Canadian Auroral Network For The OPEN Program Unified Study Research Articles

Dipolarization fronts and associated auroral activities: 1. Conjugate observations and perspectives from global MHD simulations

Earthward‐propagating dipolarization fronts (DFs) are often found to be associated with magnetic reconnection and bursty bulk flows (BBFs) in the magnetotail. Recent THEMIS (Time History of Events and Macroscale Interactions During Substorms) probe observations have shown a DF propagating over 10REfrom the mid‐tail region to the near‐Earth tail region, and THEMIS All‐Sky Imager data show a north‐south auroral form and intensification of westward auroral zone currents. In this study, we examine THEMIS in situ observations of DFs in the magnetotail and simultaneous observations of the proton aurora from ground‐based CANOPUS (the Canadian Auroral Network for the OPEN Program Unified Study) Meridian Scanning Photometers (MSPs). We find that earthward‐moving DFs are often associated with intensification of proton aurora when the THEMIS probes are conjugate to the meridian of the MSP. The proton auroral intensifications are transient and in some cases detached from the background proton precipitation. Just before the DFs, the ion distribution is anisotropic in the field‐aligned direction (mostly earthward) and the ion energy increases. These observations suggest that plasma sheet protons can be reflected and energized by earthward‐moving DFs as they propagate through the magnetotail. We postulate that this population of ions is the source of the proton auroral intensification observed on the ground. This conjecture is tested using our global MHD simulation results, where the proton precipitation is calculated with the field‐line curvature (FLC) model. The MHD simulation results show that proton precipitation enhancement can be caused by compression of plasma by approaching DFs/BBFs, which is consistent with ion reflection at DFs. Thus, using the conjugate observations from THEMIS spacecraft and MSP in this study, we are able to directly link the magnetotail dynamics, i.e., dipolarization fronts, with ground auroral activities. However, understanding of DF‐associated ion energization requires detailed test‐particle simulations with an analytical magnetotail model, such as those in our companion paper.

The Upgraded CARISMA Magnetometer Array in the THEMIS Era

This review describes the infrastructure and capabilities of the expanded and upgraded Canadian Array for Realtime InvestigationS of Magnetic Activity (CARISMA) magnetometer array in the era of the THEMIS mission. Formerly operated as the Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) magnetometer array until 2003, CARISMA capabilities have been extended with the deployment of additional fluxgate magnetometer stations (to a total of 28), the upgrading of the fluxgate magnetometer cadence to a standard data product of 1 sample/s (raw sampled 8 samples/s data stream available on request), and the deployment of a new network of 8 pairs of induction coils (100 samples per second). CARISMA data, GPS-timed and backed up at remote field stations, is collected using Very Small Aperture Terminal (VSAT) satellite internet in real-time providing a real-time monitor for magnetic activity on a continent-wide scale. Operating under the magnetic footprint of the THEMIS probes, data from 5 CARISMA stations at 29–30 samples/s also forms part of the formal THEMIS ground-based observatory (GBO) data-stream. In addition to technical details, in this review we also outline some of the scientific capabilities of the CARISMA array for addressing all three of the scientific objectives of the THEMIS mission, namely: 1. Onset and evolution of the macroscale substorm instability, 2. Production of storm-time MeV electrons, and 3. Control of the solar wind-magnetosphere coupling by the bow shock, magnetosheath, and magnetopause. We further discuss some of the compelling questions related to these three THEMIS mission science objectives which can be addressed with CARISMA.

Electrodynamics of magnetosphere‐ionosphere coupling and feedback on magnetospheric field line resonances

We present a new dynamic model that describes coupling between standing inertial or ion‐acoustic‐gyroradius‐scale shear Alfvén waves, compressional modes, and auroral density disturbances. The model is applied to the excitation of field line resonances (FLRs) in dipolar and stretched geomagnetic fields in Earth's magnetosphere. Magnetosphere‐ionosphere coupling is included by accounting for the closure of magnetospheric field‐aligned currents (FACs) through Pedersen currents in the ionosphere. A second new aspect is that the height‐integrated Pedersen conductivity is treated as a dynamic parameter by electrodynamically coupling the two‐dimensional finite element wave model “Topo” to the ionospheric ionization model “Global Airglow Model (GLOW).” We demonstrate that field line stretching brings the equatorial plasma β above unity, where the reduced MHD formulism for low‐frequency plasma breaks down. As an application of our model, we study a specific FLR event observed on 31 January 1997, when the NASA FAST satellite was over the Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) Gillam station. Using geomagnetic fields computed from the T96 magnetic field model, we show that auroral electron precipitation produces strong Pedersen conductivity enhancements that control the final amplitude and width of the excited FLR, along with the amplitude of associated density fluctuations. The predictions of the model are generally consistent with observations of this event.

On the equatorward motion and fading of proton aurora during substorm growth phase

Using 50 high‐quality events from the Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) Gillam Meridian Scanning Photometer over 10 a (1989–1998), we show that proton aurora in the substorm growth phase exhibits a systematic tendency to fade. The average pattern of fading consists of a period of relatively stable proton aurora brightness and then a period of 15–20 min before onset during which the brightness decreases by an average of 15%. We interpret the observed proton aurora brightness variation in terms of the magnetic field stretching in the near‐Earth magnetosphere; in particular, the fading is interpreted as a result of the central plasma sheet magnetic field lines having stretched to such a degree that the loss cone closing effect dominates precipitation due to nonadiabatic proton motions.

Ground‐based optical determination of the b2i boundary: A basis for an optical MT‐index

The equatorward boundary of the proton aurora corresponds to a transition from strong pitch angle scattering to bounce trapped particles. This transition has been identified as the b2i boundary in Defense Meteorological Satellite Program (DMSP) ion data [Newell et al., 1996]. We use ion data from 29 DMSP overflights of the Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) Meridian Scanning Photometer (MSP) located at Gillam, Canada, to develop a simple algorithm to identify the b2i boundary in latitude profiles of proton auroral (486 nm) brightness. Applying this algorithm to a ten year set of Gillam MSP data, we obtain ∼250,000 identifications of the “optical b2i,” the magnetic latitude of which we refer to as b2iΛ. We intercompare ∼1600 near‐simultaneous optical and in situ b2iΛ, concluding that the optical b2iΛ is a reasonable basis for an optical equivalent to the MT‐index put forward by Sergeev and Gvozdevsky [1995]. Using ∼17,000 simultaneous measurements, we demonstrate a strong correlation between the optical b2iΛ and the inclination of the magnetic field as measured at GOES 8. We develop an empirical model for predicting the GOES 8 inclination, given the universal time, dipole tilt, and the optical b2iΛ, as determined at Gillam. We also show that in terms of information content, the b2i boundary is an optimal boundary upon which to base such an empirical model.

Field line resonances in a stretched magnetotail: CANOPUS optical and magnetometer observations

Ground‐based observations support the existence of 1–4 mHz field line resonances (FLRs) at auroral latitudes. The low frequencies suggest that many nightside FLRs are excited on stretched magnetic field lines in the inner plasma sheet. Data from the Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) magnetometers and photometers on 21 February 1993 indicate a FLR with a peak at 2.4 mHz, and that of 31 January 1997 indicate a peak at 1.3 mHz. In this paper, models of FLRs and models of stretched magnetotail topologies deduced from optical data are compared with the observations. The FLR model is independently constrained by a parametric near‐Earth magnetotail model that allows us to make direct comparisons with the observed FLR frequency. For the two case studies mentioned above, we are able to show that the theoretical FLR spectrum, computed on stretched field lines, is comparable to the low‐frequency experimental observations.

Observations of highly correlated near‐simultaneous magnetic field perturbations at contraposed ground stations

A merged data set from Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) and Magnetometers‐Ionospheric Radars‐All‐Sky Cameras Large Experiment (MIRACLE) magnetometer stations is used to study variations observed at widely separated locations. Three years of 10‐s data are analyzed using canonical correlation techniques to determine the temporal offset of similar fluctuations observed at pairs of sites. Nearly simultaneous (<2‐min delay) signatures are often found between sites separated by 12 hours magnetic local time, with no evidence of more slowly propagating events (> 10‐min delay). Correlated episodes occur more often (roughly 10% of the time) when sites are in the dawn/dusk sectors and less often (∼5%) for noon/midnight configurations.

Dynamics of the substorm expansive phase

Meridian‐scanning photometer (MSP) and magnetometer data from the Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) ground‐based array have been used to study the dynamics of the substorm expansive phase. The relative latitudinal motions of the MSP 630.0, 557.7, and 486.1 nm emissions for eight isolated events have been studied. The data show the expansive phase to comprise three stages: an explosive (tens of seconds) onset; rapid poleward motion of 557.7 nm emissions of the order of a few minutes; and a period of slower (tens of minutes) poleward moving 630.0 nm emissions. We interpret the rapid poleward motion of the 557.7 nm data in terms of a region of instability, expanding rapidly down the magnetotail, possibly accelerating plasma sheet electrons as it proceeds. All events show that lobe flux reconnection occurs after near‐Earth onset, with typical time delays between 1 and 5 min. The extremely short time interval between near‐Earth onset and the beginning of lobe flux reconnection presents a severe observational constraint on any substorm model which attempts to explain the substorm expansive phase. Magnetometer data from the CANOPUS Churchill line have been used to study the dynamics of the substorm electrojets during the growth and early expansive phases of the substorm. The borders of these electrojets closely follow the motion of the MSP data during the growth and expansive phases. At the instant of expansive phase onset the growth phase electrojets (as observed along the Churchill meridian line) disappear, being replaced by the substorm westward electrojet associated with the current wedge. The equatorward border of the substorm westward electrojet is embedded within the proton aurora emissions and appears at expansive phase onset ∼2° equatorward of the growth phase electrojets, further showing the near‐Earth proximity of expansive phase onset. No indication of reconnection on closed field lines prior to expansive phase onset was observed in either the magnetometer or the photometer data.

Poleward progressing quasiperiodic disturbances at cusp latitudes: The role of wave processes

Observations from various magnetometer networks, including Magnetomer Array for Cusp and Cleft Studies (MACCS), Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS), the Greenland Coastal Chain, and the U.S./Russia Antarctic Array, made on January 5, 1995, show an event of very long period pulsations (30‐ to 40‐min period) in the cusp region. Pulsations were driven by quasiperiodic Interplanetary Magnetic Field (IMF) By variations which were observed by the Wind spacecraft. Disturbances of this type are commonly interpreted as poleward moving, east–west oriented ionospheric currents (e.g., intensification of the DPY current system), stimulated by reconnection processes at the dayside magnetopause. However, the temporal/spatial, ground‐based structure of such disturbances can better be described if this picture is augmented by the inclusion of transient wave processes, for example, distortions of the Alfven phase front, transmitting a disturbance from an assumed reconnection region to the ionosphere. In general, this type of disturbance, which we suggest calling PDPY6 pulsations, is shown to be a manifestation of the modulation of the high‐latitude ionosphere electrodynamics by a large‐scale Alfven wave in the solar wind under favorable IMF orientation.

On the use of photometer data to map dynamics of the magnetotail current sheet during substorm growth phase

The Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) is a large‐scale ground‐based instrument array of remote sensing equipment monitoring the high‐latitude ionosphere from the north central to the northwest portion of North America. It provides realtime coverage of auroral events such as magnetospheric substorms. In this paper the variation of magnetotail thickness is estimated during substorm growth phase by using a simple magnetic field model and data from the meridian scanning photometers in the CANOPUS array. An important parameter in the modeling process is the κ parameter, defined as the square root of the minimum curvature radius to maximum Larmor radius ratio. Through use of this parameter we are able to grossly determine the expected proton precipitation regions in the ionosphere. Two examples presented (February 9, 1995, and March 9, 1995) show that crosstail current sheet thinning from 2–0.1 RE can occur during the course of a growth phase. Within the framework of our model it is demonstrated that the magnetotail field line stretching, associated with thinning of the current sheet, is correlated with the equatorward motion of Hβ auroral emissions. Also demonstrated is the Earthward movement of the inner edge of the plasma sheet to within 6 RE.

The November 9, 1993, traveling convection vortex event: A case study

On November 9, 1993, at around 1715 UT, a strong and well‐structured traveling convection vortex (TCV) event occurred and was observed by the Magnetometer Array for Cusp and Cleft Studies (MACCS) and the Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) magnetometer chains. The Greenland chain, which is located farther to the east, observed only a very weak signature for the same event. We studied the propagation characteristics of the two‐dimensional vortical TCV current patterns. The TCV event is a series of four vortical currents propagating westward. We found that individual vortices propagate with different speeds. The vortices are created in the early postnoon region. They accelerate as they strengthen, and some decelerate, weaken, and disappear within the 6 hours of magnetic local time of the field of view of the ground magnetometers. The strongest, main vortex of the event accelerates until it moves out of the field of view and, more than likely, reaches well into the nightside. We studied the correlated solar wind and IMF signatures as observed by the IMP 8 spacecraft, sitting in the far dawnside and outside the bow shock. We found that the transient currents in the ionosphere are the result of sharp, short‐lived pressure pulses hitting the magnetopause during times of quiet and northward IMF that is primarily radial. We find that the pressure pulses are more than likely created just upstream of the bow shock by the interaction of the quasi‐parallel shock with IMF orientation changes and are not intrinsic features of the upstream solar wind. We also analyze the transient signatures in the inner magnetosphere by studying the magnetic field data in the GOES 6 and 7 satellites. We suggest that a series of five successive compression and depression peaks in the GOES magnetic field data are well correlated with the set of solar wind pressure pulses. We observe a propagation velocity of the transient event from the GOES 6 spacecraft to the GOES 7 spacecraft that agrees well with the propagation velocities that we calculate from the ground magnetometer stations.

Plasma mass density in the plasmatrough: Comparison using ULF waves and CRRES

Observations of ULF field line resonances detected by seven magnetometers in the Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) array were used to estimate equatorial cold plasma mass densities in and around the plasmatrough region as a function of both radial distance and time of day. The magnetometers ranged in latitude from Pinawa (L=4.3) to Rankin Inlet (L=12.4). The equatorial cold plasma mass densities were estimated using the T89 magnetic field model, an R−4 plasma density distribution function and the wave equation from cold plasma MHD theory. These ground based plasma mass density estimates were compared to those obtained from the CRRES spacecraft. Typical values obtained from the ground based measurements were ∼2 H+cm−3 at L ∼10 to ∼10 H+cm−3 at L ∼ 7. The data recorded on 3 October, 1990 showed excellent agreement with the spacecraft observations for 8.5 < L < 10. It was found that the differences between ground based estimates of plasma mass density and the values derived from spacecraft observations increased in regions of large plasma mass density spatial gradients. This result may be explained in terms of the spacing of magnetometers in the ground based array.

Latitudinal dynamics of auroral roar emissions

Auroral roar, a narrowband (δƒ/ƒ < 0.1) emission near 2 and 3 times the ionospheric electron gyrofrequency (2ƒce and 3ƒce), is observed with a meridional chain of LF/MF/HF radio receivers located in northern Canada spanning 67° to 79° invariant latitude. Observations of these emissions are compared with the auroral electrojet location inferred from the Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) magnetometer array. Variations in the intensity of the observed auroral roar emissions and in the invariant latitude of the most intense emissions are correlated with movements of the poleward boundary of the electrojet. For example, substorm onsets, which appear as rapid poleward expansions of this boundary, result in screening of the emissions from the underlying ground stations because of precipitation‐induced ionization in the lower ionosphere. In four of the five study days the peak emission intensity is located 0°–9° poleward of the poleward electrojet boundary inferred from the magnetometers. In one case the peak emission intensity is up to 10° equatorward of the poleward electrojet boundary. In all cases, there is a tendency for the latitude of the most intense auroral roar emissions to track the movements of the electrojet location inferred from the magnetometer data. For two examples, the footprint of the Fast Auroral Snapshot (FAST) satellite passes within 3° of one or more of the ground stations, and the satellite detects unstable electron populations in the polewardmost auroral arc, reinforcing the scenario that auroral roar emissions are generated by these electrons in the polewardmost arc and propagate into the polar cap where conditions are often favorable for their detection at ground level.

A study of omega bands and Ps6 pulsations on the ground, at low altitude and at geostationary orbit

We investigate the electrodynamic coupling between auroral omega bands and the inner magnetosphere. The goal of this study is to determine the features to which omega bands map in the magnetosphere. To establish the auroral‐magnetosphere connection, we appeal to the case study analysis of the data rich event of September 26, 1989. At 6 magnetic local time (MLT), two trains of Ps6 pulsations (ground magnetic signatures of omega bands) were observed to drift over the Canadian Auroral Network For the OPEN Program Unified Study (CANOPUS) chain. At the same time periodic ionospheric flow patterns moved through the collocated Bistatic Auroral Radar System (BARS) field of view. Similar coincident magnetic variations were observed by GOES 6, GOES 7 and SCATHA, all of which had magnetic foot points near the CANOPUS/BARS stations. SCATHA, which was located at 6 MLT, 0.5 RE earthward of GOES 7 observed the 10 min period pulsations, whereas GOES 7 did not. In addition, DMSP F6 and F8 were over‐flying the region and observed characteristic precipitation and flow signatures. From this fortunate constellation of ground and space observations, we conclude that auroral omega bands are the electrodynamic signature of a corrugated current sheet (or some similar spatially localized magnetic structure) in the near‐Earth geostationary magnetosphere.

Association between Geotail plasma flows and auroral poleward boundary intensifications observed by CANOPUS photometers

Poleward boundary intensifications are nightside geomagnetic disturbances that have an auroral signature that moves equatorward from the poleward boundary of the auroral zone. They occur repetitively, so that many individual disturbances can occur during time intervals of ∼1 hour, and they appear to be the most intense auroral disturbance at times other than the expansion phase of substorms. We have used data from three nightside conjunctions of the Geotail spacecraft in the magnetotail with the Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) ground‐based array in central Canada to investigate the relation between the poleward boundary intensifications and bursty plasma sheet flows and to characterize the bursty flows associated with the disturbances. We have found a distinct difference in plasma sheet dynamics between periods with, and periods without, poleward boundary intensifications. During periods with identifiable poleward boundary intensifications, the plasma sheet has considerable structure and bursty flow activity. During periods without such poleward boundary intensifications, the plasma sheet was found to be far more stable with fewer and weaker bursty flows. This is consistent with the intensifications being the result of the mapping to the ionosphere of the electric fields that give rise to bursty flows within the plasma sheet. Two different types of plasma sheet disturbance have been found to be associated with the poleward boundary intensifications. The first consists of plasma sheet flows that appear to be the result of Speiser motion of particles in a localized region of thin current sheet. The second, seen primarily in our nearest‐to‐the‐Earth example, consists of energy‐dispersed ion structures that culminate in bursts of low‐energy ions and isotropic low‐energy electrons and are associated with minima in magnetic field and temperature and maxima in ion density and pressure. Both types of plasma sheet disturbance are associated with localized regions of enhanced dawn‐to‐dusk electric fields and appear to be associated with localized enhanced reconnection. Our analysis has shown that poleward boundary intensifications are an important aspect of geomagnetic activity that is distinct from substorms. In addition to their very distinct auroral signature, we have found them to be associated with a prolonged series of ground magnetic Pi 2 pulsations and ground X component perturbations, which peak at latitudes near the ionospheric mapping of the magnetic separatrix, and with a series of magnetic Bz oscillations near synchronous orbit. Like substorms, the tail dynamics associated with the poleward boundary intensifications can apparently extend throughout the entire radial extent of the plasma sheet. Color versions of figures are available at http://www.atmos.ucla.edu/∼larry/geotail.html.

Coordinated ISTP satellite and ground observations of morningside Pc5 waves

This paper reports the result of a coordinated data analysis of a morningside Pc5 event observed at different altitudes in the magnetosphere and also on the ground. The event took place during 1400–1500 UT of April 29, 1993. The Geotail satellite was located in the boundary region and observed a 5‐min quasi‐periodic magnetic oscillation. The oscillation was mostly transverse to the background magnetic field. A 90° phase lag between the magnetic field and electric field variations was not clear, suggesting that the oscillation was not a standing wave and that Geotail was located in or close to the excitation region. The plasma flow vector rotated clockwise on the equatorial plane viewed from the north as expected for a magnetospheric surface wave on the morningside. At geosynchronous altitude, the GOES satellites also observed a 5‐min magnetic oscillation but with a significantly smaller amplitude than at the Geotail position. Five‐minute magnetic oscillations were also detected at Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) and Magnetometer Array for Cusp and Cleft Studies (MACCS) ground stations in the same local time sector as the satellites, even equatorward of a region 2 field‐aligned current observed by the Freja magnetometer data. From the phase analysis of ground signatures, the wave is inferred to propagate westward (antisunward) at a velocity of 18° in longitude per minute. The propagation speed mapped to the equator, 400 km/s, is in the range of the expected flow speed of the magnetosheath. It is inferred that in the present event, the Kelvin‐Helmholtz instability at the magnetopause, rather than at the inner edge of the boundary layer, excited an oscillation at the single frequency in a large area from the boundary region to deep inside the magnetosphere.

Flux rope structures in the magnetotail: Comparison between Wind/Geotail observations and global simulations

On December 11, 1994, Geotail observed several reconnection (negative Bz) events at a position 30–40 RE down the tail over a 24 hour period. This period is also interesting because the solar wind conditions were well monitored by Wind, which was only about 20 RE in front of the bow shock, and because IMP 8 was at nearly the same x value as Geotail, but at high latitudes. Global fluid simulations that incorporate higher‐order corrections to Ohm's law are used to produce a three dimensional picture of the reconnection site in conjunction with the Geotail data. It is shown that many of the negative Bz events have appreciable core magnetic fields, which in the modeling appear to form around 20 RE about 10 min earlier than observed by Geotail. The predicted time of the formation coincides approximately with increased ionospheric activity observed by the Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) magnetometer chain. The flux ropes also appear to be highly localized (2–4 RE), particularly in x and z when they are first created. In addition, they have a distinctive bipolar core magnetic field, but the probability that a spacecraft observes a unidirectional or bipolar signature is highly dependent on the position of the center of the current sheet relative to the spacecraft. Bipolar signatures are predicted to be most easily seen when the spacecraft is near the center of the current sheet, while unidirectional signatures are most likely to be observed at more distant encounters.

A comparison between Pc 3–4 pulsations observed by GOES 7 and the CANOPUS magnetometer array

We have analyzed 1 month of GOES 7 and Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) magnetometer data and compared Pc 3–4 wave activity seen at geosynchronous orbit with simultaneous ground Pc 3–4 pulsations at auroral latitudes. We found that at geosynchronous orbit the majority of coherent Pc 3–4 events were azimuthally or radially polarized. A number of transverse events revealed clear field line resonance (FLR) signatures in the latitudinal power, polarization, and phase variations. However, the FLR characteristics often appeared to be masked by ionospheric screening and/or wave propagation effects. The azimuthal phase propagation was found to be antisunward for most events, although sunward phase propagation has also been observed. The ground Pc 3–4 occurrence, polarization, and longitudinal phase characteristics associated with transverse waves were found to be similar to those of auroral latitude Pc 5 pulsations. These similarities between transverse Pc 3–4 and auroral latitude Pc 5 characteristics may point to a common source mechanism for these pulsations. One event appeared consistent with the excitation of multiple waveguide modes. Most compressional Pc 3–4 activity appeared to be relatively broadband and virtually unpolarized. However, the center frequency of the dominant band was found to be well correlated with the interplanetary magnetic field magnitude. Power spectra from the ground stations usually contained significant signal power within the frequency band of the compressional perturbation seen by GOES 7. It is shown that at some frequencies, the broadband compressional waves may excite local field line resonances, whereas at other frequencies the compressional signal is transmitted to the ground without obvious resonance structures. The ground Pc 3–4 occurrence, spectral, polarization, and longitudinal phase characteristics associated with compressional waves were found to differ in several respects from the signal characteristics associated with transverse waves. These differences between transverse and compressional Pc 3–4 waves argue against upstream waves being the sole source of auroral latitude Pc 3–4 pulsations.

Variation of plasmatrough density derived from magnetospheric field line resonances

The diurnal variation of ULF field line resonant frequencies has been calculated using the cross phase of data from the north‐south components recorded at seven latitudinally spaced ground magnetometers in the Canadian Auroral Network for the OPEN Program Unified Study (CANOPUS) array. CANOPUS magnetometers range in latitude from Rankin Inlet (L = 12.4) south to Pinawa (L =4.3). Using cold plasma MHD theory, an R−4 plasma density function, and the T87 magnetic field model, the variation of plasma density in the equatorial region has been calculated from the experimentally determined resonant frequencies. Consecutive, adjacent magnetometer pairs provide six local daytime spatial estimates of the variation in plasma mass density between 4 and 11 RE. Typical values are 1–20 H+cm−3 for the plasmatrough and 50–200 H+cm−3 for the plasmasphere. The data recorded on June 7, 1990, shows an afternoon increase in density near geosynchronous orbit in agreement with convection models of the magnetosphere. The ground‐based measurements of plasma mass density have been compared with data from the Los Alamos Magnetospheric Plasma Analyser on board the 1989‐046 geosynchronous spacecraft. These comparisons show that the ground‐based technique should allow a robust procedure for calculating dayside, time‐dependent mappings of the equatorial plasma mass density from the plasmapause to the magnetopause in near real time.

Pc 5 ULF waves observed simultaneously by GOES 7 and the CANOPUS magnetometer array

Applying pure state filtering techniques to 1 month (March 1990) of GOES 7 and CANOPUS (Canadian Auroral Network for the OPEN Program Unified Study) magnetometer data, we have detected several dominantly compressional Pc 5 events in the satellite data which were well correlated with the pulsation signal recorded by the ground stations. On the ground, near the GOES 7 magnetic footprint, these pulsations were seen with polarization characteristics similar to those observed at geosynchronous orbit, i.e., without ionospheric rotation of the transverse polarization major axis. Using data from 13 stations of the CANOPUS array, we show that these compressional wave events seen by GOES 1 at L = 6.67 were associated with field line resonances (FLRs) at higher L values (L ≥ ∼8), propagating antisunward with small azimuthal wave numbers. Besides these compressional waves, we identified a number of azimuthally (east–west) polarized transverse Pc 5 events at geostationary orbit. Near the subsatellite point these appeared on the ground with the polarization major axis aligned in about the north–south direction, i.e., rotated through ∼90° between magnetosphere and ground. The spatial characteristics, as derived from the ground array data, show that these transverse events represent FLRs on or near the satellite's L shell. By means of selected events with clear FLR characteristics in the CANOPUS data, we determine the resonance position from the ground array data and examine the variation of the signal characteristics at geosynchronous orbit with increasing radial distance of GOES 7 from the resonant field line. We find that the transverse wave characteristics, dominant near the resonance, decay on relatively small‐scale lengths and are transformed into compressional wave characteristics within about 1–2 RE from the resonance position. The diurnal variation of the signal parameters for these events suggests that compressional Pc 5 waves at geosynchronous orbit associated with FLRs at L > 8 are preferentially generated near local noon. Finally, via consideration of solutions of the uncoupled toroidal mode wave equations based on the observed signal frequencies and resonance positions, we discuss the field‐aligned harmonic of these resonances.