Content uploaded by Steve Milan
Author content
All content in this area was uploaded by Steve Milan
Content may be subject to copyright.
Available via license: CC BY 4.0
Content may be subject to copyright.
CUTLASS Finland radar observations of the ionospheric signatures
of ¯ux transfer events and the resulting plasma ¯ows
G. Provan, T. K. Yeoman, S. E. Milan
Department of Physics and Astronomy, University of Leicester, University Road, Leicester LE1 7RH, UK
Received: 1 December 1997 / Revised: 23 March 1998 / Accepted: 25 March 1998
Abstract. The CUTLASS Finland radar has been run
in a two-beam special scan mode, which oered excel-
lent temporal and spatial information on the ¯ows in the
high-latitude ionosphere. A detailed study of one day of
this data revealed a convection reversal boundary
(CRB) in the CUTLASS ®eld of view (f.o.v) on the
dayside, the direction of plasma ¯ow either side of the
boundary being typical of a dawn-cell convection
pattern. Poleward of the CRB a number of pulsed
transients are observed, seemingly moving away from
the radar. These transients are identi®ed here as the
ionospheric signature of ¯ux transfer events (FTEs).
Equatorward of the CRB continuous backscatter was
observed, believed to be due to the return ¯ow on closed
®eld lines. The two-beam scan oered a new and
innovative opportunity to determine the size and veloc-
ity of the ionospheric signatures associated with ¯ux
transfer events and the related plasma ¯ow pattern. The
transient signature was found to have an azimuthal
extent of 1900 900 km and an poleward extent of
250 km. The motion of the transient features was in a
predominantly westward azimuthal direction, at a
velocity of 7.5 3 km.
Key words. Magnetospheric physics (auroral
phenomena; magnetopause, cusp and boundary layers;
magnetosphere - ionosphere interaction)
1 Introduction
The two-cell convection pattern observed in the high-
latitude ionosphere was initially explained by Dungey's
(1961) reconnection cycle. Dungey (1961) stated that
during periods of southwards interplanetary magnetic
®eld (IMF), reconnection at the dayside would lead to
the creation of open magnetic ¯ux. This open ¯ux would
convect over the poles and then be destroyed by
reconnection on the nightside. This circulation of ¯ux
was originally assumed to be a steady-state phenome-
non, with the rate of creation and destruction of open
¯ux being equal at any one instant of time.
In 1972, Russell sketched the ionospheric ¯ows
resulting from a non-steady substorm cycle; reconnec-
tion on the dayside and in the tail became viewed as two
separate time-dependent processes resulting in non-
steady plasma ¯ows in the high-latitude ionosphere
(Russell and McPherron, 1973; Siscoe and Huang, 1985;
Lockwood et al., 1990). Dayside reconnection events are
often categorised as either continuous (quasi-steady) or
pulsed (impulsive) (Russell and Elphic, 1978, 1979;
Elphic and Russell, 1979). Controversy surrounds the
exact nature of these two processes, and the dierences,
if any, between them. In this study we are primarily
interested in the ground-based signatures of impulsive
dayside reconnection.
Impulsive dayside reconnection is believed to be the
primary mechanism for the transfer of ¯ux from the
Earth's magnetosheath to the magnetosphere and epi-
sodes of such ¯ux transfer are referred to as ¯ux transfer
events (FTEs) (Russell and Elphic, 1978, 1979). Much
theoretical work has been developed to explain and
model FTEs at the dayside magnetopause and the
resulting plasma ¯ows created in the high-latitude
ionosphere (Paschmann et al., 1982; Siscoe and Huang,
1985; Cowley, 1984; Southwood, 1985,1987; Lee and
Fu, 1985). However, currently there are insucient
high-time and high-spatial resolution observations of
FTEs and high-latitude convective ¯ows to verify the
various theoretical models.
The ®rst observation of impulsive dayside reconnec-
tion was published by Haerendel et al. (1978) using data
from the HEOS-2 satellite. Russell and Elphic (1978,
1979), analysed data from the ISEE -1 and -2 satellites
and oered a detailed description of ¯ux transfer events
`in which reconnection starts and stops in a matter of
Correspondence to: G. Provan
Ann. Geophysicae 16, 1411±1422 (1998) ÓEGS ± Springer-Verlag 1998
minutes or less, resulting in the ripping o of ¯ux tubes
from the magnetosphere', and oered a theoretical
model for the observed magnetic signatures. The FTEs
were observed between 9 and 15 MLT and during each
observed FTE the magnetosheath ®eld had a southward
B
z
component. Estimated calculations from the satellite
data suggested that the ionospheric signature of ¯ux
transfer events should have scale size of about 200 km,
while the recurrence rate of the FTE events should be 7±
10 min.
The ®rst ground-based observations of ¯ux transfer
events were published by Van Eyken et al. (1984) and
Goertz et al. (1985). Goertz used data from the STARE
radar (a VHF coherent scatter radar sounding the E
region ionosphere, Greenwald et al., 1978). They detected
a convection boundary in the radar's ®eld of view.
Poleward of this boundary the ¯ow was observed to be
antisunward, with an occasional signi®cant north-south
component. Flows across the convection boundary
occurred sporadically in spatially localised regions with
scale sizes between 50±300 km, and with a repetition rate
of the order of minutes. In one particular event, Goertz et
al. (1985) reported observing a poleward moving tran-
sient which grew and decayed over a 4 min time interval.
Goertz and co-workers (1985) interpreted the transient as
the result of newly reconnected ¯ux tubes being pulled
poleward by the magnetosheath ¯ow. The event occurred
immediately after a turning of the east-west (B
y
) compo-
nent of the IMF from positive to negative.
Rapid velocity transients are commonly observed in
the line-of-sight velocity measurements detected by the
CUTLASS Finland radar. These features are detected a
few hours either side of local noon, at approximate
magnetic latitudes of between 70°to 80°, and are found
to be moving away from the radar. They are interpreted
here as the ionospheric signatures of FTEs. Conven-
tional theories of FTEs predict that, as newly opened
¯ux tubes are pulled anti-sunward by the magnetosheath
¯ow, ground-based instruments should be able to detect
the ionospheric end of the ¯ux tube moving poleward
(Goertz et al., 1985), as long as plasma ¯ows are
generated which are enhanced relative to the back-
ground ¯ow (Southwood, 1987, 1989). We have studied
a single interval of data detected by the CUTLASS
Finland radar in the pre-noon sector of the 14 August,
1995. During this interval poleward moving velocity
transients were observed in the upper range gates of the
radar beams, seemingly enhancing the convective plas-
ma ¯ow. The CUTLASS radar was operating in special
scan mode on this day to optimise temporal resolution
while maintaining good spatial resolution. The scan
mode employed has allowed characterisation of the
ionospheric signatures of FTEs and the resulting
enhanced convective ¯ow in more detail than has been
possible in previous studies.
2 Instrumentation
The ionospheric convection velocities in this study are
provided by the CUTLASS Finland radar. CUTLASS
is a bistatic HF coherent radar, with stations in
Finland and Iceland, and forms part of the interna-
tional SuperDARN chain of HF radars (Greenwald et
al., 1995). Each radar of the system is a frequency agile
(8±20 MHz) radar, routinely measuring the line-of-
sight (l-o-s) Doppler velocity and spectral width of, and
the backscattered power from, ionospheric plasma
irregularities. The radars each form 16 beams of
azimuthal separation 3.24°. Each beam is gated into
75 range bins, each of length 45 km in standard
operations, when the dwell time for each beam is 7 s,
giving a full 16 beam scan, covering 52°in azimuth and
over 3000 km in range (an area of over 4 ´10
6
km
2
),
every 2 min. Common-volume data from the two
stations can be combined to provide convection
velocities perpendicular to the magnetic ®eld, although
no common volume backscatter was detected during
the interval under study here. During the interval under
study here the CUTLASS Finland radar was operating
in a non-standard scan mode. In this mode, instead of
the usual anti-clockwise sweep through the beams, the
radar was con®ned to sampling on beams 5 and 12
only, i.e. a sequence of 5, 12, 5, 12 etc. This resulted in
a 14 s time resolution for each beam, with a 45 km
spatial resolution along the beam. At the higher range
gates where the data in this study was observed, beam
5 lies to the west of the magnetic meridian while beam
12 lies to the east, there is approximately 34°between
the two beams. This dataset is supplemented by
upstream IMF data from the WIND satellite (Lepping
et al., 1995).
3 Data presentation
3.1 Radar data
Figure 1 shows the latitude-time-velocity (LTV) plots
for beams 5 and 12 between 0500 and 1010 UT on the
14 August, 1995. The Doppler velocity is colour coded
with positive velocities indicating motion towards the
radar and negative velocities indicating motion away
from the radar, while grey indicates backscatter from
the ground. Beam 5 detects ionospheric backscatter
between the magnetic latitudes of 72°to 79°(altitude
corrected geomagnetic, AACGM, co-ordinates, based
on PACE geomagnetic co-ordinates, Baker and Wing,
1989), while beam 12 observes backscatter between 73°
and 81°.
Figure 1 illustrates that numerous distinct areas of
backscatter are detected by the radar beams, separated
by intervals of little or no scatter. The areas of
backscatter are observed in general in beam 12 before
beam 5, the ®rst region of backscatter being observed in
beam 12 from 0531 until 0653 UT and in beam 5 from
0534 until 0657 UT. During the ®rst half hour of the
interval only poleward l-o-s velocities are observed by
beam 12, while beam 5 detects slower equatorward l-o-s.
velocities. From 0600 to 0700 UT, poleward line-of-
sight velocities are detected by beam 5 at high ranges
but equatorward velocities at lower ranges. The l-o-s
1412 G. Provan et al.: CUTLASS Finland radar observations of the ionospheric signatures
velocities observed by beam 12 for this time interval are
oppositely directed to beam 5: here the l-o-s velocity is
directed equatorward in the higher range gates and
poleward in the lower. From 0700 to 0800 UT,
backscatter containing equatorward l-o-s velocities is
seen in the higher ranges of beam 12 which is pointing
towards the local noon sector, while beam 5 continues to
observe positive l-o-s velocities at lower ranges and
negative l-o-s velocities at higher ranges. From 0850 to
1010 UT beam 5 observes negative l-o-s velocities, with
backscatter predominantly located at higher latitudes of
between 74.6°and 81°. Beam 5 is at noon MLT at 1000
UT, whereas, beam 12 is in the early afternoon sector
for this interval and does not observe any ionospheric
backscatter at all. WIND data presented in panels 3 and
4 of Fig. 1 will be discussed in sec 3.2.
Figure 2 is a schematic representation of the ¯ows
detected by the two radar beams during the interval
0600 to 0700 UT. The small arrows depict the direction
of the l-o-s velocity observed by each beam, while the
large arrows indicate the direction of convection ¯ows
consistent with the observed velocities. The backscatter
is detected in the pre-noon sector and indicates west-
ward, anti-Sunward ¯ow at higher ranges and eastward,
Sunward ¯ow at lower ranges, with a convection
reversal boundary (CRB) in-between. Plasma velocity
vectors have been calculated by a beam-swinging
technique. The technique assumes that the transient
signature is of a scale size such that there is spatial
uniformity between the two beams. Considering the
angle between the two beams and the magnetic merid-
ian, the l-o-s velocity observed at the two beams were
Fig. 1 Latitude-time-velocity
(LTV) plots for beams 5 and 12
between 0500 and 1010 on the 14
August 1995. The Doppler ve-
locity is colour-coded with posi-
tive velocities indicating motion
towards the radar and negative
velocities away from the radar.
The diagram also shows the B
y
and B
z
components of the IMF
for the time interval 0500 to 1010
UT, measured by the WIND
satellite. The calculated lag time
between when the IMF is ob-
served by the satellite and when
it impinges on the magnetopause
has been added onto the time-
scale of the IMF plots. Four
dashed lines mark the point of
signi®cant change of the B
y
component
G. Provan et al.: CUTLASS Finland radar observations of the ionospheric signatures 1413
combined to gain a value of the poleward and azimuthal
velocity components of the plasma ¯ow. Poleward of the
CRB, the meridional velocity component of the plasma
¯ow during the transient features is 2 1 km s
)1
in a
westward azimuthal direction and 0.5 0.25 km s
)1
in
an equatorward direction.
The convection reversal boundary was detected by
one or both radar beams during most of the interval
0600 to 0830 UT. At 0610 UT the equatorward edge of
the ¯ow reversal boundary was at a magnetic latitude of
77°in beams 5 and 12, by 0650 UT the CRB had moved
to a magnetic latitude of 72.5°in beam 5 and to a
magnetic latitude of 74.5°in beam 12, this represents an
average equatorward motion of the boundary of 180 m
s
)1
. Little or no backscatter was detected by the two
beams between 0650 and 0705 UT. At 0705 UT, the
equatorward edge of the ¯ow reversal boundary can be
detected at a latitude of 74°in beam 5 and at a latitude
of 75.5°in beam 12. The CRB then moves poleward,
until 0730, when the ¯ow reversal boundary is at a
magnetic latitude of 76°in beam 5 and 77.5°in beam 12.
This poleward motion of the ¯ow reversal boundary
occurs at an average velocity of 220 m s
)1
. At 0730, the
¯ow reversal boundary starts moving equatorward
again at a velocity of 170 m s
)1
.
Figure 3 presents an LTV plot for the two beams for
the time interval 0600 to 0700 UT. Equatorward of the
¯ow reversal boundary backscatter is continuous, the
magnitude of the l-o-s velocity having a quasi-periodic
nature. This is especially noticeable in the l-o-s velocity
of beam 12 where the velocity reaches a maximum of
between 600 and 800 m s
)1
with a periodicity of 10 min
or so. Poleward of the CRB both radar beams observe a
number of high velocity transients. Some of these
transients have been highlighted in Fig. 3, with over-
plotted black and white lines tracing the motion of
individual transients. Three transients which are well
de®ned in both beams have been labelled A, B and C in
Fig. 3.
The transients appear as pulsed red stripes in the
LTV plot of beam 5 and pulsed blue stripes in the LTV
plot of beam 12. The transients are initially observed
approximately 2 min 30 s earlier in beam 12 com-
pared to beam 5 at an estimated 0.5°±1°higher magnetic
latitude, in both beams the transients appear to be
moving away from the radar. The transients have an
average repetition frequency of 7±8 min, and there is no
backscatter poleward of the CRB in between these
transients. The motion of the transients with respect to
the two beams can be determined from a study of the
`gradient' of the transients in the LTV plots, represent-
ing their motion to higher range gates with time. The
transients have an almost vertical gradient in beam 12,
but a much more oblique gradient in beam 5. The
transients are observed in both beams simultaneously
for an estimated two minutes.
Figure 4 shows the latitude-time-spectral-width plot
for the two beams between 0600 and 0700 UT. The
spectral width is colour-coded with red indicating the
largest spectral width values and dark blue the smallest.
The plot suggests that the largest spectral width values
are detected at range gates furthest away from the radar
where the backscatter has a pulsed nature, while nearest
range gates where the backscatter is continuous have the
smallest spectral widths.
3.2 WIND data
Figure 1 shows the B
z
(north-south) and B
y
(east-west)
components of the IMF for the time interval 0500 to
1010 UT, measured by the WIND satellite. The satellite
was located near X
gsm
75 R
E
, Y
gsm
6.9 R
E
and
Z
gsm
)7.2 R
E
. A 13 1 min lag-time between when
the IMF is observed by the satellite and when it
impinges on the subsolar magnetopause has been
calculated using the method outlined by Lester et al.
(1993). This time lag has been added onto the time scale
of the IMF plots, allowing for direct comparison
between these and the LTV plots. Four vertical dashed
lines mark the times intervals of signi®cant changes in
the IMF B
y
component.
From 0600 to 0830 UT the B
y
component is generally
very stable at about 7 nT with a few negative excursions,
while the B
z
component ¯uctuates. Comparing the IMF
components with the observed l-o-s velocities some clear
correlations can be noted. The B
y
component changes
abruptly from )5 nT to 3 nT at approximately 0545,
while B
z
turns positive. This corresponds closely to the
time of initial backscatter observations by the two radar
beams, but no pulsed transients are observed poleward
of the CRB until 0600 UT when B
y
increases to 7 nT,
while B
z
turns negative.
At 0647 UT, the B
y
component of the IMF drops
from 7 nT to )2 nT and does not return to its previous
value before 0705, simultaneously the B
z
component
changes from )4 nT to 7 nT and also remains at this
Fig. 2 Schematic representation of the ¯ows detected by the two
radar beams during the interval 0600 to 0700 UT. The small arrows
depict the direction of l-o-s velocity observed by each radar beam,
while the large arrows point in the direction of convection ¯ows
consistent with the observed velocities
1414 G. Provan et al.: CUTLASS Finland radar observations of the ionospheric signatures
value until 0705. During these 20 min little or no
backscatter is observed by the two radar beams. Pulsed
transients are again observed by the two beams between
0705 and 0800 UT. No transient features are then
observed by beam 12 after 0800 UT (1230 MLT), while
beam 5 continues to observe backscatter until 0830 UT
when the B
y
component of the IMF decreases from 6 nT
to )4 nT, while B
z
increases from )6 nT to 5 nT. No
transients are then observed by beam 5 until 0848 UT, at
this time B
y
is 4 nT while B
z
is again negative. Beam 5
then observes backscatter seemingly moving away from
the radar, separated by intervals of no backscatter until
1010 (1210 MLT), during this time interval B
y
is mainly
positive and ¯uctuating around 3 nT while B
z
is
predominantly below )4 nT.
The ¯ow reversal boundary moves equatorward
during the time interval 0605±0650 and 0730±0800 UT.
During these intervals B
y
is predominantly stable at
around 6±7 nT, while B
z
, although ¯uctuating between
positive and negative values during the ®rst interval, is
increasingly negative. From 0710 to 0730 UT the ¯ow
reversal boundary is moving in a poleward direction,
again B
y
is approximately 7 nT, while B
z
has low
positive values (between 0 and 2 nT).
4 Discussion
A convection reversal boundary is detected in the
CUTLASS radar data, poleward of the CRB the plasma
¯ows in an westward (anti-Sunward) direction while
equatorward of the CRB the plasma ¯ow is in an
eastward (Sunward) direction. Equatorward of the CRB
the observation of backscatter is constant, the magni-
tude of the l-o-s velocity having a quasi-periodic nature.
Poleward of the CRB, pulsed backscatter moving away
from the radar is observed with a repetition rate of
about 7 min. During the intervals of pulsed backscatter
the CRB appears to move to lower latitudes. The
direction of plasma ¯ow either side of the CRB is
consistent with the radar observing a typical southward
IMF dawn convection cell with plasma rotating in an
Fig. 3 The LTV plot for beams 5
and 12 for the time interval 0600
to 0700 UT on the 14 August 1995.
Overplotted black and white lines
trace the motion individual tran-
sients. Three transients which are
well de®ned in both beams have
been labelled A,Band C
G. Provan et al.: CUTLASS Finland radar observations of the ionospheric signatures 1415
anti-clockwise direction (Dungey, 1961). The convection
cell is observed for approximatley four and a half hours
of univeral time, from 0530 UT when it is ®rst detected
by beam 5, until 1010 UT, this corresponds to radar
coverage of approximately seven hours of magnetic local
time. When the convection cell is ®rst observed the far
ranges of beam 5 are located at 0730 MLT, while the
far ranges of beam 12 are observing at approximately
1000 MLT. At 1000 the far ranges of beam 5 are located
at 1200 MLT, and the far ranges of beam 12 are
located at 1430 MLT.
During the intervals when transients were observed
by the radar beams, they had an approximate average
recurrence rate of 7 to 8 min, this is close to the average
recurrence rate of ¯ux transfer events ®rst reported by
Rijnbeek et al. (1984). It is our belief that the transients
observed poleward of the CRB are the ionospheric
signature of ¯ux transfer events at the magnetopause
boundary. Continuous backscatter observed equator-
ward of the CRB is due to return ¯ow on closed ®eld
lines, the periodic enhancements of the l-o-s velocity
being a response to the transient nature of reconnection
at the subsolar point. Poleward of this radar observa-
tions are not continuous, with the inferred dawn-cell
convection pattern only detected by the radar when the
conditions are met for the generation of F region
irregularities from which the radar backscatters. The
excellent spatial information oered by the two-beam
radar scan allows a detailed investigation of the spatial
and temporal characteristics of the ionospheric signa-
tures of FTEs. This interpretation of the radar data is
presented in more detail later.
Individual transients are initially observed on average
2 min 30 s earlier in beam 12 than in beam 5, and
are detected in both beams simultaneously for 2 min.
Beam 12 is at a later MLT than beam 5 and is therefore
closer to local noon during the earlier part of the
interval studied. The transients are initially observed
approximately 1°of magnetic latitude nearer the pole in
beam 12 compared to beam 5. At the latitudes that the
transients are observed the two beams have an azimuth-
al separation of between 750 and 950 km. As the
Fig. 4 Latitude-time-spectral-
width plot for beams 5 and 12
between 0600 and 0700 on the 14
August 1995. The spectral width is
colour-coded with red velocities
indicating the largest spectral
width and dark blue representing
the smallest spectral width
1416 G. Provan et al.: CUTLASS Finland radar observations of the ionospheric signatures
transient signature takes 2 min 30 s to travel from
one beam to the next, the transients have an azimuthal
velocity component of 7.5 3 km s
)1
and a velocity
component towards the radar of 0.8 0.4 km s
)1
. The
transient signatures extend 1900 900 km in an azi-
muthal direction and 250 km in a poleward direction
(between 5±6 range gates). The estimated average
magnetic latitude at which the transients were observed
was 77°, one hour of MLT at a latitude of 77°is equal to
approximately 375 km in an azimuthal direction. So the
transient signature covers approximately 5 2.5 hours
of MLT. Previous work by Elphic et al. (1990), using
data from the ISEE 1 and 2 satellites in conjunction with
data from the EISCAT radars and photometers and an
all-sky camera, have inferred that the ionospheric
footprint of FTEs extended between 400±1000 km in
the east-west direction and between 100±200 km in the
north-south direction. Comparing these results with the
size of the ionospheric footprint of FTEs observed by
us, we can see that although the scale sizes observed by
Elphic et al. (1990) are smaller in both the north-south
and east-west extent, the two sets of results are in
agreement within the large error ranges given.
As the transient signature can be observed by both
beams for several minutes, we have used a beam-
swinging technique, which assumes spatial uniformity
between the two beams, to calculate that the plasma
¯ow during the transient features have an azimuthal
velocity component. The equatorward component of the
plasma ¯ow velocity during the transient features is
0.5 0.25 km s
)1
, while the azimuthal component is
2 1 km s
)1
. The error range for these velocities is so
large mainly because of diculty in tracing the position
of the HF radar beams at very high latitude. Since the
two beams have are separated by 34°it is unlikely that
the region between them is totally uniform, leading to
further uncertainty in the measured velocities. However,
this is an inherent problem of the beam-swinging
technique and it is impossible to estimate the scale of
these errors. The equatorward velocity component of
the transients is equal to the equatorward velocity of the
plasma ¯ow, within the given error range. The azimuthal
velocity component of the transient would appear to be
somewhat greater than the azimuthal velocity compo-
nent of the plasma ¯ow.
The `gradient' that the transients make in the
latitude-time-velocity (LTV) plots of the two beams,
suggest that the transients move almost perpendicular to
beam 12 while moving away from the radar in beam 5.
Taking these mentioned factors into account, it is clear
that the transient features move in a predominantly
westward azimuthal direction, being detected in beam 12
before beam 5. Between the two beams the transients
have a small velocity component towards the radar,
however it would appear that the poleward component
of motion of the transients changes direction, as the
transients appear to be moving away from the radar
when observed in beam 5. The motion of the transients
seem best described as being along a curved trajectory
on open ®eld lines, possibly consistent with motion
along a notional L shell.
Figure 5 is a polar plot showing the position of the
radar beams with respect to geomagnetic latitude at
0630 UT. Overlaid is a schematic illustration of our
interpretation of the motion of the ¯ux transfer event
and related plasma convection ¯ow. The elliptical `blob'
represents the patch of irregularities probably associated
with soft particle precipitation on newly opened ®eld
lines created by the ¯ux transfer event. The black line
marks the open/closed ®eld line boundary, while the
grey-hatched zone represents the region of continuous
backscatter observations. The newly opened ®eld lines
initialise the convective ¯ow of plasma. Equatorward of
the opened/closed ®eld line boundary, steady energetic
electron precipitation associated with the low-latitude
boundary layer is known to produce continuous back-
scatter (Baker et al., 1995), allowing us to observe
continuously the modulated velocity within the convec-
tive plasma ¯ow. Poleward of the boundary, the area of
newly opened ¯ux will move predominantly azimuthal-
ly, in a direction almost perpendicular to beam 12.
Precipitation from this patch will create irregularities
from which the radar can backscatter, highlighting the
background westward convective ¯ow associated with
plasma being dragged anti-Sunward. The net eect of a
period of ¯ux transfer events is the motion of the open/
closed ®eld line boundary to lower latitude. The
direction of motion of the transient signature, the
creation of return ¯ow on closed ®eld line and the
motion of the open/closed ®eld line boundary to lower
latitudes is very similar to the model of dayside
reconnection suggested by Cowley et al. (1992) and
Cowley and Lockwood (1992). Similar observations of
high-latitude plasma ¯ow variations accompanied by
bursts of return ¯ow at lower latitudes have previously
Fig. 5 Polar plot showing the position of the radar beams with
respect to geomagnetic latitude at 0630 UT. Overlaid is a schematic of
the most likely motion of the poleward moving transients in the polar
cap and the region of transient velocity response in the return ¯ow
G. Provan et al.: CUTLASS Finland radar observations of the ionospheric signatures 1417
been observed by Moen et al. (1995), using satellite data
combined with optical observations and data from the
EISCAT radar.
Baker et al. (1990, 1995) have identi®ed the HF radar
signature of the low-latitude boundary layer (LLBL)
and the cusp, using the spectral width distribution of
backscattered l-o-s velocities observed by the Halley HF
radar and low-latitude particle precipitation data from
the DMSP spacecraft. They stated that the distribution
of spectral widths in the cusp is generally of a Gaussian
form with a peak at 220 m s
)1
, while the distribution of
spectral widths in the LLBL is more likely to be an
exponential distribution, with the peak of the distribu-
tion occurring at 50 m s
)1
. Figure 6 shows a plot of the
spectral width distribution of the positive and negative
line-of-sight velocities for beams 5 and 12 between 0500
and 0800 UT. At lower range gates positive (negative)
velocities observed by beam 5 (beam 12) have a spectral
width distribution with a modal value of approximately
300 (250) m s
)1
, while at higher range gates beam 5
(beam 12) negative (positive) velocities have a distribu-
tion with a modal value of about 350 (350) m s
)1
.
Comparing Baker's results with the spectral width
distribution of the l-o-s velocities observed by the
CUTLASS radar between 0500 and 0830 UT, both
poleward and equatorward of the ¯ow reversal boun-
dary, it can be seen that the distributions are remarkably
similar. Both radar beams observed a spectral width
distribution of a general Gaussian form poleward of the
¯ow reversal boundary. The form of the spectral width
distributions observed equatorward of the CRB for both
beams was similar to the shape of the spectral width
distribution Baker et al. (1995) identi®ed in the LLBL
region, although the modal values of the spectral width
distributions observed by the CUTLASS radar are
somewhat higher than those detected by Baker et al.
(1995).The shape of the spectral width distributions are
so similar that we feel con®dent that, based on the Baker
de®nition, we can identify the polar cap boundary in the
pre-noon sector in the CUTLASS f.o.v, with anti-
Fig. 6 The spectral width distribution of the positive and negative line-of-sight velocities for beams 5 and 12 between 0500 and 0800 UT. In
general the negative (positive) velocities are detected poleward (equatorward) of the CRB in beam 5 and are oppositely directed in beam 12
1418 G. Provan et al.: CUTLASS Finland radar observations of the ionospheric signatures
Sunward ¯ows on open ®eld lines within the polar cap
and Sunward ¯ows, at lower latitudes on closed ®eld
lines. The FTEs add open ¯ux to the polar cap which
enhances the overall convective ¯ow.
Drawing conclusions about the eect of the IMF B
y
and B
z
components on the observed FTE signatures and
the background convection pattern is complicated in
this case study, as a change in one component is often
accompanied by a variation in the other. This is the case,
for example, in the interval 0645 to 0705 UT, when a
sudden decline in the amount of backscatter detected, is
accompanied by a drop in the B
y
component of the IMF
(from 7 nT to )2 nT) and also by a sudden increase in
the B
z
component of the IMF ( from )4 nT to 7 nT).
However, since the changes in the B
y
and B
z
components
correspond so well with a variation in the radar
observations, this is con®rmation that any error in the
calculated lag-time of the IMF between the WIND
satellite and the Earth is within the given margins.
During the ®rst two intervals, when poleward moving
transients are detected, the B
y
component is predomi-
nantly above 6±7 nT (0600±0645 and 0705±0830 UT)
while the B
z
component ¯uctuates greatly between
positive and negative values. During the third interval
of transient observations (0845 to 1010 UT) the B
y
component ¯uctuates around 3 nT while B
z
is mainly
below 4 nT. The motion of the FTEs in a westwards
azimuthal direction can be explained by the Svalgaard-
Mansurov eect (Smith and Lockwood, 1996, and
references therein), where a positive IMF B
y
component
in the Northern Hemisphere exerts a curvature force on
magnetic ®eld lines, forcing them to move towards
dawn. No FTE signatures are observed during extended
intervals of high positive B
z
(0645±0705 UT, B
z
7 nT),
but dayside reconnection does occur during intervals of
low, positive B
z
(0705±0730 UT, B
z
0±2 nT).
The motion of the ¯ow reversal boundary adheres to
conventional reconnection theory (Dungey, 1961, 1963)
and can be explained by the model of dayside recon-
nection proposed by Cowley et al. (1992) and Cowley
and Lockwood (1992). During intervals of predomi-
nantly negative B
z
(0605±0647 UT and 0730±0800 UT),
the ¯ow reversal boundary moves equatorward. While
during intervals of positive B
z
(0710±0730) the ¯ow
reversal boundary moves poleward. This is consistent
with newly opened ¯ux being added to the polar cap due
to dayside reconnection during intervals of negative B
z
,
causing the polar cap boundary to expand, whilst during
intervals of positive B
z
, reconnection on the nightside
leads to the destruction of open ¯ux, and the polar cap
shrinks.
4.1 Comparison with previous studies
The FTEs were created during a period of mainly
positive B
y
. The ionospheric signature of the FTEs is
similar to the theoretical model predicted by Cowley
et al. (1991,1992) and Cowley and Lockwood (1992) for
a period of positive B
y
in the northern hemisphere.
Cowley and co-workers proposed that the FTEs would
create a convection pattern in the ionosphere at the
front of the polar cap boundary. Initially, the convec-
tion pattern would move azimuthally at an approximate
velocity of 2 km s
)1
for six minutes, resulting in a total
east-west displacement of about 700 km. As the system
moved towards its equilibrium con®guration, the mainly
azimuthal motion of the patch will give way to poleward
motion and a twin vortical ¯ow will be excited over the
auroral zone. The time scale for the excitation and decay
of this ¯ow pattern is predicted to be about 10±15 min.
The time scale, direction of motion and velocity of
the ionospheric signature of the ¯ux transfer events
observed by the CUTLASS radar correspond well to the
theoretical model proposed by Cowley and co-workers
(Cowley et al., 1991, 1992; Cowley and Lockwood,
1992). The CUTLASS radar also observed return ¯ow
on closed ®eld lines, which is central to the conceptual
framework postulated by Cowley and colleagues. There
are a number of points to make about the CUTLASS
observations in view of theoretical model developed by
Cowley and co-workers. First of all, backscatter asso-
ciated with newly opened magnetic ¯ux can only be
observed poleward of the CRB, this is probably because
the patch of newly opened backscatter is observed in the
pre-noon sector, only after it has moved across the
opened/closed ®eld line boundary. Secondly, the tran-
sient signature observed by the CUTLASS radar has an
azimuthal extent of 1900 900 km and a poleward
extent of about 250 km. This is more than twice the size
of the transient signature predicted by Cowley. How-
ever, as the radar is observing the signature in the pre-
noon sector, it is likely that the signature has been
`stretched' by the magnetosheath ¯ow by this stage, as
also predicted by the Cowley model.
Poleward of the CRB the plasma velocity is
2 1 km s
)1
in a westward azimuthal direction and
0.5 0.25 km s
)1
in an equatorward direction. Sand-
holt et al. (1992) observed periodic auroral events at the
midday polar cap boundary, seemingly the ionospheric
signature of ¯ux transfer events. These events moved
eastward near the noon meridian with a velocity of
1.5 km s
)1
. This is within the error range of the
azimuthal component of the plasma ¯ow observed
poleward of the CRB by CUTLASS. The transients
observed by the CUTLASS radar moved with a velocity
of 7.5 3 km s
)1
in an azimuthal direction and with a
velocity of 0.8 0.4 km s
)1
towards the radar. Cowley
and colleagues (Cowley et al., 1991, 1992; Cowley and
Lockwood, 1992) predicted that the velocity of the area
of newly opened ¯ux would be the same as the velocity
of the convective ¯ow. The CUTLASS observations
indicate that the transients' velocity appears to be larger
than the velocity of the plasma ¯ow, especially in the
azimuthal direction. It is possible that the transients are
actually moving with the same velocity as the convective
¯ow and that any discrepancy between the two mea-
sured velocities is due to the size of the error margins
being underestimated. Alternatively, it may be that the
observed transients are not the ionospheric footprints of
stable, newly opened ®eld lines but are created by
precipitation due to active reconnection still taking place
G. Provan et al.: CUTLASS Finland radar observations of the ionospheric signatures 1419
at the magnetopause boundary. The observed velocity
of the transients would then not be the plasma velocity
but a phase velocity in the ionosphere. However at the
present time it is not possible to de®nitively verify
whether the observed transients are signatures of newly
opened ®eld lines or the signatures of ®eld lines on which
reconnection is still active.
Pinnock et al. (1995) and Rodger and Pinnock (1997)
used data from the Halley PACE HF radar to study the
ionospheric signature of ¯ux transfer events, speci®cally
the early evolution of these transients. They reported
observations of poleward moving high velocity features
very similar to those observed by the CUTLASS radar
and identi®ed them as ¯ux transfer events. They
concentrated especially on the early evolution of the
transient signature in relation to the theory of FTEs.
The transient features increased the l-o-s Doppler
velocity by up to 0.5 km s
)1
and were superimposed
upon a background poleward ¯ow of about 300 m s
)1
.
The transients themselves were observed to have broad
spectral widths and were ®rst observed within regions of
low spectral width (identi®ed as the LLBL), up to three
degrees equatorward of a region of high spectral width
(identi®ed as the cusp), into which they propagated over
a 10 min period. The boundary between the low and
high spectral width regions was moving equatorward at
a velocity of 82 m s
)1
. No IMF data was available for
the interval studied.
Both the CUTLASS and Halley radars observe
transient signatures, indicative of FTE activity, moving
away from the radar and superimposed upon back-
ground convection ¯ow, but there are some fundamental
dierences between the CUTLASS and Halley results.
Firstly, whereas the transients signatures indicative of
open ®eld lines are observed by the Halley radar a few
degrees equatorward of the CRB and propagate to a few
degrees poleward of it, the transients observed by the
CUTLASS radar are only detected poleward of the
CRB. Secondly, whereas the background convective
¯ows observed by Halley were directed poleward either
side of the boundary between high and low spectral
widths, the ¯ows detected by CUTLASS were directed
anti-Sunward on the poleward side of the CRB and
Sunward on the equatorward side. Thirdly, Pinnock et
al. (1995) suggested that the transients were moving
poleward, whereas observations made by the CUTLASS
radar shows that the motion of the transients is
predominantly azimuthal.
In comparing the present observations with those
made by Pinnock et al. (1995) we did consider whether
the quasi-periodic nature of the magnitude of the l-o-s
velocity observed equatorward of the CRB, especially in
beam 12, was due to the regions of open ¯ux propagat-
ing poleward. However, studying the spectral widths of
these regions of increased l-o-s velocity in Fig. 4, we can
conclude that these features do not have an increased
spectral width compared to the surrounding area and
are most likely caused by the enhancement of the return
¯ow due to the addition of open ¯ux. They are thus
most probably a response to the FTE rather than a
direct signature of it.
It is possible to explain the ionospheric signatures of
FTEs made by the Halley and CUTLASS radars within
the same theoretical framework. The observations by
Pinnock et al. (1995) were made between 10 and 13
MLT, close to local noon, while CUTLASS made its
observations in the pre-noon sector. Therefore, Halley
would observe the area of newly opened ¯ux when it
had just been added onto the polar cap boundary but
was still equatorward of the main CRB, whereas we
have already mentioned that CUTLASS did not
observe the regions of newly opened ¯ux until after
they crossed the opened/closed ®eld line boundary. As
Halley is located in the noon sector, it is likely that the
radar is looking down the convection throat at anti-
Sunward ¯ow. The radar would thus observe the
boundary between high and low spectral widths, with
the l-o-s velocities directed in the same direction either
side of the boundary. Similar transient signatures
moving in a poleward direction only, are observed by
CUTLASS beam 5 between 0850 and 1010 UT in
Fig. 1. During this period beam 5 is observing data
between 11±12 MLT. The transients observed in beam
5 during this period is then a result of the radar looking
down the convection throat at anti-Sunward ¯ow.
Finally, the transients observed by the Halley radar
seemed to be moving poleward, due to the fact that the
radar was only observing in a high-time resolution
mode on one beam. The CUTLASS radar was observ-
ing on two beams. When observing the transients in
each beam separately, they appear to be moving
poleward. However, combining the data from both
beams shows that the transients were moving in a
predominant azimuthal direction.
5 Summary and conclusion
The CUTLASS radar, run in a two-beam high-time
resolution mode, has been able to oer unrivalled spatial
and temporal information on plasma ¯ows in the high-
latitude ionosphere. The time-averaged radar signatures
are typical of a dawn-cell convection pattern. Equator-
ward of the CRB backscatter is observed continuously
with the magnitude of the l-o-s velocity having a quasi-
periodic nature, while poleward of the boundary pulsed
transients are observed. We believe that these pulsed
transients are the ionospheric signature of ¯ux transfer
events. The two-beam mode allows us to make the
following observations on FTEs and the related ¯ows in
the ionosphere:
1. The transients had a recurrence rate of between 7 to 8
minutes.
2. The ionospheric signature of the FTEs have an
approximate azimuthal extent of 1900 900 km
and an estimated poleward extent of 250 km.
3. The motion of the transients is predominantly
azimuthal, the transients move in an westward
azimuthal direction with a velocity of 7.5 3 km
s
)1
and towards the radar with a velocity of
0.8 0.4 km s
)1
.
1420 G. Provan et al.: CUTLASS Finland radar observations of the ionospheric signatures
4. Poleward of the CRB plasma velocity is 2 1 km
s
)1
in a westward azimuthal direction and 0.5
0.25 km s
)1
in an equatorward direction.
5. Although the velocity of the transients is almost equal
to the velocity of the plasma ¯ow within the large
error range given, it would appear that the velocity of
the transients is somewhat larger than the velocity of
the convective ¯ow. It is possible that we are not
actually observing the ionospheric footprints of newly
opened ®eld lines, but precipitation from regions of
still-active reconnection at the magnetopause boun-
dary.
6. The magnitude of the l-o-s velocity equatorward of
the CRB has a quasi-periodic nature, reaching a
maximum value every 10 min or so. This period is
approximately the same as the recurrence rate of the
FTEs. It would appear that the addition of open ¯ux
at the subsolar point leads to an detectable enhance-
ment in the magnitude of the return ¯ow
7. Using a spectral width criterion for identifying the
HF signature of the cusp and LLBL (Baker et al.,
1995), we can identify the region of pulsed anti-
Sunward convecting backscatter as mapping to open
®eld lines and the region of continuous Sunward
convecting backscatter as mapping to closed ®eld
lines.
8. Our results suggest that the ionospheric signature of
FTE were only detected by the CUTLASS radar
during intervals when the B
y
component of the IMF
was greater than 3 nT, and while the B
z
component
was less than 2 nT. The problem of deconvolving the
separate eects of the B
y
and B
z
components of the
IMF on the observed FTEs and background convec-
tion pattern, means that these results will have to be
viewed as initial ones, further research on this topic
will follow.
9. Further work using greater temporal and spatial
resolution measurements from the CUTLASS radar
will we hope in the future be able to make unambig-
uous measurements of the velocities of plasma con-
vection and FTE signatures in the high-latitude
ionosphere.
Acknowledgements. Topical Editor D. Alcayde' thanks J. Moen
and R. C. Elphic for their help in evaluating this paper.
References
Baker, K. B., and S. Wing, A new magnetic coordinate system for
conjugate studies at high latitudes, J. Geophys. Res., 94, 9139±
9142, 1989.
Baker, K. B., R. A. Greenwald, J. M. Ruohoniemi, J. R. Dudeney,
M. Pinnock, P. T. Newell, M. E. Greenspan, and C.-I. Meng,
Simultaneous HF-radar and DMSP observations of the cusp,
Geophys. Res. Lett.,17, 1869, 1990.
Baker, K. B., J. R. Dudeney, R. A. Greenwald, M. Pinnock, P. T.
Newell, A. S. Rodger, N. Mattin, and C.-I. Meng, HF radar
signatures of the cusp and low-latitude boundary layer,
J. Geophys. Res., 100, 7671, 1995.
Cowley, S. W. H., Solar wind control of magnetospheric convec-
tion, in Achievement of the International Magnetospheric Study,
IMS, p. 483, Noorwijk, Netherlands, 1984.
Cowley, S. W. H., M. P. Freeman, M. Lockwood, and M. F. Smith,
The ionospheric signature of ¯ux transfer events, in Cluster ±
Dayside Polar Cusp, ESA SP-330., 105, 1991.
Cowley, S. W. H., and M. Lockwood, Excitation and decay of solar
wind-driven ¯ows in the magnetosphere-ionosphere system,
Ann. Geophysicae,10, 103, 1992.
Cowley, S. W. H., J. P. Morreli, M. P. Freeman, M. Lockwood, and
M. F. Smith, Excitation and decay of ¯ows in the magneto-
sphere-ionosphere system due to magnetic reconnection at the
dayside magnetopause and in the geomagnetic tail, in Proc.
International Conference on Substorms,ESA SP-335, p. 117,
1992.
Dungey, J. W., Interplanetary ®eld and the auroral zones, Phys.
Res. Lett., 647±648, 1961.
Dungey, J. W., The structure of the ionosphere, or adventures in
velocity space, in Geophysics: The Earth's Environment, Ed. C.
DeWitt, J. Hiebolt and A. Lebeau, p. 526, Gordon and Breach,
New York, 1963.
Elphic, R. C., and C. T. Russell, ISEE-1 and -2 observations of the
magnetopause, in Magnetospheric Boundary Layers,ESA SP-
148, 1979.
Elphic, R. C., M. Lockwood, S. W. Cowley, and P. E. Sandholt,
Flux transfer events at the magnetopause and in the ionosphere,
Geophys. Res. Lett.,17, 2241±2244, 1990.
Goertz, C.K., E. Nielsen, A. Korth, K.-H. Glassmeier, C. Haldoupis,
P. Hoeg, and D. Hayward, Observations of a possible signature
of ¯ux transfer events, J. Geophys. Res., 90, 4069±4078, 1985.
Greenwald, R. A., W. Weiss, E. Nielsen and N. R. Thomson,
STARE a new radar auroral backscatter experiment in northern
Scandinavia, Radio Sci., 13, 1021±1039, 1978.
Greenwald, R. A., K. B. Baker, J. R. Dudeney, M. Pinnock, T. B.
Jones, E. C. Thomas, J.-P. Villain, J.-C. Cerisier, C. Senior,
C. Hanuise, R. D. Hunsucker, G. Sofko, J. Koehler, E. Nielsen,
R. Pellinen, A. D. M. Walker, N. Sato, and H. Yamagishi,
DARN/SuperDARN: a global view of the dynamics of high-
latitude convection, Space Sci. Rev.,71, 761, 1995.
Haerendel, G., G. Paschmann, N. Sckopke, H. Rosenbauer and
P. C. Hedgecock, The frontside boundary layer of the magne-
topause and the problem of reconnection, J. Geophys. Res.,83,
3195, 1978.
Lee, L. C., and Z. F. Fu, A theory of magnetic ¯ux transfer at the
Earth's magnetopause, Geophys. Res. Lett., 12, 105, 1985.
Lester, M., O. de la Beujardie
Áre, J. C. Foster, M. P. Freeman,
H. Lu
Èhr, J. M. Ruohoniemi, W. Swider, The response of the
large scale ionospheric convection pattern to changes in the
IMF and substorms: results from the SUNDIAL 1987 cam-
paign, Ann. Geophysicae,11, 556±571, 1993
Lepping, R. P., et al., The WIND magnetic ®eld investigation,
Space. Sci. Rev.,71, 207±229, 1995.
Lockwood, M., P. E. Sandholt, A. D. Farmer, S. W. H. Cowley,
B. Lybekk and V. N. Davda., Auroral and plasma ¯ow
transients at magnetic noon, Planet. Space. Sci.,38, 973±993,
1990.
Moen, J., P. E. Sandholt, M. Lockwood, W. F. Denig, U. P.
Lùvhaug, B. Lybekk, A. Egeland, D. Opsvik and E. Friis-
Christensen, Events of enhanced convection and related dayside
auroral activity, J. Geophys. Res.,23, 917±934, 1995.
Paschmann, G., G. Haerendel, I. Papamastorakis, N. Sckopke, S. J.
Bame, J. T. Gosling, and C. T. Russell, Plasma and magnetic
®eld characteristics of magnetic ¯ux transfer events, J. Geophys.
Res., 87, 2157±2168, 1982.
Pinnock, M., A. S. Rodger, J. R. Dudeney, F. Rich, and K. B. Baker,
High spatial and temporal resolution observations of the
ionospheric cusp, Ann. Geophysicae,13, 919±925, 1995.
Rijnbeek, R. P., S. W. H. Cowley, D. J. Southwood, and C. T.
Russell, A survey of dayside ¯ux transfer events observed by the
ISEE-1 and -2 magnetometers, J. Geophys. Res.,89, 786±800,
1984.
Rodger, A. S., and M. Pinnock, The ionospheric response to ¯ux
transfer events: the ®rst few minutes, Ann Geophysicae,15, 685±
691, 1997.
G. Provan et al.: CUTLASS Finland radar observations of the ionospheric signatures 1421
Russell, C. T., The con®guration of the magnetopshere, in Critical
Problems of Magnetospheric Physics, Ed E.R. Dyer, National
Academy of Sciences, Washington, D. C.,1972.
Russell, C. T., and R. L. McPherron, The magnetotail and
substorms, Space Sci. Rev.,15, 205, 1973.
Russell, C. T., and R. C. Elphic, Initial ISEE magnetometer
results: magnetopause observations, Space Sci. Rev.,22, 681,
1978.
Russell, C. T., and R. C. Elphic, ISEE observations of ¯ux transfer
events at the dayside magnetopause, Geophys. Res. Lett.,6, 33,
1979.
Sandholt, P. E., J. Moen, and D. Opsvik, Periodic auroral events at
the midday polar cap boundary: implications for solar wind-
magnetosphere coupling, Geophys. Res. Lett.,19, 1223±1226,
1992.
Siscoe, G. L., and T. S. Huang, Polar cap in¯ation and de¯ation,
J. Geophys Res,90, 543±547, 1985.
Smith, M. F., and M. Lockwood, Earth's magnetospheric cusps,
Rev. Geophys;34, 233±260,1996.
Southwood, D. J., Theoretical aspects of ionosphere-magnetops-
here, solar wind coupling, in Physics of the Ionosphere-Magne-
topshere,Adv. Space. Res.,5, 4, 1985.
Southwood, D. J., The ionospheric signature of ¯ux transfer events,
J. Geophys. Res.,92, 3207±3213, 1987.
Southwood, D. J., Magnetopause coupling processes and the
ionospheric response: a theoretical perspective, Philos. Trans.
R. Soc. London. Ser. A.328, 79, 1989.
Van Eyken, A. P., H. Risbeth, and M. A. Saunders, Initial
observations of plasma convection at invariant latitudes of
70±77°,J. Atmos. Terr. Phys., 46, 635, 1984.
1422 G. Provan et al.: CUTLASS Finland radar observations of the ionospheric signatures