Upload
cr-chappell
View
217
Download
3
Embed Size (px)
Citation preview
The adequacy of the ionospheric source in supplyingmagnetospheric plasma
C.R. Chappell a,*, B.L. Gilesb, T.E. Mooreb, D.C. Delcourtc, P.D. Cravend,M.O. Chandlerd
aPhysics Department, Vanderbilt University, 708 Baker Building, Nashville, TN 37203, USAbNASA/Goddard Space Flight Center, Greenbelt, MD, USA
cCentre de Recherches en Physique de l'Environnement/Centre National d'Etudes des Telecommunications, Saint-Maur des Fosses,
FrancedNASA/Marshall Space Flight Center, AL, USA
Abstract
More than 30 years after the prediction of the polar wind out¯ow from the high latitude ionosphere, the exact
magnitude and ultimate fate of the ionospheric plasma supply remains unknown. Estimates made more than adecade ago suggested that the polar ion out¯ow might well be of su�cient strength to populate the di�erent regionsof the Earth's magnetosphere. Direct measurements in the high altitude magnetosphere became possible only with
the launch of the Polar spacecraft. The combination of the Thermal Ion Dynamics Experiment and the PlasmaSource Instrument has revealed the presence of low energy (<10 eV) ions moving through the polar regions andinto the lobes of the magnetotail. These ions would have been invisible to previous un-neutralized satellites becauseof the high positive spacecraft potentials. Through the use of a recently developed single particle trajectory and
energization code, the movement and energy transformation of these measured particles can be estimated. They arefound to move into the plasma sheet region and to be energized to typical plasma sheet energies. The magnitude ofthe ¯ux of the highly variable out-¯owing ions mapped to 1000 km altitude is 1 ÿ 3 � 108 ions/cm2 s in agreement
with the original estimates. Future observations by the TIDE/PSI instruments will be required to determine theextent of the total ionospheric contribution. 7 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction
One of the most fundamental issues remaining in the
understanding of the Earth's magnetosphere is the
source of its plasma. Early space measurements around
the Earth were made with instruments which were sen-
sitive only to high energy particles. When these par-ticles were found, it established a trend of thinkingrelated to the Sun and solar wind as the principal
source. Since the solar wind plasma has energies of theorder of kilo-electron volts and the ¯are particles fromthe Sun exhibit energies of millions of electron volts,
the earthspace plasmas found in this energy regimewere thought of as being particles of solar originwhich had gained access across magnetospheric bound-
aries. The lack of mass resolution capability on theearly instrumentation also made it di�cult if not
Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436
1364-6826/00/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved.
PII: S1364-6826(00 )00021-3
* Corresponding author. Tel.: +1-615-343-6794.
E-mail address: [email protected] (C.R. Chap-
pell).
impossible to di�erentiate the solar from the terrestrial
components.With the development of the next generation of
instrumentation which could distinguish the mass of
the ions, it was found that the terrestrial source was abona ®de contributor. The oxygen ion measurements
of Shelley et al. (1972) established the Earth's iono-sphere as a viable contributing source for the plasmasheet and ring current regions and forced a fundamen-
tal reconsideration of the origin of magnetosphericplasmas. Early measurements connected the energeticoxygen ions with the aurora and led to initial iono-
spheric source models that were related to auroralenergization processes.
In the years immediately preceding the energeticoxygen measurements, Axford (1968) and Banks andHolzer (1968) theorized that conditions in the top of
the ionosphere should be favorable to stimulate anupward ¯ow of ions and electrons into the magneto-sphere above. The ion component was predicted to be
mainly H� with energies of less than 1 eV thatincreased with altitude because of acceleration by the
ambipolar electric ®eld. The ¯uxes of these particleswere expected to be quite large, 3 � 108 ions/cm2 s,and were expected to be found throughout the iono-
sphere from the polar regions down to the inner mag-netosphere. The out¯ow was named the polar windand its future measurement was predicted by space-
craft that would carry mass spectrometers sensitive toplasma energies of one electron volt or lower.
The idea of an up-¯owing plasma with energies of 1eV matched the whistler measurements of the plasma-sphere Ð a region of plasma of the same energy that
was found to be present in the inner parts of the mag-netosphere out to L values of 4±6 (Park, 1974). Per-
haps the polar wind phenomenon would be themechanism by which the ionosphere could ®ll thisroughly corotating portion of the magnetosphere.
Because of the similarity of energy and composition ofthe plasmasphere and the ionosphere, the two regionswere typically studied by ionospheric scientists, while
the more energetic processes of the aurora, plasmasheet, ring current and radiation belts were studied by
the vast majority of magnetospheric physicists.Because of the obvious connections between the
magnetosphere, the ionosphere, and the atmosphere, a
mission was proposed to measure the three regionssimultaneously. The Dynamics Explorer missionbrought the ionospheric and magnetospheric commu-
nities together in mission design, development and op-eration and stimulated broader thinking about the
coupling of the ionosphere and magnetosphere. Theoverlapping nature of the phenomena under studyforced the development of new trans-regional instru-
ments. These instruments could operate both in theionosphere where plasmas had energies of less than 1
eV and spacecraft potentials were typically slightly
negative to the magnetosphere where particle energiesexceeded 1 MeV and spacecraft potentials reached tensof volts positive. The Retarding Ion Mass Spec-
trometer (Chappell et al., 1981) was one such instru-ment which combined the qualities of a retardingpotential analyzer designed for the ionosphere with the
ionosphere/magnetosphere capabilities of an ion massspectrometer. By using three separate sensor heads on
a spacecraft spinning in a cartwheel mode, the RIMSinstrument could sample all of the positive ions thatcould overcome spacecraft potential and reach the ana-
lyzer heads which, together, covered the complete unitsphere of look directions in one spacecraft spin period.
The RIMS instrument was designed to study thedetails of the plasmasphere and its connection to theionosphere. Its operational regime was to be in the
inner magnetosphere and ionosphere where the plas-mas were predominantly dense and cold and wherespacecraft potential remained negative or only slightly
positive. The apogee of the DE-1 high altitude space-craft was 4.5 earth radii. Because of a limitation on
the downlink telemetry, the data from the DE-1 satel-lite instruments could not be received at all times anddecisions had to be made by the investigator team
regarding the portions of the orbit in which the di�er-ent instruments would be operated. There were two
schools of interest divided along the lines of the iono-spheric and magnetospheric scientists with the formerexpressing more interest in the lower latitudes than the
latter. Indeed, since the goals of the RIMS instrumenthad been to study the plasmasphere/ionosphere inter-action, the portions of the orbits which passed through
the plasmasphere were the optimum ones and thepolar portions of the orbits seemed far less desirable.
Long discussion and compromise led to an operationalapproach which covered each region of interest duringdi�erent times.
To the great surprise of the RIMS investigator team,the polar portion of the orbit revealed a substantial
¯ow of ionospheric ions upward into the magneto-sphere including the direct measurement of the theor-etically-predicted polar wind, (Nagai et al., 1984).
And, characteristic of the serendipity of science, thisout¯ow phenomenon became arguably the most signi®-cant discovery of the investigation along with the orig-
inally-sought information about the ionosphere/plasmasphere connections. When combined with the
more energetic measurements of the Energetic IonComposition Spectrometer (Yau et al., 1985), the totalupward ¯owing plasma with energies above the posi-
tive spacecraft potential could be measured and it wassigni®cant. In fact it was of such a magnitude that thestrength of the ionospheric source moved into the cat-
egory of a contender as a supplier of magnetosphericplasmas. It appeared that these out¯ow measurements
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436422
could be combined with crude trajectory calculationsto estimate the density of magnetospheric plasmas that
could be supplied by the ionosphere (Chappell et al.,1987).During the completion of the RIMS instrument and
the Dynamics Explorer satellites, the magnetosphericcommunity developed the concept of a set of space-craft to make multi-point measurements of the solar
wind and magnetosphere. Originally called the Originof Plasmas in the Earth's Neighborhood (OPEN), theset of four and later three spacecraft were made avail-
able for proposal. Still interested in the interconnectionof the plasmasphere and ionosphere, a broadenedRIMS team submitted a proposal for the Thermal IonDynamics Experiment (TIDE) and an accompanying
Plasma Source Instrument (PSI) for spacecraft poten-tial control (Moore et al., 1995). The TIDE instrumentpermitted the achievement of better angular resolution
and higher sensitivity to low ion ¯uxes while the PSIinstrument o�ered control of spacecraft potential atthe higher altitudes. It was proposed for the Polar
Plasma Laboratory and Equatorial Magnetotail Lab-oratory of the OPEN mission. This mission withoutthe Equatorial Magnetotail Laboratory was to ¯y
more than 15 years after the original proposal andwould be in an ideal position to make the next gener-ation of measurements on the strength of the iono-spheric source. At the time of the proposal this source
had not even been measured for the ®rst time byDynamics Explorer.As a result of the TIDE/PSI instruments on what is
now called the International Solar Terrestrial Physics(ISTP) program, we are in a position to review theionospheric source strength as measured by Dynamics
Explorer and to give a second generation look at thesource strength now being measured by a satellitewhich was proposed but not developed prior to thetime of the ®rst generation measurements. This paper
reviews the original estimates of the contribution ofthe ionospheric source to the magnetosphere and re-examines it with the new and more capable TIDE/PSI
measurements. In addition, the development of a codeby Delcourt et al. (1992, 1994), which calculates singleparticle trajectories and energization, permits the more
accurate prediction of the ultimate magnetospheric lo-cation of the up¯owing ionospheric plasma and itsenergization to become the higher energy plasma that
was initially thought to be entirely of solar origin. Thecode uses only the fundamental energization mechan-isms linked to the large scale convection electric ®eldand gravity.
2. Review of the original ionosphere source estimates
Stimulated by the surprising strength of the ion
out¯ow measured by Dynamics Explorer, Chappell
et al. (1987) estimated the total ¯ux of ions movingout of the ionosphere and used this out¯ow to esti-mate the magnetospheric plasma densities that
would result in the di�erent regions such as theplasmasphere, plasma trough and plasma sheet.
Table 1, which is taken from the Chappell et al.(1987) paper, shows the strength of the di�erentionospheric sources in units of 1025 ions/s. These
data combined theoretical estimates of the polar windwith limited measurements and added the measuredsource strength of the cleft ion fountain, the auroral
zone, and the polar cap from the RIMS instrumentand the Energetic Ion Composition Spectrometer
reported by Yau et al. (1985). Note that the estimatedpolar wind ¯ux is the dominant one, ranging from 10to 46� 1025 ions/s. These estimates utilize the theoreti-
cal polar wind ¯uxes of 1±3 � 108 ions/cm2 s andassume that there is a polar wind up-¯ow for all invar-iant latitudes greater than 518, the boundary of the
inner/outer plasmasphere regions. Flux tubes throughL-shells above this boundary are thought to be not
®lled to di�usive equilibrium and hence available toreceive a strong upward polar wind ¯ux. This largearea of the up-¯ow strongly in¯uences calculations of
the total upward ¯ux.The out-¯owing ions could move into many
di�erent regions of the magnetosphere depending on
their L-shell and magnetic local time location aswell as their subsequent corotation and convective
drifts as they move up along the magnetic ®eldline. Fig. 1 shows a schematic of the out-¯owingions as envisioned in the Chappell et al. (1987)
paper. Note that the particles can move directlyupward to populate the plasmasphere and plasmatrough above or they can convectively drift across
the polar cap, through the lobes of the tail andinto the plasma sheet where they are available to
be energized by neutral sheet processes.This entry approach across the polar cap and
through the lobes of the magnetotail had not been
strongly considered in the past because out-¯owingplasma in the polar caps and tail lobes had not
been observed and because the background densitieswere low. The low densities were misleading in twoways. First, because of the high parallel velocity of
the polar wind streams, it is possible to have sub-stantial parallel ¯uxes of plasma ions even in a lowdensity background which had caused this region of
substantial ¯ux to be overlooked. Second, the space-craft charging in these low density regions can be
high (10s of volts positive) and this would prohibitthe measurement of all ions with energies less thanspacecraft potential. This low energy population of
ions would hence have been invisible on all previousspacecraft (Chappell, 1988). Only the addition of a
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436 423
spacecraft potential control device like the PSI
would enable the observation of these out-¯owing
ions.
Using measured convection electric ®eld values for
quiet and active times, the trajectories of the up-¯ow-
ing ions were estimated and their destinations in the
di�erent regions of the magnetosphere determined. It
is indeed the case that these crude trajectories pre-
dicted a dominant contribution not only to the plas-
masphere and plasma trough of the inner
magnetosphere as expected, but to the plasma sheet as
well. The calculated densities, using the ionospheric
out¯ow, an estimated volume of the region and a
characteristic residence time in the region, matched the
range of observed densities in all of these principal
magnetosphere regions. Of particular signi®cance were
the predicted densities of 0.4 to 6 ions/cm3 for the
plasma sheet and 0.03 to 0.08 ions/cm3 for the magne-
Table 1
Total ionospheric source strengtha
Quiet Acive
H� He� O� H� He� O�
Polar wind
Solar maximum 15.0(1� 108)b 1.1(7.05� 106) 10.0(6.5� 107) 0.40(2.61� 106)
Solar maximum 46.0(3� 108) 0.59(3.8� 106) 31.0(2� 108) 0.28(1.82� 106)
Cleft ion fountion
Solar maximum 0.33 1.6 0.43 4.8
Solar minimumc 0.63 0.73 0.43 1.9
Auroral zone
Solar maximum 2.1 1.6 2.6 7.7
Solar minimumc 1.7 1.0 3.3 3.2
Polar cap
Solar maximum 0.24 0.24 0.61 2.5
Solar minimumc 0.43 0.39 1.0 1.5
a Each entry is to be multiplied by 1025 ions sÿ1.b The numbers in parentheses represent the polar wind ¯uxes in ions per square centimeter per second that were used for the
di�erent solar and magnetic conditions.c The Yau et al. (1985) DE data were taken in two sets; one near solar maximum and one about half way between solar maxi-
mum and minimum (see Fig. 1).
Fig. 1. A schematic representation of ionospheric ¯ow into the magnetosphere from Chappell et al., (1987) showing `invisible'
polar out-¯ow through the magnetotail lobes to the plasma sheet.
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436424
totail lobes. The former is more than adequate to
account for the measured plasma sheet ion densitiesand the latter con®rms the very low density of the taillobes which would cause substantial positive charging
of any un-neutralized spacecraft in that region. Thispositive charging would make the low energy ions in-
visible.The spacecraft potential in the high altitude polar
cap and entry interface to the tail lobes has beenmeasured by the Electric Field Instrument on thePolar spacecraft (Harvey et al. 1995) and is shown in
the upper panel of Fig. 2. Note that it reaches 40 Vpositive with respect to the surrounding plasma. The
lower panel of Fig. 2 shows the e�ect of the PSI oper-ation which clamps the Polar spacecraft potential to
just over 2 V positive, thereby permitting the obser-vation of the low energy out-¯owing ions with energiesgreater than 2 eV.
It was the above-mentioned calculations of the den-sities which led to the general conclusion of the orig-
inal Chappell et al. (1987) paper that the ionosphere isa fully adequate source of plasma for the magneto-
sphere. But what pieces were missing in this earlier
work? First, the out¯ow measurements were made at
lower altitudes below the 4.5 RE apogee of the DE-1spacecraft. There were no direct measurements of theparticles at higher altitudes in the polar cap entering
the lobes of the magnetotail for transport to theplasma sheet region. Second, the trajectories calculated
for the up-¯owing ions were very crude based onboundary estimates of the di�erent regions and not the
exact single particle trajectory calculations. Third,there was no calculation of the energization of the par-ticles as they moved upward into the magnetosphere.
This omission is not as important in the plasmasphereand plasma trough where the cold particle energies are
similar to their ionospheric source. It is important,however, in the plasma sheet where energy enhance-
ments of a factor of 1000 would be required to boostthe ionospheric particles up to the observed plasmasheet energies. Fourth, the theoretical ion out¯ow con-
tribution of the polar wind used only the classicalpolar wind models which did not include the variety of
energization processes such as ion heating in the polarcleft which can dramatically change the polar wind
energy and composition. Finally, there was no direct
0
10
20
30
40
50
15:00:00 16:00:00 17:00:00 18:00:00 19:00:00 20:00:00
1997 August 6 15:00-20:00 UT
PO
LAR
EF
IS
pace
craf
t Pot
entia
l (V
)
Time (UT)
0
1
2
3
4
11:00:00 13:00:00 15:00:00 17:00:00 19:00:00 21:00:00
1997 August 15 11:00-22:00 UT
PO
LA
R E
FI
Spa
cecr
aft P
ote
ntia
l (V
)
Time (UT)
Spacecraft Potential for typical agogee pass of POLAR
Spacecraft Potential for typical agogee pass of POLAR with PSI in operation
Fig. 2. Two plots of Polar spacecraft potential as the spacecraft moves across the high altitude polar cap. In the upper panel with-
out the plasma neutralizer the potential reaches 40 V positive with respect to the surrounding plasma. The lower panel with the
Plasma Source Instrument operating shows a potential of only 2.5 V.
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436 425
measurement of the ion out¯ow actually reaching theplasma sheet region of the tail due to the elimination
of the Equator spacecraft from the program. We canhope that the Polar mission will be extended su�-ciently to permit the line of apsides of the orbit to
rotate into the magnetotail region so that the space-craft will be able to cross the tail lobe/plasma sheetinterface.
3. The ISTP era
In the decade since the original ion source estimates,
there have been a number of measurements of iono-spheric out¯ow. See the excellent review articles byYau and Andre (1997) and Andre and Yau (1997). In
addition, there have been two important developmentsÐ the improved instrumentation and the single par-ticle trajectory codes, (see for example, Cladis, 1986;
Delcourt et al., 1994; Ashour-Abdalla et al., 1999).The development of the TIDE instrument with its highsensitivity, and high angular resolution combined withthe PSI for spacecraft potential control has e�ectively
removed the blinders which made the low energy ionout¯ow invisible to earlier high altitude missions. Thesingle particle trajectory code of Delcourt et al. (1994)
has enabled the more accurate calculation not only of
the path of the ion transit into the magnetosphere but
of the energization of the particles as they move in thelarge scale convection ®eld all the way from the iono-sphere to the inner plasma sheet. In addition to the
out¯ow trajectories, the Delcourt code has beenupgraded for this study to permit an integration back-
wards in time using for the ®rst time a tilted dipolemagnetic ®eld model to trace the particles measured
high in the polar cap back to their point of origin inthe ionosphere below.
The importance of the ISTP program to magneto-spheric physics is well known. The ability to make
multi-point coordinated measurements throughout geo-space opens up new avenues of investigation of particleand energy ¯ow. The unique orbit of the Polar space-
craft as shown in Fig. 3 samples the high altitudenorthern polar cap of the Earth and covers the ®eld
lines along which out-¯owing ions would enter thenorthern lobes of the magnetotail. When the TIDE/
PSI combination is operated through this region, theout-¯owing ions are very evident as shown in Fig. 4which is a set of passes across the high altitude polar
cap. The plots display ¯ux levels in a color scaleplotted versus time on the X-axis and spin angle on
the Y-axis. The line across each plot is the ion out¯owmoving close to the magnetic ®eld line direction. The
PSI instrument is turned o� before the start of the
Fig. 3. A typical orbit of the Polar spacecraft shown in solar-magnetospheric coordinates.
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436426
PSI ONPSI OFF
Fig.4.A
seriesofpasses
ofthePolarsatellite
acro
ssthehighaltitu
depolarcapwhich
show
the¯ux(co
lorsca
le)plotted
versu
s
timeontheabscissa
andspin
angle
ontheordinate
ofeach
plot.These
measurem
ents
were
madebeginningjust
befo
reapogee
at
8.9
REgeocen
tricwith
thePSIneutralizer
opera
tingonallexcep
tthelast
twopanels.
C.R.Chappell
etal./
JournalofAtm
ospheric
andSolar-T
errestrialPhysics
62(2000)421±436
427
third panel from the bottom. In the last three panels
only the higher energy ions which can overcome space-craft potential are seen. The TIDE/PSI combination isonly operated for two weeks every two months. In the
course of the total mission, the number of passes isbuilding up and the ion out¯ow statistics are beginningto improve.
Before examining the characteristics of the ion out-¯ow measured by the TIDE instrument, it is instructive
to look at where the ions will travel as they move outinto the magnetosphere. To get an idea of their desti-nation, a point has been selected every hour along the
satellite pass through the high altitude polar cap for allof the passes in which the TIDE/PSI combination has
been operated. The mean energy of each of the ionout¯ows has been determined as well as its location inthe polar cap. These parameters have been used as
input to a three-dimensional test particle simulationcode along with the 3-h Kp that existed at the time ofthe measurement and the appropriate dipole tilt lo-
cation for that period. The code then calculates theparticle trajectory for the ion as it moves out into the
magnetosphere. Of the runs made, about two-thirds ofthe ions move into the plasma sheet region of the mag-netotail; the remainder precipitate in the northern or
southern hemisphere or are lost out of the model mag-netic ®eld 70 RE down the tail.The particle trajectory code is an adaptation of that
of Delcourt and co-workers (Delcourt et al., 1992,1993, 1994, 1997). For this application the model uses
the Tsyganenko (1989) c model magnetic ®eld and theVolland (1978) model of the ionosphere potential dis-tribution for the convection electric ®eld. In addition,
to more closely re¯ect the conditions of each hourlyobservation, the model ®elds are matched to the mag-netic activity (Kp) level and magnetic dipole tilt angle
of the observing UT time.It should be noted that single particle calculations
cannot be considered exact. As with magnetosphericmodels of all types, the calculations contain approxi-mations. For example, the Tsyganenko 89c and Vol-
land models do not provide self-consistent magneticand electric ®elds. However, these models do re¯ect by
design the average ¯ow characteristics observed in theplasma sheet (Angelopoulos et al., 1993). We are treat-ing the ions in this study as test particles in the mod-
eled electric and magnetic ®elds. Eventually, one has totreat the e�ects of these particles back on the ®elds,but this is beyond the scope of this study. A recent
®rst look at the e�ects of the ionospheric source onthe plasma sheet can be found in this volume (Winglee,
2000). In addition, although processes such as waveparticle interactions are not included in the model, theresults of particle energization and trajectory desti-
nations are remarkably similar to measured plasmasheet particle characteristics. Thus, the modeled trajec-
tories provide further quantitative grounds on which
to view the ionosphere as a signi®cant plasma sheetsource.Fig. 5 shows one of the trajectory calculations. The
top panel shows the X±Z plane of the magnetosphere,the middle panel the X±Y equatorial plane, the lower
left panel the Y±Z plane looking toward the sun withthe lower right panel showing the energy of the particleas it moves along the trajectory. The initial observed
energy of the particle is 11 eV. In the top panel thepoint of transit is shown for 1.5 and 2 h transit timeÐ before and after entry into the plasma sheet region.
These times may be matched with the horizontal axisof the lower right panel. Notice the dramatic energiza-
tion of the particle as it enters the plasma sheet chan-ging from an energy of less than 100 eV to an energyof 3 keV in less than 10 min. This energization is a
result only of single particle motion in the distendedmagnetic ®eld of the neutral sheet region and thecross-tail convection electric ®eld.
Spatially the particle moves from the high altitudepolar region through the low density lobe of the geo-
tail where its energy remains below a few tens of eVand then into the plasma sheet region where it is ener-gized to plasma sheet energies and moves toward the
Earth eventually ending up in the dusk ring currentregion having acquired an energy in excess of 10 keVtypical of the ring current particles. Hence, we see a
very low energy out-¯owing ion being transformedinto di�erent parts of the energetic magnetospheric
particle population in the precise places where particlesof the calculated energies are found. Note also that ittakes the out-¯owing particles one to two hours to tra-
vel from the ionosphere to the plasma sheet. Thesetimes are somewhat longer than the duration of a typi-cal substorm cycle, but typical of storm development
scales. DE-1/RIMS and recent POLAR/TIDEmeasurements, Moore et al. (1999), indicate that day-
side out¯ows are strongly enhanced by solar windpressure ¯uctuations. Thus it may be that the mainphase of storms is enabled or facilitated by the arrival
of enhanced ionospheric out¯ows at the plasma sheet.Figs. 6±8 show three other examples of calculated
trajectories for out-¯owing ions with initial energies of13.3, 4.8, and 7.4 eV respectively. In each of theseexamples, the ion transits the northern tail lobe at very
low energy (less than 20 eV) and is then rapidly ener-gized to energies of several keV as it enters the plasmasheet. The particle subsequently drifts toward the
Earth where it reaches energies greater than 10 keV atthe inner edge of the inward ¯ow.
If the ions are traced backwards from their obser-vation points in the polar cap where guiding center-based arguments are valid, their foot points of origin
in the ionosphere can be calculated. In the upper twopanels of Fig. 9, an example of these downward trajec-
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436428
tories is shown in the X±Z and Y±Z planes. In the
example shown the ion energy was 11.2 eV. If all of
the trajectory footprints from the data are plotted ver-
sus invariant latitude and magnetic local time, the plot
at the bottom of the page is the result. This plot
suggests that the origin of the upward ¯ux of ions is
spread across the polar cap and extends to lower lati-
tudes in the dawn dusk sector. The satellite coverage
of these out¯ows is not uniform and is decreased
within two hours either side of the noon midnight mer-
idian. This non-uniform coverage may well distort the
picture of the ionospheric source and should be ®lled
in with future measurements.
There is, however, an uncertainty in the location of
these foot points which comes from an assumption
used in the trajectory calculation. Since the trajectory
code does not include wave-particle interactions or
energization mechanisms other than those related to
large scale convection and gravity, the backwards tra-
jectory from the point of observation in the high alti-
tude polar cap back to the ionospheric foot point
assumes a particle energy equal to the energy measured
at the satellite. This energy of 11.2 eV for the case
shown in the top of Fig. 9 may be misleading. If the
particle began its journey as a low energy (1 eV) classi-
cal polar wind up-¯owing particle, it could have drifted
farther in an anti-sunward direction as a very low
energy particle before becoming energized to 11.2 eV
through one or a series of acceleration processes as-
sociated with the polar cleft, for example. Hence, the
1.5 hours
2 hoursSunward
Dawn
Dusk
Observation @ 6Jun1997 1345UT8.2 RE, 9.17MLT, 80.51MLAT, Kp=2
Fig. 5. Calculated particle trajectories which trace the path and energization of a measured out-¯owing ion as it moves through the
magnetotail lobe into the plasma sheet. The top two panels show the motion in solar-magnetospheric coordinates. The bottom two
panels show solar-magnetospheric coordinates on the left and energy versus time of travel on the right.
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436 429
points plotted in the lower part of Fig. 9 probably rep-
resent the most anti-sunward location for the foot
point of ionospheric origin with the more probable lo-
cation being on the dayside equatorward of the polar
cleft region.
To get an estimate of the lower latitude limit of the
point of origin on the dayside, we have calculated the
backward trajectories of several of the sunward points
near noon using an energy of 1 eV and assuming that
the particles moved up the ®eld line with this lower
velocity. Hence, they would have drifted farther in an
anti-sunward direction before receiving the additional
10 eV of energy just before reaching the spacecraft.
This calculation gives initial foot points in the latitude
range of 55±688.Determining the precise point of origin of the out-
¯owing ion is made even more di�cult by the variety
of energization processes that can take place as it
leaves the ionosphere. Initially, perhaps it is a classical
polar wind ion which is subsequently acted upon by
many `non-classical' energization processes. Fig. 10
from Schunk and Sojka (1997) illustrates several of the
processes that cause changes not only in the energy of
the out¯ow but in its composition mixture as well.
Many of these occur when the particles drift through
the auroral regions such as the polar cleft and the
nightside auroral zone. The drift path is often quite
involved as shown in Fig. 11 which also comes from
the work of Schunk and Sojka (1997). In their coupled
model of the ionosphere and the resulting out¯ow, a
large variety of ¯uxes, energies and compositions are
possible. Therefore, a great deal of detailed modeling
must be brought into play before accurate backward
tracing of out-¯owing particles can be accomplished.
Observation @ 17Mar1997 1558UT8.7 RE, 19.8MLT, 79.7MLAT, Kp=1
Fig. 6. Calculated particle trajectories as shown in the previous ®gure.
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436430
The Schunk and Sojka (1997) model, predicts a peak
out¯ow of 3 � 1025 ions/s out of the total ionosphere
for the magnetic storm which they modeled. This is a
factor of three below the lower end of the range of
out¯ow from the Chappell et al. (1987) predictions.
Fig. 12 shows the out-¯ux of all ions adjusted to
1000 km as measured by the TIDE/PSI instrument
combination in the high altitude polar cap. The ¯ux
magnitude is in the same range as the theoretical polar
wind ¯uxes of 1±3 � 108 ions/cm2 s used in the Chap-
pell et al. (1987) paper. These measured ¯uxes come
from 14808 two minute averages made from apogee
passes between March 1997 and September 1999 in
which the TIDE/PSI instrument combination was
operating. It is clear that the range of upward traveling
ion ¯ux matches or exceeds the original estimates and
can reach 8� 108 ions/cm2 s at high Kp.
Although the individual ¯uxes are comparable to or
greater than the polar wind numbers used in the pre-
vious paper, the total out¯ow that results across the
whole ionosphere remains uncertain because the area
of up-¯ow has not been completely measured. Due to
a limitation in the operational range of the TIDE
instrument, the high altitude polar wind out¯ow for L-
values less than about eight cannot be measured.
Hence the areal breadth of the up-¯ux cannot be
directly determined at high altitudes. It is expected
that information on polar wind up-¯ow from the peri-
gee passes in the southern hemisphere will be import-
ant in addressing this issue.
A second area of future study will be the variation
Observation @ 20Mar1997 0100UT8.4 RE, 2.98MLT, 72.9MLAT, Kp=0+
Fig. 7. Calculated particle trajectories as shown in the previous ®gure.
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436 431
in ion out¯ow with changing solar wind parameters.Initial studies on this important subject have been
made by Moore et al. (1999) and Giles et al. (1998)and show a surprising apparent control by the solarwind pressure ¯uctuations. Further measurements to
build adequate statistics will be contingent on the con-tinued operation of the TIDE/PSI instruments on thePolar spacecraft.
4. Conclusions
The advances in instrumentation and computer
modeling of the past decade have given us an enhancedability to measure and predict the contribution of theionospheric source in supplying magnetospheric plas-
mas. The predictions of Chappell et al. (1987) have
been con®rmed in several areas. First, the idea of an
out-¯owing plasma in the polar regions which moves
`invisibly' toward the lobes of the magnetotail to
supply the plasma sheet has been demonstrated by the
TIDE/PSI measurements of out-¯owing plasma near
the polar cap/tail lobe interface. The single particle tra-
jectory calculations provide insights into the destiny of
this ionospheric plasma in the plasma sheet. Second,
these same calculations show the straight-forward ener-
gization of these invisible out-¯owing ions as they
enter the neutral sheet region Ð an energization that
matches the energy of plasma sheet particles that have
been observed there for decades.
Thirdly, the magnitude of the ion out¯ow measured
by TIDE/PSI at 1±3 � 108 ions/cm2 s matches the
Observation @ 20Mar1997 0240UT7.8 RE, 2.28MLT, 65.9MLAT, Kp=O+
Fig. 8. Calculated particle trajectories as shown in the previous ®gure.
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436432
¯uxes used in the original paper to predict the strength
of the ionospheric source. This ¯ux can be highly vari-
able in both spatial location and in response to chan-
ging solar wind and ionospheric solar illumination
conditions and requires continued observation with the
TIDE/PSI combination as the Polar mission pro-
gresses. The accurate determination of the total ion
out¯ow requires a subsequent study to characterize the
spatial breadth of the out¯ow across the full range of
ionospheric invariant latitudes. In the original paper
the polar wind out¯ow was assumed to exist in all
un®lled ¯ux tubes at L-values above the inner/outer
plasmasphere boundary at L � 2:5: The TIDE/PSI
high altitude measurements to date have been limited
to L-values greater than eight and hence have left a
gap in measured out¯ow that needs to be investigated
further to facilitate the determination of the total ion
out¯ow. Even with the limited observations currently
in hand, we can state that the regions of out¯ow that
have been observed by POLAR/TIDE are essentially
1997 June 9 01:10 UT
Footpoints of alltrajectories
Typical trajectory tracing
12
18 6
0 MLT
8070
6050
Fig. 9. A calculated particle trajectory moving backward in time to estimate the location of the origin footprint of a measured out-
¯owing ion. The top two panels show the trajectory in solar-magnetospheric coordinates and the lower panel shows the footprints
of all the measured particle traces in an invariant lattitude-magnetic local time plot.
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436 433
Fig. 10. A schematic drawing of the out-¯ow processes that in¯uence the energy and motion of polar wind up-¯owing ions. From
Schunk and Sojka (1997).
Fig. 11. A calculated convection trajectory of a ¯ux tube of
plasma moving in magnetic local time and magnetic latitude
during changing magnetic activity. The tic marks show the
time in hours. From Schunk and Sojka (1997).
Fig. 12. A plot showing the average magnitude of the parallel
up-ward ¯ux as a function of invariant latitude from 14,808
samples measured by the TIDE/PSI instrument combination
between March 1997 and September 1998. The lines through
each square show the standard deviation for each invariant
latitude summed over all magnetic local times and corrected
to the equivalent ¯ux at 1000 km.
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436434
those that feed directly through the lobes to the plasmasheet. The lower latitude regions we have missed are
largely those that are on more dipolar ®eld lines andare therefore within the closed inner magnetosphericcirculation pattern except during enhanced convection
periods during higher Kp.In the future, in addition to continued TIDE/PSI
measurements to understand the ion out¯ow variations
with changing solar and solar wind conditions, theexciting opportunity exists to measure the dramaticchanges in energy of the low energy ions as they move
from the lobe of the magnetotail into the neutral sheetregion. This eagerly anticipated measurement can bemade with TIDE/PSI when the apogee of the Polarspacecraft has precessed su�ciently toward the equator
to give an orbital track which traverses the innerplasma sheet. This will replace some of the measure-ments that were lost with the cancellation of the
Equator spacecraft and can shed further light on thedetails of the dominant ionospheric supply process forthe central plasma sheet that has been predicted
recently by the multi-¯uid magnetospheric simulationsof Winglee (1998).In summary, the predictions of the adequacy of the
ionospheric source strength of a decade ago still lookintriguing and reasonable according to the newmeasurements. More TIDE/PSI observations arerequired and more study in several speci®c areas must
be done. As of this point in time the ionosphericsource appears to be rich in its mixture of ions andimpressive in its strength in a variety of magneto-
spheric conditions.
Acknowledgements
The authors would like to recognize the considerablee�orts of the TIDE/PSI engineering and software
teams at NASA's Marshall Space Flight Center. Thisresearch was supported by the Global GeospaceScience program at NASA's Goddard Space Flight
Center.
References
Andre, M., Yau, A., 1997. Theories and observations of ion
energization and out¯ow in the high latitude magneto-
sphere. Space Science Reviews 80, 27.
Angelopoulos, V., Kennel, C.F., Coroniti, F.V., Pellat, R.,
Spence, H.E., Kivelson, M.G., Walker, R.J., Baumjohann,
W., Feldman, W.C., Gosling, J.T., Russell, C.T., 1993.
Characteristics of ion ¯ow in the quiet state of the inner
plasma sheet. Geophysical Research Letters 20 (16), 1711.
Ashour-Abdalla, M., El-Alaoui, M., Peroomian, V., Walker,
R.J., Raeder, J., 1999. Source distributions of substorm
ions observed in the near-Earth magnetotail. Geophysical
Research Letters 26 (7), 955.
Axford, I., 1968. The polar wind and terrestrial helium bud-
get. Journal of Geophysical Research 73 (21), 6855.
Banks, P.M., Holzer, T.E., 1968. The Polar Wind. Journal of
Geophysical Research 73, 6846.
Chappell, C.R., 1988. The terrestrial plasma source: a new
perspective in solar-terrestrial processes from Dynamics
Explorer. Reviews of Geophysics 26, 229.
Chappell, C.R., Fields, S.A., Baugher, C.R., Ho�man, J.H.,
Hanson, W.B., Wright, W.W., Hammack, D.H., Carignan,
G.R., Nagy, A.F., 1981. The retarding ion mass spec-
trometer on Dynamics Explorer Ð A. Space Science
Instrumentation 5, 477.
Chappell, C.R., Moore, T.E., Waite Jr, J.H., 1987. The iono-
sphere as a fully adequate source of plasma for the Earth's
magnetosphere. Journal of Geophysical Research 92, 5896.
Cladis, J.B., 1986. Parallel acceleration and transport of ions
from polar ionosphere to plasma sheet. Geophysical
Research Letters 13, 893.
Delcourt, D.C., Moore, T.E., Sauvaud, J.A., Chappell, C.R.,
1992. Nonadiabatic transport features in the outer cusp
region. Journal of Geophysical Research 97 (A11), 16833.
Delcourt, D.C., Sauvaud, J.A., Moore, T.E., 1993. Polar wind
ion dynamics in the magnetotail. Journal of Geophysical
Research 98 (A6), 9155.
Delcourt, D.C., Moore, T.E., Chappell, C.R., 1994.
Contribution of low-energy ionospheric protons to the
plasma sheet. Journal of Geophysical Research 99 (A4),
5681.
Delcourt, D.C., Sauvaud, J.A., Moore, T.E., 1997. Phase
bunching during substorm dipolarization. Journal of
Geophysical Research 102 (A11), 313.
Giles, B.L., Moore, T.E., Chappell, C.R., Delcourt, D.C.,
1998. Spacecraft potentials and measured ionospheric out-
¯ows. Transactions of the American Geophysical Union.
Harvey, P., Mozer, F.S., Pankow, D., Wygant, J., Maynard,
N.C., Singer, H., Sullivan, W., Anderson, P.B., Pfa�, R.,
Aggson, T., Pedersen, A., Falthammar, C.-G.,
Tanskannen, P., 1995. The electric ®eld instrument on the
Polar satellite. Space Science Reviews 71, 583.
Moore, T.E., Chappell, C.R., Chandler, M.O., Fields, S.A.,
Pollock, C.J., Reasoner, D.L., Young, D.T., Burch, J.L.,
Eaker, N., Waite Jr, J.H., McComus, D.J., Nordholt, J.E.,
Thomsen, M.F., Berthelier, J.J., Robson, R., 1995. The
thermal ion dynamics experiment and plasma source
instrument. Space Science Reviews 71, 409.
Moore, T.E., M.O. Chandler, C.R. Chappell, R.H. Comfort,
P.D. Craven, D.C. Delcourt, H.A. Elliott, B.L. Giles, J.L.
Horwitz, C.J. Pollock, Su, Y.-J. 1999. Polar/TIDE results
on polar ion out¯ows. In: Sun±Earth Plasma Connections,
Geophysical Monograph, vol. 109.
Nagai, T., Waite Jr, J.H., Green, J.L., Chappell, C.R., Olsen,
R.C., Comfort, R.H., 1984. First measurements of super-
sonic polar wind in the polar magnetosphere. Geophysical
Research Letters 11, 669.
Park, G., 1974. Some features of plasma distribution in the
plasmasphere deduced from Antarctic whistlers. Journal of
Geophysical Research 79, 169.
Schunk, R.W., Sojka, J.J., 1997. Global ionosphere-polar
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436 435
wind system during changing magnetic activity. Journal of
Geophysical Research 102, 11625.
Shelley, E.G., Johnson, R.G., Sharp, R.D., 1972. Satellite ob-
servations of energetic heavy ions during a geomagnetic
storm. Journal of Geophysical Research 77, 6104.
Tsyganenko, N.A., 1989. A magnetospheric magnetic ®eld
model with the warped tail current sheet. Planetary and
Space Science 37, 5.
Volland, H., 1978. A model of the magnetospheric electric
convection ®eld. Journal of Geophysical Research 83,
2695.
Winglee, R.M., 1998. Multi-¯uid simulations of the magneto-
sphere: the identi®cation of the geopause and its variation
with IMF. Geophysical Research Letters 25 (24), 4441.
Winglee, R.M., 2000. Mapping of ionospheric out¯ows into
the magnetosphere for varying IMF conditions. Journal of
Atmospheric and Solar-Terrestrial Research 62, 527.
Yau, A.W., Shelley, E.G., Peterson, W.K., Lenchyshyn, L.,
1985. Energetic auroral and polar ion out¯ow at DE 1
altitudes: magnitude, composition, magnetic activity
dependence, and long-term variations. Journal of
Geophysical Research 90, 8417.
Yau, A.W., Andre, M., 1997. Sources of ion out¯ow in the
high latitude ionosphere. Space, Science Reviews 80, 1.
C.R. Chappell et al. / Journal of Atmospheric and Solar-Terrestrial Physics 62 (2000) 421±436436