Electric field measurements from the Dynamics Explorer 2 satellite have been analyzed to extend the empirical models previously developed from dawn‐dusk OGO 6 measurements (J. P. Heppner, 1977). The analysis embraces large quantities of data from polar crossings entering and exiting the high latitudes in all magnetic local time zones. Paralleling the previous analysis, the modeling is based on the distinctly different polar cap and dayside convective patterns that occur as a function of the sign of the Y component of the interplanetary magnetic field. The objective, which is to represent the typical distributions of convective electric fields with a minimum number of characteristic patterns, is met by deriving one pattern (model BC) for the northern hemisphere with a +Y interplanetary magnetic field (IMF) and southern hemisphere with a −Y IMF and two patterns (models A and DE) for the northern hemisphere with a − Y IMF and southern hemisphere with a +Y IMF. The most significant large‐scale revisions of the OGO 6 models are (1) on the dayside where the latitudinal overlap of morning and evening convection cells reverses with the sign of the IMF Y component, (2) on the nightside where a westward flow region poleward from the Harang discontinuity appears under model BC conditions, and (3) magnetic local time shifts in the positions of the convection cell foci. The modeling above was followed by a detailed examination of cases where the IMF Z component was clearly positive (northward). Neglecting the seasonally dependent cases where irregularities obscure pattern recognition, the observations range from reasonable agreement with the new BC and DE models, to cases where different characteristics appeared primarily at dayside high latitudes. The analysis of these differences for strong +Bz conditions, using passes through different MLT zones, led to finding that the deviations from both the BC and DE models could be modeled by extending the evening cell foci toward, and beyond, noon with a deformation that roughly resembles a rotational twist of the extended evening cell. The morning cell accommodation to the model BC deformation produces a narrow tongue of eastward convection near 75° invariant latitude on the dayside. The model DE deformation compresses the morning cell toward the night hours such that the sunward convection in the morning cell is diverted to polar latitudes in the 0400 to 0700 MLT sector. In both cases, the deformations of the two‐cell patterns lead to sunward convection in dayside polar regions while maintaining the integrity of the nightside convection pattern. Thus the nightside dilemmas that plague three‐ and four‐cell models designed to explain sunward convection in polar regions under +Bz conditions do not appear.

[1] We present a comparison between a simple but general model of solar windmagnetosphere-ionosphere coupling (the Hill model) and the output of a global magnetospheric MHD code, the Integrated Space Weather Prediction Model (ISM). The Hill model predicts transpolar potential and region 1 currents from environmental conditions specified at both boundaries of the magnetosphere: at the solar wind boundary, electric field strength, ram pressure, and interplanetary magnetic field direction; at the ionospheric boundary, conductance and dipole strength. As its defining feature, the Hill model predicts saturation of the transpolar potential for high electric field intensities in the solar wind, which accords with observations. The model predicts how saturation depends on boundary conditions. We compare the output from ISM runs against these predictions. The agreement is quite good for non-storm conditions (differences less than 10%) and still good for storm conditions (differences up to 20%). The comparison demonstrates that global MHD codes (like ISM) can also exhibit saturation of transpolar potential for high electric field intensities in the solar wind. We use both models to explore how the strength of solar wind-magnetosphere-ionosphere coupling depends on the strength of Earth's magnetic dipole, which varies on short geological timescales. As measured by power into the ionosphere, these models suggest that magnetic storms might be considerably more active for high dipole strengths. [2] Total region 1 current, I 1 , and transpolar potential, È pc , epitomize solar wind-magnetosphere-ionosphere (SW-M-I) coupling. Progress in understanding this subject can almost be measured by how well the field predicts these quantities. (Region 2 currents, which this paper does not treat, are also an important aspect of the story. In section 7 we discuss how they might affect results presented here.) First models of SW-M-I coupling, reviewed by Reiff and Luhmann [1986], assumed one-way coupling from the solar wind to the ionosphere in which magnetic reconnection at the magnetopause taps a fraction of the solar wind potential across the magnetosphere, È sw , to yield an available magnetospheric convection potential È m . È m is then impressed via equipotential magnetic field lines onto the ionosphere, where it becomes the È pc that generates region 1 currents. The envisioned process was therefore linear. Empirical formulas based on this linear assumption work fairly well, except they tend to overpredict È pc for big values of È sw . This tendency has been called saturation of the transpolar potential at high values [Reiff and Luhmann, 1986;Russell et al., 2000].[3] Hill et al. [1976] presented a model of SW-M-I coupling that manifests saturation intrinsically and at about the observed value. (Hill [1984] developed the implications of the model further. We therefore refer to it as the Hill model.) Saturation is a nonlinear process that, in the Hill model, results from a feedback in which the magnetic field generated by region 1 cu...

[1] It has been known that the fluctuations in the interplanetary magnetic field (IMF) may be oriented in approximately planar structures that are tilted with respect to the solar wind propagation direction along the Sun-Earth line. This tilting causes the IMF propagating from a point of measurement to arrive at other locations with a timing that may be significantly different from what would be expected. The differences between expected and actual arrival times may exceed an hour, and the tilt angles and subsequent delays may have substantial changes in just a few minutes. A consequence of the tilting of phase planes is that predictions of the effects of the IMF at the Earth, on the basis of IMF measurements far upstream in the solar wind, will suffer from reduced accuracy in the timing of events. It has recently been shown how the tilt angles may be determined using multiple satellite measurements. However, since the multiple satellite technique cannot be used with real-time data from a single sentry satellite, then an alternative method is required to derive the phase front angles, which can then be used for more accurate predictions. In this paper we show that the minimum variance analysis (MVA) technique can be used to adequately determine the variable tilt of the plane of propagation. The number of points that is required to compute the variance matrix has been found to be much higher than expected, corresponding to a time period in the range of 7 to 30 min. The optimal parameters for the MVA were determined by a comparison of simultaneous IMF measurements from four satellites. With use of the optimized parameters it is shown that the MVA method performs reasonably well for predicting the actual time lags in the propagation between multiple spacecraft, as well as to the Earth. Application of this technique can correct for errors, on the order of 30 min or more, in the timing of predictions of geomagnetic effects on the ground.

Regions of enhanced cold plasma, isolated from the main plasmasphere along the Explorer 45 (S³‐A) orbit in the equatorial plane, have been detected by using the sheath‐induced potentials seen by the electric field experiment. The occurrence of these regions has a strong correlation with negative enhancements of Dst, and their locations are primarily in the noon‐dusk quadrant. The data support the concept that changes in large‐scale convection play a dominant role in the formation of these regions. Plasma tails that are predicted from enhancements of large‐scale convection electric fields in general define where these regions may be found. More localized processes are necessary to account for the exact configuration and structure seen in these regions and may eventually result in detachment from the main plasmasphere.

The theta aurora is a remarkable configuration of auroral and polar cap luminosities for which a generally sun‐aligned transpolar arc extends contiguously from the dayside to nightside sectors of the auroral oval. Four individual occurrences of theta aurora over earth's northern hemisphere are examined in detail with the global auroral imaging instrumentation on board the high‐altitude, polar‐orbiting spacecraft DE 1. Simultaneous measurements of fields and plasmas with this high‐altitude spacecraft and its low‐altitude, polar‐orbiting companion, DE 2, are examined in order to establish an overview of auroral and polar cap phenomena associated with the appearance of the theta aurora. For these series of observations, two general states of the polar cap are found corresponding to (1) a bright, well‐developed transpolar arc and (2) a dim or absent transpolar arc. During periods of a relatively bright transpolar arc the plasma convection in the polar cap region associated with the transpolar arc is sunward. Elsewhere over the polar cap the convection is antisunward. The convection pattern over the auroral zones and polar cap is suggestive of the existence of four cells of plasma convection. Field‐aligned electron acceleration into the polar atmosphere and field‐aligned current sheets are present in the transpolar arc plasmas. This electron precipitation and these current sheets are relatively absent over the rest of the polar cap region. The transpolar arc plasmas exhibit similar densities and ion compositions relative to those plasmas observed simultaneously over the poleward zone of the auroral oval. The ion compositions include hot H+, He++, and O+ ions and thus are of both ionospheric and solar wind origins. Principal hot ions in the remainder of the polar cap region are H+ and He++, indicating access from the magnetosheath for these ions. Low‐energy electrons identified with a magnetosheath source are also present in this region. The dominant thermal ions in the polar cap region are O+ ions flowing upward from the ionosphere. These thermal ions are heated along magnetic flux tubes within the transpolar arc plasmas. Pairs of current sheets with oppositely directed current densities occur in the transpolar arc region and with magnitudes similar to those associated with the poleward zones of the auroral oval. The upward currents are carried by electrons accelerated by a field‐aligned potential. Funnel‐shaped auroral hiss and broadband electrostatic noise are associated with the presence of the transpolar arc plasmas. Energetic solar electrons are employed to show that the magnetic field lines threading both the transpolar arc and the poleward zone of the auroral oval are probably closed. In contrast, the accessibility of these electrons to the remainder of the polar cap indicates that these polar regions are characterized by a magnetic topology that is connected directly to field lines within the interplanetary medium. Thus the overall character of the transpolar arc region appears to be very similar to that obser...

The characteristics of the hrge-scale electrodynamic parameters, field-aligned currents (FACs), electric fields, and electron prec/pitation, which are associated with auroral substorm events in the nighttime sector, have been obtained through a unique analysis which places the ionospheric measurements of these parameters into the context of a generic substorm determined from global auroral images. A generic bulgetype auroral emission region has been deduced from auroral images taken by the Dynamics Explorer I (DE 1) satellite during a number of isolated substorms, and the form has been divided into six sectors, based on the peculiar emission characteristics in each sector: west of bulge, surge horn, surge, middle surge, eastern bulge, and east of bulge. By comparing the location of passes of the Dynamics Explorer 2 (DE 2) satellite to the simultaneously obtained auroral images, each pass is placed onto the generic aurora. The organization of DE 2 data in this way has systematically clarified peculiar characteristics in the electrodynamic parameters. An upward net current mainly appears in the surge, with litfie net current in the surge horn and the west of bulge. The downward net current/s distributed over wide longitudinal regions from the eastern bulge to the east of bulge. Near the poleward boundary of the expanding auroral bulge, a pair of oppositely directed FAC sheets is observed, with the downward FAC on the poleward side. This downward FAC and most of the upward FAC in the surge and the middle surge are associated with narrow, intense antisunward convection, corresponding to an equatorward d/rected spikelike electric field. This pair of currents decreases in amplitude and latitudinal width toward dusk in the surge and the west of bulge, and the region 1 and 2 FACs become embedded in the sunward convection region. The upward FAC region associated with the spikelike field on the poleward edge of the bulge coincides well with intense electron precipitation and aurora appearing in this western and poleward portion of the bulge. The convection reversal is sharp in the west of bulge and surge horn sectors, and near the high-latitude boundary of the upward region I FAC. In the surge, the convection reversal is near the low-latitude boundary of the upward region 1, with a near stagnation region often extending over a large interval of latitude. In the eastern bulge and east of bulge sectors, the region I and 2 FACs are located in the sunward convection region, while a spikelike electric field occasionally appears poleward of the aurora but usually not associated with a pair of FAC sheets. In the eastern bulge, magnetic field data show complicated FAC distributions which correspond to current segments and filamentary currents. INTRODUCTION Auroral substorms are defined by a systematic sequence of auroral motions and magnetic disturbances [Akasofu, 1964]. Since field-aligned currents (FACs), ionospheric electric fields or convection, charged particle precipitation, and resulting ionospheric conductivity are closely...

[1] Simultaneous measurements of the interplanetary magnetic field (IMF) are obtained at various locations with four spacecraft, ACE, Wind, IMP-8, and Geotail. We have devised a technique whereby the exact propagation delay time between ACE, at the L1 orbit, and each of the other three spacecraft can be derived from these measurements. This propagation delay is determined as a continuously varying function of time; when this measured delay is applied to all three components of the IMF measured by ACE, they will match the other satellites' IMF to a degree that is much better than expected. However, the actual time delays can vary by nearly an hour in either direction from the expected advection delays, and the lag times have significant changes that can occur on a timescale of a few minutes. These results are interpreted as due to the effects of tilted phase fronts that are changing orientation with time. We have used the delay measurements between multiple satellites to calculate the three-dimensional orientation and temporal variations of the phase front. The best fit phase front plane usually lies within 4 R E or less from the four-point measurements, indicating a lag resolution of a minute or less. Computer animations of the time-varying phase fronts are used to illustrate their behavior. Orientations can change on short timescales. Our findings have implications for both basic research and ''space weather'' predictions. These results give a high confidence that the same IMF that is measured near L1 will most likely impact the Earth's magnetosphere, providing ample justification for use of spacecraft data in halo orbit at L1 for monitoring the upstream solar wind prior to its interacting with the magnetosphere. However, there is strong uncertainty in the timing of the arrival of the detailed IMF structures, and these delays will need to be considered.

The electrodynamics of the inner magnetosphere near times of substorm onsets have been investigated using CRRES measurements of magnetic and electric fields, energetic electron fluxes, in conjunction with ground‐based observations. Six events were studied in detail, spanning the 2100 to 0000 MLT sector and L values from 5 to 7. In each case the dawn‐dusk electric field was enhanced over typical background electric fields, and significant, low‐frequency pulsation activity was observed. The amplitudes of the pulsations were larger than the background electric fields. Dusk‐dawn excursions of the cross‐tail electric field often correlated with changes in currents and particle energies at CRRES and with ULF wave activity observed on the ground. Variations of the electric field and Poynting vectors with periods in the Pi 2 range are consistent with bouncing AlfVén waves that provide electromagnetic communication between the ionosphere and plasma sheet. Magnetic signatures of field‐aligned current filaments directed away from the ionosphere, presumably associated with the substorm current wedge, were observed during three orbits. In all cases, ground signatures of substorm expansion were observed at least 5 min before the injection of electrons at CRRES. Field‐aligned fluxes of counter‐streaming, low‐energy electrons were detected after three of the injections. We develop an empirical scenario for substorm onset. The process grows from ripples at the inner edge of the plasma sheet associated with dusk‐dawn excursions of the electric field, prior to the beginning of dipolarization. Energy derived from the braking of the inward plasma convection flows into the ionosphere in the form of Poynting flux. Subsequently reflected Poynting flux plays a crucial role in the magnetosphere‐ionosphere coupling. Substorms develop when significant energy (positive feedback?) flows in both directions, with the second cycle stronger than the initial. Pseudobreakups occur when energy flow in both directions is weak (negative feedback?). “Explosive‐growth‐phase” signatures occur after onset, early in the substorm expansion phase. Heated electrons arrive at the spacecraft while convection is earthward, during or at the end of electromagnetic energy flow away from the ionosphere.