Solar wind/magnetosheath plasma in the magnetosphere can be identified using a component that has a higher charge state, lower density and, at least soon after their entry into the magnetosphere, lower energy than plasma from a terrestrial source. We here survey observations of He2+ ions made by the Magnetospheric Ion Composition Sensor (MICS) of the Charge and Mass Magnetospheric Ion Composition Experiment (CAMMICE) instrument aboard POLAR taken over 3 years. The occurrence probability of these solar wind ions is then plotted as a function of Magnetic Local Time (MLT) and invariant latitude (Lambda) for various energy ranges. For all energies observed by MICS (1.8-21.4 keV) and all solar wind conditions, the occurrence probabilities peaked around the cusp region and along the dawn flank. The solar wind conditions were filtered to see if this dawnward asymmetry is controlled by the Svalgaard-Mansurov Effect (and so depends on the BY component of the Interplanetary Magnetic Field, IMF) or by Fermi acceleration of He2+ at the bow shock (and so depends on the IMF ratio BX/BY). It is shown that the asymmetry remained persistently on the dawn flank, suggesting it was not due to effects associated with direct entry into the magnetosphere. This asymmetry, with enhanced fluxes on the dawn flank, persisted for lower energy ions (below a "cross-over" energy of about 23 keV) but reversed sense to give higher fluxes on the dusk flank at higher energies. This can be explained by the competing effects of gradient/curvature drifts and the convection electric field on ions that are convecting sunward on re-closed field lines. The lower-energy He2+ ions ExB drift dawnwards as they move earthward whereas the higher energy ions curvature/gradient drift towards dusk. The convection electric field in the tail is weaker for northward IMF. Ions then need less energy to drift to the dusk flank, so that the cross-over energy, at which the asymmetry changes sense, is reduced.

We use Equator-S magnetic field measurements at 64 or 128 Hz to examine the small scale structure of dawn-side magnetopause crossings. Using a subset of exceptionally clear crossings with a well defined rise in field magnitude we examine (1) the relative location of the field rotation and field magnitude rise under northward and southward IMF and (2) the occurrence of weak electromagnetic waves at the lower hybrid frequency. We find that under southward IMF the field rotation and ramp in |B| are separated, as expected from magnetopause models, and that in a significant number of cases the field rotation is associated with a deep minimum in |B|. Under northward IMF the rotation and |B| ramp are more often co-located. There are several clear cases where weak electromagnetic waves occur in the Earthward part of the ramp in |B|. We use plasma data measurements, where available, to examine the location of the waves relative to the density ramp. First results suggest that the waves are located within the density ramp.

A case study has been performed on one day of data from six of the northern hemisphere SuperDARN radars, allowing the high-latitude ionosphere to be studied simultaneously over 12 hours of magnetic local time. During this interval reconnection signatures were observed on the dayside, and substorm signatures on the nightside. Combining the observations from the SuperDARN radars with data from WIND and DMSP spacecraft, and using modelling techniques to map the position of the polar-cap boundary (Lewis et al., 1998) and the high-latitude plasma flow (Ruohoniemi and Baker, 1998), has resulted in a detailed study on the voltages across the dayside and nightside merging gaps, the cross polar-cap potential and the size, motion and boundary of the polar cap during a complete Dungey cycle of the creation and destruction of open magnetic flux on the dayside and nightside of the magnetosphere, respectively.

On the 18th August 1999, the Cassini spacecraft performed an Earth swing-by manoeuvre in order to achieve a trajectory to Saturn. During the Cassini Earth encounter a world-wide programme of data-gathering was undertaken whose purpose was to define the prevailing interplanetary and geophysical conditions. This programme included in situ observations in the interplanetary medium, in the outer magnetosphere, at geostationary orbit, remote sensing of the UV aurora, and ground-based measurements of geomagnetic disturbance and ionospheric flow. These data show that during the Cassini outbound passage through the geomagnetic tail the magnetosphere underwent two complete *classic* substorm cycles from growth, through expansion onset and intensification, to recovery. We show that the global data are able to set the in situ Cassini magnetic field data into a rather precise spatial and temporal context which allows a much fuller interpretation of the observations. Specifically, we show that the first indication of the presence of hot flowing plasma at Cassini, sensed at a down-tail distance of ~18RE via the onset of depressed variable fields, occurred in association with an impulsive intensification of the first substorm which resulted in the expansion of the disturbance into the morning local time sector of the spacecraft. Dipolarisation of the field was then observed at geostationary orbit in the same sector, while the field at Cassini itself remained *tail-like*, and did not relax for a further ~35min, until the substorm recovery phase. During recovery, significant magnetic activity was present at high latitudes (~70°-80°) on the nightside, generally expanding from dusk to dawn. The relaxation of the Cassini field was found to occur as this disturbance reached the local time of the spacecraft in the morning sector, and is taken to represent the expansion across the tail from dusk of a final episode of open flux closure and tail deflation. During the second substorm, the field strength at Cassini, now at distances ~30-50RE, showed a general rise during the growth phase, levelled off during expansion (while the IMFBz field was still negative), and declined steadily during an extended ~80min recovery phase. The recovery phase was again characterised on the ground by an interval of high-latitude impulsive disturbances. The decline in the tail field observed by Cassini is a direct indication of the net closure of open lobe flux during this period. At the end of the decline, Cassini made a first transition into the magnetosheath, which we thus infer was related to the recovery phase deflation of the tail diameter.

The ionosphere responds almost immediately to magnetospheric forcing, however, the neutral atmosphere has inertia owing to its far greater mass density. As a consequence the neutral atmosphere can significantly modify the ionosphere-magnetosphere coupling processes such as the energy transferred between them. Thermospheric neutral temperatures measured by FPI from airglow are presented from both auroral and polar cap sites. These are compared to both ISR derived values and modelled temperatures from MSIS. The large differences evident in these comparisons are discussed and the reaction of the measured values to geomagnetic conditions are shown from both sites. We demonstrate the importance of accurate neutral temperatures in monitoring the dissipation of energy and in the derivation of ionospheric parameters.

The characteristics and propagation effects of gravity waves, particularly at high polar latitudes, is investigated using a Global Ray Tracing model (GROGRAT). Using CIRA-86 and ECMWF operational analyses as the background wind fields, gravity waves of varying horizontal wavelengths and horizontal phase speeds are projected upwards from an altitude of 5km. The ray paths, spatial distribution and average flux of the waves at selected altitudes were then examined to help determine geographical distribution and possible source regions for the waves. These initial results aim to support the observations of gravity waves in Antarctica obtained using an airglow imager, recently installed at Halley.

Tidally induced variations in exothermic chemical heating in the upper mesosphere - lower thermosphere region are often not included in the energetics schemes of models that do not incorporate self-consistent composition. A general circulation model is used to investigate the contribution to mean energetics associated with such variations.

The thermosphere of Saturn's moon Titan, due to be explored in detail by the Cassini orbiter in 2004, has a number of similarities with the terrestrial thermosphere, but also many important differences. Previous results from numerical simulations with the Titan General Circulation Model, a joint effort between University College London and Boston University, have found some of Titan's unique temperature and wind features. Self-consistent calculations of global gas diffusion have now been implemented into the model and show that similarities between the diffusion time scales and length of day on Titan lead to unique effects in the diurnal variability of global composition. These latest results are presented and discussed.

Averaged values of the azimuthal component of the magnetic field observed outside the jovian middle magnetosphere equatorial current sheet have been used to derive radial profiles of the radial current intensity over the jovicentric distance range 20-50RJ. Data from four spacecraft flybys have been used, spanning the dawn sector from ~0100 to ~0900MLT (i.e. inbound Pioneer11, and outbound Pioneer10, and Voyagers1 and 2). These profiles have been combined with a recent empirical model of the azimuthal current intensity to estimate the total divergence of the current in the current sheet along the trajectory, and hence the density of the field-aligned current that couples the current sheet and the ionosphere. For the Voyager passes the inferred field-aligned current flows from the ionosphere into the current sheet, with (j||/B)~2-10 x 1013Am2nT1. The inferred field-aligned currents at the ionosphere require significant field-aligned acceleration of thermal magnetospheric electrons (of 2-3keV thermal energy), through voltages of ~40-150kV. The latitude of these currents, together with the estimated electron energy flux deposited in the ionosphere (0.01-0.1Wm2), suggests a direct connection with the main jovian auroral oval. Related currents are present on the Pioneer passes as well, but lie equatorially inside of ~20RJ. Between 20 and 50RJ, the inferred field-aligned currents on these passes are either small (Pioneer-11), or reversed in sense with similar densities (Pioneer-10). The radial profiles of the radial current associated with magnetosphere-ionosphere coupling have also been used to derive radial profiles of the angular velocity of the magnetospheric plasma, for given values of the effective Pedersen conductivity of the jovian ionosphere. Reasonable profiles are obtained for effective conductivities of several tenths of a mho. The Voyager data then indicate slowly falling values from near-rigid corotation at ~20RJ, to ~50% of rigid corotation at ~50RJ. For the same values of the conductivity the Pioneer data indicate smaller angular velocities in the inner region (~70% of rigid corotation), remaining either constant or even increasing with distance. This behaviour may either indicate the presence of some magnetospheric dynamic process operating in the outer magnetosphere on these passes, or could alternatively be due to falling conductivities with distance in the conjugate ionosphere.

The presence of heavy ions in a plasma introduces additional characteristic frequencies one of which is a crossover frequency between each adjacent pair of gyrofrequencies of the ions present. The crossover frequency is controlled by the fractional ion charge densities. The phase speed of the L and R wave modes have the same phase velocity at this frequency. For oblique propagation waves become linearly polarized at the crossover frequency. They reverse their natural polarization from left to right (or vice versa) as the wave frequency transverses the crossover frequency. Here we present polarization analysis of a portion of the Ulysses magnetometer observations in the middle Jovian magnetosphere during the inbound pass and close to a plasma sheet approach. The analysis show transverse waves with frequency between the gyrofrequencies of $SO^+$ and $S^+$. We observe polarization reversal at this frequency, which could be an indication of wave observations at the crossover frequency. On this assumption these observations can then be used to derive heavy ion composition of the resident plasma.

By making use of spare duty cycle in the existing CUTLASS transmitters and by implementing a second receiver channel, we have developed a new system which allows CUTLASS to sound and receive on two frequencies nearly simultaneously. This upgrade will enable us to undertake two different experimental modes at the same time, thereby improving this already powerful system. The first Stereo system has recently been deployed at the CUTLASS Finland site. In this paper we shall discuss the principles behind Stereo, present some of the first results and discuss some of the potential experiments which the UK community will be able to undertake with this new facility.

The CUTLASS statistical database has been used to investigate the spectral width parameter on the nightside measured by the radars in Finland and Iceland. Two distinct distributions are seen in the Finland data which are dependent upon the latitude of observation. The distributions are similar in nature to those seen from the low-latitude boundary layer and cusp on the dayside. A preliminary analysis of the data is given.

The solar wind is a highly turbulent and intermittent medium at frequencies between 10-3 and 10-1 Hz. Various models have been put forward in an attempt to understand this process better, and tested against solar wind data. We have used the Ulysses fast latitude scan data to perform a wide-ranging comparison of three intermittency models - the well-known P model, the lognormal cascade model and a model adapted from atmospheric physics, the G infinity model. They were tested using fits to graphs of the structure function exponents g(q), comparison with a non-linear measure of the deviation of g(q) from the non-intermittent straight line, and using extended self similarity techniques, over a large range of heliolatitudes. Tests of all three models indicated a high level of intermittency in the fast solar wind, and showed a varied structure in the slow wind, indicating that the slow wind has no uniform origin. While the lognormal model performed the best out of the three models, the G infinity model fitted the results surprisingly well since it has no theoretical base in fluid mechanics, and showed that there may very well be important universal geometrical aspects of intermittency over many physical systems.

Two-site observations of interplanetary scintillation provide accurate measurements of solar wind velocity over a wide range of heliocentric distances which are inaccessible to other techniques. The observations can also be used to study the evolution of the microstructure of the solar wind which produces the scintillation. In this paper we present recent results from IPS measurements made using the EISCAT and MERLIN facilities, including measurements of fast and slow wind acceleration, Alfven wave flux in the fast solar wind and the evolution of solar wind microstructure in the fast and slow winds.

During its mission around the poles of the Sun, the Ulysses spacecraft encountered a number of well developed corotating interaction regions (CIRs) at mid-latitudes. In this paper we discuss the character of magnetohydrodynamic (MHD) turbulence within defined regions of these CIRs. We examine magnetic field power spectra at a fixed wavenumber, and the normalised cross-helicity, in order to describe solar wind fluctuations. We find that each CIR has a complicated turbulent signature. Magnetic field power spectra show evidence for the turbulent mixing of plasma at large distances from the Sun, and there are possible signs of instabilities arising at stream interfaces. Statistical analyses of our data set show that there does not appear to be any clear dependence of turbulence in the CIRs upon heliographic latitude, but that there is a clear trend in magnetic field power and Alfvenicity with CIR structure. We also find that the variation of the solar cycle during Ulysses' observations of these CIRs has a clear effect upon the turbulent character of the solar wind, and thus provides further evidence that the source of the majority of turbulent fluctuations in the solar wind is the Sun.

The Ulysses spacecraft has encountered a highly unusual heliospheric magnetic field enhancement of several hours' duration in the high-latitude solar wind. This class of event has been detected in the past only near Venus and the Earth. It has been suggested that the events could represent the crossings of wakes in the solar wind downstream of dust streams. Our analysis of the Ulysses event however suggests that the event's signatures are consistent with the crossing of a comet's ion tail, though no known comet was aligned with Ulysses at the time. We demonstrate a technique for constraining the orbital parameters of the putative comet which could be applied to the data from other events. If shown to be comet tail crossings, such events are more common than previously thought. .

Unlike previous missions to the magnetotail (Imp 6 ISEE 1, ISEE 3, and Geotail), which effectively made observations in the lobe at a single downtail distance, the CASSINI earth swingby allowed, for the first time, near continuous observations covering a range of downtail distances. Bi-directional electrons are found in the northern lobe, constant with previous studies. Enhancements in the electron fluxes are seen in a boundary layer between the lobe and sheath. These enhancements are accompanied by enhancements in ion fluxes travelling tailwards. While generally these results confirm previous understanding that the source of the bi-directional electrons is the strahl component of the solar wind, the lack of direct observation of strahl electrons in the sheath is a problem.

Ionospheric flow signatures of dayside reconnection are a widely observed and accepted phenomenon. Presented here, are data from an interval on November 20 1999, in which convection features which appear to be associated with tail reconnection are observed. These were seen in SuperDARN data during an extended period of northward IMF as an enhanced convection cell in the pre-midnight sector. The context of these observations has been investigated using data from the DMSP and LANL satellites and the SAMNET magnetometer array.

Comparative analyses of CUTLASS SuperDARN HF radar data and ultraviolet images of the dayside aurora from the Polar UVI and VIS Earth Camera instruments have revealed that the GSM y component of the interplanetary magnetic field (By) affects the relative positions of ultraviolet auroral emissions and the HF radar signature of the cusp in the postnoon sector during episodes of flux transfer at the magnetopause. When By is positive the region of strong auroral emissions is found to be at the same latitude as the radar footprint of the cusp. When By is negative the radar cusp is found to be poleward of the strong auroral emissions. These results are consistent with models of the ionospheric convection and field-aligned current response to reconnection at the magnetopause under positive and negative By conditions.

Observations of cusp ion velocity dispersions made by the TIMAS instrument on the Polar spacecraft during intervals of northward interplanetary magnetic field (IMF) reveal a clear ordering with magnetic local time (MLT). Between 1100 and 1300MLT the injected ion velocity increased with spacecraft invariant latitude for 95% of the cusp crossings, between 0900 and 1100MLT this percentage reduced to 5% and between 1300 and 1500MLT the percentage reduced to 25%. No systematic variation in the dipole tilt angle was observed which could explain these observations. The IMF |Bz/By| ranged from less than 0.5 to 16, despite this variation the MLT dependency can be best explained by lobe reconnection in a sub-alfvenic magnetosheath flow resulting in sunward convection over the polar cap with tailward convection on open field lines displaced from the noon meridian.

On 24th November the Ulysses spacecraft will be passing over the southern solar pole at its highest latitude (~80.2º). This unique orbit, combined with recent enhanced solar activity, allows in-situ observations of magnetic clouds for the first time at high latitudes. Magnetic clouds are associated with coronal mass ejections (approximately 1/3 from in-ecliptic observations) and consist of smooth magnetic field rotations, enhanced magnetic field magnitudes and low proton temperatures. We have used this unique data set to examine how occurrence rates of magnetic clouds vary with solar latitude and activity. We also present some example events, using a simple force free flux rope model as a guide to describing their structure and properties.

Motivated by WIND observations and earlier low resolution electrostatic simulations, 1 1/2 D electromagnetic particle-in-cell simulations of the lunar wake have been undertaken. By using a sufficient number of particles per cell, we are able, for the first time, to resolve the full phase space dynamics of both electrons and ions. The simulation begins immediately downstream of the moon, and evolves in the solar wind rest frame. Computational restrictions do not allow a full scale simulation to be run, therefore a number of simulations have been produced with different initial void widths (from 128 to 512 Debye lengths). In all the simulations, electrons are initially trapped by potentials at the edge of the wake, but then counter-streaming electron beams are generated. By using a sufficiently wide initial void, the counter-streaming electrons cause a two-stream instability in the centre of the wake. The simulations are consistent with both WIND observations and earlier lower resolution electrostatic simulations. Initial results from two-dimensional simulations may also be presented here.

The overwhelming frequency of fast-mode collisionless shock observations compared to slow shocks remains a puzzle. To some extent this observational fact has been explained by the effectiveness of Landau damping in removing slow-mode perturbations in the majority of solar wind conditions. However an extensive study of Ulysses observations over 33 solar rotations in the years 1990 - 1993 [Kilmurray, 1999] showed that there were many periods of time when solar wind parameters were in a region suitable for the growth of slow wave perturbations. No slow shocks were actually identified in this study. A possible explanation for this is that the periods where conditions are right only lasted for a few hours which may not be long enough for slow shocks to form or at least make them unlikely to be observed. In the work presented here, three slow shock candidates are identified from a survey of 130 interplanetary shock waves observed by Ulysses. The magnetic and plasma data for each of the three slow shock candidates are analysed in detail, to confirm (or otherwise) their classification. Two of them are found to be consistent with slow shocks and are therefore classified as such. There are some discrepancies in the analysis of the third event, which makes its classification uncertain. The results of this work demonstrates that the obstacles to finding slow shocks in interplanetary space are not only due to their natural scarcity, but also in their identification using the available analytical tools. The theoretical predictions for slow shocks are clear and well defined, but the added complexities observed in magnetic and plasma data, at the time resolution presently available, make the completely unambiguous identification of possible slow shocks very challenging.

The accurate prediction of Earth arrival times of Southward magnetic field turnings using L1 monitor data requires knowledge of their orientations. Recent work has suggested that most solar wind discontinuities, and Southward turnings, are tangential rather than rotational. In this case, their orientations can be estimated with single spacecraft data by taking the cross product of up and downstream magnetic fields. Using ACE L1 data and Wind as a proxy for the Earth, we compare this method with other estimators of the orientations, including minimum variance, and show that the cross product is by far the most accurate. The results suggest that this can provide a reliable and accurate predictor, with around 86% of events at L1 that arrive at the Earth being predicted within 10 minutes of the actual arrival time.

We survey MeV proton data from the Anisotropy Telescopes (ATs) instrument on board the Ulysses spacecraft, for the period from day 310 of 1997 to day 209 of 2000. We search for two types of extreme anisotropy events: 1) events with very large and long duration first order anisotropy, indicating strong particle streaming along magnetic field lines, and 2) events with very constant close to zero first and second order anisotropies, i.e. isotropic particle distribution in the solar wind frame. We find 5 events of the first type, all at times when the Ulysses spacecraft was close to the ecliptic plane. On the contrary 3 of the 4 zero anisotropy events identified are seen at heliographic latitudes greater than 42 degrees south. We put forward possible interpretations of these results.

The polar cap absorption event of April 1998 is examined using widebeam and imaging riometers covering a number of L-shells. The measured absorption is compared with GOES proton flux data over a range of energies. The absorption cutoff in invariant latitude is studied and compared with proton flux cutoffs measured by the SAMPEX spacecraft. Preliminary results are presented.

Regions of He+ dominance have been observed by the DMSP F10 satellite around 800 km altitude at 09:00 LT. Modelling simulations using the Sheffield University Plasmasphere Ionosphere Model have been performed in which a basic simulation is made to a geomagnetic storm. The model reproduces the qualitative features seen in the observations, although the modelled regions of He+ dominance are observed at a greater altitude than shown in the DMSP data. Extensions to this first approximation are planned for the future.

Observational evidence for scale free regimes in auroral indices and the distribution of patches of activity in POLAR UVI auroral images have been cited as evidence that "internally" the coupled solar wind-magnetospheric system is a highly correlated, complex system. This, coupled with the bursty nature of transport in the magnetotail have led to the suggestion that the system is in a (forced) self organised critical state, exhibiting dynamical properties that can be explored with reference to avalanche (sandpile) models as reduced descriptions for the system. Complex systems can often be characterized by a lack of scale, and in particular, by the exponents of the power law probability distributions of patches of activity in the system. Examples of these include energy dissipated in avalanches in sandpiles and in turbulent fluids. Determining the exponents is nontrivial, but is needed to provide quantitative constraints on theory. We consider what may be a more readily accessible measure, a global average quantity such as the total energy dissipation across the system which can be observed remotely rather than in- situ. The probability of this global measure is non Gaussian and here we derive its functional form in terms of the distribution of the individual energy dissipation events in the system. This 'global' probability distribution yields a single parameter that importantly distinguishes energy dissipation events that are highly correlated from an uncorrelated (Gaussian) process, and yields the exponent.