Solar-wind-driven Pc5 field line resonances and pulsed
magnetic reconnection at the dayside magnetopause

P. Prikryl1, R. A. Greenwald2, G. J. Sofko3,  J. P. Villain4, and C. W.
S. Ziesolleck5 

1Communications Research Centre, Ottawa, ON
2The Johns Hopkins University Applied Physics Laboratory, Laurel, MD
3Institute of Space and Atmospheric Studies, Department of Physics, University of
Saskatchewan, Saskatoon, SK
4Laboratoire de Physique et Chimie de l'Environnement/CNRS, Orléans, France
5Herzberg Institute of Astrophysics, National Research Council of Canada, Ottawa,
ON

Abstract. The field line resonances (FLRs) associated with enhanced
convection flows were observed by the West Iceland SuperDARN radar
measuring ionospheric plasma drifts near the ionospheric footprint of the
cusp that was identified by large spectral widths. The enhanced flows
showed characteristics of flow channel events (FCEs) which are believed
to be the ionospheric signatures of flux transfer events (FTEs) in the cusp.
The FCEs evolved from, and were structured similarly to, the FLR
associated flows in the ionospheric footprint of the low latitude boundary
layer (LLBL). The FCEs recurred at a rate which was determined by the
FLR frequency excited on the outermost closed magnetic flux tubes that
were about to be eroded by reconnection at the dayside magnetopause.
The radar observations of spectral widths indicated that the polar cap
expanded as a result of a series of FTEs. It is suggested that the
resonating magnetic shells perturbed the magnetopause and modulated the
magnetic reconnection into pulses by means of the FLR Alfvén wave
varying the reconnection electric field along the X line. In the solar wind,
large amplitude IMF fluctuations on the scale from 20 min to a few hours
superposed by many short period oscillations preceded and continued
throughout the pulsation event observed by the radar. Low resolution
(~0.2 mHz) Fourier transform (FFT) of a time series (~98 min)
spanning the radar event resolved several dominant periodicities, e.g.,
~12, 7 and 5 min which correspond to typical FLR frequencies (1.4, 2.4
and 3.3 mHz). The FLRs on magnetic shells adjacent to the dayside
magnetopause were driven by these and other IMF oscillations. The IMF
and ground magnetometer data that were obtained during similar events
previously discussed in literature show further evidence for solar wind
driven FLRs. The multi-peak FFT power spectra with higher spectral
resolution (~0.1 mHz) strongly suggest the presence of solar oscillation
vestiges in the solar wind IMF. The discrete peaks in the spectra are
tentatively associated with dominant modes of solar oscilations (for
intermediate spherical harmonic degree l~50). These results are
consistent with recently published evidence that solar oscillations
propagate through the interplanetary medium.  As a topic for future
research, it is suggested that spectral measurements of the solar and/or
IMF oscillations on the scale of a few hours or more combined with a
study of the magnetospheric response may be used to predict the
conditions for magnetic merging at the dayside magnetopause and to test
applicability of the current models of reconnection.



1. Introduction

  Magnetic reconnection [Dungey, 1961] is regarded as important for the
transfer of energy and momentum from the solar wind to the
magnetosphere-ionosphere system. Several models of reconnection have
been proposed (see, e.g., Kan et al., [1996] and references therein) but
the discussion of their correctness and/or relative importance still
continues. While the early observations by satellites (ISEE 1 and 2)
provided evidence for quasi-steady reconnection [e.g., Paschmann et al.,
1979], the merging of the interplanetary magnetic field (IMF) at the
dayside magnetopause, at times, appears to be pulsed with pulses
separations of several minutes [Russell and Elphic, 1978 and 1979;
Lockwood and Wild, 1993]. The distribution of separation times between
two successive FTEs ranging from a few minutes to a few tens of minutes
was reported [Kuo et al., 1995]. An earlier estimate of the average FTE
occurrence period often quoted is about 8.5 min [Lockwood et al., 1989]
when the IMF is continuously and strongly southward. The cause for this
pulsed behavior of magnetic reconnection is still unknown. The pulsed or
time-varying reconnection [Cowley and Lockwood, 1992] has been linked
with the concept of a pulsating cusp [Smith and Lockwood, 1990] which
is based on the flux transfer event (FTE) models of Southwood et al.
[1988] and Scholer [1988]. Unlike the Russell and Elphic [1978] FTE
model for an isolated newly-opened flux tube of roughly a circular cross-
section, the Southwood/Scholer models of FTE predicted an elongated and
much larger FTE signature in the ionosphere (up to a few thousand km in
the East-West and a few hundred km in the North-South extent). 
  The VHF, HF and incoherent scatter radars were used to identify
ionospheric signatures of FTEs [Goertz et al., 1985; Lockwood et al.,
1990, 1993; Pinnock et al., 1991, 1993]. The latter authors, using the HF
radar data combined with the DMSP satellite particle data, showed that an
enhanced convection channel, also called a flow channel event (FCE), was
positioned at the equatorward edge of the particle cusp. The FCEs have
been attributed to ionospheric signatures of FTEs. The flows, initially
westward, sharply rotated poleward (for By<0 and southern hemisphere)
giving a characteristic signature on the radar maps of line-of-sight (los)
velocities [Pinnock et el., 1993; their Figure 3b]. Recently, Pinnock et al.,
[1995] used a new sounding mode to obtain high spatial and temporal
resolution measurements of the ionospheric signatures of the cusp and the
low latitude boundary layer (LLBL). They tentatively concluded that
reconnection at the dayside magnetopause can be both intermittent and
patchy (i.e., simultaneous FTEs mapping to different longitudes in the
cusp ionosphere have been observed). A three-dimensional patchy-
intermittent reconnection model was proposed by Kan [1988].
  The class of ULF waves in the Pc5 range of periods (150-600 s) that
are often referred to as geomagnetic field line resonances (FLRs) are
thought to be driven by MHD waveguide/cavity modes possibly excited
by the solar wind through Kelvin-Helmholtz instabilities (KHI) or
impulsive stimulations on the magnetopause [cf. Walker et al., 1992;
Samson et al., 1992a]. Traditionally, FLRs have been studied using
ground and satellite-based magnetometers [e.g. Samson et al., 1971;
Barfield et al., 1972]. The VHF and HF radar measurements of the
coherent backscatter measurements of the ionospheric irregularities that
are generated by the electric field of these hydromagnetic waves are also
well documented [e.g. Walker et al., 1979; Ruohoniemi et al., 1991].
These observational results were conducive to extensive theoretical work
[Southwood, 1974; Chen and Hasegawa, 1974; Walker, 1981; Allan et al.,
1986] which successfully explained the observations. 
  More recently, ULF waves with discrete spectra, showing remarkably
robust frequency components near 1.3, 1.9, 2.7, and 3.3 mHz, were
identified in the data from the Goose Bay HF radar and CANOPUS
magnetometer array at high latitudes in the local midnight and early
morning sectors [Ruohoniemi et al., 1991; Samson et al., 1992a; Walker
et al., 1992]. ULF wave signals with discrete frequencies (including
frequencies either higher than or between previously reported values, e.g.,
1.6 and 2.4 mHz) were also observed with the CANOPUS array
throughout the dayside and dusk magnetosphere [Ziesolleck and
McDiarmid, 1994]. Frequencies <1 mHz (typically 0.8 mHz) were also
observed but these ULF waves do not fit with the FLR cavity model
[Walker et al., 1992]. Even in the noon sector which is most relevant for
the event presented in this paper, but where the KHI instability is no
longer a possible source of FLRs, the magnetic Pc5 pulsations reveal
typical FLR characteristics, namely the latitudinal dependence of the
spectral and polarization parameters [Ziesolleck and McDiarmid, 1994].
Recently, Ziesolleck et al. [1996] have reported FLRs at high L shells near
local noon, and suggested that these may be related to Pc5 waves
generated near the cusp or cleft region. 
  The characteristic radar signature of Pc5 FLRs is a sequence of
poleward progressing backscatter bands in the Range-Time-Intensity (RTI)
plot as a consequence of the phase shift with latitude of the FLR
pulsation. At times, the sign of the Doppler velocities observed in these
bands alternates from one band to the next [Walker et al., 1979] as a
result of alternating ionospheric electric fields. Most frequently, however,
the alternating electric field is superposed on a relatively large convection
field thus enhancing or reducing the strength of the latter. These electric
fields are set up in the ionosphere because the hydromagnetic waves are
associated with oscillatory field-aligned currents (FACs) which close via
Pedersen currents in the ionosphere. 
  An important question that still remains to be answered in detail is what
are the relative roles of the solar wind and the magnetosphere in the
momentum transfer across the magnetopause. It is well known that the
orientation of the IMF is playing a crucial role in this process [e.g.
Berchem and Russell, 1984]. In a recent statistical study of FTEs Kuo et
al. [1995] examined a possibilty of the solar wind control of the formation
of FTEs considering solar wind parameters other than IMF (beta, dynamic
pressure and Mach number). They concluded that "the occurrence of
FTEs is probably controlled by some intrinsic property of the
magnetospheric system itself rather than by these solar wind parameters".
In view of the resonant properties of the magnetosphere one may ask
whether field line resonances (FLRs) could represent this intrinsic
property of the magnetosphere and provide a reverse coupling from the
resonating magnetic shells near the magnetopause modulating the magnetic
reconnection at the dayside magnetopause.
  Several mechanisms of FLR generation have been proposed (see above)
including the possibility of FLRs driven by perturbations in the solar wind
dynamic pressure and IMF [Lysak et al., 1994]. Potemra et al. [1992]
invoked an FLR and solar wind pressure variations in their interpretation
of a pulsation event associated with an FTE. They emphasized that both
magnetospheric resonant processes and processes driven by solar wind
determine many varied phenomena observed in the dayside magnetosphere
and ionosphere. In the present paper we show evidence that FLRs are be
directly driven by the IMF oscillations which have origin on the sun.
  Because of the complexity of the dayside phenomena that include
production of polar patches we present the results in two papers. This
paper (paper 1) concentrates on the observations of ionospheric signatures
of the resonant response to the IMF oscillations. The observed ionospheric
signatures of FTEs indicate that the solar wind-magnetosphere-ionosphere
coupling processes on the closed (FLRs) and open (reconnected) field lines
are closely interrelated. In the second paper (paper 2), using a more
extensive ground database obtained during another event, we show that
these processes can also structure the polar ionosphere into patches. 

2. Instruments 
  The Stokkseyri HF radar that is located at the west coast of Iceland
(N63.86o, W22.02o) is a French component of an extended network of HF
radars called SuperDARN (extended Dual Auroral Radar Network)
[Greenwald et al., 1995]. The field of view (FoV) of this radar extends
from 68o up to 85o north magnetic latitude and covers up to ~4 hours in
magnetic local time. The radar employs linear phased arrays of 16 log-
periodic antennas and the operational frequency is between 8 and 20 MHz.
The radar forms a beam which is narrow in azimuth (between 2.5o to 6o
depending on the transmitted frequency) but broad in elevation (up to
~40o at 8 MHz). Usually, the beams are stepped through 16 adjacent
azimuthal positions every 120 s. In near-real time, the radar measures the
backscatter power, line-of-sight (los) velocity and Doppler spectral width
by fitting the auto-correlation functions (ACFs) for 70 range bins starting
at the slant range of 180 km (normally with the range resolution of 45
km).
  Magnetic Fields Investigation (MFI) is an instrument on the Wind
spacecraft which is part of the International Solar-Terrestrial Physics
(ISTP)/ Global Geospace Science (GGS) mission. Also, additional data
from IMP-8 satellite are used. 

3. Observations
  The Stokkeseyri radar is particularly well situated to observe the
ionospheric flows mapping into the cusp. The particle cusp under various
conditions of solar wind pressure [Newell and Meng, 1994] is expected to
be located at relatively short ranges if compared with other SuperDARN
radars in the northern hemisphere. Frequently, the enhanced cusp flows
are associated with moderate but structured flows in the ionospheric
footprint of the LLBL equatorward of the cusp. In this paper, we describe
one such event when a series of moderate to intense flow channel events
(FCEs) was observed in the ionospheric footprint of the cusp during a
period when the IMF conditions measured by Wind were favorable for
magnetic merging at the dayside magnetopause (Bz<0 and By>0). A
sequence of line-of-sight (los) velocity maps (Figure 1) shows examples
of FCEs which moved poleward and westward as expected of FTE
ionospheric signatures in the northern hemisphere for the above IMF. We
shall concentrate on the period when the FCEs were particularly intense
(1600-1640 UT). The latitude-time-velocity (LTV) plot for the radar beam
#10 (Figure 2a) shows the poleward progression of 3 major FCEs
occurring every ~11 min with the flow maximizing around 75o of
magnetic latitude after 1600 UT. As the first and least intense FCEI of the
series extended northwestward a more transient and localized FCEIa
occurred near 73o at the trailing edge of FCEI at 1608 UT. In other words,
the first of the FCEs appeared to have been structured into two separate
FCEs. At 1610 UT, after FCEI moved northwestward and faded, FCEIa
shifted poleward and extended farther northwestward. At its previous
position (at 1608 UT) it was about to be replaced by an extended return
flow progressing poleward (1610 UT). The next in the series, FCEII
evolved into the strongest westward flow with los velocity exceeding 1500
m/s at 1616 UT. Finally, the latter was replaced by another intense FCEIII
that was very similar to FCEII with los velocity at its maximum (1626 UT)
reaching up to 1500 m/s. More FCEs followed but were not so intense
(los velocities up to ~1000 m/s) and were not seen by beam #10 until
after 1700 UT. The backscatter at ranges beyond 1000 km (~73o) and
along the central beams become fragmented or entirely missing due to
absorption in the E-region (most likely due to precipitation associated with
field line resonances that are discussed below). Nevertheless, moderately
enhanced flow bursts similar to the one observed at 1612 UT were
detected by the beams #2-4 (flow bursts maximized at ~1632/1634, 1646,
1718, and 1726 UT). Some of these westward flows preceded FCEs and
several fragmented signatures of FCEs were observed by beams #10-15
(1700, 1712, 1722, 1732, 1740 and 1748 UT). However, one should not
necessarily expect a one-to-one correspondence between the flow bursts
at lower latitudes (beam #2-4) and FCEs that were observed a few minutes
later, because they were parts of a latitude-dependent FLR structure
(discussed below). The approximate times of the latter FCEs are shown
in Figure 2a. Before 1600 UT, a few FCEs occurred at a lower rate
(every ~12 min) earlier and will be briefly discussed below. 
  The radar observations of spectral widths which often exceeded 300
m/s (Figure 2b) indicated that the FCEs occurred near the ionospheric
footprint of the cusp [Baker et al., 1995]. Equatorward of the FCEs
(<73o magnetic latitude), low to moderate spectral widths were typical of
the ionospheric footprint of the LLBL. The LLBL flows were weak to
moderately enhanced and, at times, clearly structured into bands that were
approximately parallel although somewhat tilted with respect to L shells
as they progressed poleward (see e.g., radar maps at 1612/1614 UT;
Figure 1). After 1600 UT in particular, the bands of alternating (westward
to eastward) flow progressed poleward toward the radar cusp where the
(northwestward) flows intensified and rapidly evolved into FCEs flanked
by moderate but spatially extended return flows (Figure 1). These single
radar observations of plasma flow clearly suggest traveling twin vortices
(there was not sufficient backscatter from Goose Bay radar to generate
vector maps). As one FCE weakened in the ionospheric footprint of the
cusp it was abutted on a new one that evolved from weaker flows in the
LLBL footprint. The flows intensified in the radar cusp as they progressed
poleward and eventually replaced the previous FCE. Sometimes, when
FCEs recurred more frequently, two FCEs (one forming and another
fading away) were observed in the same radar map (see e.g. Figure 1b
and another event described in paper 2; Figure 15a).
  The structured flows that map to the LLBL (Figure 2) are typical of
FLRs. Figure 3 shows the results of the FFT analysis [Fenrich et al.,
1995] of the radar los velocities observed by the beam #10. The 2.6-mHz
FFT peak near 72.5o of magnetic latitude (Figure 3a) spans ~2o of
latitude (half power width of 1o). The phase decreased by 180o across the
latitudinal width of the power peak. These signal characteristics are fully
consistent with a discrete low-m FLR near a typical FLR frequency
[Ruohoniemi et al., 1991; Samson et al., 1991, 1992ab; Ziesolleck and
McDiarmid, 1994, 1995; Fenrich et al., 1995]. Poleward of the 2.6-mHz
(~6.5 min) resonance, near 76o magnetic latitude, where the FCEs
occurred, the FFT shows a peak near 1.6 mHz (~10 min). This peak
spans more than 3o degrees of latitude while the phase decreases by only
60o. The 1.6 and 2.6-mHz peaks straddle a peak near 2 mHz, which
appears to be a weak signature of another typical FLR frequency. Clearly,
at this latitude (~74o) there were large gaps in the data which were filled
by two dimensional interpolation thus the spectral estimates are not very
reliable (note that the strong low frequency component is an artifact).
However, the weak peak at 2 mHz (8 min) simply represents the time
separation between two successive FCEs (FCEIa and FCEII) that were
observed between 73o and 74o magnetic latitude (Figures 1 and 2). While
FCEII eventually evolved into the strongest flow channel as it progressed
to 76o latitude, FCEIa was more transient, spatially confined to latitudes
<74o and appeared to have been "damped" at higher latitudes where the
1.6-mHz frequency dominated the FFT spectra. Similar observations of
FLR modulated cusp flows are presented in paper 2.
  Figure 4 shows the IMF measured by the Wind spacecraft. At 0910
UT, after an extended period of northward IMF, the IMF sharply turned
southward and then swayed between southward to northward direction on
the scale of a few hours. At the same time, it oscillated on shorter time
scales. Note a large amplitude 20-25 min fluctuation between 1430 and
1600 UT (particularly in By) while the IMF was rapidly diverted from
southward to northward and back to southward again. The IMF By-
component was mostly positive and all three IMF components also
oscillated on the scale of a few minutes for the next 3 hours. Figure 5
shows the 32-point FFT power spectra of the IMF components for a
period between 1500 and 1640 UT (at this time, the available Wind data
were sampled every 92 s). This interval starts one hour before the onset
time of the intense radar event described above. At 1500 UT, Wind was
at a distance of 121 RE from the Earth or ~110 RE from the
magnetopause, positioned at (107.5, -54.4, 11.2 RE) in GSE coordinates.
At 300 km/s (solar wind speed measured by Wind at 1200 UT, just before
a gap in the SWE data) a disturbance would reach the magnetopause in
about 39 min. The FFT spectra of all three IMF components are very
similar and show three main peaks at ~0.7, 1.4, and 2.4 mHz (apart
from the peak near 0.3 mHz which has its special significance as
suggested below). Also, some spectral power appears to be concentrated
into weaker peaks (1.7, 3.3, 4.0 and 4.7 mHz). The 1.4/1.7- and 2.4-
mHz peaks approximately (within limits of the spectral resolution and
considering the fact that FFT window was significantly wider than the 
length of the main radar event) coincide with the two major peaks that
were found in the radar data (Figure 3). The IMF oscillations and their
effects on the dayside magnetosphere/ionosphere are further discussed
below.
  Now we shall compare the radar event (1600-1640 UT) with similar
but more typical and rather frequent events, when SuperDARN radar
maps and LTV plots indicated that FLRs and even FCEs (depending on
the IMF orientation) were present but fragmented. Conveniently, but not
unexpectedly, such events can be identified in Figure 2. As we noted
above, flow bursts were observed both before and after the main event.
We will concentrate on the period after 1700 UT and then briefly mention
some characterics of the ionospheric flow before 1600 UT.
  Because the propagation delay between Wind and the magnetopause
was about 40 min, the IMF conditions pertaining to the radar data after
1700 UT were those recorded from ~1620 UT. The IMF fluctuated on
the scale of a few minutes and the conditions continued to be favorable for
magnetic merging. However, only moderate and fragmented ionospheric
flows were observed by the radar as the radar cusp shifted equatorward.
The reduced spectral widths (Figure 2b) at high latitudes indicated that the
polar cap also expanded equatorward into the radar field of view. Clearly,
this was a result of a long series of FTEs (associated with the FCEs) that
significantly eroded the closed magnetic flux at the dayside
magnetosphere. Fragmented and weaker FCEs (identified above) continued
but recurred at an increased rate, namely, one FCE every ~12 min
(before 1700 UT) down to one per 8 min (after 1730 UT) as they shifted
to lower latitudes. While beam #10 (Figure 2a) did not resolve any
significant flow structure in the LLBL ionospheric footprint after 1700
UT, beam #4 (Figure 6) indicated a pulsation with another typical FLR
frequency near 3 mHz (5.5 min). Note that earlier (1600-1640 UT), the
same beam #4 also showed a strong 2.6-mHz (6.5 min) oscillation of the
FLR related flow along the beam direction. This general increase in the
FLR/FCE frequencies indicates that the reconnection process at the
dayside magnetopause continued. As more and more magnetic flux was
removed from the dayside magnetosphere due to reconnection, smaller and
smaller magnetic shells were left for the FLRs to be acted upon. As a
result, the FCEs recurred at a progressively faster rate, although the radar
FTE-like signatures weakened as the cusp gradually shifted to the west of
the radar and was eventually replaced by cleft. Later (at ~2000 UT),
some poorly defined flow bursts appeared in the field of view of
Saskatoon radar (not shown) which apparently intercepted the cusp at this
time. In contrast with the increased FCE rate after 1730 UT, the FCEs
recurred every ~12 min before 1600 UT. The backscatter between 69o
and 72o magnetic latitude that was characterized by very small spectral
widths (~50 m/s) indicated pulsations (FLRs) with oscillation period
increasing from ~4 min up to 6 min between 69o and 71o, i.e., more than
5o of latitude equatorward from the radar cusp (beam #5; not shown).
  Clearly, the observations during the two periods (1530-1600 UT and
1700-1800 UT) were very similar to the main event (1600-1640 UT)
except for the flow magnitudes and clarity of observations. Also, several
other events that were analyzed showed characteristics quite similar to
those described above. The majority of these dayside events were
observed with the Stokkseyri radar. Unfortunately, many factors can cause
the HF radar data appear confused and fragmented. Below we summarize
some of these factors. Probably the most important factor is the HF
propagation which can be quite uncertain in a disturbed ionosphere. In
particular, the FLRs are likely to be associated with electron precipitation
into auroral arcs [Samson et al., 1992b] that can be distorted. In addition,
the arcs are often associated with electron density depletions adjacent to
arcs [Opgenoorth et al., 1990]. As a result, absorption and anomalous
refraction can be quite severe at times. After 1630 UT, as discussed
above, the backscatter become rather fragmented or entirely missing,
particularly along the central beams. This is an evidence of a strong E-
region absorption. Furthermore, radar beams (e.g., the southernmost
beams of the Stokkseyri radar in particular) that intersect auroral arcs at
oblique angles are subject to HF or even VHF anomalous refraction
[Prikryl et al., 1992]. Occasionally, it was observed that the backscatter
along a couple of the southern beams was suddenly extended to several
range gates. This could be explained by anomalous refraction in distorted
and structured auroral arcs (similarly to the VHF observations presented
by the latter authors). Ground scatter that was not a serious problem
during the event presented in this paper can also fragment the data and
affect the observed los velocities and spectral widths but this, in principle,
can be rectified by computing the Doppler spectra. In the normal mode of
SuperDARN operation, the temporal resolution is insufficient to resolve
FLR frequencies >4 mHz. Problems with temporal aliasing of rapidly
moving phenomena such as fast moving or frequently recurring FTE
signatures and insufficient spatial resolution can be alleviated or improved
by various scanning techniques [e.g., Pinnock et al., 1995]. In the E-
region, the relation between the magnitudes of measured velocities and the
ExB motions can be uncertain because of difficulties with the
interpretation of the E-region Doppler spectra [St.-Maurice et al., 1994].
In general, the coherent backscatter radar measurements significantly
underestimate the E-region drift velocity, particularly for large velocities
[Nielsen and Schlegel, 1983]. Most of the backscatter at close ranges
(<1000 km in Figure 2a) could have been from the E-region (note the
sharp cut off near 73o). Any of the above factors can contribute to
problems with interpretation or affect the availablility of the SuperDARN
data.


4. Discussion

4.1. The IMF modulation of dayside reconnection by FLRs
  Two main observations presented in this paper should be emphasized
because we believe that they will lead to better understanding of the
coupling processes between the solar wind and magnetosphere/ionosphere
system. First, the above observations show that FCEs are closely related
to FLRs in the LLBL near the dayside magnetopause. FCEs can be
viewed as an extension of FLRs to the outermost closed flux tubes that are
about to be open by magnetic reconnection modulated into pulses (FTEs)
by FLRs. Second, the IMF data showed oscillations with the frequencies
that were similar to those of FLRs/FCEs indicating that the FLRs were
driven by the IMF oscillations in the solar wind. The IMF disturbance,
after reaching the magnetopause, could have propagated to the resonant
L-shells via fast compressional mode and then couple to the shear Alfvén
mode driving the resonance. We suggest that the resonating magnetic
shells adjacent to the magnetopause perturbed the magnetopause (Figure
7) thus modulating the magnetic and electric field topology and location
of the reconnection line (also called a singular line).
  There may be one potential difficulty with the above interpretation of
the FLR/FCE observations. The magnetospheric magnetic field
predominantly oscillates in the azimuthal direction (toroidal mode) while
the radial (compressional) oscillations (poloidal mode) are expected to be
heavily damped. However, the observations suggest that these directly
driven FLR oscillations (particularly those at the outermost magnetic shells
adjacent to the magnetopause) have a significant radial component. The
flow channels (ionospheric signatures of FLRs on the field lines about to
be merged with the IMF) as well as the FLR associated flows in the
ionospheric footprint of the LLBL were tilted with respect to L-shells thus
indicating a radial magnetic component. Also, the results presented in
paper 2 showed that the amplitudes of the corresponding ground magnetic
oscillations exceeded 100 nT in the north-south direction (X component)
but reached up to 50 nT in the east-west direction (Y component).
Similarly, the IMF oscillations, while usually most pronounced for By,
were often present in other two components. Because the magnetospheric
oscillations were continuously driven (as opposed to impulse driven but
damped oscillations) by the IMF oscillations, the electric field component
of a resonant Alfvén wave on the outermost magnetic shells could have
modulated the electric field along the reconnection line. It is well known
that this electric field is the most important measure of the reconnection
rate. It was recently emphasized by Dungey [1995] that "reconnection is
effected solely by the electric field along a singular line". The electric
field component of an FLR Alfvén wave of only a few mV/m along an X
line a few RE long would result in a significant contribution to the average
reconnection voltage [Lockwood, 1996]. The FLR associated radar los
velocities typically varied by about ±300 m/s in the LLBL footprint
indicating that the corresponding electric field alternated by more than
±15 mV/m. However, at times, the amplitudes of the FLR associated
electric field were significantly larger. For example, Figure 1 shows FLR
related los velocities near 71o varying from +500 m/s (1610 UT) to -700
m/s (1612 UT) indicating electric field variations of more than ±30
mV/m about the background electric field. At higher latitudes (near the
equatorward boundary of the cusp), just before the FLRs broke loose and
evolved into FCEs as the field lines were opened (reconnected), the
observed flows become more enhanced suggesting even stronger electric
fields. These electric fields mapped to the reconnection region near the
magnetopause and modulated the reconnection electric field along the X
line. 
  Conversely, the oscillation in the IMF (possibly associated with an
Alfvén wave) could have directly modulated the reconnection process, and
the compressional mode wave associated with the pulsed reconnection may
have excited FLRs. Taguchi et al. [1993] discussed a reconnection model
for By>0 studying the IMF By-dependent FACs in the cleft that could
explain such direct modulation of the reconnection process from the solar
wind. However, at least for one other event that we analyzed (not shown)
the amplitudes of the IMF By oscillations were very small (~1 nT or less)
and it seems rather unlikely that they could strongly modulate the
reconnection, yet these small amplitude fluctuations in the solar wind were
sufficient to drive a resonance which was observed. During another event
supported by an extensive ground-based data (paper 2), the satellite-to-
ground delays were smallest for the equatorward edge of the observed
FLRs (see Figure 10 in paper 2). Also, in paper 2, it was shown that the
multi-peak FFT power spectra of the IMF (similar to those presented in
Figures 5 and 8) were best correlated with the FFT power of the ground
magnetic components measured on closed field lines. In particular,
virtually all of the oscillation modes observed in the IMF By were
identified in the ground magnetic data from the LLBL footprint. These
two cross-correlation results (in the time and frequency domains) clearly
suggest that FLRs were excited first and modulated the reconnection
process (FTEs) as discussed above. Furthermore, we have identified FLRs
when the IMF conditions were not favorable for magnetic merging (no
FCEs were observed), yet the resonance was clearly driven by the IMF
oscillations.
  While the above observations show that the magnetic reconnection can
be pulsed (at FLR frequencies) this does not exclude a possibility of
patchy reconnection that has been reported in the literature [Pinnock et
al., 1995]. The characteristics of FLRs may vary with longitude and this
could result in localized FCEs occurring simultaneously at different
longitudes (flux tubes). However, the theoretical work [Cowley and
Lockwood, 1992] predicts a large longitudinal extent of FTEs. The
observations, such as those that are presented in paper 2, are consistent
with the latter. Furthermore, FTEs of a very large longitudinal extent
were also reported in literature, particularly for By*0 [e.g., Lockwood et
al., 1990; Pinnock et al., 1993; see also paper 2]. It remains to be seen
whether the patchy appearence of FTEs can be supported by the theory
after the reverse coupling from the resonant magnetospheric cavity to the
reconnection region is studied in more detail and/or included in the theory
of reconnection.
  A multiple X line reconnection model [Lee and Fu, 1985] was invoked
as a possbile explanation of multiple brightenings of transient dayside
auroral forms [Fasel et al., 1992]. It is quite conceivable that directly
driven multi-mode oscillations (or FLR harmonics) near the magnetopause
could result in multiple reconnection sites (X lines) along the same flux
tube (Figure 7). The observations of two nearly simultaneous FCEs (FCEIa
and FCEI) discussed above (Figures 1 and 2a) may suggest such
interpretation. The radar observations presented here and in paper 2 are
very similar to the dayside auroral transients (poleward moving auroral
forms, PMAFs) reported in literature [Sandholt et al., 1990; Fasel et al.,
1992; Fasel, 1995] and some of the similarities are discussed in paper 2.
The transient 557.7-nm emissions [Fasel et al. 1992; their Figures 1 and
2] occurred at a much higher rate in the auroral oval than the reconnection
related transients at the poleward edge of the emission band (presumably
in the cusp). For example, in their Figure 2, transient a is followed by
another transient b after ~11 min and then the recurrence rate of these
most poleward transients increased in two steps (each associated with a
equatorward step in latitude) to one transient per 7-8 min and eventually
to 3-5 min between the transients (see Fasel et al., [1992], their Figure
1). At the same time equatorward edge of the optical emissions was
structured by pulsations with frequencies higher than those observed near
the cusp. These optical observations clearly suggest FLRs and FLR driven
reconnection events (FTEs) in good agreement with the radar observations
presented in this paper and paper 2.
 
4.2. Solar-wind-driven modulation of dayside DPY currents and FLRs
  Poleward progressing dayside ionospheric disturbances associated with
long-period geomagnetic pulsations due to the IMF By-modulated DPY
currents [Friis-Christensen and Wilhjelm, 1975] are frequently observed
in the cusp and the dayside polar cap just poleward of the cusp [Clauer et
al., 1995; Stauning et al., 1994 and 1995]. Stauning et al. [1995]
presented results of a detailed analysis of time delays of these disturbances
that were strongly coupled to the solar wind via connecting open field
lines. In their model explaining the IMF By dependency of DPY currents
they assumed that the "open" magnetic field lines of the ionospheric
footprint which they were studying have just merged with the IMF.
Papitashvili et al. [1995] interpreted the observed phenomena as a direct
ground-based evidence of the IMF By component reconnection at the
dayside magnetopause and discussed a suitable reconnection model by
Taguchi et al. [1993].
  While we have not presented any ground magnetometer data for the
event discussed in the present paper the DPY current intensifications are
coincident with FCEs (see paper 2). We believe that a discussion of DPY
currents and their relation to the IMF By-driven FLRs/FCEs is appropriate
here because the DPY currents are important for understanding the
coupling between the solar wind and the magnetosphere/ionosphere
system. In this paper we presented evidence that the coupling processes
on the open and closed field lines are closely interrelated. The
observations of the By-driven FCEs and the related ground signatures
(paper 2) are similar to observations of poleward progressing dayside
convection disturbances poleward of the cusp [Stauning et al., 1995]. The
study by Stauning et al. [1995] was mainly concerned with the long-period
disturbances on the open field lines connecting the polar ionosphere to the
oscillating IMF. Stauning et al. [1995] concluded that during a period of
southward IMF there is a strong and direct coupling between the IMF and
the polar ionosphere. The large amplitude IMF By variations with long
periods (~20-30 min) coupled to the polar ionosphere and modulated
polar convection pattern. The progression of IMF By-related field-aligned
currents feeding the DPY (Hall) currents produced characteristic poleward
progressing geomagnetic perturbations. However, it should be noted that
these disturbances were extended equatorward to the southernmost ground
stations (FHB and NAQ; their Figures 3 and 7). Presumably, these
stations were equatorward but near the cusp. This was clearly manifested
by the abundance of pulsations with shorter periods for both events (Aug
2/3, 1991) presented by Stauning et al. [1995] and we consider these
events similar to those described here and in paper 2. 
  Because the ground magnetometers respond to integrated ionospheric
currents over a large area Stauning et al. [1995] smoothed the IMP-8 data
before correlating it with the IMF. While we agree with this argument in
principle, such smoothing removes from the data the higher frequency
components which are most relevant for the IMF coupling to the closed
field lines through FLRs. A close examination of the FHB and NAQ
ground magnetograms [Stauning et al. 1995; see their Figures 3 and 7]
reveals many discrete pulsation bursts which we will now briefly discuss.
After 1300 UT, on Aug 2, 1991, the Z component at FHB showed a burst
of 10-12 min (~1.5 mHz) pulsation followed by another wave packet with
a period ~9 min (1.9 mHz). Between 1200 and 1400 UT, one can
identify pulsations in the NAQ Z-component with periods of ~7 min (2.4
mHz), ~4.5 min (3.7 mHz) and ~8 min (2 mHz) which are also found
in the NAQ H-component. On Aug 3, 1991, the FHB and NAQ ground
magnetometers showed very complex pulsations throughout this second
dayside event. The most conspicuous pulsations (H and Z-components)
seem to have had periods of ~12 min (1.4 mHz) followed by ~9 min
(1.9 mHz) at FHB and ~6 min (2.8 mHz) followed by ~5 min (3.3
mHz) at NAQ. Other periodicities were clearly present but impossible to
determine with certainty from these figures. The latitude dependent
periods of these pulsations (longer periods at FHB than in NAQ) suggest
that they were due to FLRs. The unsmoothed IMF data for these two
events (Figures 6 and 9 presented by Stauning et al. [1995]) also showed
oscillations on the scale of a few minutes. We have computed 64-point
FFT power spectra of the IMF components for two 130-min periods
(Figure 8). These FFTs (By in particular) show multi-peak spectra that are
similar to those presented in Figure 5 and in paper 2. From a
magnetospheric point of view, it should be noted that the discrete peaks
in the IMF spectra are at, or near, the so called "magic" frequencies that
were associated with FLRs (see references cited above). All the
periodicities that we identified with several pulsation bursts in the ground
magnetic data (above) can be directly associated with individual peaks in
the FFT power spectra (Figure 8). Specific long periods corresponding to
the low frequency peaks (0.7-0.8 mHz) in the IMF spectra could be
added. These observations are very similar to the strong correlation (0.84)
between the IMF in the solar wind and ground magnetic data on the
closed field lines reported in paper 2 (see Figure 2bc therein) and provide
further evidence that FLRs with several specific frequencies (not a
continuum of possible frequencies) can be directly driven by the IMF
oscillations. The frequencies that are found in the IMF and ground data
are the same frequencies (within 10% of reproducibility [Samson et al.,
1991]) that were associated (except for those near 0.8 mHz) with cavity
or waveguide mode frequencies [Samson et al., 1992b; Walker et al.,
1992].
  The point to be made here regarding the DPY currents is that the IMF
By in addition to modulating the Hall currents poleward of the cusp
footprint similarly modulate the ionospheric currents (and the associated
convection disturbances) equatorward of, and near, the ionospheric
footprint of  the cusp. The coupling of the IMF oscillation to the auroral
ionosphere is accomplished by FLRs driven by oscillations in the solar
wind. Both long [Stauning et al., 1995] and short (see above and also
paper 2) period IMF By oscillations modulate the DPY currents. Thus we
suggest that there is no fundamental difference between the IMF By-related
disturbances on the open and closed field lines except that it is the FLR
associated Alfvén wave and corresponding field aligned currents (FACs)
that modulate and feed the Hall currents on the closed field lines. On the
"open" (newly merged) field lines directly connecting the polar ionosphere
to the oscillating IMF [Stauning et al., 1995], interplanetary currents
flowing into (and out of) the magnetosphere could explain the progressing
polar ionospheric disturbances [Stauning et al., 1994]. Such currents could
also be a manifestation of Alfvén waves on these "open" magnetic field
lines.
  The Hall current intensities are expected to maximize near the merging
gap according to the Clauer and Banks, [1986] model and such
intensification of the DPY currents is manifested by FCEs, ionospheric
signatures of FTEs. As the newly merged field lines convect into the polar
cap the IMF By continues modulating the DPY currents which will
progress poleward (for Bz<0). On the closed field lines, where the IMF
oscillations modulate the currents through the FLR coupling, the DPY
currents are modulated with progressively lower (cavity/waveguide mode)
frequencies with increasing latitude but a modulation by the low
frequencies (*1 mHz) similar to but weaker than the DPY modulation
poleward of the cusp are also observed (see Stauning et al. [1995]; their
Figures 3 and 7, and paper 2). It should be noted that although the low
frequency IMF oscillations do not result in a resonance they would still
couple to the magnetosphere and drive the ULF waves which in turn
modulate the DPY currents. In summary, we suggest that, in the
ionospheric footprint of the cusp, poleward progressing DPY current
disturbances and FCEs are two equivalent ground/ionosphere signatures
of the pulsed magnetic reconnection at the dayside magnetopause. They 
are a result of the resonant response of the magnetosphere to IMF in
which FCEs/DPY currents are extensions of latitude dependent FLR
structure from LLBL (closed field lines) into the cusp (field lines about to
be merged with IMF). In the region spanning the LLBL, cusp and polar
cap poleward of the cusp, the dayside magnetosphere and ionosphere
respond to progressively longer periodicities of many discrete IMF modes
modulating the field aligned and ionospheric currents while only one or
two of these IMF modes and corresponding FLRs modulate the magnetic
reconnection into pulses. 

4.3. Observations of Solar Oscillation Vestiges in the Solar Wind and
the Earth Magnetosphere/Ionosphere
  The presence of coherent IMF oscillations in the solar wind naturally
raises the question of their source. The multi-peak power spectra of the
IMF time series are in sharp contrast to a common belief that solar wind
is a turbulent medium with a continuous spectrum of modes. Recently,
Thomson et al. [1995] reported results of time-series analysis of charged
particle fluxes measured by Ulysses and Voyager spacecraft in the
interplanetary space. In addition to many discrete low frequency spectral
components (0.001-0.140 mHz) they identified spectral lines in the higher
frequency band (1-4 mHz) and found them consistent with so g (gravity)
and p (pressure or acoustic) modes, respectively, of solar oscillations.
Intermediate (f or surface) modes between these two spectral bands were
also detected. These results imply that solar oscillation modes propagate
into the interplanetary medium as fluctuations of the IMF that in turn can
modulate the particle fluxes.
  The multi-peak power spectra of the IMF presented in the present
paper (Figures 5 and 8) are consistent with the above observations
[Thomson et al. 1995]. Clearly, the spectral resolution of these spectra
(~0.15 mHz) is inferior when compared to the spectral estimates obtained
by the latter authors who used a special spectrum estimation method to
identify periodic components of the particle fluxes with specific solar
oscillation modes. However, when a wider (double) FFT window were
used the spectral peaks (Figures 5 and 8) have been further resolved into
narrow components which can be associated with dominant modes of solar
oscillation, namely, modes with intermediate spherical harmonic degree
l (l is one of three eigenmode parameters that are used to describe solar
oscillations; the other two being spherical harmonic order m, and radial
order n). A detailed discussion of these results is beyond the scope of this
paper and will be published elsewhere. For the purpose of this paper it is
sufficient to state a tentative conclusion of this analysis, namely, that the
FFT spectra (e.g., Figure 8) represent partly resolved dominant modes
(l~50) of solar oscillation propagated in solar wind. Figure 9 shows one
example of an FFT power spectrum using 256-min window (doubling the
window that is used in Figure 8; top). Clearly, there are more peaks
resolved and the correlation between the three IMF components is
substantially improved (e.g., note peaks near 1.4, 1.6 and 1.9 mHz as
well as some weaker peaks between 4 and 5 mHz). Also, eigenfrequencies
of solar oscillations (p, f and g modes) for l=50 and a typical solar model
[Christensen-Dalsgaard et al., 1985] are shown for comparison. They
represent the best match between the spectral peaks and the model
eigenfrequencies (p, f and g modes) out of all possible values degree l,
give or take a few. Apart from the fact that some modes (p8, p10/11, and
p15/16) and some subsidiary peaks (particularly in the region where f
modes are expected) the agreement is remarkably good (including the
unresolved band of g modes). Some subsidiary peaks (including those that
have not been resolved, e.g., at ~2 mHz) could be due to p- or f-modes
for other values of degree l but these peaks are weaker. Thus it is
tentatively concluded that dominant modes of solar oscillations (l~50)
were transmitted into solar wind IMF. Similar arguments could be made
for other events discussed in this paper and paper 2. It should be noted
that this tentative conclusion is not needed to support the main conclusions
of this paper, namely, those about the relationship between solar wind
driven FLRs and pulsed reconnection. However, mentioning these results
here allows us to place the conclusions of papers 1 and 2 into a broader
perspective and make some general statements about a possible source of
discrete FLR frequencies and pulsed magnetic reconnection at the dayside
magnetopause. 
  From a magnetospheric point of view, the fact that solar oscillations
propagate through the interplanetary medium as coherent fluctuations of
the IMF (possibly as Alfvén waves) indicate that the magnetosphere of the
Earth is inundated with waves in the solar wind that have a discrete
spectrum of frequencies. Therefore, the resonant magnetospheric cavity
is subjected to these multi-mode IMF oscillations. The results presented
here, in paper 2, as well as further results of a preliminary analysis of
several other events show that some of the oscillation frequencies tend to
dominate the IMF power spectra and appear to be more reproducible than
other. As noted above, preliminary analysis indicated that these IMF
oscillations may have been caused by dominant solar oscillation modes.
Such dominant modes would explain puzzling discrete frequencies
identified in the FLRs [Samson et al, 1992a]. This may require a revision
of the current theories of FLR generation mechanisms/sources. For
example, most of the FLR observations have been reported from time
zones other than local noon where the KHI is commonly invoked as a
source of FLRs. Because the magnetic reconnection at the dayside
magnetopause and FLRs are found closely interrelated we believe that the
solar oscillations transmitted by the IMF play an important role in the
coupling of the solar wind to the magnetosphere and thus condition other
magnetospheric and ionospheric processes which structure the ionosphere,
including the DPY currents and polar patches. The latter are discussed in
paper 2. 
  On the other hand, an IMF signal that is rich in spectral content
(containing a multitude of closely spaced solar oscillation modes and no
dominant modes) would be equivalent to a continuum of frequencies, in
the first approximation. Therefore, one would expect a corresponding
continuum of resonant frequencies driving FLRs in the magnetosphere and
a steady-state reconnection at the dayside magnetopause. However, it is
quite conceivable that a heavily structured solar wind IMF could also
support the patchy-intermittent reconnection model based on the same
conceptual model of reverse coupling proposed above. We suggest that the
answer to the question whether the reconnection is pulsed (transient or
bursty), patchy-intermittent or quasi-steady may ultimately depend on how
the solar oscillation characteristics vary with time, namely, whether or not
dominant oscillation modes exist on the sun (provided that the distribution
of oscillation frequencies is not drastically modified while they propagate
through the interplanetary medium). We believe that future studies of solar
and/or IMF oscillation variability on the scale of a few hours or more will
lead to better understanding of the solar wind coupling to the
magnetosphere. Also, it is likely, that a similar approach to the study of
solar phenomena may also help in solving some of the unresolved
problems regarding the causes of solar activity.

5. Conclusions

  We have presented evidence that the solar wind-magnetosphere-
ionosphere coupling processes on the open and closed field lines are
closely interrelated. It was observed that enhanced convection flows (flow
channel events, FCEs) evolved from and were structured similarly to the
flows associated with field line resonances (FLRs) in the low latitude
boundary layer (LLBL). The latter flow structure progressed poleward
toward the ionospheric footprint of the cusp that was identified by large
spectral widths. In view of previously published observations which
associated FCEs with flux transfer events (FTEs) and reports that
identified discrete frequencies FLRs, we conclude that the FCE/FTE rate
was determined by the FLR frequency excited on the outermost closed
magnetic flux tubes that were about to be eroded by FTEs. In other
words, FCEs are extensions of FLR flow signatures to the field lines that
are being reconnected with the IMF in solar wind. It was found that the
oscillating IMF signal contained resonant frequencies that were observed
in the ionosphere by the Stokkseyri SuperDARN radar including the low
frequency component that determined the recurrence rate of FCEs. Since
the latter have been associated with FTEs, these observations constitute
evidence that the pulsed magnetic reconnection at the dayside
magnetopause was caused by solar wind-driven FLRs modulating the
reconnection into pulses.
  Multi-peak Fourier transform power spectra indicate that the most
reproducible discrete peaks in the IMF spectra are found at, or near, the
most commonly observed Pc5 pulsation frequencies previously refered to
as "magic" FLR frequencies. Also, these IMF spectra are consistent with
recently published evidence that solar oscillations propagate through the
interplanetary medium. Based on the preliminary results of the spectral
analysis using wider FFT windows (to be published elsewhere), it is
suggested that the most reproducible spectral peaks that are found in the
IMF and are associated with FLRs are vestiges of dominant modes of
solar oscillations in solar wind and magnetosphere. Also, it is suggested
that spectral measurements of the solar and/or IMF oscillations on the
scale of a few hours or more combined with a study of the magnetospheric
response may be used to predict the conditions for magnetic merging at
the dayside magnetopause and to test applicability of the current models
of reconnection.

  Acknowledgements. The Stokkseyri radar is a French component of
SuperDARN which is a collaboration of Canada, France and United
States. The Stokkseyri radar is supported by ???? Grant ????. We
acknowledge the contributions by R. Lepping (NASA, Goddard Space
Flight Center) who made available the Wind MFI and IMP-8 IMF data.
One of the authors (PP) would like to acknowledge the contribution by F.
Fenrich (University of Alberta) making available the FLR analysis
software for SuperDARN data. We also benefited from discussions with
G. Atkinson, D.R. McDiarmid, J.W. MacDougall, I.F. Grant, H.G.
James, A.S. Rodger, J.C. Samson, and A.D.M. Walker. 


References
Allan, W., S.P. White, and E.M. Poulter, Impulse-excited hydromagnetic cavity and field line
 resonances in the magnetosphere, Planet. Space Sci., 34, 371-385, 1986.
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 layaer,
 J. Geophys. Res., 100, 7671-7695, 1995.
Barfield, J.N., R.L. McPherron, P.J. Coleman, and D.J. Southwood, Storm-associated Pc 5 
 micropulsation events observed at the synchronous equatorial orbit, J. Geophys. Res., 77, 143,
 1972.
Berchem, J., and C. T. Russell, Flux transfer events on the magnetopause: spatial distribution
 and controlling factors, J. Geophys. Res., 89, 6689-6703, 1984.
Clauer, C.R., and P.M. Banks, Relationship of the interplanetary electric field to the 
 high-latitude ionospheric electric field and currents: Observations and model simulation, J.
 Geophys. Res., 91, 6959-6971, 1986.
Clauer, C.R., P. Stauning, T.J. Rosenberg, E. Friis-Christensen, P.M. Miller, and R.J. 
 Sitar, Observations of a solar-wind-driven modulations of the dayside ionospheric DPY current
 system, J. Geophys. Res., 100, 7697-7713, 1995.
Chen, L., and A. Hasegawa, A theory of long-period magnetic pulsations, 1, Steady state 
 excitation of field line resonance, J. Geophys. Res., 79, 1024-1032, 1974.
Christensen-Dalsgaard, J., D.O. Gough, and J. Toomre, Science, 229, 923-931, 1985.
Cowley, S.W.H, and M. Lockwood, Excitation and decay of solar-wind driven flows in the 
 magnetosphere-ionosphere system, Annales Geophys., 10, 103-115, 1992.
Dungey, J.W., Interplanetary magnetic field and the auroral zones, Phys. Rev. Lett., 6, 47, 
 1961.
Dungey, J.W., Origin of the concept of reconnection and its application to the magnetopause:
 A historical view, Physics of the magnetopause, eds., P. Song, B.U.O. Sonnerup, and M.F.
 Thomsen, AGU monograph 90, 17-19, 1995
Fenrich, F.R., J.C. Samsom, G. Sofko, and R.A. Greenwald, ULF high and low-m field line
 resonances observed with SuperDARN, J. Geophys. Res., 100, 21535-21547, 1995.
Fasel, G.J., Dayside poleward moving auroral forms: A statistical study, J. Geophys. Res., 100,
 11891-11905, 1995.
Fasel, G.J., J.I. Minow, R.W. Smith, C.S. Deehr, and L.C. Lee, Multiple brightenings of 
 transient dayside auroral forms during oval expansions, Geophys. Res. Lett., 12, 2429-2432,
 1992.
Friis-Christensen, E., and J. Wilhjelm, Polar cap currents for different directions of the 
 interplanetary magnetic field in the Y-Z plane, J. Geophys. Res., 80, 1248-1260, 1975.
Goertz, C.K., E. Nielsen, A. Korth, K.H. Glassmeier, C. Haldoupis, P. Hoeg, and D. 
 Hayward, Observations of a possible ground signature of flux transfer events, J. Geophys.
 Res., 90, 4069-4078, 1985. 
Greenwald, R.A., K.B. Baker, J.R. Dudeney, M. Pinnock, T.B. Jones, E.C. Thomas, J.-C. 
 Villain, J.-C. Cerrisier, 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, 761-796, 1995.
Kan, J.R. A theory of patchy and intermittent reconnection for magnetospheric flux transfer 
 events, J. Geophys. Res., 93, 5613, 1988.
Kan, J.R., C.S. Deehr, L.H. Lyu, and P.T. Newell, Ionospheric signatures of patchy-
 intermittent reconnection at the dayside magnetopause, J. Geophys. Res., 101, 10939-10945,
 1996.
Kuo, H., C.T. Russell, and G. Lee, Statistical studies of flux transfer events, J. Geophys. Res.,
 100, 3513-3519, 1995.
Lee, L.C. and Z.F. Fu, Atheory of magnetic flux transfer at the Earth's magnetopause, 
 Geophys. Res. Lett., 12, 105-108, 1985.
Lockwood, M., The case for transient magnetopause reconnection, Eos, Transactions, 77, 246,
 1996.
Lockwood, M., and M.N. Wild, On the quasi-periodic nature of magnetopause flux transfer 
 events, J. Geophys. Res., 98, 5935-5940, 1993.
Lockwood, M., P.E. Sandholt, S.W.H. Cowley, and T. Oguti, Interplanetary magnetic field 
 control of dayside auroral activity and the transfer of momentum across the dayside
 magnetopause, Planet. Space Sci., 67, 1347-1365, 1989. 
Lockwood, M., S.W.H. Cowley, P.E. Sandholt, and R.P. Lepping, The ionospheric signatures
 of flux transfer events and solar wind dynamic pressure changes, J. Geophys. Res., 95, 17113,
 1990.
Lockwood, M., W.F. Denig, A.D. Farmer, V.N. Davda, S.W.H. Cowley, and H. Luhr, 
 Ionospheric signatures of pulsed reconnection at the Earth's magnetopause, Nature, 361,
 424-428, 1993.
Lysak, R.L., Y. Song, and D.-H. Lee, Generation of ULF waves by fluctuations in the 
 magnetopause position, AGU Monograph on Solar Wind Sources of Magnetospheric Ultra-
 Low-Frequency Waves, 81, 273-281, 1994.
Newell, P.T. and Ch.-I. Meng, Ionospheric projections of magnetospheric regions under low
 and high solar wind pressure conditions, J. Geophys. Res., 99, 273-286, 1994.
Nielsen, E. and K. Schlegel, A first comparison of STARE and EISCAT electron drift velocity
 measuremenmts, J. Geophys. Res., 88, 5745-5750, 1983. 
Opgenoorth, H.J., I. Haggstrom, P.J.S. Williams, and G.O.L. Jones, Regions of strongly 
 enhanced perpendicular electric fields adjacent to auroral arcs, J. Atmos. Terr. Phys., 52,
 449-458, 1990.
Papitashvili, C.R. Clauer, A.E. Levitin and B.A. Belov, Relationship between the observed 
 and modeled modulation of the dayside ionospheric convection by the IMF By component, J.
 Geophys. Res., 100, 7715-7722, 1995.
Paschmann, G., et al., Plasma acceleration at the Earth's magnetopause: Evidence for
reconnection, Nature, 282, 243, 1979.
Pinnock, M., A.S. Rodger, J.R. Dudeney, R.A. Greenwald, K.B. Baker, J.M. Ruohoniemi, An
 ionospheric signature of possible enhanced magnetic field merging on the dayside
 magnetopause, J. Atmos. Terr. Phys., 53, 201-212, 1991.
Pinnock, M., A.S. Rodger, J.R. Dudeney, K.B. Baker, P.T. Newell, R.A. Greenwald, and 
 M.E. Greenspan, Observations of an enhanced convection channel in the cusp ionosphere, J.
 Geophys. Res., 98, 3767-3776, 1993. 
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. Geophys., 13, 919-925, 1995. 
Potemra, T.A., L.J. Zanetti, R. Elphinstone, J.S. Murphree, and D.M. Klumpar, The pulsating
 magnetosphere and flux transfer events, Geophys. Res. Lett., 19, 1615-1618, 1992.
Prikryl, P., D. André, J.A. Koehler, G.J. Sofko, J.-P. St.-Maurice, D.D. Wallis and F. 
 Creutzberg, Type 4 VHF radio aurora observations during low to moderate geomagnetic
 activity,  Planet. Space Sci., 40, 1371-1389, 1992. 
Ruohoniemi, J.M., R.A. Greenwald, K.B. Baker, and J.C. Samson, HF radar observations of
 Pc5 field line resonances in the midnight/early morning MLT sector, J. Geophys. Res., 96,
 15697-15710, 1991.
Russell, C.T., and R.C. Elphic, Initial ISEE magnetometer results: Magnetopause observations,
 Space Sci. Rev., 22, 681-715, 1978.
Samson, J.C., J.A. Jacobs, and G. Rostoker, Latitude-dependent characteristics of long-period
 geomagnetic micropulsations, J. Geophys. Res., 76, 3675, 1971.
Samson, J.C., R.A. Greenwald, J.M. Ruohoniemi, T.J. Hughes, and D.D. Wallis,
 Magnetometer and radar observations of magnetohydrodynamic cavity modes in the earth's
 magnetosphere, Can. J. Phys., 69, 929, 1991. 
Samson, J.C., B.G. Harrold, J.M. Ruohoniemi, and A.D.M. Walker, Field line resonances 
 associated with MHD waveguides in the magnetosphere, Geophys. Res. Lett., 19, 441-444,
 1992a.
Samson, J.C., D.D. Wallis, T.J. Hughes, F. Creutzberg, J.M. Ruohoniemi, and R.A. 
 Greenwald, Substorm intensifications and field line resonances in the nightside magnetosphere,
 J. Geophys. Res., 97, 8495-8518, 1992b. 
Sandholt, P.E., M. Lockwood, T. Oguti, S.W.H. Cowley, K.S.C. Freeman, B. Lybekk, A. 
 Egeland, and D.M. Willis, Midday auroral breakup events and related energy and momentum
 transfer from the magnetosheath, J. Geophys. Res., 95, 1039, 1990.
Scholer, M., Magnetic flux transfer at the magnetopause based on single X line bursty 
 reconnection, Geophys. Res. Lett., 15, 291-294, 1988. 
Smith, M.F., and M. Lockwood, The pulsating cusp, Geophys. Res. Lett., 17, 1069-1072, 1990.
Southwood, D.J., Some features of field line resonances in the magnetosphere, Planet. Space
 Sci., 22, 483-491, 1974.
Southwood, D. J., C. J. Farrugia, and M. A. Saunders, What are flux transfer events?, Planet.
 Space Sci., 36, 503-508, 1988.
Stauning, P., E. Friis-Christensen, O. Rasmussen, and S. Vennerstrom, Progressing polar 
 convection disturbances: Signature of a open magnetosphere, J. Geophys. Res., 99, 11303-
 11317, 1994.
Stauning, P., C.R. Clauer, T.J. Rosenberg, E. Friis-Christensen, and R. Sitar, Observations of
 solar-wind-driven progression of interplanetary magnetic field By-related dayside ionospheric
 disturbances, J. Geophys. Res., 100, 7567-7585, 1995.
St.-Maurice, J.-P., P. Prikryl, D.W. Danskin, A.M. Hamza, G.J. Sofko, J.A. Koehler, A. 
 Kustov, and J. Chen, On the origin of narrow non-ion-acoustic coherent radar spectra in the
 high-latitude E region, J. Geophys. Res., 99, 6447-6474, 1994.
Taguchi, S., M. Sugiura, J.D. Winningham, and J.A. Slavin, Characterization of the IMF By-
 dependent field-aligned currents in the cleft region based on DE 2 observations, J. Geophys.
 Res., 98, 1393-1407, 1993.
Thomson, D.J., C.G.Maclennan, and L.J. Lanzerotti, Propagation of solar oscillations through
 the interplanetary medium, Nature, 376, 139-144, 1995.
Walker, A.D.M., The Kelvin-Helmholtz instability in the low-latitude boundary layer, Planet.
 Space Sci., 29, 1119-1133, 1981.
Walker, A.D.M., R.A. Greenwald, W.F. Stuart, and C.A. Green, STARE auroral radar 
 observations of Pc 5 geomagnetic pulsations, J. Geophys. Res., 84, 3373-3388, 1979.
Walker, A.D.M., J.M. Ruohoniemi, K.B. Baker, and R.A. Greenwald, and J.C. Samson, 
 Spatial and temporal behavior of ULF pulsations observed by the Goose Bay HF radar, J.
 Geophys. Res., 97, 12187-12202, 1992. 
Ziesolleck, C.W.S., and D.R. McDiarmid, Auroral latitude Pc 5 field line resonances: 
 Quantized frequencies, spatial characteristics, and diurnal variation, J. Geophys. Res., 99,
 5817-5830, 1994.
Ziesolleck, C.W.S., and D.R. McDiarmid, Statistical survey of auroral latitude Pc5 spectral and
 polarization characteristics, J. Geophys. Res., 100, 19299-19312, 1995.
Ziesolleck, C.W.S., Q. Feng, and D.R. McDiarmid, Pc5 ULF waves observed simultaneously
 by GOES 7 and the CANOPUS magnetometer array, J. Geophys. Res., in press, 1996.


Figure captions

Figure 1.  A series of radar velocity maps showing a sequence of flow channel events. 

Figure 2.  Line-of-sight velocity (a) and spectral width (b) for beam #10. 

Figure 3.  Contour plots of 32-point fast Fourier transform (FFT) power spectra (left panels) of
line-of-sight velocity time series at several range gates for beam #10. The right panels show
latitude profiles of the spectra power and phase at 1.5 and 2.5 mHz.

Figure 4.  The interplanetary magnetic field measured by Wind.

Figure 5.  32-point FFT and MEM power spectra of the IMF components measured by Wind
on January 24, 1996.

Figure 6.  Line-of-sight velocity for beam #4 showing the main FLR event followed by a
pulsation with a shorter period (indicated by arrows).
                                                        
Figure 7.  Schematic representation of the solar wind-magnetosphere coupling via the magnetic
reconnection at the dayside magnetopause and field line resonances (FLRs) on the closed field
lines near the magnetopause. The solar-wind-driven FLRs are shown modulating the
reconnection into pulses. The inset shows the ionospheric signatures of the solar wind-
magnetosphere coupling in the auroral oval, cusp and polar cap for the IMF By>0. 

Figure 8.  64-point FFT power spectra of the IMF components measured by IMP 8 on August
2/3, 1991.

Figure 9.  128-point FFT power spectra of the IMF components measured by IMP 8 on August
3, 1991. Eigenfrequencies (p, f and g modes) of a solar model for spherical harmonic degree
l~50 are shown for comparison.