вход по аккаунту



код для вставкиСкачать
Magnetospheric Dynamics and the Proton Aurora
E. Donovan, E. Spanswick, J. Liang, J. Grant, B. Jackel, and M. Greffen
Department of Physics and Astronomy, University of Calgary, Calgary, Alberta, Canada
On the nightside, the bright proton aurora forms a several degrees in latitude band
of diffuse aurora near the equatorward boundary of the electron auroral oval. The
precipitation is due to strong pitch angle diffusion that is thought to be the result of
nonadiabatic motion of sub-keV to tens of keV central plasma sheet (CPS) protons
as they traverse the tightly curved magnetic field topology in the vicinity of the
neutral sheet. In this paper, we provide an overview of the relationship between the
spatiotemporal evolution of the proton aurora and magnetospheric dynamics. We
focus on the equatorward boundary of the proton aurora, the latitude of which has
been shown to be strongly correlated with magnetic field line stretching in the inner
CPS, and provide the first-ever identification of the position of the ion isotropy
boundary relative to equatorial magnetospheric spacecraft.
The term aurora refers to the light emitted from the upper
atmosphere resulting from the precipitation of electrons and
protons from outer space. Precipitating electrons collide with
upper atmospheric particles, leaving them in excited states,
relaxation from which produces the photons comprising
the “electron aurora.” Precipitating protons undergo chargeexchange collisions with atmospheric particles, creating a
neutral hydrogen atom in an excited state: subsequent relaxation leads to the photons that comprise the “proton aurora.”
There are fundamental differences between the electron
and proton aurora. Early in the twentieth century, Vegard
[1939] showed conclusively that hydrogen emissions in the
auroral light were Doppler shifted indicating that the proton
aurora was caused by particles precipitating down on the
atmosphere from above. Since there is no such Doppler shift
of emissions in electron aurora (the photons come from
heavy atmospheric particles that have been excited by light
incoming electrons), it was the proton aurora that provided
the first concrete evidence of the precipitation of particles
Auroral Phenomenology and Magnetospheric Processes: Earth and
Other Planets
Geophysical Monograph Series 197
© 2012. American Geophysical Union. All Rights Reserved.
from outer space causing the aurora. As we discuss below,
this proton precipitation results from pitch angle scattering,
rather than field-aligned electric fields, so the proton auroral
evolution and structure is especially valuable for studying the
time-evolving magnetospheric topology and dynamics. The
focus of our chapter is divided between review of some wellknown (though not all published) relationships between proton auroral phenomenology and magnetospheric processes,
and some new results related to mapping the proton auroral
distribution to the magnetotail.
Here we provide only a very cursory introduction related
to how the precipitating protons interact with the atmosphere
to produce the emissions and how proton aurora is observed.
Precipitating protons undergo charge exchange collisions
with atmospheric particles such as N2, O2, and O. Subsequent collisions reionize and reneutralize (taking those terms
from Eather [1967]) the hydrogen atom/proton. The proton
auroral photons are emitted from the (excited) neutral, and
since this ionization/neutralization process can occur many
tens of times for a given incident proton, a single proton can
give rise to many photons with standard H emission wavelengths [see Eather, 1967, and references therein]. While the
proton is in a neutral hydrogen atom, it is not magnetized.
These ionization/neutralization sequences thus lead to the
drifting of the proton away from the field line that it was
incident on (not to be mistaken for convection), leading to
the smearing of fine spatial features in the precipitating
proton distribution as first described by Davidson [1965].
There has been significant work done on how the distribution of incoming protons translates to vertical profiles of
emission in the various resulting hydrogen emissions [see,
e.g., Galand and Chakrabarti, 2006, and references therein]. This, together with spectrographic studies of the proton
aurora, show broadening of the emission lines (blueward
and redward) from which information can be inferred regarding the incident energy of the protons [see, e.g., Galand
and Chakrabarti, 2006, and references therein].
The brightest emissions are the hydrogen Balmer α and β
and Lyman α emissions. Lyman α is the brightest of the three,
but is not useful for observations from the ground because of
atmospheric absorption [see, e.g., Galand et al., 2004]. It has
been observed from space, notably with the Spectroscopic
Imager (SI) on the NASA IMAGE satellite, which provided
the only global images of the proton aurora to date [Mende et
al., 2000, 2001]. For ground-based observations, the Balmer
α is brighter than the Balmer β (herein Hβ), but is relatively
close to other (N2 1P bands) bright electron auroral emissions and so has not been widely targeted for observation.
We refer the reader to excellent reviews that address interactions with the atmosphere and the spectroscopy of proton
auroral emissions [see, e.g., Galand and Chakrabarti, 2006,
and references therein].
From the ground, the proton aurora has been studied with
spectrographs and photometers. Spectrographs can provide
the shape of the Doppler-shifted emissions, which (as stated
above) are useful for inferring information about the characteristic energies of the precipitating protons. Photometers
with narrow band-pass filters are useful for tracking the total
intensity of one emission band. Given the complications of
observing the Balmer α, most photometer observations have
been restricted to Hβ. Most Hβ observations have been
produced by meridian scanning photometers (MSPs). The
fact is that the proton aurora is problematic to observe from
space or the ground. From the ground, spectrographs give
information relevant to the precipitating energies, but are less
capable of tracking the very dim emissions that often characterize the proton aurora. MSPs are highly sensitive, capable of detecting extremely dim emissions, but it is quite
difficult (in many cases impossible) to separate background
from the signal, and they provide no information about
energy (from a single channel). From space, the Lyman α
images obtained by the IMAGE SI provided us with our first
view of the global proton aurora, but there are difficulties
with that data. For example, the image cadence was 2 min
(dictated by the satellite spin period), and the spatial resolution was comparable to the latitudinal width of the typical
proton auroral band (see below). Furthermore, part of the
Doppler-shifted Lyman α band was excluded to mask the
geocoronal Lyman α, so the SI intensities are not always the
total due to proton aurora. In the remainder of this chapter,
we focus on MSP Hβ observations, but we caution the reader
that a truly comprehensive view of proton auroral dynamics
will need better observations that are presently available, and
with our focus, we restrict ourselves to only part of the
overall story.
Our focus here is the relationship between magnetotail
dynamics and topology and the proton aurora. To that end,
this chapter is written as both a review of some well-known
results (not all of which have been published) and as a
vehicle for new material relating to the magnetospheric
(equatorial) counterpart of the equatorward boundary of the
proton aurora. We finish with a discussion suggesting future
interesting work that should be undertaken in order to improve our ability to use the proton aurora to remote sense
magnetotail dynamics.
The keograms in Figure 1 show auroral intensities in the
oxygen “greenline” or 558 nm (top panel) and proton auroral
“Hβ” or 486.1 nm (bottom panel) emissions. The intensities
Figure 1. Merged keograms from the Rankin Inlet (MLAT = 74°)
and Gillam (MLAT = 67°) meridian scanning photometers (MSPs).
Nine hours of (top) 558 nm (oxygen “greenline”) and (bottom)
486 nm (“H-beta” proton aurora) data, with brightness indicated in
Rayleighs. Local magnetic midnight on this meridian is at roughly
06:30 UT.
are displayed “keogram”-style, with color scale indicating
brightness as a function of UT and magnetic latitude
(MLAT), where MLAT is according to Altitude Adjusted
Corrected Geomagnetic Coordinates (AACGM) [Baker and
Wing, 1989]. Each keogram is merged from data from two
MSPs located on the same “Canadian Auroral Network for
the OPEN Program Unified Study (CANOPUS) Churchill
Line” magnetic meridian, one at Rankin Inlet (MLAT = 74°)
and the other Gillam (MLAT = 67°). These provide slightly
overlapping fields of view that together span a region extending from typically poleward of the polar cap boundary
(PCB) to typically equatorward of the equatorward boundary
of the auroral oval. The data shown here is from 9 h on one
night starting roughly 3 h before and extending to roughly
6 h past local magnetic midnight.
We can understand much about the proton aurora and how
it relates to the magnetosphere by considering observations
such as these. The greenline emissions are understood to be
the result of discrete and diffuse electron aurora, as well as
secondary electrons produced by proton precipitation [see,
e.g., Eather, 1967]. The greenline keogram shows a poleward boundary (sharp decrease in brightness with low intensity poleward of high intensity) that is likely at or near the
PCB. The repeated (though not periodic) brightenings at this
boundary are commonly referred to as poleward boundary
intensifications (PBIs) [see, e.g., Lyons et al., 1999]. These
are thought to be causally related to reconnection at the
distant neutral line and often mark the formation of either
north-south structures or east-west auroral arcs that subsequently move equatorward toward the equatorward edge of
the auroral oval. These equatorward moving forms, be they
arcs or north-south structures, show up here as the equatorward propagating “streamers” in the greenline keogram.
Most of the streamers terminate near the equatorward edge
of the auroral oval and are thought to be the ionospheric
signature of earthward motion of either narrow channels
(north-south structures) of fast earthward flow or equatorward propagation of arcs [Zesta et al., 2000, 2002].
In the bottom panel, the proton aurora is the band of
luminosity with poleward and equatorward boundaries indicated by the white diamonds. These boundaries are ±1.4σ
from the latitude of peak luminosity, where σ is the standard
deviation of the latitude brightness profile (this specific
choice of 1.4σ relates to the “optical b2i” discussed below).
The same diamonds are overplotted on the greenline panel.
The keograms in Figure 1 reflect commonly (though not
ubiquitously) observed features of the proton aurora and its
relationship to the electron aurora. The most notable of
these are the following: (1) Even the “bright” proton aurora
are dim in an absolute sense, with intensities shown here
being typical (structured aurora of 1 kR brightness are just
visible to the unaided human eye, and 486 nm aurora in
excess of 200 R are exceedingly rare); (2) The proton aurora
is significantly less structured than the electron aurora (some
of this lack of structure is attributable to the cumulative
effects of the multiple charge-exchange collisions as discussed by Davidson [1965]); (3) The band of significant
proton auroral luminosity is usually near the equatorward
border of the electron auroral oval [see Zou et al., in press]
for a more thorough discussion of this relationship);
(4) Streamers emerging from PBIs terminate either within
or poleward of the band of bright proton aurora [see Kauristie et al., 2003].
The population of precipitating protons is readily identifiable in data from particle detectors on low-altitude satellites
transiting the auroral oval. Figure 2 includes two panels (top)
showing typical FAST ESA ion data from an evening sector
cut through the auroral oval. Although the ESA instrument
cannot differentiate protons from other (e.g., He+, etc.) positive ions, it is usually safe to presume that most of the ions
are protons (see below). The second panel shows the integrated (up to 30 keV) ion energy flux in the downgoing loss
cone as a function of pitch angle and magnetic latitude. The
sharp high-latitude flux change (green moderate fluxes to
blue low fluxes) is the ionospheric projection of the highlatitude boundary of the ion plasma sheet (equatorward of
but close to the PCB). The region of significant fluxes (indicated by green and yellow shading) is the ionospheric projection of the ion plasma sheet. The empty upgoing loss cone
is centered on 180°. The downgoing loss cone spans an
identical pitch angle range, but is centered on 0°. There is a
boundary between an empty (equatorward of the boundary)
and a full (poleward) downgoing loss cone that in this instance is at roughly 68° magnetic latitude. Poleward of this
“isotropy boundary” (IB), the energy flux is isotropic except
for the empty upgoing loss cone, so that fluxes inside and
outside of the downgoing loss cone are the same. The empty
downgoing loss cone equatorward of the IB indicates stably
bounce-trapped ions, indicating conservation of the first adiabatic invariant (µ) on time scales longer than a bounce
period. Poleward of the IB, the downgoing loss cone is
“full,” indicating stochastic process or processes that lead to
violation of µ and that are responsible for strong pitch angle
diffusion (meaning the loss come is filled on time scales less
than a characteristic bounce period).
The top panel in Figure 2 shows differential ion energy
flux as a function of energy and magnetic latitude for the
same pass. One can see in this plot that the characteristic
energy of the plasma sheet ions (mostly protons) generally
increases with decreasing latitudes until the IB, inside of
which the characteristic energy decreases sharply. As with
many, or even most, such meridional passes, the trends are
Figure 2. FAST ESA ion data from an evening sector pass through
the auroral zone. (top) Integrated ion energy flux as a function of
pitch angle and geomagnetic latitude. (second) Differential ion
energy flux in the downgoing loss cone. The energy flux perpendicular to the magnetic field has been subtracted to remove the
nearly isotropic signature of penetrating radiation in the outer radiation belt. The gray shading in the 55°–64° region masks an artifact
of a slight anisotropy in the radiation belt signature. (third) Sixtyfive times the downward integrated ion ESA energy flux, providing
a rough proxy for H-beta proton auroral brightness (see text);
(bottom) Differential energy flux (for the 8.6 keV ion ESA channel)
in the direction parallel to (in red) and perpendicular to the magnetic
field. The vertical dashed lines are ionospheric projections of
boundaries between magnetospheric regions, with the regions being
(1) equatorward of the inner edge of the ion plasma sheet; (2) stably
bounce trapped plasma sheet ions; (3) strong pitch angle diffusion
such that fluxes outside the upgoing loss cone are isotropic; and
(4) poleward of the ion plasma sheet. The proton aurora is the
ionospheric footprint of region 3, and the bright proton auroral band
sits near the equatorward edge of that region.
somewhat complicated beyond these statements. Here for
example, we note the emergence of a lower-energy population near the inner edge of the plasma sheet that we believe is
of plasmaspheric origin. The proton aurora is the result of the
protons in the loss cone between the IB and the poleward
boundary of the ion plasma sheet.
In the bottom panel of Figure 2, we show the meridian
profiles of differential energy fluxes from one FAST electrostatic analyzer ion energy channel (8.6 keV). The magnetic
latitude dependence of the former shows qualitatively how
different magnetotail regions impress themselves on the proton precipitation. The region poleward of the ion plasma sheet
shows up as a region of fluxes orders of magnitude lower than
those in the plasma sheet. The differential fluxes of ions in the
channel shown (8.6 keV) perpendicular to and parallel to
(meaning downward in the northern hemisphere) the magnetic
field are essentially identical from the poleward boundary
right to the IB. Equatorward of the IB, there is a sharp drop
off of downward fluxes, consistent with the IB demarking the
earthward limit of whatever process or processes fill the loss
cone. The region (indicated by “2” on the figure), where the
perpendicular fluxes are significantly larger than the poleward fluxes, corresponds to stable bounce-trapped protons.
In this case, the ion plasma sheet extends significantly earthward of the IB. The relatively large and isotropic flux values
inside of the plasma sheet (region “1” in the figure) indicate
the effects of penetrating radiation belt particles on the detector (not plasma sheet fluxes). The boundary between “1”
and “2” is either on or poleward of the ionospheric footprint
of the inner edge of the ion plasma sheet. We have shown the
differential energy fluxes from one energy channel. Examination of fluxes at other energies shows essentially identical
latitudinal dependence as do the 8.6 keV fluxes (the IB is
slightly latitude dependent, as we discuss below). Further,
the vast majority of nightside FAST passes through the
auroral zone show similar morphology as this specific case.
Figure 3a is a graph of the downward integrated energy
flux multiplied by 65 for the same FAST pass. As stated
above, this is a rough proxy for the H-beta proton auroral
brightness along the satellite track. As expected, the nonzero
proton aurora along the satellite track occurs between the IB
and the poleward boundary. The values are largest immediately poleward of the IB: this relatively narrow peak corresponds to the band of the “bright” proton aurora in Figure 1
(note this overflight is not the same time/location). The
downward fluxes indicate there would be nonzero proton
aurora all the way to the poleward boundary, but those values
are at the lower limit of what can be detected.
Low-altitude satellite passes through the proton auroral
region have demonstrated that the particles are in the downgoing loss cone as a consequence of strong pitch angle
diffusion, meaning the differential energy flux is uniform
across the downgoing loss cone [Sergeev and Tsyganenko,
1982; Sergeev et al., 1983; Donovan et al., 2003a]. Since the
diffusion is due to scattering, the proton aurora is properly
Figure 3. Keograms of 486 nm (H-beta) intensity from Gillam, Canada. Each plot spans 140 min, with (different) UT start
times as indicated in the lower left corner of each. Note the different color-level scales, as well as the overplot of magnetic
inclination as observed at GOES East (located ~2 h magnetic local time (MLT) east of the meridian scanning photometer).
The inclination is shown in red (GOES data not available for event shown in Figure 3d) with the values indicated on the
right-hand axis corresponding to degrees above the SM equatorial plane (plot range 20° to 90°). We also show three MLT
values separated by 30 min (the station is rotating under the fixed MLT coordinate system). We show these four cases to
illustrate four different manifestations of magnetospheric dynamics (or lack thereof ): (a) quiet time; (b) small substorm
(possibly pseudobreakup) with almost no evidence of dipolarization; (c) large substorm with major reconfiguration
(dipolarization) of the inner magnetospheric magnetic field topology; and (d) solar wind pressure increase with northward
interplanetary magnetic field (IMF) and subsequent (20 min later) southward turning of the IMF.
“diffuse”. Such topside in situ observations also show that
the bulk of the protons that cause the proton aurora have
energies in the keV to several tens of keV range.
Based on such in situ and auroral observations, many
authors have concluded that the precipitating protons are on
closed field lines that thread the central plasma sheet (CPS)
[e.g., Eather, 1967]. These protons are understood to enter
the CPS at large distances downtail with sub-keV energies
originating either from the solar wind or ionospheric outflow.
They move with quasi-static or bursty “EXB” convection,
gaining energy through Fermi/Betatron acceleration as they
move into ever-increasing magnitude magnetic field. The
protons are scattered into the loss cone by a process or
processes that violate conservation of µ. It is widely held
that the dominant mechanism responsible for the pitch angle
scattering are nonadiabatic effects as the protons traverse the
tightly curved magnetic field lines near the vicinity of the
neutral sheet. Test particle simulations have shown that this
process is effective enough to fill the loss cone every bounce
period for protons whose local gyroradius is larger than
approximately one ninth of the local radius of curvature of
the field line [Tsyganenko, 1982; Sergeev et al., 1983]. This
is often stated as the scattering is sufficient for strong pitch
angle diffusion if the so-called “kappa” parameter introduced
by Büchner and Zelenyi [1987] satisfies κ < 3, where κ2 is the
ratio of the radius of curvature to the gyroradius (and is hence
energy dependent). The meridional profile of proton auroral
brightness and the energy of the precipitating protons (from
topside in situ observations such as those provided by FAST
ESA) reflect the increasing proton energies with decreasing
downtail distance. The proton aurora is dim on field lines
mapping to the outer CPS due to the low energies and
densities of the protons in the outer CPS. The proton aurora
becomes brighter on field lines mapping to regions with
larger magnetic field strength.
In a study being prepared for publication, E. Spanswick and
colleagues (private communication, 2012) compare the fluxes
of ions in the loss cone as observed by FAST ESA (see e.g.,
Figure 2) with the fluxes of ions in the loss cone observed
near the magnetospheric equator by identical ESA instruments on the Time History of Events and Macroscale Interactions during Substorms (THEMIS) probes [Angelopoulos,
2008]. The total integrated energy flux in the downgoing loss
cone is directly observed by FAST ESA whose entrance
aperture angular width is smaller than the loss cone width.
The entrance aperture of THEMIS ESA is much larger than
the loss cone near the equator, but the downgoing flux can be
inferred presuming that the fluxes are relatively constant
across the loss cone, and with the knowledge that the loss
cone is a small fraction of the solid angle subtended by the
entrance aperture. The rationale behind their study is that
FAST and THEMIS are carrying ESA instruments of the same
design, and the total energy flux in the downgoing loss cone is
the same just off the equator as it would be at the ionospheric
end of that flux tube. They have in addition developed an
entirely empirical relationship between integrated ESA downgoing ion energy flux and Hβ from numerous FAST overflights of the Gillam MSP. The relationship they established is
between total energy flux (J) and 486 nm proton auroral
intensity (I), and is I = 65J, where I and J are in Rayleighs
and ergs cm 2 s 1, respectively. Given this, one can infer the
proton auroral brightness at the ionospheric end of a field line
from the THEMIS ESA observation near the equator. By
looking at large numbers of THEMIS ESA observations, and
comparing those to large numbers of MSP observations of the
proton aurora, Spanswick and colleagues are concluding that
the bright band of proton aurora corresponds to CPS protons
that are typically 7–12 Earth radii (RE) in radial distance on
the nightside.
As discussed above, the equatorward boundary of the
proton aurora corresponds to the ion IB and also the b2i
boundary in DMSP ion data [Newell et al., 1998]. The latter
is the equatorward edge of significant downward ion fluxes
observed by the DMSP ion detector [Newell et al., 1996].
DMSP cannot infer the IB location as it does not observe
transverse fluxes, but in general, this boundary is close to if
not at the IB [see, e.g., Newell et al., 1998]. Donovan et al.
[2003a] chose the name “optical b2i” for the equatorward
boundary of the proton aurora as it is identified in the same
way, namely, a termination of downward fluxes rather than a
transition in the ratio between downward and transverse
The fact that the scattering underlying the precipitation is
thought to be caused by nonadiabatic motions in the tightly
curved field lines around the neutral sheet means that the IB
should correspond to the earthward limit of that scattering
condition. If that limit moves earthward, then one expects the
IB to move equatorward. Sergeev and Gvozdevsky [1995]
showed this is indeed the case by demonstrating that the IB
geomagnetic latitude (adjusted for the magnetic local time
(MLT) dependence) is strongly correlated with the inclination of the magnetic field lines at geosynchronous orbit near
midnight. Presuming that, in general, the proton aurora maps
to beyond geosynchronous distance, then this increased
stretching at geosynchronous distance with decreasing magnetic latitude of the IB is consistent with the idea that the
precipitation is caused by pitch angle scattering related to
field line curvature. It is important to note, however, that this
is just a consistency check and does not prove that the
scattering is related to field line curvature.
Since a more equatorward IB means more stretched field
lines in the inner magnetosphere, Sergeev and Gvozdevsky
[1995] argued that this is an ionospheric measurement that
could be the basis of a proxy for the state of the inner CPS. In
that paper, they presented the “MT index,” which was inferred from topside ionospheric in situ observations of the
ion IB. The MT index was the best estimate, based on an IB
latitude made at some local time, of the IB latitude at magnetic midnight. They further demonstrated that the MT index
is strongly correlated with the magnetic field inclination at
geosynchronous orbit near midnight.
Referring to Figure 2, the ion IB will, in general, correspond to the equatorward boundary of the proton aurora.
Motivated by this, and the MT index, Donovan et al.
[2003a] used simultaneous observations of the proton aurora
from the Gillam MSP and the magnetic field inclination at
GOES 8 (then GOES “East”), to show that the equatorward
boundary of the proton aurora is an excellent indicator of
how stretched the magnetic field is at geosynchronous distance. They developed an empirical relationship between the
latitude of the equatorward boundary of the proton aurora,
which they referred to as the “optical b2i,” and the magnetic
inclination at GOES 8, which was located roughly 2 h MLT
east of Gillam and on average ~0.8 RE above the magnetic
equator. Even with the 2-h MLT separation, they found that
their relationship could “predict” the inclination at GOES
8 within 10° almost always. Significant differences between
the inferred (from the proton aurora) and actual inclinations
were often due to azimuthally propagating disturbances such
as dipolarizations.
In developing the optical b2i, Donovan et al. [2003a]
experimented with a number of possible boundaries that
could be inferred from the proton aurora. Examples included
the latitudes of the peak in brightness and variously defined
equatorward and poleward boundaries. They carried out
regression analysis, which allowed quantitative assessment
of how well each boundary could predict the inclination at
GOES 8. They concluded that of a large family of simple
boundaries, the best predictor was obtained by a specific
equatorward boundary defined to be Λb2i = Λpeak
where Λpeak is the magnetic latitude of the peak in brightness, and σ is the standard deviation of the best fit Gaussian
to the latitude profile of the proton aurora. By comparing ion
data from DMSP overflights of the Gillam MSP and the
proton auroral data, they showed that Λb2i is typically within
0.5° of the b2i, which is the DMSP boundary that most
closely corresponds to the ion IB [see, e.g., Newell et al.,
1996, 1998]. It was this correspondence between the DMSP
b2i and Λb2i that led Donovan et al. [2003a] to use the term
optical b2i for their parameter. It is interesting to speculate on
what it might mean that this particular boundary is the best
indicator of the inclination at geosynchronous of any simple
boundary that can be inferred from the proton aurora. One
possibility is that on average the equatorward boundary of
the proton aurora maps to near geosynchronous distance.
This warrants further investigation.
Figure 3 is a stack plot of four keograms of proton auroral
(486 nm) intensities from the CANOPUS (now NORSTAR)
MSP at Gillam Manitoba. The time duration for each panel is
140 min, with different start times for each. The magnetic
latitude range for each is the same (63°–70° AACGM). Note
that magnetic midnight for Gillam is at 06:30 UT. We have
chosen these four time periods to illustrate how four different
magnetospheric activity scenarios are typically manifested in
the proton auroral evolution. Further, in three of the four
cases, we have also shown the magnetic inclination at GOES
East for the corresponding time period (in red). There is no
GOES East data for the time corresponding to Figure 3d.
Figure 3a (27 December 1992 04:40–07:00 UT) corresponds to a very quiet time, quiescent auroral oval (both
proton and electron, the only activity evident are some weak
poleward boundary intensifications in the electron aurora
well poleward of the proton aurora). This keogram shows
the proton aurora during a period during which the magnetosphere was remarkably quiet based on ground magnetometer as well as GOES East and West magnetic field data. The
GOES East magnetic inclination shows only diurnal variation, with values that correspond very closely to what would
be expected for a dipole field (an Earth-centered dipole field
would have an inclination of ~70° above the magnetic equator at the location of GOES East). The “bright” proton
auroral band is, in this case, dim and shows only the diurnal
variation expected from statistical studies [see, e.g., Creutzberg et al., 1988]. This is the typical proton auroral behavior
during times of little, if any, magnetic activity.
Figures 3b and 3c show two substorm events. Figure 3b
shows the proton aurora during a relatively small substorm.
As it always does during expansion phase, the proton aurora
brightens and expands in latitude; however, the equatorward
boundary of the proton aurora does not change in the initial
20 min following expansion phase onset. As well, the inclination of the magnetic field at GOES East undergoes only a
minor change. Figure 3c is another substorm, although significantly larger. In this case, the proton aurora brightens and
expands, and the equatorward boundary of the proton aurora
moves significantly poleward following expansion phase
onset (in that meridian). As well, the GOES East magnetic
field inclination shows a significant dipolarization some
30 min before the onset in the MSP meridian.
Figure 3d (18 February 1999 01:40–04:00 UT) shows the
proton aurora during a period of different magnetospheric
activity. In this case, the changes in the proton aurora are due
to changes in the solar wind driver, with a northward interplanetary magnetic field (IMF) pressure pulse hitting the magnetosphere at roughly 02:50 UT. The proton aurora is very
dim prior to the effects of the pressure pulse. At ~02:52 UT,
the proton aurora becomes significantly brighter (at the
same time there is a significant increase in the electron
aurora, but that response is more dynamic than is the response in the proton aurora), but the location of the equatorward boundary does not change at that time. This is
confirmed by FAST ESA ion data evening sector passes that
are just prior to and just after the pressure pulse arrival (they
show the ion IB does not change as a consequence of the
pressure pulse). Roughly 20 min later, the proton aurora
begins to move rapidly equatorward. This southward motion
corresponds to the arrival of a rotation from northward to
southward IMF.
These four keograms demonstrate a broad range of geomagnetic activity as impressed on the proton aurora. To
begin with, the proton aurora is an ever-present feature of
the terrestrial environment, even during the quietest geomagnetic periods. The data shown here is consistent with
the fact that the equatorward boundary of the proton aurora
is an excellent proxy for the magnetic inclination in the
inner CPS (see discussion around the MT index and optical
b2i in the previous section). The very quiet proton aurora
corresponds to an inner CPS with little cross-tail current
and, hence, an essentially dipolar field at geosynchronous
orbit. There must be stretching of the magnetic field leading
up to the small substorm shown in Figure 3b, but the
magnetotail dynamics apparently happen well beyond geosynchronous orbit and translate to no dipolarization at geosynchronous and no poleward motion of the equatorward
boundary of the proton aurora. The sudden brightening that
comes with the expansion phase is due to energization of
CPS protons by induced electric fields as the magnetic field
increases in the dipolarization.
The stretching at geosynchronous is more pronounced for
the larger substorm in Figure 3c, and there is a dipolarization
at geosynchronous as well as a pronounced poleward motion
of the equatorward boundary of the proton aurora. The time
lag between the dipolarization (occurring at a time of roughly
030 on the plot) and poleward motion (occurring at 060 on
the plot) is a consequence of the time it takes the dipolarization to evolve across the 2 h of MLT separating GOES East
from the MSP at Gillam.
The brightening in Figure 3d is the proton auroral signature of a sudden increase in solar wind dynamic pressure.
This is a consequence of energization of the CPS proton
auroral population during the compression of the entire magnetosphere. Since the loss cone in the proton auroral band is
full before and after the arrival of the pressure pulse, this
brightening is fundamentally different than what has been
proposed to happen to the diffuse electron aurora during such
events [see, e.g., Zhou et al., 2003]. That is to say, the
brighter proton aurora is not the result of an increase in the
effectiveness of the scattering mechanism, since the CPS
protons in the proton auroral region were already under
strong pitch angle diffusion. Furthermore, the sudden equatorward motion of the proton aurora that begins roughly
20 min after the brightening indicates a sudden and dramatic
increase in stretching of the magnetic field in the inner CPS
(again, refer back to the MT index and optical b2i discussion
of the previous section). As stated above, this equatorward
motion starts roughly with the arrival of a southward turning
of the IMF and, hence, a rapid increase of energy input into
the magnetosphere, which, in general, should lead to increased stretching of the magnetic field in the nightside
magnetosphere [see, e.g., Daglis et al., 2003].
Our primary intention in this section is to convey the value
of the proton aurora as an indicator of dynamics in the inner
CPS. Equatorward (poleward) motion of the proton auroral
equatorward boundary over and above diurnal variation has a
one-one correspondence with stretching (relaxation) of the
magnetic field topology. Brightening indicates energization
of the CPS proton population due to increasing magnetic
field strength in the inner CPS. While we have not shown
examples, sudden decreases in the solar wind dynamic pressure lead to sudden decreases in the proton auroral brightness. This affords us a powerful tool for studying magnetotail
dynamics, if we have proton auroral observations at a number of MLTs simultaneously: we can use the proton aurora to
study the 2-D (meaning “XY ”) spatio-temporal evolution of
the inner CPS magnetic field topology. Nicholson et al.
[2003] presented several event studies wherein they tracked
the azimuthal evolution of the dipolarization with simultaneous proton auroral observations from MSPs in Canada and
Alaska. More recently, Gilson et al. [2011] used IMAGE
SI12 global proton auroral image sequences to explore the
azimuthal (longitudinal) “splitting” of the proton aurora and
its relationship to the spatiotemporal evolution of the inner
CPS magnetic field topology during growth phase. The significant limitation here is the lack of multiple meridian observations of the proton aurora with the time and space
resolution to adequately capture the effects of the evolution
of the magnetic field topology on those meridians. The
IMAGE SI12 data was obtained at 2-min cadence, with
several hundred-kilometer resolution (inadequate for tracking small changes in latitude), while there are only a few
MSPs providing proton auroral observations, and those are
located at widely separated locations. Denser deployments of
MSPs and/or meridian imaging spectrographs will hopefully
come in time and, with them, the opportunity to study the
inner CPS dynamics in greater detail.
As discussed above, the proton auroral precipitation is the
result of strong pitch angle diffusion, which is generally held
to be a consequence of violation of µ due to tight field line
curvature in the vicinity of the neutral sheet. Also as stated
above, this is based on test particle simulations that have
shown the pitch angle scattering to be sufficient to fill the loss
cone on time scales comparable to typical proton bounce
periods if the “kappa” parameter satisfies the relationship
κ < 3. Although there is the possibility of local minima in
magnetic field strength in the neutral sheet, the magnetic field
strength, in general, increases with decreasing distance from
the Earth. Along with that, the local radius of curvature in the
neutral sheet increases with decreasing distance. Thus, for a
specific energy, the kappa parameter increases with decreasing distance. At any one location, the kappa parameter is
smaller for larger energies. Thus, presuming the pitch angle
scattering that fills the loss cone is due to field line curvature,
one expects the IB to be closer to the Earth for larger
energies. This is illustrated conceptually in Figure 4, using
an empirical magnetic field model to infer equatorial radii of
curvature and gyroradii.
Models such as that used to produce Figure 4 can give a
heuristic picture of how gyroradius, radius of curvature, and
hence kappa depend on the distance from the Earth. These
models cannot, however, be expected to reproduce the
Figure 4. (bottom) Magnetic field lines traced with the T89 [Tsyganenko, 1989] Kp = 3 magnetic field model (using a
dipole for the Earth’s internal field). The blue and red shaded regions are meant to illustrate the ion CPS and thin current
sheet regions, with the latter corresponding to curvature sufficient to violate the conservation of µ. (middle) Three panels
showing field line traces in the vicinity of the neutral sheet (in red) with an inscribed circle (in gray) indicating the radius of
curvature of the field line at the neutral sheet. Also shown are the gyroradius and kappa values for 10 keV protons. (top)
The kappa parameter (square root of the ratio of the radius of curvature to the gyroradius) for 5, 10, and 30 keV protons.
magnetic field strength and topology around the (thin current) neutral sheet region; hence, the locations of the IB
inferred from such an exercise should not be trusted. The
loss cone in the vicinity of the neutral sheet is too small to
allow for the determination of whether or not the loss cone
is full, so that a direct identification of passage across the
IB using particle observations such as those provided by
THEMIS ESA is not possible. In this section, we present
what we believe is the first-ever observation from which the
passage of the proton κ = 3 boundary for energies in the
range that cause the proton aurora across a pair of closely
spaced satellites. Provided that field line curvature is the
primary source of scattering that fills the loss cone, then this
is the passage of the proton IB across those satellites. This
observation corroborates, at least in this case, the Spanswick
and colleagues statistical picture of where the proton aurora
maps to in the magnetotail.
There was a small substorm with onset at roughly 02:20 UT
on 28 February 2009. The expansion phase onset occurred
when THEMIS A and E were in the late evening sector CPS,
near the neutral sheet, ~8.1 RE from the Earth. Lui [2011] has
presented the THEMIS in situ and contemporaneous (electron) auroral observations in a study of the onset mechanism.
Here we use observations from THEMIS A and E to explore
the proton kappa parameter in the vicinity of the probes in
order to infer the location of the probes relative to the ion IB.
THEMIS A and E were located at roughly the same XY
location in GSM coordinates, but were separated by ~0.8 RE
in Z during the minutes around onset. Further, from the
magnetic field measurements, it is clear that THEMIS E was
near the neutral sheet. This situation affords us a remarkable
opportunity to investigate the local curvature of the magnetic
field lines while at the same time having knowledge of the
magnetic field strength in the vicinity of the neutral sheet. To
do so, we assume that the field line curvature can be represented by a circle. In support of this step, the magnetic field
at the two satellites is close to coplanar. Further, if there is a
region between the two satellites of even tighter curvature
than we infer, then it would lead us to overestimating the
radius of curvature. While this might be true in the late
growth phase, it is almost certainly not so after the dipolarization, so our conclusions (which relate to kappa before and
after the dipolarization) would be strengthened rather than
In Figure 5, we summarize our results. The top panels
show the magnetic field at the two satellites at a time before
and after the dipolarization. In those panels, we show with a
gray thick curve the circle that has the same radius of curvature that we inferred from the magnetic field direction at the
two satellites. We infer this radius of curvature from the
THEMIS FGM data throughout the late growth and expan-
sion phases. The radius of curvature is shown in the top time
series panel. The two vertical red lines indicate the times
corresponding to the two panels above. The sudden increase
in radius of curvature around 02:23 UT is the beginning of
dipolarization that marks the onset at the satellites (note we
would not want to make a connection between this event and
any particular onset signature in the ionosphere). At any
given time, we have the radius of curvature and magnetic
field strength in the neutral sheet (from THEMIS E). With
these two numbers, we can determine the energy of proton
for which the kappa parameter is exactly equal to 3. We call
this the “critical energy,” and provided that field line curvature is the cause of loss cone filling, we can infer that protons
with energies larger than the critical energy are in strong
diffusion, while those with lower energy are stably bounce
trapped. We show this critical energy with the middle graph.
On the bottom graph, we overplot this critical energy on the
combined THEMIS E ion ESA/SST spectrogram.
From the results shown on the bottom graph of Figure 5,
we can infer that in the late growth phase, the critical energy
is below the energies of the bulk of the ion distribution
observed by ESA/SST. Although the energies of the ions
increase along with the dipolarization, we can see that within
2 min of the beginning of the dipolarization at this location,
the critical energy has climbed above the bulk of the ion
population at that location. From this, we can conclude that
assuming that field line curvature is responsible for the loss
cone filling, then throughout the late growth phase, the IB is
earthward of the THEMIS satellite pair and that by 2 min
after the start of dipolarization, the IB is tailward of the two
satellites. Based on this simple observation, we can infer that
in the late growth phase in this event, the equatorward
boundary of the proton aurora maps to inside of 8 RE radial
distance in the late evening sector and that it has moved out
beyond that distance after the onset. This mapping is consistent with the results of Spanswick and colleagues, who are
finding that on average the bright proton aurora maps to
outside of 7 but inside of 10 RE under most conditions. It is
unfortunate that we cannot gauge how far the IB is inside (or
outside) of the satellites nor how radially thick the IB is as it
crosses the satellites. Such work will need to wait for a larger
constellation of satellites, but these results nevertheless offer
us hope that we are starting to build up a reasonable picture
of this mapping.
The proton aurora arises as a consequence of convection,
which brings protons earthward in the nightside magnetosphere, gradient, and curvature drifts, which moves the protons
in the cross-tail electric field, thereby increasing their energy,
Figure 5. On 28 February 2009, THEMIS A and E were located very close in XY, and separated by ~0.8 RE in Z. As well,
THEMIS E was very near the neutral sheet. Simultaneous observations of the magnetic field at the two satellites give us the
opportunity to estimate the local radius of curvature. The two panels on top illustrate this for two separate times bracketing
a small substorm onset that occurred around 02:23 UT (as seen at the satellites). The three time series plots show, in order
(top) the inferred radius of curvature at the satellites, (middle) the “critical energy” (see text) or energy of a proton with κ =
3 based on the local radius of curvature and magnetic field strength at THEMIS E, and (bottom) the critical energy plotted
over a merged THEMIS ESA/SST spectrogram.
and pitch angle scattering that fills the loss cone on a time
scale shorter than the bounce period. One can also view this
process as energization as a consequence of conservation of µ
as the particles move into increasing magnetic field strength
in the inner magnetosphere. The larger magnetic field in the
inner magnetosphere accounts for the energization and also
slows the convection leading to increased proton densities.
The region of strong pitch angle scattering extends from near
the open-closed field line boundary inward to the energy-
dependent IB. This boundary is usually tailward of the inner
edge of the ion plasma sheet, but may at times be collocated
with it (this separation is spanned by the region labeled “2” in
Figure 2).
The loss cone is small enough in the plasma sheet that it
cannot be resolved by existing in situ plasma instruments.
Hence, we cannot directly observe the IB in the magnetotail.
However, the inner edge of the plasma sheet can be readily
observed in the magnetotail [see, Runov et al., 2008]. On
the other hand, the equatorward boundary of the proton
aurora (the “optical b2i” as discussed above), which is
readily observed in optical observations [Donovan et al.,
2003a] corresponds to the ionospheric projection of the IB,
whereas the ionospheric projection of the plasma sheet inner
edge is not discernible in the optical observations. Thus, the
readily observable boundary in the ionosphere is not the
same as the readily observable boundary in the magnetotail.
More research needs to be carried out to explore the spatial
relationship between the inner edge of ion plasma sheet and
the ion IB and how that relationship depends on magnetotail
The proton aurora provides us with an excellent tool for
remote sensing certain aspects of magnetotail dynamics. The
bright proton aurora paints the ionospheric footprint of the
transition between the highly stretched tail and the lessstretched inner magnetosphere. We can expect that many fast
CPS flows terminate in this “nightside transition region”
(NTR), as evidenced by streamers that propagate into and
terminate within the proton aurora (see Figure 1). The latitude of the equatorward boundary (the MT index or optical
b2i) of the proton aurora is an excellent proxy for the magnetic topology in the inner CPS. Equatorward and poleward
motion of the equatorward boundary correspond one-one
with stretching and dipolarization in that MLT sector. Sudden
brightening of the proton aurora corresponds to energization
of the CPS proton population, often through conservation of
µ during rapid increases in magnetic field strength such as
during dipolarization and sudden compressions. These observations offer the possibility of tracking the 2-D (meaning
“XY ”) inner CPS dynamics as they pertain to magnetic field
topology, something that is not otherwise possible. That
being said, the observational facilities that would enable such
monitoring of the CPS dynamics are not available at present.
Given the value such observations would have for understanding the relationship between convection, magnetic field
topology, and plasma sheet particle populations, more
ground-based instruments capable of observing the proton
aurora would be a significant step forward for our field.
In the vast majority of situations where knowledge of the
proton aurora is used to infer quantitative information about
the magnetosphere, the presumption is made that the pitch
angle scattering is due to the field line curvature criterion
(κ < 3) being satisfied. There are other mechanisms that
might also cause the scattering [see, e.g., Gvozdevsky et al.,
1997]. As discussed above, it is true that the inclination of
the magnetic field at geostationary orbit is very well correlated with the ionospheric latitude of the ion IB (as inferred
from the equatorward boundary of the proton aurora). The
fact is, however, that this is to be expected if there is a strong
correlation between the distance between the ion plasma
sheet and the Earth and the latitude of the ion IB, regardless
of what causes the pitch angle scattering. The ion IB latitude
is observed to depend on energy. In a few cases, that latitudinal dependence has been shown to be consistent with what
is expected if the dominant source of pitch angle scattering is
the field line curvature. However, this has been done with
ions of higher energy than those typically responsible for the
proton aurora or very high-energy electrons [see, e.g., Imhof
et al., 1977]. Donovan et al. [2003b] explored this dependence of ion isotropy boundary latitude on energy using
FAST ESA ion data, whose energy range is, in general,
appropriate for exploring the proton aurora. They found that
the vast majority of cases (not all) in the evening sector
showed the dependence expected if field line curvature is
causing the pitch angle scattering (namely, the IB is at lower
latitude for increasing energy). Interestingly, they found that
a significant fraction of the morning sector IBs showed the
opposite. They concluded that it is often the case that at the
equatorward boundary of the proton aurora in the morning
sector, the proton precipitation is caused by a mechanism
other than pitch angle scattering due to field line curvature
(note that the converse is not necessarily true in the evening
sector). As observations of the electron and proton aurora are
seeing increasing use in remote sensing the magnetosphere,
this needs to be explored more. Certainly, we have to exercise caution in interpreting the proton aurora while assuming
that the pitch angle scattering is always caused by field line
When considering the “kappa” condition for scattering, it
is usually assumed that this is a sharp boundary. Such interpretation is all we can do at present, but clearly we must
again use caution. As pointed out by Gilson et al. [2011], it is
not likely the case that the pitch angle diffusion is strong if
kappa is slightly less than 3 and exactly zero if kappa is
slightly greater than 3. The test particle simulations upon
which the κ = 3 interpretation is based were carried out in the
late 1970s and early 1980s and must of necessity have used
relatively few test particles. Since that time, readily available
computer power has increased orders of magnitude, and our
understanding of the relevant magnetic field topology has
increased significantly. This, too, points toward some important future research, namely, a revisiting of the test particle
simulations in order to refine our understanding of, for example, how radially thick the IB ought to be in typical
magnetic field topologies.
More than 30 years ago, Lui and Burrows [1978] used the
intense proton precipitation to identify the ionospheric footprint of the transition between highly stretched and quasidipolar magnetic topologies (they referred to this NTR as the
“nighttime cusp”; see their Figure 5). Since then, this has
become widely accepted as a fact, being used as a cornerstone
in papers where mapping of specific auroral features (e.g., the
“onset arc”) and their magnetospheric counterparts is useful.
These are numerous, but the paper by Samson et al. [1992]
provides an excellent example. In that paper, this mapping of
the proton aurora to the NTR is used as follows. The bright
band of proton aurora marks the ionospheric projection of the
NTR. The onset arc is on the poleward shoulder of or just
poleward of the bright proton aurora. Therefore, the onset arc
maps close to the NTR. This argument is mostly sound and oft
repeated (another example is by Donovan et al. [2008]).
Until recently, we could not say with any real confidence
where the NTR is located and, hence, where the proton
aurora actually maps to. Our results using two THEMIS
satellites allow us to say with some certainty that in at least
that one event, the IB was inside of the satellites. We finish
with a final cautionary note that also points to necessary
future work on the proton aurora. In general, we can study
the current sheet properties tailward of the NTR using satellites such as THEMIS, Geotail, and Cluster. We have excellent observations inside of the NTR from missions such as
GOES and CRRES, and RBSP promises significantly more
observations in that region very soon. What we are lacking is
a multisatellite picture of how the thin current sheet region
blends into the inner magnetosphere. Lacking this detailed
picture of the topology of the NTR means that, for now, we
have important unanswered questions about the physics of
this region, in general, and the proton aurora specifically.
Acknowledgments. Eric Donovan acknowledges the significant
contributions of John Samson of the University of Alberta to the use
of the proton aurora for remote sensing the nightside transition
region (NTR). He is also grateful to William Liu, David Knudsen,
Larry Lyons, and Toshi Nishimura for the many insightful discussions about auroral science and the NTR that have added enormously
to this work. The authors acknowledge the CSA for its long
support of the CANOPUS and now CGSM Meridian Scanning
Photometer program. As well, the CSA has supported this research
through three Space Science Enhancement Program grants to the
University of Calgary. We are grateful to NOAA and NASA for
support of the GOES, THEMIS, and FAST missions. The authors
acknowledge J. McFadden, C. Carlson, D. Larson, and K.-H.
Glassmeier for their work on THEMIS ESA, SST, and FGM, and
FAST ESA, and Vassilis Angelopoulos for his leadership of the
THEMIS mission.
Angelopoulos, V. (2008), The THEMIS mission, Space Sci. Rev.,
141, 5–34, doi:10.1007/s11214-008-9336-1.
Baker, K. B., and S. Wing (1989), A new magnetic coordinate
system for conjugate studies at high latitudes, J. Geophys. Res.,
94(A7), 9139–9143.
Büchner, J., and L. M. Zelenyi (1987), Chaotization of the electron
motion as the cause of an internal magnetotail instability and
substorm onset, J. Geophys. Res., 92(A12), 13,456–13,466.
Creutzberg, F., R. L. Gattinger, F. R. Harris, S. Wozniak, and A. V.
Jones (1988), Auroral studies with a chain of meridian scanning
photometers, 2, Mean distributions of proton and electron aurora
as a function of magnetic activity, J. Geophys. Res., 93(A12),
Daglis, I. A., E. T. Sarris, and G. Kremser (2003), Ionospheric
contribution to the cross-tail current enhancement during the
substorm growth phase, J. Atmos. Sol. Terr. Phys., 53, 1091–
1098, doi:10.1016/0021-9169(91)90057-E.
Davidson, G. T. (1965), Expected spatial distribution of low-energy
protons precipitated in the auroral zones, J. Geophys. Res., 70(5),
Donovan, E. F., B. J. Jackel, I. Voronkov, T. Sotirelis, F. Creutzberg,
and N. A. Nicholson (2003a), Ground-based optical determination of the b2i boundary: A basis for an optical MT-index,
J. Geophys. Res., 108(A3), 1115, doi:10.1029/2001JA009198.
Donovan, E. F., B. Jackel, D. Klumpar, and R. Strangeway (2003b),
Energy dependence of the isotropy boundary latitude, Sodankylä
Geophys. Obs. Publ., 92, 11–14.
Donovan, E. F., et al. (2008), Simultaneous THEMIS in situ and
auroral observations of a small substorm, Geophys. Res. Lett., 35,
L17S18, doi:10.1029/2008GL033794.
Eather, R. H. (1967), Auroral proton precipitation and hydrogen
emissions, Rev. Geophys., 5(3), 207–285.
Galand, M., and S. Chakrabarti (2006), Proton aurora observed
from the ground, J. Atmos. Sol. Terr. Phys., 68, 1488–1501,
Galand, M., J. Baumgardner, D. Pallamraju, S. Chakrabarti, U. P.
Løvhaug, D. Lummerzheim, B. S. Lanchester, and M. H. Rees
(2004), Spectral imaging of proton aurora and twilight at Tromsø,
Norway, J. Geophys. Res., 109, A07305, doi:10.1029/2003JA
Gilson, M. L., J. Raeder, E. Donovan, Y. S. Ge, and S. B. Mende
(2011), Statistics of the longitudinal splitting of proton aurora
during substorms, J. Geophys. Res., 116, A08226, doi:10.1029/
Gvozdevsky, B. B., V. A. Sergeev, and K. Mursula (1997), Long
lasting energetic proton precipitation in the inner magnetosphere
after substorms, J. Geophys. Res., 102(A11), 24,333–24,338.
Imhof, W. L., J. B. Reagan, and E. E. Gaines (1977), Fine-scale
spatial structure in the pitch angle distributions of energetic
particles near the midnight trapping boundary, J. Geophys. Res.,
82(32), 5215–5221.
Kauristie, K., V. A. Sergeev, O. Amm, M. V. Kubyshkina, J.
Jussila, E. Donovan, and K. Liou (2003), Bursty bulk flow
intrusion to the inner plasma sheet as inferred from auroral
observations, J. Geophys. Res., 108(A1), 1040, doi:10.1029/
Lui, A. T. Y. (2011), Reduction of the cross-tail current during nearEarth dipolarization with multisatellite observations, J. Geophys.
Res., 116, A12239, doi:10.1029/2011JA017107.
Lui, A. T. Y., and J. R. Burrows (1978), On the location of auroral
arcs near substorm onsets, J. Geophys. Res., 83(A7), 3342–3348.
Lyons, L. R., T. Nagai, G. T. Blanchard, J. C. Samson, T. Yamamoto, T. Mukai, A. Nishida, and S. Kokubun (1999), Association
between Geotail plasma flows and auroral poleward boundary
intensifications observed by CANOPUS photometers, J. Geophys. Res., 104(A3), 4485–4500, doi:10.1029/1998JA900140.
Mende, S. B., et al. (2000), Far ultraviolet imaging from the IMAGE
spacecraft, 1, System design, Space Sci. Rev., 91, 243–270.
Mende, S. B., H. U. Frey, M. Lampton, J.-C. Gerard, B. Hubert, S.
Fuselier, J. Spann, R. Gladstone, and J. L. Burch (2001), Global
observations of proton and electron auroras in a substorm, Geophys. Res. Lett., 28(6), 1139–1142, doi:10.1029/2000GL012340.
Newell, P. T., Y. I. Feldstein, Y. I. Galperin, and C.-I. Meng (1996),
Morphology of nightside precipitation, J. Geophys. Res., 101
(A5), 10,737–10,748.
Newell, P. T., V. A. Sergeev, G. R. Bikkuzina, and S. Wing (1998),
Characterizing the state of the magnetosphere: Testing the ion
precipitation maxima latitude (b2i) and the ion isotropy boundary, J. Geophys. Res., 103(A3), 4739–4745.
Nicholson, N., E. Donovan, B. Jackel, L. Cogger, and D. Lummerzheim (2003), Multipoint measurements of the ion isotropy
boundary, Sodankylä Geophys. Obs. Publ., 92, 37–40.
Runov, A., V. Angelopoulos, N. Ganushkina, R. Nakamura,
J. McFadden, D. Larson, I. Dandouras, K.-H. Glassmeier, and
C. Carr (2008), Multi-point observations of the inner boundary of
the plasma sheet during geomagnetic disturbances, Geophys.
Res. Lett., 35, L17S23, doi:10.1029/2008GL033982.
Samson, J. C., D. D. Wallis, T. J. Hughes, F. Creutzberg, J. M.
Ruohoniemi, and R. A. Greenwald (1992), Substorm intensifications and field line resonances in the nightside magnetosphere,
J. Geophys. Res., 97(A6), 8495–8518.
Sergeev, V. A., and B. B. Gvozdevsky (1995), Mt-index: A possible
new index to characterize the magnetic configuration of the
magnetotail, Ann. Geophys., 13, 1093–1103.
Sergeev, V. A., and N. A. Tsyganenko (1982), Energetic particle
losses and trapping boundaries as deduced from calculations
with a realistic magnetic field model, Planet. Space Sci., 10,
Sergeev, V. A., E. M. Sazhina, N. A. Tsyganenko, J. A. Lundblad,
and F. Søraas (1983), Pitch-angle scattering of energetic protons
in the magnetotail current sheet as the dominant source of their
isotropic precipitation into the nightside ionosphere, Planet.
Space Sci., 31, 1147–1155.
Tsyganenko, N. A. (1982), Pitch-angle scattering of energetic particles in the current sheet of the magnetospheric tail and stationary distribution functions, Planet. Space Sci., 30, 433–437.
Tsyganenko, N. A. (1989), A magnetospheric magnetic field model
with a warped tail current sheet, Planet. Space Sci., 37, 5–20.
Vegard, L. (1939), Hydrogen lines in the auroral spectrum, Nature,
114, 1089.
Zesta, E., L. R. Lyons, and E. Donovan (2000), The auroral signature of earthward flow bursts observed in the magnetotail, Geophys. Res. Lett., 27(20), 3241–3244, doi:10.1029/2000GL
Zesta, E., E. Donovan, L. Lyons, G. Enno, J. S. Murphree, and L.
Cogger (2002), Two-dimensional structure of auroral poleward
boundary intensifications, J. Geophys. Res., 107(A11), 1350,
Zhou, X.-Y., R. J. Strangeway, P. C. Anderson, D. G. Sibeck, B. T.
Tsurutani, G. Haerendel, H. U. Frey, and J. K. Arballo (2003),
Shock aurora: FAST and DMSP observations, J. Geophys. Res.,
108(A4), 8019, doi:10.1029/2002JA009701.
Zou, Y., Y. Nishimura, L. R. Lyons, and E. F. Donovan (2012), A
statistical study of the relative locations of electron and proton
auroral boundaries inferred from meridian scanning photometer
observations, J. Geophys. Res., 117, A06206, doi:10.1029/
E. Donovan, J. Grant, M. Greffen, B. Jackel, J. Liang, and E.
Spanswick, Department of Physics and Astronomy, University of
Calgary, Calgary, AB T2N 1N4, Canada. (
Без категории
Размер файла
2 295 Кб
Пожаловаться на содержимое документа