and from case to case may be attributed to a large temporal variation in the surge structure, in contrast to the stable, well-defined wavy structure of omega bands; such a large temporal variation in the surge yields a variation in spatial distribution of the FACs and non-uniform ionospheric conductivity, which causes a deformation of the ionospheric current pattern.

Since the EEASs coincided quite precisely with Pi 2 pulsations near the magnetic equator and magnetic bay variations associated with the DP-1 current system, it is reasonable to assume that they occurred just after substorm onsets. Nakamura et al. The eastward-propagating aurora and the north-south-aligned aurora eventually develop into a diffuse and pulsating aurora after the expansion.

They suggested that these discrete auroral structures in the bulge should be attributed to the plasma streamlines in the magnetosphere. The auroral streamers in the poleward region and the N-S auroras shown in the present study Fig. Global auroral imager data from satellites have provided evidence that the poleward boundary intensifications PBIs and subsequent auroral streamers are followed by equatorward emergence of N-S-aligned auroras, which finally evolve into torches and omega bands Henderson We note that EEASs have not been reported from satellite observations yet, which may be because the spatial and temporal resolutions of the global imagers on board the satellites were too low to detect them.

They found that at least five current vortices, showing alternating senses of rotation i. These characteristics of the current vortices are consistent with those of EEASs.

To our knowledge, this is the only paper that has reported phenomena corresponding to the EEASs. There have been several studies on the azimuthal expansion speed of the initial brightening arcs Shiokawa et al. Ogasawara et al. Thus, it is possible that the eastward expanding auroral surges and eastward expansion fronts of a substorm current wedge are related. The fast propagation of the vortices may be alternatively explained by a surface wave on a plasma boundary in the closed magnetosphere Maltsev and Lyatsky In the frame of this model, the fast motion of auroral structures is a manifestation of the propagation of the magnetospheric wave, rather than the flow of magnetospheric plasma.

Saito showed an example of Ps 6 pulsations considered to coincide with omega bands that started simultaneously with a poleward expansion just after an auroral breakup. They also indicated that the Ps 6 onset coincided with the Pi 2 onset in the midnight sector and showed a linear time lag after the Pi 2 onset in the dawn and dusk sectors. Furthermore, Connors et al. Wild et al. Considering these previous studies and the analogy between the EEASs and omega bands, we speculate that the EEASs may be related to the generation process of the omega bands, and quite a few omega bands observed in the recovery phase may originate in the eastward expansion of the auroral bulge.

The difference in the ionospheric equivalent current between the EEASs and omega bands may be attributed to a large temporal variation in the surge structure, in which the distributions of the FACs and the ionospheric conductivity vary from moment to moment, in contrast to the more stable, well-defined wavy structure of omega bands.

One thing is the air we breathe, our atmosphere. It is really a mixture of several gases, mostly nitrogen and oxygen, with traces of hydrogen, helium and various compounds.

A Field of Earth Another thing we can't see is a magnetic field that surrounds the Earth. If you've ever played with a bar magnet and iron filings you've seen the curved patterns the filings form in the magnetic field. The next picture shows how the magnetic field around the earth's core is like the field of a bar magnet. Charged Particles A third invisible thing in the space around the Earth is a plasma , made of lots of charged particles.

There are always electrons and positive ions in the surrounding magnetic field. Charged particles in a magnetic field move in a special way: they are guided by the field. The particles travel along magnetic field lines as if they were wires, circling around the lines in a long spiral as they go. Charged particles are the "ammunition" of an aurora. Solar Powered Display The short answer to how the aurora happens is that energetic electrically charged particles mostly electrons accelerate along the magnetic field lines into the upper atmosphere, where they collide with gas atoms, causing the atoms to give off light.

But why does that happen? To find the answer, we must look further away, to the Sun. The spectacular, "great" auroras in " What do they look like? The Sun also has an atmosphere and a magnetic field that extend into space. The Sun's atmosphere is made of hydrogen, which is itself made of subatomic particles: protons and electrons.

These particles are constantly boiling off the Sun and streaming outward at very high speeds. Together, the Sun's magnetic field and particles are called the " solar wind.

Squeezing the Earth's magnetic field takes energy, just the way it takes energy to compress a balloon with air in it. The whole process is still not fully understood, but energy from the solar wind is constantly building up in the magnetosphere, and this energy is what powers auroras. The Big Push So we have the Earth's magnetosphere, with the solar wind squeezing the magnetosphere and charged particles everywhere in the field.

Solar particles are always entering the tail of the magnetosphere from the solar wind and moving toward the Sun. Now and then, when conditions are right, the build-up of pressure from the solar wind creates an electric voltage between the magnetotail and the poles, like the voltage between the two terminals of a battery.

It can reach about 10, volts!