Cross-Scale: Multi-scale Coupling in Space Plasmas ESA ...s3.amazonaws.com/.../€¦Cross-Scale: Multi-scale Coupling in Space Plasmas ESA Cosmic Vision 2015-2025 Call for Proposals

Cross-Scale: Multi-scale Coupling in Space PlasmasESA Cosmic Vision 2015-2025 Call for Proposals

The Cross-Scale Team

28 June 2007

+

+

-

-

Contact Prof. Steven J. Schwartz, Blackett Laboratory, Imperial College London, London SW7 2AZ, UK.Tel: +44 (0)20 7594 7660; Fax: +44 (0)20 7594 7772; email: [email protected].

Cross-Scale: Contents Cosmic Vision 2015-2025

Contents

1 Executive Summary 2

2 Scientific Objectives 32.1 Universal Plasma Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32.2 How do shocks accelerate and heat particles? . . . . . . . . . . . . . . . . . . . . . . . . . 62.3 How does reconnection convert magnetic energy? . . . . . . . . . . . . . . . . . . . . . . . 92.4 How does turbulence control transport in plasmas? . . . . . . . . . . . . . . . . . . . . . . . 152.5 Synergies with Other Science Endeavours . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

3 Mission Profile 193.1 Launch and Orbit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193.2 Ground Segment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203.3 Other Requirements and Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

4 Payload 214.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214.2 Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224.3 Data rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

5 Spacecraft Considerations 245.1 Spacecraft Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245.2 Electromagnetic Cleanliness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

6 Science Operations and Archiving 256.1 Spacecraft Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256.2 Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266.3 Data Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

7 Technology 267.1 Payload Readiness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267.2 Mission and Spacecraft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

8 Programmatics and Costs 278.1 SCOPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278.2 Payload Provision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288.3 Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288.4 Risks and Alternative Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

9 Communications and Outreach 32

References 32

List of Acronyms 34

Image Credits 34

Cross-Scale Community 34

28 June 2007 1

Cross-Scale: 1 Executive Summary Cosmic Vision 2015-2025

1 Executive SummaryMost of the visible universe is in the highly ionisedplasma state, and most of that plasma is collision-free. Plasma processes are at work everywhere,from radio galaxy jets and supernova explosions tosolar flares and planetary magnetospheres. Cross-Scale is an M-class mission dedicated to quantify-ing the coupling in plasmas between different phys-ical scales. This cross-scale coupling, being highlyvariable and structured, is critical in underpinningand quantifying the physical mechanisms inferredin plasmas that are difficult to observe.

As plasma regimes encounter each other, theabsence of collisions raises fundamental questionsabout how energy is shared amongst the threemain elements (electrons, ions, and overall bulkflows). These constituents, each of which operateson its own physical scale, are coupled through elec-tromagnetic fields.

Three fundamental physical processes operateto bring about the universal collisionless plasmacoupling in physical environments where momen-tum and energy transfer is important.

Shock waves guide strong flows around ob-stacles or at interfaces between two flow regimes.They are important locations for the transfer of di-rected bulk flow energy into heat, with an attendantacceleration of energetic particles.

Magnetic reconnection releases stored mag-netic energy to the plasma, and allows for ex-change of material between previously isolatedregions. Moreover, the consequent change inmagnetic topologies provides a coupling betweenplasma regions which often drives the global scaledynamics of the system.

Turbulence transports energy from largescales at which it is input to small scales where itis dissipated. In the process, it interacts strongly,and often selectively, with plasma particle popula-tions as either a source or sink (or both) of energy.

Near-Earth space is a unique laboratory forquantifying the physics of these three processes.Breakthroughs have arisen due to the high qualityof data that, unlike more distant regimes, is sam-pled directly by plasma and fields experiments onsatellites.

To date, in situ measurements have focusedon terrestrial phenomena, such as the mechanismsthat populate the van Allen belts. Dual spacecraftstudies during the 1980’s began to address the realmicrophysics. Present generation missions (Clus-

ter and MMS) utilise 4 spacecraft to sample a spe-cific volume, and hence characterise the physicsoperating on the single scale corresponding to thespacecraft separation. By the time MMS has flown,we shall have a catalogue of behaviour that rangesfrom the smallest, electron scale, to the largestfluid-like phenomena.

That knowledge is incomplete due to the am-biguity and uncertainty about the dynamics andvariability of the larger contextual scales (for theelectron and ion scales) and of the internal micro-processes that mediate the energy exchange (forthe larger scales). The complex, dynamic nonlinearcoupling of scales and physical mechanisms cannot be quantified without simultaneous informationon all scales.

Cross-Scale will target compelling and funda-mental questions, such as: How do shocks accel-erate and heat particles? How does reconnectionconvert magnetic energy? How does turbulencecontrol transport in plasmas? These address di-rectly the Cosmic Vision question “How does theSolar System work?” by studying basic processesoccurring “From the Sun to the edge of the SolarSystem.” Moreover, by quantifying the fundamentalplasma processes involved, the advances made bythe mission will extend beyond the Solar System toplasmas elsewhere in the Universe.

Cross-Scale will employ 10 spacecraft whichwill fly with two highly complementary spacecraftfrom its sister mission SCOPE provided by JAXA.Together, they will form three nested tetrahedra toseparate spatial and temporal variations simultane-ously on the three key scales for the first time.

The spacecraft, which carry a minimal payloadwith strong heritage, will be launched into a highlyelliptical Earth orbit. Over the two year missionthey will encounter various collisionless shocks, ex-plore regions of both spontaneous and stronglydriven reconnection, and investigate both nascentand highly evolved plasma turbulence.

There are no technologies that need to be de-veloped or proven for a launch in 2017, or earlier.The mission is thus low risk for high science return.It taps directly into European leadership in multi-point in situ space plasmas. This proposal is signedby researchers from around the world.

In the pages that follow, we amplify the univer-sal science objectives and present a mature, con-crete mission concept that fully meets these objec-tives.

28 June 2007 2

Cross-Scale: 2 Scientific Objectives Cosmic Vision 2015-2025

2 Scientific Objectives2.1 Universal Plasma Processes

A small number of phenomena dominate the be-haviour and effects of plasmas throughout the Uni-verse: collisionless shocks, magnetic reconnec-tion and plasma turbulence. Shock waves result-ing from supernova explosions and other ener-getic flows accelerate cosmic rays to high energies.Magnetic reconnection plays a pivotal role in the re-lease of stored energy in phenomena as diverse assolar flares and γ-ray-rich “magnetars.” Turbulencein astrophysical disks allows accretion to proceedby transporting angular momentum; it also chan-nels energy from the largest scales through a cas-cade to the smallest where it dissipates.

2.1.1 Shocks

Shock waves are formed whenever a supersonicflow encounters an obstacle of some kind, includ-ing other material. The basic process involves atransition from supersonic to subsonic flow, so thatinformation can propagate sufficiently upstream ofthe obstacle to deflect the oncoming flow. To do so,a shock must decelerate, deflect, compress, andheat the incident material. Classically, the shocktransition occurs abruptly, on scales related to thedissipative (i.e., collisional) processes.

In astrophysical contexts, however, the incidentmaterial is usually a highly ionised plasma in whichcollisions are negligibly rare. Under these circum-stances, shock waves may still exist, but the ab-sence of collisional coupling can result in highlynon-equilibrium physics that can be traced to themultiple scales associated with multiple speciesand fields.

Collisionless plasma shocks are some of themost spectacular, visually striking and energeticevents in the Universe. Generated by supernovae,stellar winds, or the rapid motion of objects suchas neutron stars, they have a number of impor-tant effects. Supernova shock waves trigger thecollapse of galactic nebulae and hence the forma-tion of planetary systems. They are responsible forheating and deflecting the surrounding plasma, andblow large-scale magnetic bubbles out of galacticdisks. Collisionless shocks also accelerate parti-cles to extraordinarily high energies.

The interaction of the fast-moving solar windwith the Earth’s magnetosphere results in a bow-shock. Its curvature means that regions of quasi-

parallel (magnetic field parallel to the shock normal)and quasi-perpendicular shock front co-exist, andthat the Mach number varies across the shock sur-face. Magnetohydrodynamic Mach numbers canreach 20, comparable to those in many astrophysi-cal scenarios. The terrestrial bowshock is thereforea useful analogue for astrophysical shocks. Theability to measure particle distributions and electro-magnetic fields around and within the shock frontmakes it possible to study the details of the colli-sionless shock transition and accompanying non-thermal phenomena. Historically, in situ measure-ments at the bow shock and interplanetary shockshave challenged existing theory and led to signifi-cant advances in computational modelling.

2.1.2 Reconnection

Magnetic reconnection is a fundamental plasmaphysics process which breaks down the barriersbetween neighbouring plasmas, releasing energyfrom their magnetic fields, transferring material andmomentum between those plasmas, and accelerat-ing a part of the plasma population to high energies.

The Universe is filled with situations where re-connection is expected to play significant roles intheir dynamical evolution, including stars (exoticand otherwise) and planetary systems at all stagesof their life cycles. Reconnection also governs theinteractions of those systems with their surroundingmedia.

In highly electrically conductive space plasmas,the constituent charged particles are unable to con-vect transverse to the local magnetic field direction;the magnetic flux is “frozen in” to the plasma ma-terial. Thus two such plasmas in regions of mag-netic flux with different sources generally cannotmix. They are separated by a sheet of electricalcurrent that exerts the forces acting to keep theplasmas apart. The plasmas remain isolated fromone another, even if pressed together by externalforces or momentum. However, exceptions to thisprinciple arise in situations in which magnetic re-connection occurs.

A key consequence of magnetic reconnectionis the linkage of magnetic fields from the two neigh-bouring plasmas, allowing material to flow betweenthe previously isolated regions (e.g., from a stel-lar atmosphere to planetary magnetosphere or be-tween different regions of the solar corona - seeFigure 1). The reconnection process locally dis-rupts the current sheet, changes the topology of the

28 June 2007 3


Case Study: SLAMS and Particle Acceleration: A universal shock process in operation

The quasi-parallel shock transition, (left) asenvisaged in early work (after [1]) based on lim-ited multi-spacecraft data. Studies showed thatthe coherent Short Large Amplitude MagneticStructures (SLAMS) sweep up suprathermal par-ticles (bottom left, from [2]) during the acceler-ation process, in contrast to the traditional pic-ture of diffusive first-order Fermi acceleration withmonotonic pressure profiles.

More recent results shown below at 1/10th(top) and 1/100th (bottom) of the dimension in-ferred from[1], reveal considerable sub-structure,calling into question some of the key assump-tions about the scales associated with shocksand Fermi acceleration under quasi-parallel con-ditions. The variability and small-scale structurehas important, and to date elusive, implicationsfor the efficiency of the quasi-parallel shock interms of energy partition and ion acceleration.(after [3, 4]).

magnetic fields, and releases its energy, e.g., intothe bulk flow and heat of the plasma.

Many in situ measurements made by space-craft at current sheets in the vicinity of Earth (prin-cipally the magnetopause and the magnetotail cur-rent sheet) have provided compelling evidence thatreconnection does indeed occur in collisionlessspace plasmas. However, these have yet to pro-vide clear indications of how it happens or a com-plete picture of what conditions are needed to initi-ate and maintain the process. For example, obser-vations suggest that a necessary, but not sufficient,condition for sustained magnetic reconnection is asignificant shear angle between the magnetic fieldvectors in the two adjacent plasma regions. Pro-

cesses occurring on much smaller scales, compa-rable to ion and probably even electron gyroradii,appear to play a key role in initiating and supportingreconnection. In turn, these localised processes re-sult in global scale consequences for the dynamicsof the plasma systems. Thus to understand the wayor ways in which reconnection arises, operates, andcontrols large scale dynamics, it is essential that weinvestigate the process across all relevant scales.

2.1.3 Turbulence

Turbulence is present in many astrophysical plas-mas such as the interstellar medium, accretiondisks, the solar wind, supernova remnants, and col-lapsing nebulae. It is also dynamically importantin the transport of energy, mass and momentum in

28 June 2007 4


Figure 1: The image on the right shows a reconnec-tion event occurring in the Sun’s corona. A study byYokoyama et al.[5] identified coronal plasma in- and out-flowing (yellow and green arrows) as expected for a re-connection event, as well as heated material appearingon closed magnetic loops formed by reconnection.

many of these environments. The accurate predic-tion of turbulent properties, and their effects on thesurrounding plasma, is therefore key to the quanti-tative analysis of many astrophysical scenarios, aswell as terrestrial plasmas such as tokamaks.

Measurements of turbulence in solar systemplasmas have provided several key insights into itsbehaviour which are difficult or impossible to ob-tain any other way: the scaling of the main energycascade, anisotropy, spatial variability, and effectson particle propagation and energisation. Recently,multi-spacecraft techniques have made it possibleto uniquely identify energised wave modes, offeringa revolution in our ability to analyse plasma turbu-lence. Such advances both validate and challengetheoretical and numerical models.

Despite these successes, many key questionsremain unanswered concerning the nature of theturbulent cascade, particularly near the ion andelectron kinetic scales: How is it driven to ananisotropic state by the magnetic field and its localenvironment? How does it spontaneously gener-ate structures from a previously uniform plasma?These issues must all be quantified if the largescale effects of space plasma turbulence are to bepredicted.

2.1.4 Multi-scale Coupling

Shocks, reconnection, and turbulence are con-trolled by dynamics which are coupled on 3 fun-damental scales simultaneously: electron kinetic,ion kinetic, and fluid. It is the nonlinear interac-tion of 3D, time-varying structures on these 3 scaleswhich produces the complex behaviour and conse-quences of these processes. Critically, most as-trophysical plasmas are collisionless, which means

that their constituents can be far from equilibriumwith each other. The resulting nonlinear dynam-ics provides diverse and exotic mechanisms for mo-mentum and energy flow and redistribution.

The complex, three dimensional nature ofplasma structures has long been recognised. Pre-vious, existing and upcoming missions have beendesigned to measure this 3D structure using mul-tiple spacecraft. A minimum of four spacecraft arenecessary to determine 3D structure: ESA’s Clus-ter and NASA’s upcoming MMS missions both usefour spacecraft for this task. A fundamental restric-tion of multi-spacecraft measurements, however, isthat they are sensitive to scales of the order of thespacecraft separation. With four spacecraft, multi-ple scales can be probed by varying this separa-tion, but only one scale can be measured at anytime.

Plasmas are not just three dimensional: theyalso contain time-varying structure on many scales,simultaneously. Different scales are affected bydifferent physical processes. It is the interplay ofthese which results in the complexity of shocks,reconnection, and other phenomena, and conse-quently in their large scale effects.

To understand the interplay of forces and dy-namics within such regions and hence predicttheir effects, it is essential to measure the time-dependent behaviour in 3D on the three key phys-ical scales – electron, ion and fluid. This canonly be achieved with spacecraft positioned suchthat some have separations comparable to each ofthese three physical scales, simultaneously. Thus4 spacecraft are required at each of the three phys-ical scales, making a complement of 12 spacecraftin total. Instrumentation on the spacecraft at eachscale must be tailored to the physical processes atthat scale. Near-Earth space, which is relativelyaccessible and contains examples of all the phe-nomena of interest, is the obvious target for such amission, which we call Cross-Scale.

While simulations of collisionless plasmas haverevealed a great deal about their dynamics andcomplexity, it is not possible to simulate the keyphenomena of interest in sufficiently large simu-lation boxes to resolve the three logarithmically-spaced physical scales in 3D. Indeed, this goal willnot be achieved within the time scale of ESA’s Cos-mic Vision 2015-2025 programme, assuming thatcomputer power increases at its historical rate. Nei-ther can laboratory plasmas be probed over the

28 June 2007 5


Table 1: Cross-Scale main science questions and multi-scale coupling physics

How do shocks accelerate and heat particles?How do shocks accelerate particles?

coherent magnetic structures; surface ripples

How is the energy incident on a shock partitioned?ion reflection; cross-shock potential; electron demagnetisation

How do shock variability and reformation influence shock acceleration?response to upstream variability; non-steady reformation

How does reconnection convert magnetic energy?What initiates magnetic reconnection?

non-adiabatic particle motion; dissipation of boundary current

How does the magnetic topology evolve?Hall currents; current sheet tearing; magnetic island coalescence; magnetic tension

How does reconnection accelerate particles and heat plasma?parallel electric field; perpendicular drifts; wave electric fields

How does turbulence control transport in plasmas?How does the turbulence cascade transfer energy across physical scales?

ion & electron dissipation scales; wavemode identification

How does the magnetic field break the symmetry of plasma turbulence?magnetic field braiding; effect of boundaries

How does turbulence generate coherent structures?ion & electron instabilities; intermittency

necessary scales. The only way to validate theo-retical concepts and limited numerical experimentsconcerning these phenomena, and so to studythem in sufficient detail, is to measure them directlyin space.

In the following subsections we describe inmore detail the specific science questions relatingto shocks, magnetic reconnection, and turbulence.These set the science objectives in their presentcontext and understanding, highlight the importantareas where there are crucial uncertainties, and re-veal how Cross-Scale measurements will replacedoubt and qualitative concepts with a definitive andquantifiable framework.

2.2 How do shocks accelerate and heatparticles?

Cross-scale coupling is integral to shocks in col-lisionless plasmas. Small scale electron dynam-ics results in a highly structured, fluctuating electric

field within the shock ramp. At scales around anorder of magnitude larger, ions gyrate through theramp, with trajectories determined by the fluctuat-ing electric field. Reflected and gyrating ions thengenerate the even larger scale reformation and rip-pling of the shock front, which in turn affects thesmall scale dynamics of the electrons - as well asbeing pivotal in the acceleration of inflowing parti-cles to high energies.

Collisionless shocks both heat the main plasmaconstituents and accelerate sub-populations of ionsand electrons to high energies. This acceleration ispossible because the physics involves more thanone scale. Acceleration is a part of the larger ques-tion about how shocks partition the bulk flow en-ergy incident upon them, which again is effectedover fluid, ion, and electron scales. Finally, varia-tions in the flow that drives them, and instabilitieswithin the transition scales, leads to transient phe-nomena that can significantly alter or enhance the

28 June 2007 6


acceleration efficiency.

2.2.1 How do shocks accelerate particles?

Diffusive acceleration

Some particles are reflected from the shock.Their propagation into the upstream plasma is un-stable to the generation of waves, which in turnscatter them – a highly nonlinear process. This ini-tiates further acceleration. However, many aspectsof this process are poorly understood, and involvespatially variable structures, both within the shockand in the upstream plasma.

Fundamental to modern acceleration theory isthe scattering length of the energised particles. Re-peated reflection and scattering (the so-called “firstorder Fermi” process) can accelerate particles tovery high energies. Recent studies[6] have forthe first time evaluated the scattering mean freepath for shock-accelerated particles. Simultaneousmeasurements of the resonant plasma waves andparticles over ion and fluid scales is required toquantify the wave modes and particle interactionsover a range of plasma conditions. Only in thisway can we predict shock acceleration efficienciesin other astrophysical environments.

Coherent acceleration

The surface of shocks is believed to be rippledby local ion and current instabilities, but the prop-erties of these ripples, e.g., amplitude and wave-length, are unknown. The ripples provide time-varying fields which can trap some particles, en-abling them to “surf” the shock front and systemat-ically pick up energy from the large-scale motionalelectric field. Such surfing is potentially importantfor both ion and electron acceleration, and may“inject” suprathermal particles into the first orderFermi process, the efficiency of which is dramati-cally improved if fed a pre-accelerated population.In order to quantify the effects of shock ripples, si-multaneous measurements are required of ripplesof the shock surface at fluid scales; of ion distribu-tion variations around and within the ripples, withvariations on the scale of an ion gyroradius; andof electron heating and acceleration at the smallestscales.

Quasi-parallel shocks are associated with ShortLarge Amplitude Magnetic Structures (SLAMS)which grow in the generally turbulent fore-shock/shock region accompanied by energetic par-ticles. Their polarisation suggests they grow due to

a hot ion instability, essentially feeding off the par-ticle pressure gradients, whereas other foreshockturbulence is beam-driven. Thus the relationshipbetween SLAMS and the general turbulence is notclear. The basic presumed structure is sketched inthe Case Study on page 4.

Again, the origin and evolution of the cycle ofturbulence and particles needs to be explored bymeasurements at the disparate scales. SLAMSexhibit internal structure down to electron scales,and their overall size, shape and porosity are corre-spondingly difficult to determine without multi-scaleobservations. Yet these characteristics are criti-cal to our picture of quasi-parallel shocks. Largescale measurements of the orientation of SLAMS,energetic particle gradients and foreshock wavesmust be combined with ion-scale structures withinSLAMS, with electron-scale fine structure, and with3D electric field measurements.

2.2.2 How is the energy incident on a shockpartitioned?

In the frame of the shock, the upstream, incom-ing plasma carries kinetic, thermal and electromag-netic energy into the shock front. This energy isthen partitioned into a number of forms. Most is dis-tributed between the downstream kinetic and ther-mal energies of the ions and electrons. Knowledgeof the fraction of energy distributed into each ofthese plasma constituents is essential for predict-ing the large scale effects of shocks. For example,the energy taken up by electron heating, togetherwith the actual shape of electron energy spectra,at astrophysical shocks is responsible for the ob-served X-ray emission. It is possible to measurethis partitioning directly at the terrestrial bowshock.However, this does not mean that the physical pro-cesses by which the redistribution of energy occursare well known. Indeed, this partitioning varies con-siderably with shock parameters and is also highlyvariable in space and time, so the application ofthese measurements to astrophysical shocks is notstraightforward without a detailed knowledge of thephysical processes which govern the energy distri-bution.

Ion reflection

The rise in electric potential through the shocklayer is responsible for decelerating the incomingions, some of which are reflected at shocks abovea critical Mach number. This reflection initiates thespread in velocities that will ultimately account for

28 June 2007 7


Figure 2: Recent measurements of the 3D electric fieldat a high-Mach number collisionless shock which revealthe presence of electric field spikes (lower panels) onscales much smaller than the ion scale variation in themagnetic field shown in the top panel. The size andvariability of such spikes is not known. Full 3D measure-ments at multiple scales are required to transform suchelectric fields into the cross-shock electric potential thatcontrols both the ion and electron dynamics. (from [8])

the rise in non-directed motion (the “heating” re-quired of the shock).

However, the potential is the integral of the elec-tric field across the shock - and this electric field isknown to be highly structured and variable on allmeasured scales down to the electron scales[7].Figure 2 demonstrates large amplitude, small-scaleelectric field spikes within the main shock transition.However, the role of such spikes and their variabilityrequires electron-scale multi-point measurementsof the 3D electric field together with larger-scalemeasurements to accurately determine the shockorientation, motion, and feedback processes.

Electron heating

Just as ion trajectories within the shock are af-fected by small scale electric field structures, soelectron trajectories and small scale electric fieldstructures are altered by the large scale shock pro-file, which is constantly varying as a result of refor-mation and ion dynamics. Very little is known aboutthe effect of large scale shock variability on the for-mation, size and lifetime of small scale electric fieldstructures, such as those illustrated in Figure 2, butwithout this knowledge, the 3D electric field struc-ture, and final electron distribution functions, cannotbe predicted. Moreover, it is the electrons them-selves that support most of the overall cross-shockpotential jump, so the electron dynamics and en-

ergy partition are strongly interlinked processes.Of key importance is the nature of the elec-

tron motion: Magnetised or unmagnetised? Dom-inated by steady fields or high frequency fluctua-tions or turbulence? What is the feedback fromthe ion scales? Electrons with different energies orgyrophases may behave differently, requiring goodcoverage over the full range of velocity space.

Heavier ion species heating

Observations of heavier ions show that they areefficiently heated and accelerated at shocks. Thisis somewhat puzzling, since the main shock fieldsare self-consistently tuned to process the dominantmomentum and energy carriers, which in the caseof the solar wind are protons. The resolution ofthis puzzle almost undoubtedly requires some in-terplay between the electric fields (at small scales)and larger variations.

2.2.3 How do shock variability and reforma-tion influence shock acceleration?

Internal variability

Supercritical shocks are fundamentally variablein time and space. They exhibit reformation, aquasi-periodic variation in the shock profile onscales comparable to the proton gyroradius. Thisresults in a non-planar, and varying, shock pro-file, with important consequences for how particlesare deflected, heated and accelerated. Computersimulations[9, 10] such as that shown in Figure 3illustrate the difficulty faced by two, or even four,spacecraft trying to disentangle the overall struc-ture and dynamics of the shock surface[9, 11].

The trajectory of an individual ion through theshock ramp is controlled by the instantaneous elec-tric field that it encounters – but since this fieldis structured and variable, different ions followmarkedly different trajectories. The number of ionsreflected by the shock, which is directly related tothe partition of energy problem, is known to vary(see Figure 4) but determining the feedback be-tween that variability and the shock energy partitionrequires simultaneous multi-scale measurements.

This variability leads to many complex trajecto-ries within the shock, some of which result in ionsbeing ejected into the upstream plasma. There theyseed the shock acceleration process and modifythe upstream plasma conditions. A knowledge ofthe intricate feedback between very fine scale elec-tric field structures and ion dynamics, and the re-

28 June 2007 8


Figure 3: Two-dimensional magnetic field maps of a sim-ulated shock. Note the structure and variability of theshock “surface.” (from [10])

Figure 4: Ion distributions near the upstream edge ofa high Mach-number shock (top) and the variability ofreflected ions observed there (bottom). Quantitative as-sessment of the role of small scale fields, variability, andlarge scale control and consequences await full multi-scale measurements. (from [12])

sulting variability in the shock profile and structure,is essential for predicting the reflection and acceler-ation of ions around the shock, and in turn the bulkeffect of shocks on the ambient medium.

Shock variability also has a strong influence onthe acceleration of electrons. In the “fast Fermi pro-cess,” incident electrons experience energy-gainingreflection that is a strong function of the angle be-tween the magnetic field and local (on an electronscale) shock normal. Trapping within ripples andsmaller variations enables electrons to reside closeto the shock where they gain energy from the shockelectric fields. Both these processes rely heavily on

internal and variable shock structure.

Externally-induced variability

Several circumstances conspire to complicatethe physics at real shocks. Some, described above,are intrinsic variability. Others are due to externalvariations which can have profound effects on theshock dynamics. For example:

Hot flow anomalies are formed when a plasmadiscontinuity with suitable orientation and parame-ters impinges on a shock front. Despite the over-all pressure balance through such a discontinuity,and its thin structure, the particle dynamics at theinteraction region can give rise to dramatic explo-sive events that create a hot cavity upstream ofand/or attached to the shock. At the bow shock,these are observed to have some of the hottest,most thermalized particle distributions seen in situanywhere in the solar system. It is not at all clearhow particles are accelerated within the cavity andsmall scale magnetic field structures of a hot flowanomaly, nor the role played by electron dynamics.As sources of energetic particles and possible trig-gers of further events, these structures require fur-ther study. Measurements of the large, fluid-scalecavity shape and evolution must be made simulta-neously with ion distributions within and around thecavity, as well as fine scale electric field and elec-tron variations, in order to understand the dynamicsand effects of these phenomena.

Shock-shock interactions occur often in realsystems, such as travelling interplanetary shockscolliding with planetary bow shocks, or forwardshocks catching up with slower or reverse travellingshocks in strongly variable astrophysical systems.Theoretically, such interactions provide conditionsfor extreme heating and particle acceleration, andwill be studied in detail for the first time by Cross-Scale.

2.3 How does reconnection convert mag-netic energy?

Magnetic reconnection is expected to occur whenmagnetic fields are sheared across relatively thincurrent layers[13]. In such current sheets the ki-netic effects of the particle populations become im-portant, and the onset of reconnection is expectedto occur on the distance and time scales of the rele-vant electron and ion gyromotions. Microscale pro-cesses can control the change of topology of themagnetic field, eventually affecting the large-scaleplasma mixing and converting magnetic energy to

28 June 2007 9


plasma energy. On the other hand, large-scale(MHD) processes can control the location and for-mation of thin current sheets, and thus directly af-fect how reconnection initiates and evolves. It istherefore essential to follow both the large-scaleand kinetic scale processes of the plasma to un-derstand the onset, the evolution, and the result ofthe magnetic reconnection.

Current sheets where reconnection takes placecan be either large scale boundaries, such as theEarth’s magnetopause, tail neutral sheet, and solarwind discontinuities, or form dynamically on smallscales, such as in magnetosheath turbulence (asshown in the Case Study on page 11) and withinvortices at the magnetopause[14]. These regionsinclude both strongly driven systems (such as thesolar wind stand-off at the magnetopause) and rel-atively calm conditions (such as convecting solarwind current sheets).

The near-Earth plasma environment is uniquein that we can make detailed measurements for dif-ferent types of current sheets, including their for-mation and their microphysical processes. We canalso examine in detail both the external conditionscontrolling the occurrence of reconnection and thelarge-scale consequences it has on the system.By simultaneously monitoring the essential scales,Cross-Scale will investigate the key questions in thereconnection as highlighted in the following subsec-tions.

2.3.1 What initiates magnetic reconnection?

In order for magnetic reconnection to occur, a re-gion must exist in which the magnetic field can dif-fuse relative to the plasma and across the currentsheet separating two plasma regimes. Spacecraftobservations suggest that this does not happen un-less the shear angle between the magnetic fieldseither side of the boundary exceeds ∼ 70◦. How-ever, most current sheets for which that conditionis satisfied remain stable much of the time, withoutany reconnection. Moreover, the conditions in thelarge scale plasma environments either side of re-connecting current sheets have been observed tobe quite diverse, showing that reconnection is ca-pable of happening under many conditions. We donot yet understand why reconnection begins, but aneffective approach would be to study why it may be-gin at a particular point in a current sheet and nota neighbouring location. A basic requirement formaking this comparison is numerous simultaneous

measurement points both in and around a currentsheet as it begins to reconnect.

According to present theory, a precondition forreconnection is that the current sheet must undergoa thinning process. Quantitative measurements ofcurrent sheet thinning have been made by the Clus-ter mission[20]. Cluster 4-point data can be used todetermine the average current through the tetrahe-dron at a given moment. If the tetrahedron is smallcompared to a current sheet, a local measurementis made but the total current is unknown; if the tetra-hedron is large compared to the sheet, the totalcurrent is measured, but its distribution within thesheet is unknown.

In order to fully characterise the development ofthinning pre-reconnection current sheets we musttherefore use a set of nested tetrahedra of space-craft, across at least three scale sizes, each opti-mised to capture a different stage in the thinningprocess. The largest should give information aboutthe magnetic flux gradient across the current sheet,and also characterise the orientation of the currentsheet. The smaller ones should measure the inten-sification and thinning of current within the overallsheet structure, and also provide information on theinternal variations in plasma characteristics to de-termine how such currents are supported. Nestedtetrahedra provide more than three sets of 4-pointdata from which current can be calculated, provid-ing valuable information about current and plasmasubstructures along and across the current sheet.

This set of multi-scale measurements will revealthe critical thickness at which a current sheet be-gins to break up. Measurements of ion and electronplasma properties put the critical thickness datainto context in terms of the ion and electron scalesizes of the system, enabling ready comparisonwith analytical theory and simulations[21] (cf. Fig-ure 5). Multi-point measurements of plasma wavesand particle distributions will facilitate the testing ofideas about current sheet instabilities which maycause the breakup and initiate reconnection.

Internal current sheet instabilities may causereconnection onset, but another possibility is the ef-fect of an external driver. Both scenarios may beoperating in different situations. Recent studies ofmagnetotail reconnection indicate that a large frac-tion of reconnection onset cases may be caused bysudden changes of the driving (global scale) elec-tric field in the solar wind; similar suggestions havebeen made for magnetopause reconnection. Re-

28 June 2007 10


Case Study: Reconnection in a turbulent plasma

Magnetic reconnection has until recently onlybeen observed at large-scale boundaries be-tween different plasma environments. Obser-vations analysed in [15] revealed, for the firsttime, in situ evidence of reconnection in a tur-bulent plasma. The turbulent environment is themagnetosheath, the solar wind downstream of

the Earth’s bowshock. The results have signif-icant implications for laboratory and astrophysi-cal plasmas where both turbulence and reconnec-tion are common. The Figure shows observationsof narrow current sheets in which reconnectiontakes place as well a schematic diagram, fromnumerical simulations[16, 17] that show the cur-rent sheet formation between coherent magneticstructures. Observations suggest that reconnec-tion contributes significantly to particle energisa-tion processes at the scales comparable to thesmall scales of current sheets or even smaller.However, much higher time resolution particle in-struments and small separation spacecraft arerequired to understand particle energisation pro-cess at these small scales. At the same time, si-multaneous observations of the largest scales arerequired to follow the evolution of coherent mag-netic structures. The most recent studies alreadydemonstrate that reconnection plays an importantrole in the development of turbulence, influencingintermittency and other properties[18].

Figure 5: A simulation of a current sheet during theonset of a reconnection event[19]. The breakup of thecurrent sheet is illustrated. A set of four electron scaleCross-Scale spacecraft is superimposed, to show howit would capture the transition from a stable sheet intorapidly changing fragmented currents which is thoughtto occur at reconnection onset. Additional spacecraft atlarger separations will quantify the orientation and con-ditions in which that onset is triggered.

cently a set of simulations using various numeri-cal schemes have investigated driven reconnectionscenarios and show consistent outcomes, increas-

ing confidence in the results. However, it is vital tomake experimental tests of these results. As withthe investigation of internal instabilities, this will re-quire measurements within the current sheet on atleast the ion and electron scales, and on the fluidscale outside the current sheet in order to capturethe driver conditions.

2.3.2 How does the magnetic topologyevolve?

Once magnetic reconnection has begun, there aremany questions about how it is maintained. Recon-nection in the solar wind can endure for hours[22];at the magnetopause can be either similarly persis-tent or very bursty on a timescale of minutes; in themagnetotail reconnection seems more often to lastonly minutes or tens of minutes. However, thesetime scales are all long compared to typical elec-tron and ion timescales, showing that reconnectioncan operate in a quasi-steady state, relative to thetimescale of the kinetic processes that are thoughtto sustain it.

28 June 2007 11


Figure 6: A simulation of a current sheet during a re-connection event. Note that the situation is more com-plex than the X-line scenario sketched in Figure 1.Such an ion/fluid scale simulation cannot simultane-ously address the electron scale (for that, see Figure 5).A possible Cross-Scale configuration is superimposedshowing how the large scale spacecraft measure thein/outflow jets and magnetic fields far from the currentsheet; the intermediate/small scales are targeted on theion/electron diffusion regions respectively.

Quasi-steady 2D reconnection

The most popular model of steady-state re-connection has anti-parallel magnetic fields eitherside of a current sheet. Plasma inflow and out-flow vectors define a plane that is perpendicular tothe current sheet and that also contains the anti-parallel magnetic fields (see the centre panel in Fig-ure 7); the reconnection electric field is normal tothis plane.

Magnetic reconnection occurs inside an elec-tron diffusion region, which is very thin, with an ex-tent normal to the current sheet of order an electrogyroradius or less (see Figure 5). This region is sur-rounded by an ion diffusion region no broader thanan ion gyro-radius along the current sheet normal.

In their respective diffusion regions theion/electrons are no longer magnetised, i.e. theyno longer gyrate about the magnetic field and arenot “frozen-in” to the magnetic flux. In the iondiffusion region, the reconnection electric fieldensures that “frozen in” electrons continue to flowinwards towards the electron diffusion region butdemagnetised ions are not affected in the sameway. The electrons accelerate as the magneticfield weakens near the current sheet.

The relative motion of ions and electrons gen-

erates a “Hall current” directed against the inflow,and an oppositely directed “Hall electric field” act-ing to oppose the charge separation. A further con-sequence is that the electrons and magnetic fluxare deflected out of the inflow/outflow plane, gen-erating a characteristic quadrupolar “Hall magneticfield” signature.

In this picture, the electron diffusion region issomewhat of a “black box” and processes within itthat demagnetise the electrons allowing reconnec-tion to occur are not well understood.

We need comprehensive multi-scale measure-ments such as those proposed for Cross-Scale inorder to validate this general picture and bring clar-ity to its grey areas. Single spacecraft observationshave been interpreted as evidence of Hall magneticfields, electric fields and outflow jets; these inter-pretations have been strengthened by 4-spacecraftmulti-point Cluster observations showing that thefields and flows are organised in patterns consis-tent with expectations of this model. A more directtest for unmagnetised ions is needed to verify thatthe Hall fields are located within the ion diffusionregion. This can only be accomplished by a com-parison of the ion flow velocities and electric fielddrift velocities, which will differ there, set in theirlarger-scale context. Such a comparison requiresvery good plasma measurements and 3D electricfield measurements. This test is an essential andunique Cross-Scale measurement.

The extents of the ion and electron diffusion re-gions are unknown, both along the external mag-netic field direction and perpendicular to the 2Dgeometry (i.e., along the neutral- or X-line). Theyare vital to theoretical ideas about the reconnectionrate and particle acceleration. These extents can-not be determined without monitoring the diffusionregions at electron/ion/fluid scales simultaneously.We need to be able to compare simultaneous ob-servations from regions where reconnection is oc-curring, with nearby regions where it is not, in orderto discover the factors which control the length ofthe neutral line.

Very little information is available about theelectron diffusion region. In order to test ideasabout how the diffusion occurs it will be neces-sary to measure the contribution to the reconnec-tion electric field made by very small scale phe-nomena including non-gyrotropic electron pressureanisotropies or bulk electron flow anti-parallel to themain current. However, reliable data on particle be-

28 June 2007 12


haviour in the electron diffusion region requires bet-ter time resolution than is available from today’s re-search spacecraft[23]. In addition, we also needto identify any waves which may scatter and de-magnetise electrons in and around their diffusionregion. For example, whistler mode wave activity inthe diffusion region may enable more efficient re-connection.

Departures from 2D

Reconnection topologies are most often morecomplicated than the picture discussed above; atthe magnetopause, the large scale magnetic fieldsseparated by the current sheet are often sheared,such that there is a component along the directionof the reconnection electric field. Many numericalsimulations have studied how this “guide field” sce-nario might differ from the simpler 2D case, but thiscannot be tested without multi-scale observations.Other departures from the symmetrical 2D picturethat need to be examined include the differencesin flow velocity, plasma density and temperatureacross the current sheet.

Cross-Scale measurements on all scales areneeded to reveal whether there are differences inthe way reconnection works at the magnetopausein comparison to the magnetotail. If so, are these todo with the electron scale, the ion scale or how theyinteract with each other and with the (asymmetric)driving at the fluid scale? Simulations suggest thatreconnection is sensitive to the outermost bound-ary conditions of the system. Similarly, compar-isons of magnetotail reconnection with and withoutsignificant oxygen ion populations are needed totest models of how the current sheet structure andreconnection rate depends on their plasma compo-sition. It would be expected that an additional scaleassociated with oxygen gyroradii will play a role incontrolling the reconnection process in this case.

Turbulent reconnection

The change in the magnetic topology due to re-connection is even more difficult to analyse withinturbulent plasma regions. Turbulence can de-velop at plasma boundaries (such as vortices atthe magnetopause[14]); changes in topology dueto reconnection can lead directly to plasma trans-port across the boundaries and formation of theboundary layer[24] (see Figure 6). Thus space-craft are required on a large scale to monitorthe boundary layer development and on smallerion/electron scales to understand the reconnection

Figure 7: Electron acceleration within a reconnection re-gion. Cold electrons are accelerated to keV energiesby very small timescale (millisecond) interactions withelectric field solitary waves[25]. Further acceleration to100’s keV involves non-adiabatic processes in the out-flow region[26].

mechanisms (again here measurements must bemade simultaneously on both ion and electron sep-aration spacecraft).

In other situations, such as magnetosheath, theplasma is turbulent in a large volume; magnetic is-lands and other coherent magnetic structures cancontinuously form and disappear due to reconnec-tion. The case study on page 11 illustrates onesuch example and also shows how Cross-Scalespacecraft at multiple scales are needed to analysethe reconnection in that case. Spacecraft at largeseparation would be able to follow the developmentof the turbulence while small separation spacecraftwould be able to follow the physics of reconnectionprocess itself. Both types of turbulent reconnectionare expected in a wide range of astrophysical plas-mas.

2.3.3 How does reconnection accelerate par-ticles and heat plasma?

Reconnection most probably occurs in a wide vari-ety of contexts beyond the Earth’s magnetosphere;most can only be studied by remote sensing of theemissions generated indirectly by energetic parti-cles. Even in the magnetosphere, energised parti-cles measured in situ are often the only sign of adistant reconnection site. In order to interpret such

28 June 2007 13


data as reconnection signatures or otherwise, andperhaps to use them to infer the properties of thereconnection site itself, we must understand howparticles are accelerated and heated.

Most of the energy released during the recon-nection process goes into the energisation of ionsand electrons. Observations show evidence of re-connection outflow jets and particle heating, and ofthe formation of field-aligned beams and energetictails in the particle distributions. However, their ori-gin is not understood in detail.

Inductive electric fields associated with rapidchanges in magnetic field may cause strongcharged particle acceleration, such as in a recon-necting current sheet with a non-steady reconnec-tion rate. To verify this with observations we needmultiple spacecraft in the vicinity of the reconnec-tion site to provide a combined data set that re-solves space-time ambiguities and confirms thatthe magnetic configuration is rapidly varying. Lo-cal electric fields in 3-dimensions and particle datafor a broad range of energies at high time resolu-tion are needed to infer where in the system theacceleration occurs.

The reconnection electric field is expected tohave a component along the local magnetic field di-rection in the diffusion region, which will readily ac-celerate charged particles[27]. A long neutral linecan in principle accelerate particles to energies lim-ited only by its length (e.g., 10’s of keV in the mag-netotail). A test of whether this acceleration mecha-nism really operates efficiently in nature requires aCross-Scale constellation to make measurementsof electric and magnetic fields, as well as particlepopulations, at several points along a significantportion of the length of a neutral line and aroundit. The upper part of Figure 7 shows data from aCluster study which identified large amplitude elec-trostatic solitary waves seen only for milliseconds atthe separatrices between in- and out-flowing mate-rial, together with a narrow 5 keV electron beam.The waves are consistent with an “electron hole”phenomenon predicted by simulations to occur inassociation with acceleration of inflow region elec-trons at the reconnection neutral line. Proper test-ing of this model requires simultaneous observa-tions of this kind throughout the diffusion region andin the surrounding plasma.

Waves with a broad range of wavelengths andfrequencies are generated in and around reconnec-tion sites. Wave-particle resonant or diffusive in-

teractions can both tap and enhance particle en-ergy, with different wave types likely to affect dif-ferent particle populations. Full characterisation ofwavevectors and modes requires multipoint mea-surements on the corresponding scale. Whenanalysing the gamut of relevant plasma waves,e.g., lower hybrid waves, whistler-mode waves, andelectron cyclotron waves, we require separationssimultaneously covering scales from tens of ionsscales down to the electron scales. Confirma-tion of wave-particle interactions also requires well-resolved ion and electron velocity distributions at anappropriate cadence.

Strong wave activity is often reported in outflowjets, and energetic (100’s keV) electrons are alsooften seen there, especially near the separatrices.One scenario for generating these electrons pro-poses that they are heated to a few 10’s of keVnear the reconnection site, and are further ener-gised to 100’s of keV by a combination of diffu-sion across the reconnection electric field togetherwith strong pitch angle scattering that violates adi-abatic invariants[28]. Betatron and Fermi acceler-ation are also expected on the contracting mag-netic field lines in the reconnection outflow jets.Measurements on ion and fluid scales are neededto capture the evolution of particle energy with in-creasing distance away from the reconnection site,and to measure the magnetic field and other pa-rameters to test whether the degree of energisationis consistent with that expected from these mech-anisms. A partial example of such a test is seenin Figure 7 where three Cluster spacecraft locatedat increasing distances from the inferred diffusionregion see electrons of progressively higher ener-gies.

Significant fluxes of electromagnetic energy lib-erated at reconnection sites have also been ob-served to propagate away in the form of waveswith an Alfvenic character. Their associated wave-particle interactions are known to cause local par-ticle energisation. The generation mechanism ofthese waves is not clear. To make significantprogress in understanding these waves and theirrelation to reconnection processes requires space-craft that are suitably separated at several scalesdictated by the nature of the waves: sub-ion scalesto study the wave properties across the magneticfield, a few ion scales to study the environment ofreconnection site and wave properties along themagnetic field and finally, hundreds of ion lengths

28 June 2007 14


away along the magnetic field to see the develop-ment in the wave propagation.

2.4 How does turbulence control trans-port in plasmas?

The ubiquitous nature and important conse-quences of turbulence in space plasmas make it animportant target for the Cross-Scale mission. Tur-bulence is responsible for the transport of manyphysical quantities: energy, both between scalesand across space; momentum; and energetic par-ticles through the resulting complex, tangled mag-netic fields. It covers a vast range of scales, frominter-galactic to below the electron gyroradius.

While significant advances in simulation, theoryand observations have been made, the highly com-plex, non-linear and multi-scale nature of plasmaturbulence means that many important questionsremain: How is energy transferred between scales,particularly at scales at which kinetic particle dy-namics are important? How do the unique proper-ties of plasmas alter the traditional hydrodynamicview of turbulence? What is the origin of the dis-crete structures that are observed in turbulent plas-mas? All of these issues impact directly on theeffects of turbulence on the surrounding plasma;without detailed knowledge of them we cannot pre-dict the large scale effects of space plasma turbu-lence.

Cross-Scale offers a unique capability to ad-dress these fundamental questions about the na-ture and effects of turbulence in space plasmas.By measuring the 3D turbulent structure at threescales simultaneously, it will be possible for the firsttime to measure the energy transfer process asit actually occurs, at both fluid and kinetic scales.It will also sample plasma turbulence in a num-ber of dramatically different regimes: the solarwind, where turbulence is well developed and en-ergy inputs are steady; the highly disturbed mag-netosheath, where fluctuations are highly drivenby shocks and compressions; and the magnetotail,which is dominated by the magnetic field and there-fore highly anisotropic.

The following sections describe key spaceplasma turbulence questions, and how Cross-Scalewill address them.

2.4.1 How does the turbulent cascade trans-fer energy across physical scales?

Cascades, dissipation, and particle (non-fluid)scales

A defining characteristic of turbulence is thenonlinear transfer of energy between scales, a pro-cess that leads turbulence analysis to seek uni-versal scaling laws. However, scaling laws canonly exist where the energy transfer process isthe same over a range of scales, such as withinthe plasma fluid (magnetohydrodynamic) regime.Some of the most important effects of turbulence,including plasma heating, particle scattering andthe acceleration of stellar winds, are intimately re-lated to fluctuations on the gyroscales of ions orelectrons. On these scales, the turbulence is muchmore complex and is not scale invariant. In ad-dition, many of the assumptions that are made tostudy fluid range turbulence, e.g., that the plasmaflows past spacecraft much faster than the internalwave speeds, or that the plasma is incompress-ible, cannot be used: many more wave modes arepresent and they interact highly nonlinearly.

The four Cluster spacecraft have made it pos-sible, using novel analysis techniques, to measurethe wavevectors of turbulent energy in space plas-mas for the first time[29]. However, this analysis isrestricted to an order of magnitude in scales abovethe average spacecraft separation.

The small scale termination of the turbulencecascade is controlled by dissipation processes.While these are rather straightforward in collisionalfluids where viscosity is dominant, in collisionlessplasmas there are several possible mechanisms,on both ion and electron scales, and many wavemodes, both electromagnetic and electrostatic, thatcan participate in the transfer of turbulent energyinto particle heating. The multi-scale nature of thissituation means that we have no satisfactory the-oretical or observational understanding of it – andindeed it may be different in different plasma en-vironments. However, without such a picture, wecannot hope to predict the effects of turbulent heat-ing on space plasmas.

It is essential that the three physical scales ofthe cascade – fluid, ion and electron – are mea-sured simultaneously with sufficient high quality in-strumentation (electric and magnetic fields, ionsand electrons) if we are to determine the proper-ties of the active turbulent cascade. Cross-Scale,

28 June 2007 15


electron scale

d~10 km

B²

(k)

k

10-1

100

101

102

103

0,1 1 10

-----y = k-2,60+/-0.04

B2[n

T2 (rd/km

)-1]

kvr

B²~k-8/3

B²~?

B²~?Sahraoui et al., PRL 2006

L ~2000 kmmax

L ~200 kmmin

ion scale

d~100 km

fluid scale

d~1000 km

Figure 8: A wavevector spectrum of turbulent magneticenergy in the highly disturbed magnetosheath derivedusing multi-point measurements to filter the time se-ries data into their joint space-time spectra. The fittedpower law is shown in red for the range of scales, cover-ing around a decade, which were accessible to the fourCluster spacecraft at this time. Electron and fluid scales,marked in green, would only be accessible at the sametime with a three-scale formation mission such as Cross-Scale.

with its three tetrahedra targeted at these scales,will measure over this entire range for the first time(see Figure 8).

Theory suggests that in addition to the familiarenergy cascade to small scales, a “reverse” cas-cade exists in quantities such as the magnetic helic-ity, making it possible to generate larger scale mag-netic structures. However, such a cascade cannotbe measured without multi-scale data. Cross-Scalewill measure the helicity spectrum and its develop-ment, with the aim of identifying this reverse cas-cade for the first time.

Wave characteristics and particle interactions

Cross-Scale will not only measure turbulenceover three decades of scale simultaneously, it willdo so with highly targeted instrumentation. In par-ticular, full ion and electron distribution functions willbe measured at the relevant scales, as well as mag-netic and electric fields. This will make it possiblefor the first time to track the energy flow from fluidto ion and electron scales and determine the directeffect on the particles themselves, such as reso-nances, wave generation and heating. Only in thisway can we study the process by which turbulentenergy is finally partitioned between fields and par-ticles.

2.4.2 How does the magnetic field break thesymmetry of plasma turbulence?

Neutral fluid turbulence is usually isotropic. Onlynear a boundary layer or other compression orshear is the isotropic symmetry broken so that tur-bulent properties become different in different di-rections. In contrast, the presence of a local mag-netic field direction means that plasma turbulenceis always anisotropic, even far from boundariesor shears. Indeed, theoretical work suggests thatthe turbulent cascade progresses differently paral-lel and perpendicular to the local field, a result thathas been confirmed with spacecraft data in the so-lar wind[30].

Recently, multi-spacecraft data have made itpossible to study anisotropy in the highly dis-turbed and turbulent plasma near the magne-topause boundary layer (see the Case Study onpage 17). These reveal simultaneous anisotropiesrelative to two directions – the local magnetic fieldand the boundary layer – a scenario unique toplasma turbulence.

The anisotropy of plasma turbulence has im-portant consequences for the transport of massand energy through boundary layers. For instance,the velocity shear layer between the streaming so-lar wind and the stagnant magnetospheric plasmacan become Kelvin-Helmholtz unstable and leadto highly anisotropic turbulent vortical flows, asrevealed by recent 4-spacecraft Cluster observa-tions. Structures with a scale comparable to the iongyroradius can be created through cascading, orthrough secondary instabilities excited in the vor-tex. These small scale dissipative structures canmix the plasma contained in the vortex, and lead todiffusion through the shear layer.

The anisotropy of solar wind turbulence has im-portant implications for the propagation of energeticparticles throughout the solar system, as well asthe large scale connectivity of the magnetic field:the turbulence can “braid” or tangle the magneticfield, resulting in field-line wandering far from theaverage direction (see the Case Study on page 17).These effects have been successfully used to rec-oncile scattering mean free paths of solar ener-getic particles with measured turbulence levels[30].However, it is not clear how this anisotropy devel-ops from the large scale where energy is injected,and how it passes between the fluid, ion and elec-tron scales. In addition, the details of the energytransfer process itself are very poorly understood.

28 June 2007 16


Case Study: Anisotropy and particle transport

Measurements of turbulence in space plas-mas have revealed the presence of signifi-cant anisotropy. Recently, multi-spacecraft datademonstrated that turbulence in the highly dis-turbed plasma downstream of the Earth’s bow-shock scales differently in different directions[29](see top figure). The power in wavevectors par-allel to the field scales in a dramatically differ-ent way to that along the flow or perpendicular tothe local plane of compression. This anisotropyappears to be due both to compression of theplasma next to the magnetopause boundary layer,and to the fundamental differences in energytransfer parallel and perpendicular to the mag-netic field.

The anisotropy is of more than academic in-terest. Such anisotropy can have dramatic effectson the large scale connectivity of the magneticfield, as shown in the simulation box (bottom left),where field lines “wander” perpendicular to thefield. The transport of energetic particles alongthis braided structure results in an enhanced dif-fusion process, which has important implicationsfor the acceleration and propagation of, for exam-ple, cosmic rays, throughout the Universe.

Cross-Scale will make it possible to study forthe first time how this anisotropy develops and istransferred between scales.

Cross-Scale will let us track the development ofanisotropy between scales for the first time, pro-viding information on how energy is transferredthroughout space and time in the turbulent cas-cade. By relating this anisotropy to that observed insheared or compressed hydrodynamic turbulence,we will gain insight into anisotropy as part of theuniversal process of turbulence. Recent advancesbased on the available data are limited to onlyaround a decade in scale and at one location ata time, making it impossible to determine the evo-lution of anisotropy in scales or in space. OnlyCross-Scale can make the multi-scale, multi-pointmeasurements that are required, on both fluid andkinetic scales, to measure, understand and pre-dict the effects of the anisotropy that is unique toplasma turbulence.

2.4.3 How does turbulence generate coher-ent structures?

In both neutral fluids and plasmas, turbulence gen-erates discrete structures. Spacecraft data revealthe presence of large magnetic/density structuresknown as mirror modes[30, 31], Kelvin-Helmholtzwaves[14] and current sheets[32], which appear tobe involved in, or the result of, the energy trans-fer process. Fully-developed turbulence is gener-ally not spatially uniform, and can exhibit strongspatio-temporal “burstiness,” commonly known asintermittency (see Figure 9).

Intermittency has consequences on the trans-port of energy over scales, through the modificationof the scaling in the inertial range, and on its dissi-pation at the small scales[33]. It is frequently pro-posed to explain the heating of the solar corona. It

28 June 2007 17


is also expected to strongly affect scattering and dif-fusion of plasma particles[34]. Statistical methodsbrought from hydrodynamics (e.g., probability den-sity functions and structure functions) have beenapplied to single spacecraft data but they cannotprovide information on the 3D nature of intermittentstructures in plasma turbulence.

Cross-Scale will make measurements at thefluid, ion and electron scales simultaneously, facil-itating the analysis of the statistical properties ofspatial and temporal inhomogeneity. That analy-sis will reveal how intermittency is generated andtransmitted down the cascade. More than this,however, with Cross-Scale’s 12 measuring pointswe will move beyond statistical analyses and for thefirst time actually “image” these structures, answer-ing vital questions such as: How large are they?What is their 3D shape? How do they link together?

The recent identification of reconnection drivenby turbulence (see the Case Study on page 17)demonstrates the importance of current sheets inturbulent flows. Their 3D structure and develop-ment is not well known, but without this knowledgewe cannot predict the statistical effects of the result-ing multiple reconnection sites. Additional observa-tions of bursty reconnection in the Earth’s magneto-tail suggests that even in this very low β plasma, inwhich the strong magnetic field tends to resist theparticle pressure, broad-band fluctuations can drivereconnection, emphasising the multi-scale natureof this process. These results show that turbulenceand reconnection are intimately linked and a multi-scale approach is essential in order to fully quantifytheir effects.

Cross-Scale will not only measure three scalessimultaneously: the 12 spacecraft formation canalso be considered as multiple intermediate-scaletetrahedra at several locations. By performinga similar analysis at multiple locations simultane-ously, for the first time it will be possible to mea-sure unambiguously the growth and developmentof these structures as they travel past the forma-tion.

2.5 Synergies with Other Science Endeav-ours

Cross-Scale (together with SCOPE) forms a stand-alone mission that addresses universal plasma pro-cesses. SCOPE is described further in Section 8.1.Modelling and simulational work will be an impor-tant contributor to all Cross-Scale science objec-

Figure 9: A 2D simulation of electron MHD turbulencereveals that plasma currents are spatially localised intothin sheets and whorls. Many-point measurements,such as those from Cross-Scale, are required to mea-sure such structures in 3D in space plasmas over arange of scales not accessible to simulation[35].

tives. The capabilities and operations of the mis-sion will provide numerous instances of synergywith other activities, some of which are describedhere.

2.5.1 Magnetospheric physics

The availability of data from the outer magneto-sphere provides a natural synergy with studies ofthe Sun-Earth coupling all the way to the iono-sphere. Simultaneous global-scale observationswill be made by networks of ground-based iono-spheric radars, all-sky cameras, and riometers to-gether possibly with satellite-borne imagers. Sincemagnetospheric regions map down to the iono-sphere, those facilities gather a remotely-sensed2D map that in effect adds a larger global scalecontext to the three fundamental scales coveredby Cross-Scale. Thus global-scale measurementsboth enhance the multi-scale physics investigationsand facilitate the exploration of their applicability,linking cause and effect, to the magnetosphere.

2.5.2 Interplanetary physics

At the other extreme, solar and interplanetaryplasma measurements by spacecraft at the L1 La-grange point, and elsewhere, extend observationsof the drivers of the coupling processes to still

28 June 2007 18

Cross-Scale: 3 Mission Profile Cosmic Vision 2015-2025

larger scales. Combinations of spacecraft at L1and, for example, at the Moon, together with Cross-Scale will extend the study of turbulence to scalesthat approach those of source regions on the Sun.

2.5.3 Solar and astrophysical plasmas

As all of the plasma processes to be investigated byCross-Scale are operating in more distant plasmas,the mission objectives are relevant to a very widecommunity. At the same time, interaction with thatwider community will be essential to pose questionsto the data, and phrase the resulting answers, in away that is most accessible to scientists outside thecore mission community.

2.5.4 Laboratory plasmas

Turbulence and reconnection are responsible fortransport processes and disruption in laboratoryplasmas, and form an important aspect of futureexperiments such as ITER. Thus Cross-Scale pro-vides a vital incentive for these two relatively sepa-rate disciplines to cross-fertilise by addressing thephysics rather than the context.

3 Mission ProfileTwelve spacecraft are required to observe simulta-neously three disparate scales. This optimum con-figuration ensures that the measurements at eachsmaller scale are centred within their larger context.The payload demands at particularly the electronscale result in a spacecraft bus in the 100 kg class,with an overall dimension of ∼ 1.2 − 1.5 m. Forreasons of economy, interchangeability, and oper-ations, the multiple use of a single bus design isan effective strategy. Soyuz-Fregat launch capabili-ties can deliver 10 suitably-instrumented spacecraftinto an Earth orbit where they can address the sci-ence objectives posed in Section 2; this forms theCross-Scale ESA mission.

Cross-Scale’s partner, the SCOPE mission tobe provided by JAXA, complements the minimalconfiguration with (at least) two additional space-craft, one with a comprehensive suite of high-resolution instruments that enhances the finely-targetted Cross-Scale capabilities at the smallestscales. SCOPE is described more fully in Sec-tion 8.1. The combination thus meets the optimal12 spacecraft configuration with an optimal pay-load.

Table 2: Cross-Scale mass budget[36] incl. margin.Item No. Mass∗ P/L∗∗ ∆V† Fuel

(kg) (kg) (m/s) (kg)Fluid s/c 4 130 13 585 39Ion s/c 4 156 39 235 17Electron s/c 2 164 47 120 9Dispenser 1 427 1215 1020Total wet s/c 1713Tot. dispenser 1447Margin 23% 725SF-2 max.†† 3885∗ Dry mass per item, includes payload∗∗ Incl. ranging sys. + 23% average margin† Includes de-orbit requirement†† Soyuz-Fregat-2 (1b) to 20,000 km

3.1 Launch and Orbit

3.1.1 Overall Characteristics

The 10 ESA spacecraft will be accommodated ona single dispenser module, with the entire assem-bly launched using a Soyuz-Fregat 2 from Kourou.The launch mass budget is shown in Table 2. JAXAwill provide a separate launch for SCOPE. The fi-nal apogee is of order 25 Re with a low perigee(1.4 Re geocentric). This roughly equatorial (14◦ in-clination) orbit (see Figure 10) enables Cross-Scaleto reach the regions where the targetted physicsis occurring, including the dayside bowshock (12–20 Re), reconnection at the magnetopause (8–12 Re) and in the “tailbox” region of the geomag-netic tail (10–20 Re), and turbulence in the solarwind (12-25 Re), magnetosheath, and geomagnetictail/plasma sheet. The orbit suffers a radiation doseof approximately 12 krad per year (assuming 4 mmof Al) and maximum eclipses of 8 hours. The mis-sion can satisfy international space debris agree-ments through controlled de-orbiting of all space-craft and the dispenser.

An alternative 10 × 25 Re orbit would also meetthe science objectives, but is difficult to reach withinthe launch capability for the 10 spacecraft configu-ration. A change in the number or size of space-craft, or utilisation of additional launcher capacity(e.g., JAXA), could make such an orbit attractive interms of both science and space debris control.

Payload mass estimates include intersatelliteranging systems (∼ 5 kg) where needed. The pay-load mass margins on individual spacecraft, basedon the instrument configurations summarised in Ta-ble 5, vary from 8% to 35%. The total payload masscarried by the 10 spacecraft (including ranging) is

28 June 2007 19

Cross-Scale: 3 Mission Profile Cosmic Vision 2015-2025

Tailbox: 25 x 10 x 4 Re

Sun Ecliptic Plane

Magne topaus e

Bow Sh ock

Magne tos h e athSolar W ind

Figure 10: Cross-Scale orbit showing the location of thetargetted regions. Note that the orbit precesses annu-ally.

258 kg, to which a mass margin of 17% has beenadded to reach the total launch wet mass in Table 2;a further 23% margin has then been added to thefinal launch stack.

Propulsion requirements for deployment, con-stellation maintenance, and limited reconfigurationranges from 10 m/s per spacecraft at the smallestscale, 125 m/s for the ion scales, and 475 m/s atthe fluid scales[37]. These allow for corrections dueto perturbations and drift. Masses in Table 2 in-clude that required to de-orbit each spacecraft. Adispenser capable of delivering all 10 spacecraft tothe high-apogee orbit is illustrated in Figure 11 withspecifications given in Table 2.

The constellation of spacecraft will evolvearound the orbit, essentially stretching out throughperigee. All of the key science objectives areaddressed with observations at interim to largegeocentric distances, so that station-keeping canbe kept to a minimum consistent with health andsafety of the spacecraft and restricted operationsresources. The constellation configuration is sum-marised in Table 3. In particular, the reconstruc-tion of the spacecraft separation and relative tim-ings between events observed on different space-craft needs to be adequate to resolve the spatialand temporal variations on the scale of the differentspacecraft separations. At the smallest, electronscale, these require dedicated spacecraft rangingsystems. The baseline mission includes rangingon the ion scale spacecraft for accuracy and redun-dancy purposes.

Table 3: Constellation parameters and accuracyScale Separation Accuracy Timing

(km) (ms)electron 2–100 125 m 0.25ion 50–2000 1% 0.25–2fluid 3000–15000 1% 2

Figure 11: Dispenser vehicle with the 10 Cross-Scalespacecraft in their launch fairing[36].

3.2 Ground Segment

The ground segment is based on a minimal ESAMission Operations Centre for commanding andoccasional manoeuvres together with a ScienceOperation Centre that will coordinate instrumentcommanding and perform data handling and dis-semination activities, including event identificationand selection of data stored onboard for downlinkthe following orbit.

Comunication will utilise two existing ESA track-ing stations, Perth and Maspalomas, equippedwith 15 m antennas operating in the X-band.Each spacecraft will communicate directly with theground station. There could be cost-benefit advan-

28 June 2007 20

Cross-Scale: 4 Payload Cosmic Vision 2015-2025

tages in S-band equipment which should be stud-ied further. Shared ground segments with SCOPEand/or other international partners could be em-ployed to increase the data return or possibly re-duce overall costs. Data volumes and operationsare discussed more fully below in Sections 4.3 and6. Payload and spacecraft operations, and down-links, need a high degree of autonomy to minimisethe required ground segment resources.

3.3 Other Requirements and Issues

3.3.1 Orbit selection and mission lifetime

Final orbit selection will be agreed between theCross-Scale and SCOPE elements. Commission-ing and interspacecraft calibration will require 6months followed by a nominal mission of 2 years.This will allow each of the relevant regions to bevisited twice, permitting alternative spacecraft sep-aration strategies/scales to be explored based onthe science objectives and experience during thefirst year. This results in a total radiation dose of30 krad.

3.3.2 Other agencies

Various possibilities exist to involve scientists andinitiatives from other Agencies in the Cross-Scaleconcept. Examples are given below. These willneed to be fully explored during the AssessmentPhase. Letters of Commitment accompany thisproposal.

Spacecraft provision

The Cross-Scale approach is scalable to in-corporate additional spacecraft provided by newpartners. These offer capabilities to enhance therange of scales and/or increase the coverage be-yond the minimal four spacecraft at any particularscale. Such additions would be particularly helpful,and straightforward, at the fluid and larger scales.New partner missions could also provide comple-mentary instrumentation and a level of redundancyto mitigate partial or total failure of a Cross-Scalespacecraft. Finally, should descoping be neces-sary, partner missions would be able to step into complete the full multi-scale fleet. The currentCross-Scale/SCOPE launch capability could ac-commodate additional spacecraft (see Section 8.1).

Potential spacecraft-providers include Roscos-mos, who have considered launching and deploy-ing one or more spacecraft, and NASA. Other part-ners may also come forward during the Assess-ment Phase.

Instrument participation

Scientists from the USA have expressed an in-terest, through NASA, in participating in the missionat least at the level of payload. Similarly, Canadianscientists have coordinated potential involvement inflight hardware and science synergies. Given thesubstantial hardware to be flown, the resultant datavolume to be analysed, and the contribution Cross-Scale can make to the science goals of many agen-cies, there are strong arguments to welcome theinternational aspects of the mission.

Ground-based systems and global imaging

Synergies in terms of magnetospheric physicsexist between the Cross-Scale orbits and pay-load on the one hand and magnetospheric facili-ties on the other. The Canadian network of mag-netometers, established in concert with NASA’sThemis mission, would provide global coverage ofthe large-scale current systems and their dynamics,as would international radars (e.g., SuperDarn).Possible satellite-borne global imaging initiativeswould enhance these diverse and complementarydatasets. The Cross-Scale community includes sci-entists from all these disciplines.

4 Payload4.1 Overview

The payload consists almost entirely of proventechnology, with heritage in recent missions (e.g.,Cluster, Polar, Themis, STEREO) and well-advanced concepts (e.g., MMS). Although thespacecraft bus and subsystems are identical,spacecraft devoted to different scales need differentpayload in terms of, e.g., time resolution, particledistributions vs. gross characteristics, and DC/ACfields. The largest demands come from the small-est scale (electron) spacecraft, with those at thefluid scale more modest.

To achieve this while benefiting from theeconomies of identical bus systems, the payloadcan be accommodated in universal mountings,such as modular bays or simple mounting plates.These plates present a common mechanical andelectrical interface to the spacecraft, and are ar-ranged on a single instrument shelf located at thetop of the spacecraft. Figure 12 illustrates the con-cept, which provides ease of assembly, integration,and verification and facilitates replacement of faultyinstruments. A thermal and electrical cover of some

28 June 2007 21


ABH

CG

DFE

O

Figure 12: Instruments mounted on universal plates[36].For reasons of economy, Cross-Scale will utilise a sin-gle spacecraft bus design. Since the science payloadis not identical for all spacecraft, a modular approachis adopted to accommodate different instruments in thesame position. For ease of assembly, integration, andverification, a universal plate will be provided to instru-ment teams. Plate locations used in Table 5 are letteredfor convenience.

kind is needed, which could be a full or partial topto the entire spacecraft or specific to the individualbay.

Table 5 provides an overview of the instrumentcomplement on each spacecraft, shown pictoriallyin Figure 13. The main spacecraft resources areprovided in Table 6.

4.2 Instruments

The instrument complement on each spacecraft istargetted in terms of type of measurement, timeresolution, and level of detail to match the highestscience priorities (see [38] for details). The overallcharacteristics are shown in Table 4 with the de-ployment on the various spacecraft encapsulatedin Table 5 and Figure 13.

DC magnetometers (dual-sensors) will be de-ployed on all spacecraft, returning field vectorssampled at up to ∼ 100 Hz on the electron scale,and 1–10 Hz on the ion scale spacecraft. This dataprovides a robust roadmap of the 3D multi-scalemorphology/connectivity, fundamental componentsof the overall turbulence spectrum, and specificfluctuations responsible for particle scattering andenergisation.

AC search coil magnetometers extend wave-form measurements up to ∼ 500 Hz and spectralinformation up to several kHz. These complete thecharacterisation of waves responsible for particlescattering and extend the turbulence analysis intothe electron dissipation range.

Electric field antennae will be used to coverthe range from DC to 100 kHz on the inner space-

craft, where wave scattering and energisation ofions and electrons is a key priority for shocks andreconnection, and for the dissipation of turbulence.Electric measurements complement the magneticones to identify the electrostatic components of thewaves and turbulence. The antennae will also beemployed to measure the ambient DC electric fieldand the spacecraft floating potential. These, to-gether with an Electron density sounder, providea high time resolution, high quality total electrondensity measurement that is central to the shockand reconnection science objectives, and that un-derpins the calibration of the particle instruments.Apart from the fluid-scale spacecraft, all satelliteswill carry a crossed pair of wire boom antennae. Atthe electron scale all three components of the cru-cial reconnection electric field and the cross-shockelectric fields will be completely characterised bythe addition of rigid, dual-axial booms.

Electrostatic particle detectors of various de-signs will be carried on all spacecraft. These pro-vide not only the basic plasma parameters (den-sity, velocity, temperature, etc.) but also the full3D velocity distribution functions. Collisionlessplasma processes are driven by particle beams,temperature anisotropies, and other features thatare not accessible to the low-order moments. En-ergy partition (heating and acceleration) amongstthe various ion components requires some mass-resolving capability employing time-of-flight tech-niques in concert with the basic electrostatic de-sign that is utilised for electrons. The comprehen-sive instrument suite on the SCOPE mother space-craft (see Table 8) enables Cross-Scale to deploya small number of core instruments to address themulti-scale aspects.

Modern analysers utilise a top-hat design whichhas a narrow field of view in one dimension. Full 3Dvelocity measurements require a full satellite spinperiod for a single sensor, which is adequate at thefluid scale. Multiple sensors, as either dual-headswithin a single unit or multiple units, are deployedon the ion and electron scale spacecraft to in-crease the resolution and provide full 3D capabilityat the corresponding ion or electron time scales de-manded by the science objectives. Additionally, de-flector plates can be employed to increase the fieldof view for sub-spin resolution and reduce the num-ber of sensors required. A combined electron/ionelectrostatic analyser would greatly enhance thescience return within payload mass restrictions by

28 June 2007 22


Figure 13: Accommodation and fields of view of the scientific payload on the electron (left), ion (two middle), andfluid-scale (right) spacecraft[36]. Table 5 identifies the instruments and bay locations, which are lettered in Figure 12.

Table 4: Cross-Scale Instruments - see Table 5 for key to the subset of these deployed on the different spacecraftInstr Mass∗ Power∗ Volume∗∗ Measurement Recent Heritage TRL

(kg) (W) cm3

MAG 1.5 0.5 11 × 5 × 5(×2 units)

Boom-mounted DC vector mag-netic field

VEX, Double Star 9

ACB 1.75 0.1 10 × 10 × 10 Boom-mounted AC vector mag-netic field 1Hz–2kHz (spectra +waveform)

Themis, Demeter 9

E2D 8.0 3.0 20 × 30 × 15(×4 units)

30–50m wire double probe 2Delectric field (DC + spectra) 0–100kHz (DC + spectra)

Themis, MMS, Deme-ter

8

E3D 4.0 2.0 10 cm dia. Dual axial 5m antennae; (DC +spectra)

Polar, Themis 7

EDEN 0.2 0.25 - electron density sounder Cluster 9ACDPU 1.5 2.0 23 × 19 × 6 Common processor & electron-

ics for ACB, E2D, E3D, EDENCluster 7

EESA 2.0 3.0 26 × 15 × 26 Dual-head thermal 3D electronelectrostatic analyser 3eV–30keV

MMS 6

CESA 2.5 2.0 15 × 20 × 15 Combined 3D ion/electron elec-trostatic analyser 3eV–30keVions,electrons

Medusa/Astrid 6

ICA 5 6 25 × 45 × 25 3D ion composition <40 keV Cluster 8ECA 1 0.5 10 × 15 × 15 3D energetic multi-species ion

analyser 100 keV–1 MeVSTEREO 8

HEP 1.2 1 15 × 5 × 15 Solid-state high energy particledetector >30 keV

Themis, Demeter 9

CPP 3.0 2.0 10 × 20 × 5 Centralised payload processor Themis, SpaceWire 5ASP 1.9 2.7 19 × 16 × 17 Active potential control Cluster, MMS 8

∗ Total (all units); excluding most margins; excludes booms∗∗ Sensor volume only for MAG, ACB; MAG electronics 19 × 16 × 10

avoiding separate instruments for each species.Energetic ion and electron detectors are re-

quired to investigate acceleration processes. Thelarger gyroradii and scales associated with theseparticles implies that not all spacecraft need to beso equipped, and the deployment pattern shown inTable 5 represents a minimal configuration at eachscale in a way which returns essential informationon energetic ions, electrons, 3D ion velocity distri-butions, and ionic species.

Centralised onboard processing representsan attractive means of coordinating the payload,sharing (and hence reducing) resources. A dedi-cated wave processor has been successfully usedin the past to coordinate common commanding,data taking, signal processing, and telemetry forelectric and AC magnetic field instruments. Asimilar approach to particle instruments may beadopted. Alternatively, a distributed system takingadvantage of modern communication and comput-

28 June 2007 23

Cross-Scale: 5 Spacecraft Considerations Cosmic Vision 2015-2025

Table 5: Cross-Scale payload by spacecraft. Instrumentlocations are given by lettered bays, and shown in Fig-ure 12; blanks indicate instruments not flown on that par-ticular spacecraft.

Instr Spacecraft/Location on which Deployed

electron ion fluide 1-2 i 1 i 2-4 f 1-4

MAG F F F FACB B B B BE2D ACEG ACEG ACEGE3D OEDEN Z Z ZACDPU Z Z ZEESA BDFHCESA BF BDFH DICA DHECA AEHEP CG HASP OCPP Z Z Z Z

Z refers to internal electronics

ing capabilities, such as SPACEWIRE, will be stud-ied during the assessment phase. There are obvi-ous and significant mass savings to be gained byreducing or eliminating one or several instrument-dedicated data processing units. This strategyneeds to be developed in tandem with the multi-instrument bays in the current spacecraft design.

Active spacecraft potential control will beflown on the electron scale. This limits the con-tamination of the electron measurements by photo-electrons of spacecraft origin, and enables the lowenergy population to be resolved without large andoften uncertain corrections.

The basic characteristics of the instruments aredescribed in Table 4. All instruments have recentflight heritage. Ongoing developments in mass,power, and science performance will be utilised tooptimise the actual mission hardware.

4.3 Data rates

The science requirements met by the payload com-plement on the various spacecraft lead to a sub-stantial data volume. That volume is estimated inTable 7, and used to allocate adequate spacecraftresources, telemetry, and ground station support.

The total mission data rate is 1.6 M-sciencewords per second. A 16-bit science word (gener-ous for the particle instruments) and a compressionratio of 10 (conservative given modern compres-sion techniques for particle data, which dominates

the data volume) would imply a mission data-rateof 2.6 Mbps of which 31% can be accommodatedin a nominal total downlink capability of 800 kbs.The actual downlink budget is driven by the sub-set of the orbit corresponding to the key target re-gions and boundary layers, so the full science ob-jectives can be comfortably accommodated by thisdata rate. In fact, the majority of the data volume isgenerated by the electron scale spacecraft. Basedon Table 7, these accumulate roughly 61 Gbytesof compressed data in two orbits (6 days). If ad-equate centralised onboard storage is not avail-able for these spacecraft, a number of strategiesexist to avoid compromising the science, includ-ing internal instrument memory, autonomous orpredicted event triggers, higher compression, anddown-sampling in energy-angle-time resolution.

5 Spacecraft Considerations5.1 Spacecraft Characteristics

For reasons of economy, all 10 spacecraft will sharea common bus with instrument “bays” capable ofaccommodating different instruments as illustratedin Figure 12. The instrument suite on an individualspacecraft will target the appropriate key measure-ments for one physical scale (electron, ion, or fluid),with some limited redundancy between scales.

The spacecraft design [37, 36] is ∼ 1.5 m in di-ameter with a total dry mass (excluding payload)of ∼110 kg. AOCS is provided by combined sunand star sensors together with mono-propellantthrusters (bi-propellant systems have been con-sidered, but existing flight-proven mono-propellantunits offer similar overall performance and can pro-vide the necessary levels of thrust). The spacecraftare spin-stabilised at 20–40 rpm. Faster spin-ratesimprove data resolution at the fluid scales, but atthe electron scales stability of the spin-axis electricantennas limit the spin-rate to these values.

Intersatellite ranging is required on at least theelectron scale spacecraft, and will be carried onthe ion scale spacecraft to guarantee the accuracydemanded by the mission and to provide redun-dancy. This would utilise the X-band transponders,although a separate S-band system could be em-ployed. This area will require close coordinationwith the SCOPE to ensure that compatible systemsare flown on both missions.

The maximum science payload, excluding mar-gins, that can be accommodated by the bus design

28 June 2007 24

Cross-Scale: 6 Science Operations and Archiving Cosmic Vision 2015-2025

Table 6: Mass and power requirements of the payload complements on the different spacecraft, excluding margin.

Mass Power e 1-2 Mass Power i 1 Mass Power i 2-4 Mass Power fluid Mass Power(kg) (W) No. No. No. No

MAG 1.5 1 1 1.5 1 1 1.5 1 1 1.5 1 1 1.5 1ACB 1.75 0.1 1 1.75 0.1 1 1.75 0.1 1 1.75 0.1 1 1.75 0.1E2D 8 3 1 8 3 1 8 3 1 8 3 0 0 0E3D 4 2 1 4 2 0 0 0 0 0 0 0 0 0EDEN 0.2 0.25 1 0.2 0.25 1 0.2 0.25 1 0.2 0.25 0 0 0ACDPU 1.5 2 1 1.5 2 1 1.5 2 1 1.5 2 1 1.5 2EESA 2 5.5 4 8 22 0 0 0 0 0 0 0 0 0IESA 1.5 2 0 0 0 0 0 0 0 0 0 0 0 0CESA 2.5 2 0 0 0 2 5 4 4 10 8 1 2.5 2ICA 5 6 0 0 0 2 10 12 0 0 0 0 0 0ECA 1 0.5 0 0 0 0 0 0 2 2 1 0 0 0HEP 1.2 1 0 0 0 0 0 0 2 2.4 2 1 1.2 1CPP 3 2 1 3 2 1 3 2 1 3 2 1 3 2ASP 1.9 2.7 1 1.9 2.7 0 0 0 0 0 0 0 0 0Ranging 5 * 1 5 * 1 5 * 1 5 * 0 0 0

Totals 13 34.85 35.05 11 35.95 24.35 15 35.35 19.35 6 11.45 8.1* Included in spacecraft system power budget

is 40 kg consuming 40 W[36]. Less mass is avail-able on the ion and fluid spacecraft due to the ad-ditional fuel requirements. Each spacecraft will re-quire a storage device of ∼ 256 Gbits to hold two fullorbits of data; the two electron spacecraft will re-quire additional storage within the EESA electron-ics and/or onboard selection algorithms (see Ta-ble 7). Communications will be direct to groundstations using an X-band transponder such as thatbaselined for GAIA. This provides variable data-rates up to ∼6.5 Mbps. Alternative low-mass sys-tems are expected to become available in Europein time to be considered for the mission.

5.2 Electromagnetic Cleanliness

Electromagnetic field measurements form a corepart of the data from all spacecraft, and a compre-hensive cleanliness programme, together with ap-propriate boom considerations, will be necessaryto meet the science requirements. Previous mis-sion experience shows that standard techniques tominimise noise through appropriate wiring strate-gies, judicious use of materials and parts, and rigidboom mounting of magnetic sensors will be ade-quate. Continuous surface conductivity will also beimportant to provide electrostatic cleanliness, en-abling accurate measurements of low-energy par-ticles to be made. In flight calibration and inter-

calibration will provide definitive characterisation ofthe stray DC fields and AC contamination.

6 Science Operations and Archiv-ing

6.1 Spacecraft Operations

Operations will be centralised and, where possible,autonomous, e.g., in terms of instrument modesand data taking. The full dataset will be too largeto transmit to ground. Data will be stored onboardeach orbit, with a representative subset (e.g., ev-erything except the highest resolution data from theelectron scale spacecraft) telemetred to the MissionOperations Centre. The Science Operations Cen-tre, which could be virtual or co-located at a sciencedata centre, will use that data to select periods of in-terest with regard to the science objectives of suffi-cient duration to fill the downlink budget. Data fromthose intervals, from all spacecraft, will be teleme-tred during the next contact period(s).

A fallback automatic selection based on aver-age locations of the boundaries and processes andthe orbital position will be in place. Onboard eventselection based on suitable triggers provides an al-ternative and reduces the onboard storage require-ments.

The operation of 10 spacecraft, and the joint

28 June 2007 25

Cross-Scale: 7 Technology Cosmic Vision 2015-2025

Table 7: Cross-Scale data volumes

Dim

1D

im2

Dim

3D

im4

Tim

ee1

-2R

ate

Vol

i1R

ate

Vol

i2-4

Rat

eVo

lFl

uid

Rat

eVo

lE

,com

pth

,fre

qph

isp

ecie

sS

tam

pN

o.(H

z)W

ords

/sN

o.(H

z)W

ords

/sN

o.(H

z)W

ords

/sN

o(H

z)W

ords

/s

MA

G3

11

11

110

0.0

400

110

0.0

400

110

0.0

400

150

.020

0A

CB

31

11

11

500.

020

001

500.

020

001

500.

020

001

200.

080

0E

2D(d

c)2

11

11

150

0.0

1500

150

.015

01

50.0

150

010

.00

E2D

(ac)

232

11

11

100.

065

001

10.0

650

110

.065

00

0.5

0E

3D(d

c)1

11

11

150

0.0

1000

050

.00

050

.00

010

.00

E3D

(ac)

132

11

11

100.

033

000

10.0

00

10.0

00

0.5

0E

DE

N1

11

11

150

0.0

1000

150

.010

01

50.0

100

010

.00

AC

DP

U0

00

00

10.

00

10.

00

10.

00

10.

00

EE

SA

3012

321

14

50.0

5760

500

10.0

00

10.0

00

0.5

0C

ES

A30

1232

11

050

.00

21.

011

521

42.

023

042

10.

557

61IC

A30

1232

121

00.

50

20.

569

121

00.

50

00.

50

EC

A30

1232

121

00.

50

00.

50

20.

569

121

00.

50

HE

P30

1232

11

00.

50

00.

50

20.

557

611

0.5

5761

AS

P1

11

11

150

.010

00

0.5

00

0.5

00

0.5

0

Dat

a-ra

te:

k-w

ords

/(s/c

)/s59

1.9

83.9

101.

212

.5To

talk

-wor

ds/s

/cla

ss11

83.7

83.9

303.

750

.1C

ompr

esse

dG

bits

/(s/c

)in

2or

bits

(6da

ys)

490.

969

.684

.010

.4To

talc

ompr

esse

dG

bits

in2

orbi

ts(6

days

)98

1.8

69.6

251.

941

.5

operations with JAXA, to appropriately position thespacecraft in the nested, multi-scale configurationis a challenge. However, ESA has ample experi-ence in multi-spacecraft operations from which toevolve an effective strategy. The overall configura-tion is relatively stable; once it has been achieved,only minimal station-keeping will be required tocompensate for any small drifts. The science ob-jectives can be met without regular adjustments to

the spacecraft separations; a single reconfigurationat the end of the first year of operations is plannedto fine-tune the separation strategy in the light ofthe initial results.

6.2 Calibration

The large number of instruments will demandnew approaches to instrument calibration activities.Preliminary calibrations will be performed on theground to the extent possible, though it may benecessary and/or efficient to fully calibrate only asubset of each instrument type. Inflight intercal-ibration will be part of the commissioning phase,preferably while spacecraft are relatively close toothers. Ongoing routine calibrations will requiresemi-autonomous procedures based on algorithmsdeveloped pre-flight and evaluated using existingmulti-spacecraft datasets (e.g., Cluster) that havelargely been calibrated using more labour-intensivemethods.

6.3 Data Access

The mission will adopt a modern approach to dataaccess, with the full dataset available to the Cross-Scale community, and open to the entire scientificcommunity after a suitable delay (e.g., 6 months)to ensure adequate calibration and quality control.The Cluster experience of distributed data centresand an Active Archive will serve as a guide, but willform part of the mission concept from the outset. Alead role, e.g., at PI level, identified at the time ofpayload selection would provide the appropriate re-source and schedule for coordination amongst theinstrument teams and for the construction of thenecessary software, standards, and data facilities.

7 TechnologyThe mission has no obvious technological obsta-cles. An ESA TRS [37] has confirmed its overallfeasibility at mission, spacecraft, subsystem, pay-load, and operations levels. A follow-up industrialstudy [36] has provided further concepts and de-tails that are largely consistent with the conclusionsof the TRS.

7.1 Payload Readiness

Most, if not all, the instrumentation has sufficientheritage that it could be built and flown now with-out further development. Realistic instrument TRLsare given in Table 4. Improvements in mass andpower requirements, increased autonomy, speedof duty cycle, and demonstrable multi-functionality

28 June 2007 26

Cross-Scale: 8 Programmatics and Costs Cosmic Vision 2015-2025

Figure 14: The original SCOPE concept: Mother, neardaughter, and three intermediate daughters

(e.g., combined ion/electron detectors, capabilityfor both hot and cold/solar wind ion distributions,etc.) will help to increase both the margins and thescience return. Data compression, calibration anddata processing are areas where efficiency will beparamount.

7.2 Mission and Spacecraft

No novel spacecraft engineering is required. Futurestudies will focus on the control and reconstructionof the interspacecraft positioning and synchronisa-tion. This requires effective use of existing technol-ogy for intersatellite ranging. The dispenser moduleis based on existing commercial buses (SpaceBus4000 or Eurostar 3000) and thus has a high TRL.

8 Programmatics and CostsCross-Scale will work in partnership with its sistermission, SCOPE, to be provided by JAXA. Belowwe summarise the characteristics of SCOPE be-fore detailing the ESA cost estimates, alternativemission strategies, and risk management.

8.1 SCOPE

As noted in Section 3, the Cross-Scale/SCOPEcombination not only meets the optimal 12-spacecraft configuration to fulfil the mission ob-jectives, but also provides comprehensive high-resolution capability at the electron scale. SCOPEis a mature concept within JAXA. Its profile, pay-load, programmatics, and status are summarisedbelow.

8.1.1 SCOPE Profile

The science objectives of SCOPE are closelyaligned with those of Cross-Scale, addressing fun-damental and universal plasma processes by usingmulti-point measurements to investigate the cou-

pling between physical scales. Specifically, themission objectives include magnetic reconnection,shocks in space, plasma mixing at boundaries, andthe physics of current sheets[39]. The missionbuilds on JAXA/ISAS expertise in reconnection andassociated turbulence investigated with GEOTAIL.The mother and a near daughter spacecraft arewell-suited to studying the small scale electron pro-cesses within the tailbox region. The capabilitiesof the instrument suite have been designed to alsostudy small-scale dayside phenomena (the bowshock, driven magnetopause reconnection, and tur-bulence).

The other three daughters, of comparable sizeto the Cross-Scale spacecraft, were designed toprovide basic plasma parameters at intermediate(100–1000 km) scales to address the cross-scalecoupling aspects of the key science objectives. Thefull complement continues to be under considera-tion by JAXA, but is dependent on the overall re-sources available for the mission. It is possible thatup to two of the intermediate daughters may beconstructed by JAXA.

At a minimum, the mother and near daughterwould be provided by JAXA, with Cross-Scale com-pleting and extending the multi-scale constellation.The spare launch capacity released by the removalof the three daughters could be used to launch two150 kg spacecraft provided by international part-ners.

The mother payload is more ambitious than thataccommodated by the Cross-Scale mission, andthus provides comprehensive and complementarydatasets. Most notable are the intermediate energyparticle instruments. The near daughter servesboth to provide spatial information at very small(1–10 km) scales and, spinning orthogonal to themother, employs long wire electric booms to cap-ture, in combination with those of the mother, the3D electric fields in a novel and totally unbiasedway.

8.1.2 SCOPE Payload

The SCOPE payload is summarised in Table 8.Most of the instruments have heritage in previousmissions or are in an advanced breadboard stage.The mother carries 87 kg of science payload andhas a total dry mass of 464 kg. The near-daughtercarries 14 kg of science payload with a total drymass of 79 kg. An intersatellite ranging systemis employed for accurate determination of space-

28 June 2007 27


craft separation vectors (1% at separations downto 1 km); this would be integrated with a similar sys-tem on Cross-Scale at the electron scales.

8.1.3 SCOPE Launch and Operations

SCOPE will be launched on an H2A rocket fromTNSC, Japan, either directly into a highly ellipticalhigh apogee orbit or via GTO; both scenarios havebeen verified. The launcher is capable of deliveringthe complete SCOPE mission (including all daugh-ters) into a high-perigee (9 Re) orbit.

The studied orbit is inclined to spend approx-imately one month per year in the tailbox re-gion (offset from the ecliptic to be placed withinthe reconnection regions) during winter. Cross-Scale mission requirements[37] are virtually identi-cal. The mothership has hydrazine, bi-propellant,and mono-propellant systems for orbit insertion,manoeuvres, and attitude control; the daughter isequipped with mono-liquid hydrazine propulsion.

Communications will use X-band for both up-and downlink, with the intersatellite communica-tions and ranging system operating in S-band.Ground segments are based on a 64 m dish at theUsuda Deep Space Center and a 34 m dish at theUchinoura Space Center, both in Japan.

8.1.4 SCOPE Status and Schedule

SCOPE studies began in earnest in 2003; the con-cept is now mature. Key sub-systems, including in-tersatellite ranging, spin axis antenna design, andnovel instruments have been under developmentsince 2004; most of the science payload has strongflight heritage.

SCOPE will stay in Pre-Phase A for the next twoyears, during which the development of the Inter-satellite Ranging and Clock Synchronisation Sys-tem, spin axis antenna, and medium energy rangeparticle spectrometers will be completed.

In late 2009, SCOPE will be reviewed by theSpace Science Committee of ISAS/JAXA, afterwhich it will move to Phase B. The earliest possi-ble launch date is 2016 but this can be adjustedquite easily to match the Cross-Scale launch datein 2017.

8.1.5 SCOPE Budget

Assessment Phase

The budget for the Pre-Phase A term to the endof 2009 is guaranteed.

Cost to Completion

The total cost of the mission is estimated to staywithin the envelope for a science mission managedby ISAS/JAXA.

8.2 Payload Provision

The number of spacecraft in Cross-Scale demandsa large number of instruments, which following nor-mal practice will be provided by the national agen-cies. This is feasible provided economies of massproduction are fully exploited. A programme ofstreamlined ground calibration (plus in-flight inter-calibration) will enable instruments to be deliveredin a timely fashion.

8.3 Costs

The overall mission costs to ESA are shown in Ta-ble 10, based on ESA guideline amounts, estimatesbased on the study reported in [37], and estimatesprovided by industry. Two scenarios, of the severaldiscussed in Section 8.4.2, are costed: a full imple-mentation of the mission concept utilising the full 10ESA spacecraft together with the mother-daughterSCOPE pair, and a minimal configuration compris-ing 6 ESA spacecraft together with 4 JAXA/otheragency-supplied spacecraft.

We have estimated the costs to national agen-cies in providing the necessary instrument suitesin Table 9. That table includes a salary elementfor scientists and technicians in the various labo-ratories although such costs are not explicit in thefunding model employed in some countries.

Apart from the overall payload costs, the costsof non-ESA participants are not available. In par-ticular, while the overall costs for SCOPE arenot publically available, JAXA has confirmed thatthe mother, near daughter, and two intermediatedaughters fit within the possible financial envelope.

8.4 Risks and Alternative Strategies

8.4.1 Risk Assessment

Cross-Scale is a relatively low-risk mission basedlargely on flight-proven technology and instrumen-tation. Operations and data processing will bechallenging within a restricted set of resources,but manageable. The optimum mission includesthe SCOPE elements to be provided by JAXA; adescope to 10 spacecraft in which one corner isshared by all three scales (electron, ion, and fluid)would meet the core objectives. Table 11 sets outthe main risks and the mitigation strategy.

28 June 2007 28


Table 8: SCOPE Payload SummaryAcronym Instrument Measurement characteristics

Mother

FESA Fast electron spectrum analyser 10 eV–40 keV electrons (8 for ∼2–8 ms resolution) @ 32×8×16

MESA Medium energy electron spectrumanalyser

2–100 keV electrons @ 1/2 spin @ 32 × 8 × 16

HEP-ele High energy particles - electrons 3D electrons 30–700 keV @ spinFISA Fast ion spectrum analyser 5 eV/q–40 keV/q ions (4 for 1/8 spin resolution) @ 32×8, 32×

16, 64IMSA Ion energy mass spectrum anal-

yser5 eV/q–25 keV/q ions @ 1/2 spin @ 32× 8× 16 for 8 massesin range 1-20

MIMS Medium energy ion mass spectrumanalyser

10–200 keV/q ions @ 1/2 spin @ 16, 32×8×16 for 5 masses

HEP-ion High energy particles - ions 30–1000 keV 3D ions @ 12 energies @ 6 mass groupsMGF Magnetic field dual sensors mid- and tip of 5 m boom for vector magnetic

field at 64,128 HzOFA/WFC-B Onboard frequency analyser/

Wave-form capture - magnetic fieldtri-axial search coil wave spectra and waveforms < 20 kHz

EFD Electric field detector 30 m wire + 15 m tip-to-tip spin axis rigid dipole antennae fordc & low frequency electric field waveform and spacecraftpotential

OFA/WFC-E Onboard frequency analyser/Wave-form capture - electric field

< 100 kHz electric field

HFR High frequency receiver single component electric spectrum 20 kH-10 MHzDWPC Digital wave-particle correlator correlation index between plasma waves and particles

Near daughter

MGF as for motherEFD as for motherWFC-E as for mother

Particle resolutions given in number of energy × polar × azimuthal bins.Mother and daughter spin rate is 0.3 Hz

8.4.2 Alternative mission strategies

The mission objectives of simultaneous character-isation of the three key scales (electron, ion, andfluid) are uncompromisingly met by the full fleet of10 ESA-supplied Cross-Scale spacecraft in part-nership with the 2 (minimum) JAXA SCOPE space-craft. In particular, this configuration enables mea-surements at each scale to be centred in the largerones. The full context (e.g., upstream and down-stream) in which the smaller scale is driven by, andfeedbacks into, the larger scale is retrieved.

Nevertheless, for both financial reasons and in-flight spacecraft failure, it is important to assess theextent to which the science objectives could be ad-dressed with fewer spacecraft, and to explore otherscenarios. A subset of these is presented here,with the expectation that some will be explored incollaboration with any prospective partners duringthe Assessment Phase.

Eight ESA Cross-Scale spacecraft plus fourSCOPE

This scenario retains the full science capabil-ity at reduced mass and cost to ESA. At this level,and especially in the further reduced scenarios de-scribed below, the mission relies on a firm ESA-JAXA partnership.

Ten ESA Cross-Scale spacecraft

In the event that SCOPE is unavailable, orfails, most of the ground-breaking science objec-tives could be met by configuring the 10 availablespacecraft into a shared-corner configuration, assketched in Figure 15. Some objectives, such asthe acceleration and energy partition at shocks,could not be performed together as both the up-stream and downstream regions would not be fullysampled. Thus reducing the total number of space-craft to 10 requires more events for analysis, with

28 June 2007 29


Table 9: Payload costs (M€) incl. staff

Instr. Dev->EM Per Unit Num TotalMAG 1.0 0.3 10 4.0ACB 1.0 0.3 10 4.0E2D 1.5 0.4 6 3.9E3D 1.0 0.3 2 1.6EDEN 0.2 0.1 6 0.8ACDPU 0.5 0.3 10 3.5EESA 3.0 0.5 8 7.0CESA 3.0 0.5 18 12.0ICA 3.0 0.5 2 4.0ECA 3.0 0.5 6 6.0HEP 2.0 0.3 10 5.0CPP 0.5 0.3 10 3.5ASP 2.0 0.5 2 3.0Data Fac. 0.5 2.0 1 2.5Totals 22.2 101 60.8

Table 10: Cross-Scale ESA mission costs

Full 10 spacecraft fleetActivity Cost Margin Tot Tot

M€ % M€ %

ESA study + internal 35 15 40 10MOC and SOC 54 20 65 17Industrial activities:s/c design to EM 30 10 33 9s/c build 150 20 180 47Dispenser 20 15 23 6Soyuz Fregat-2B 39 10 43 11

Total 328 17 384 100

Minimum 6 ESA spacecraftActivity Cost Margin Tot Tot

M€ % M€ %

ESA study + internal 32 15 37 12MOC and SOC 50 20 60 20Industrial activities:s/c design to EM 27 10 30 10s/c build 90 20 108 36Dispenser 20 15 23 8Soyuz Fregat-2B 39 10 43 14

Total 258 16 300 100

implications for the mission lifetime and more de-tailed constellation planning. Additionally, the morecomprehensive instrument suite on the SCOPEmother would be unavailable. While this does not

Figure 15: The shared-corner configuration for 10spacecraft. This covers all 3 scales, but loses the con-text on one side.

compromise the primary objective of cross-scalecoupling, provided a core dataset as envisaged forthe electron-scale spacecraft is available, it wouldhave a measurable impact on the total science re-turn.

Eight Cross-Scale spacecraft plus two SCOPESix Cross-Scale spacecraft plus four SCOPE

These scenarios would save ESA costs andlaunch mass. They would utilise the shared-cornerconfiguration described above, with the benefit ofthe SCOPE instrument suite. Otherwise, they offersimilar science capability and would meet the pri-mary science objectives to a satisfactory level.

Eight Cross-Scale spacecraftSix Cross-Scale spacecraft plus two SCOPE

With 8 spacecraft, 3D coverage of 3 scales isnot feasible. However, it would be possible to covertwo scales completely and hence to address themission objectives by a sequence of constellations.A large fraction of the objectives require good un-derstanding of the coupling between the ion andelectron scales, and this coupling would be a pri-ority in an 8-spacecraft mission. Turbulence stud-ies and particle acceleration would benefit by somecorner sharing here so that at least a single point atthe fluid scale could be obtained simultaneously.

8.4.3 Options

The science objectives and mission concepts ofCross-Scale lend themselves well to creative andcollaborative options which would save resourceat the risk of adding international and operationalcomplexity. While it is too early to lay down detailed

28 June 2007 30


Table 11: Risk AssessmentRisk Likelihood Severity Mitigation

ESA cost over-run H M Descope to fewer ESA spacecraft; add internationalpartner missions; use EM spacecraft as 10th; re-duce redundancy (increases risk elsewhere); re-move ion-scale ranging system

SCOPE unavailable M L Put Cross-Scale in shared-corner configurationNational agency cost over-run M M Widen international participationNational agency inability to fund fullpayload

M M Widen international participation

Single spacecraft failure M M Adjust configuration to shared corner; add interna-tional partner missions for redundancy. Ensure min-imal level of payload redundancy between differentscales

Instrument failure H M Ensure adequate payload redundancy between dif-ferent scales

Instrument Development (e.g.,Dual-head EESA, CESA withadequate geometrical factor)

M L Heritage instrumentation can be flown in place ofany potential modified designs

Total mass excess M M Some payload descope is possible which would af-fect secondary science goals; utilise SCOPE sparelaunch capacity; use low mass X-band system(lower TRL) or S-band

Science risk from alternative mis-sion strategies

L M Evaluate science capability under alternatives;widen international participation

H = High; M = Medium; L = Low

scenarios, it is useful to know that they exist andshould be considered if and when the need arises.

Dispenser and orbit

The dispenser that places the 10 Cross-Scalespacecraft into their near-final orbits needs suffi-cient subsystems to perform that task and then de-orbit itself from the nominal, low-perigee missionorbit. In fact, it needs power, communications, andAOCS. Two possibilities have emerged in a Cross-Scale industrial study[36]. One is to turn the dis-penser into the 10th spacecraft to occupy one ofthe fluid-scale positions, say, by the addition of sci-entific payload, thereby making much better andlonger use of the subsystems the dispenser mustcarry. Alternatively, use one of the 10 Cross-Scalespacecraft to provide the subsystems (power, com-munications) for the dispenser, which then is littlemore than a motor. The Cross-Scale spacecraftwould then need to place the dispenser in an or-bit that would decay and manoeuvre itself back tojoin the constellation.

Much of this is driven by the space debrisrequirement to de-orbit such low inclination, lowperigee spacecraft. This also has implications for

the fuel, mission lifetime, and reliability of all 10Cross-Scale spacecraft. A higher perigee (10 Re)orbit would fully meet the science objectives, andwould even be advantageous for studying drivenreconnection at the magnetopause. Such an orbitdoes not appear to be reachable by the 10 Cross-Scale spacecraft on a Soyuz-Fregat launch vehi-cle. Moving some spacecraft to the daughter lo-cations within the JAXA SCOPE launcher, or other-wise reducing the total launch mass, could enablethe remaining ESA spacecraft to reach that higherperigee orbit, from which they and their dispenserwould not need to de-orbit.

JAXA SCOPE daughters

The number of JAXA-built daughters in addi-tion to the SCOPE mother-near daughter pair is un-der study. The original SCOPE mission includedthree such daughters; current estimates show thattwo would be feasible within the financial enve-lope. These would provide a comfortable level of re-dundancy for Cross-Scale and/or enable the over-all Cross-Scale resource envelope to be reducedshould that prove necessary (see Section 8.4.2).

28 June 2007 31

Cross-Scale: References Cosmic Vision 2015-2025

Multi-agency collaborations

As noted in Section 3.3.2 there is already con-siderable international interest in the mission. Thatinterest includes participation in the flight hardwareand science objectives, which are well-aligned withthose of many international space agencies.

Additionally, some agencies, such as Roscos-mos or NASA, may be in a position to provide theirown spacecraft. Depending on the size and scale,these would either be independently launched orpossibly integrated into the SCOPE mission asdaughter spacecraft.

Less redundancy/increased risk

The level of resource required is strongly de-pendent on the level of redundancy and reliabilitydemanded.

9 Communications and OutreachCross-Scale is an ambitious undertaking target-ted at quantifying fundamental, complex processes.The goal of any outreach programme should there-fore capture the essence and universality of thekey phenomena. The spacecraft will be studyingshocks, reconnection, and turbulence within thenear-Earth environment. Thus solar storms, geo-magnetic activity, and their consequences (spec-tacular aurorae, telecommunications and technicalsystems outages, and bio-hazards) provide visualand tangible vehicles to convey the mission mes-sage. A familiar blend of web resources and mediacoverage will be employed.

References[1] Schwartz, S. J. & Burgess, D. Quasi-parallel

shocks - A patchwork of three-dimensional struc-tures. Geophys. Res. Lett. 18, 373–376 (1991).

[2] Giacalone, J., Schwartz, S. J. & Burgess, D. Ob-servations of suprathermal ions in association withSLAMS. Geophys. Res. Lett. 20, 149–152 (1993).

[3] Lucek, E. A. et al. Cluster magnetic field observa-tions at a quasi-parallel bow shock. Annales Geo-physicae 20, 1699–1710 (2002).

[4] Burgess, D. et al. Quasi-parallel Shock Structureand Processes. Space Science Reviews 118, 205–222 (2005).

[5] Yokoyama, T., Akita, K., Morimoto, T., Inoue, K. &Newmark, J. Clear Evidence of Reconnection In-flow of a Solar Flare. Astrophys. J. Lett. 546, L69–L72 (2001).

[6] Kis, A. et al. Multi-spacecraft observations of dif-fuse ions upstream of Earth’s bow shock. Geophys.Res. Lett. 31, 20801 (2004).

[7] Walker, S., Alleyne, H., Balikhin, M., Andre, M.& Horbury, T. Electric field scales at quasi-perpendicular shocks. Annales Geophysicae 22,2291–2300 (2004).

[8] Bale, S. & Mozer, F. Measurement of large paralleland perpendicular electric fields on electron spa-tial scales in the terrestrial bow shock. Phys. Rev.Letters in press (2007).

[9] Burgess, D. Interpreting multipoint observations ofsubstructure at the quasi-perpendicular bow shock:Simulations. Journal of Geophysical Research(Space Physics) 111, 10210 (2006).

[10] Hellinger, P., Travnıcek, P. & Matsumoto, H. Ref-ormation of perpendicular shocks: Hybrid simula-tions. Geophys. Res. Lett. 29, 87–1 (2002).

[11] Moullard, O., Burgess, D., Horbury, T. S. & Lucek,E. A. Ripples observed on the surface of theEarth’s quasi-perpendicular bow shock. Journal ofGeophysical Research (Space Physics) 111, 9113(2006).

[12] Lobzin, V. V. et al. Nonstationarity and reformationof high-Mach-number quasiperpendicular shocks:Cluster observations. Geophys. Res. Lett. 34, 5107(2007).

[13] Biskamp, D. Magnetic Reconnection in Plasmas(Magnetic reconnection in plasmas, Cambridge,UK: Cambridge University Press, 2000 xiv, 387p. Cambridge monographs on plasma physics,vol. 3, ISBN 0521582881, 2000).

[14] Hasegawa, H. et al. Transport of solar wind intoEarth’s magnetosphere through rolled-up Kelvin-Helmholtz vortices. Nature 430, 755–758 (2004).

[15] Retino, A. et al. In situ evidence of magnetic re-connection in turbulent plasma. Nature Physics 3,236–238 (2007).

[16] Dmitruk, P. & Matthaeus, W. H. Structure ofthe electromagnetic field in three-dimensional Hallmagnetohydrodynamic turbulence. Physics ofPlasmas 13, 2307 (2006).

[17] Matthaeus, W. H. & Lamkin, S. L. Turbulent mag-netic reconnection. Physics of Fluids 29, 2513–2534 (1986).

[18] Sundqvist, D., Retino, A., Vaivads, A. & Bale, S.Dissipation in turbulent plasma due to reconnec-tion in thin current sheets. Phys. Rev. Lett. in press(2007).

28 June 2007 32

Cross-Scale: References Cosmic Vision 2015-2025

[19] Tanaka, K. G., Shinohara, I. & Fujimoto, M. Param-eter dependence of quick magnetic reconnectiontriggering: A survey study using two-dimensionalsimulations. Journal of Geophysical Research(Space Physics) 111, 11 (2006).

[20] Nakamura, R. et al. Dynamics of thin currentsheets associated with magnetotail reconnection.Journal of Geophysical Research (Space Physics)111, 11206 (2006).

[21] Daughton, W., Scudder, J. & Karimabadi, H. Fullykinetic simulations of undriven magnetic reconnec-tion with open boundary conditions. Physics ofPlasmas 13, 2101 (2006).

[22] Phan, T. D. et al. A magnetic reconnection X-lineextending more than 390 Earth radii in the solarwind. Nature 439, 175–178 (2006).

[23] Henderson, P. D. et al. Cluster PEACE observa-tions of electron pressure tensor divergence in themagnetotail. Geophys. Res. Lett. 33, 22106 (2006).

[24] Fujimoto, M., Nakamura, T. K. M. & Hasegawa,H. Cross-Scale Coupling Within Rolled-Up MHD-Scale Vortices and Its Effect on Large ScalePlasma Mixing Across the Magnetospheric Bound-ary. Space Science Reviews 122, 3–18 (2006).

[25] Cattell, C. et al. Cluster observations of elec-tron holes in association with magnetotail recon-nection and comparison to simulations. Journal ofGeophysical Research (Space Physics) 110, 1211(2005).

[26] Hoshino, M. Electron surfing acceleration in mag-netic reconnection. J. Geophys. Res. 110, 10215(2005).

[27] Drake, J. F., Swisdak, M., Che, H. & Shay, M. A.Electron acceleration from contracting magnetic is-lands during reconnection. Nature 443, 553–556(2006).

[28] Imada, S. et al. Energetic electron accelerationin the downstream reconnection outflow region. J.Geophys. Res. 112, 3202 (2007).

[29] Sahraoui, F. et al. Anisotropic Turbulent Spectra inthe Terrestrial Magnetosheath as Seen by the Clus-ter Spacecraft. Physical Review Letters 96, 075002(2006).

[30] Bieber, J. W., Wanner, W. & Matthaeus, W. H.Dominant two-dimensional solar wind turbulencewith implications for cosmic ray transport. J. Geo-phys. Res. 101, 2511–2522 (1996).

[31] Narita, Y. et al. Low-frequency wave characteris-tics in the upstream and downstream regime of the

terrestrial bow shock. Journal of Geophysical Re-search (Space Physics) 111, 1203 (2006).

[32] Knetter, T., Neubauer, F. M., Horbury, T. & Balogh,A. Four-point discontinuity observations usingCluster magnetic field data: A statistical survey.Journal of Geophysical Research (Space Physics)109, 6102 (2004).

[33] Frisch, U. Turbulence. The legacy of A.N.Kolmogorov (Cambridge: Cambridge UniversityPress, 1995).

[34] Veltri, P., Zimbardo, G. & Pommois, P. Particle prop-agation in the solar wind: anomalous diffusion ofmagnetic field lines in turbulent magnetic fields. Ad-vances in Space Research 22, 55–58 (1998).

[35] Laval, J.-P. http://www.atmos.ucla.edu/ ∼laval /Tur-bulence/ turbulence.html (2007).

[36] EADS Astrium Ltd, UK. Cross-Scale: UK STFC-funded Study. internal report (2007).

[37] v den Berg, M. The Cross-Scale Tech-nology Reference Study Documentation athttp://sci.esa.int/ science-e/www/ object/ in-dex.cfm?fobjectid=38982. http://sci.esa.int/(2007).

[38] The Cross-Scale Team. Cross-Scale SciencePriority Document. http:// www.cross-scale.org/Documents/ CrossScaleSciPriorityDocV1.3a.pdf(2007).

[39] Fujimoto, M. et al. The SCOPE mission. In Fa-vata, F., Sanz-Forcada, J., Gimenez, A. & Battrick,B. (eds.) ESA Special Publication, vol. 588 of ESASpecial Publication, 249 (2005).

[40] Runov, A. et al. Current sheet structure near mag-netic X-line observed by Cluster. Geophys. Res.Lett. 30, 33–1 (2003).

[41] Vaivads, A. et al. Structure of the Magnetic Re-connection Diffusion Region from Four-SpacecraftObservations. Physical Review Letters 93, 105001(2004).

[42] Wygant, J. R. et al. Cluster observations of an in-tense normal component of the electric field at athin reconnecting current sheet in the tail and itsrole in the shock-like acceleration of the ion fluidinto the separatrix region. Journal of GeophysicalResearch (Space Physics) 110, 9206 (2005).

28 June 2007 33

Cross-Scale: Community Cosmic Vision 2015-2025

List of AcronymsEM Engineering ModelGTO Geostationary Transfer OrbitISAS Inst. of Space Astronautical Sci.

(Japan)ITER International Thermonuclear Experi-

mental ReactorJAXA Japan Aerospace Exploration AgencyMHD MagnetoHydroDynmaicsMMS Magnetospheric Multi-Scale MissionMOC Mission Operations CenterRe Radius of EarthSCOPE Scale COupling in the Plasma uni-

versESOC Science Operations CenterSTFC Sci. & Tech. Facilities Council (UK)TNSC Tanegashima Space CenterTRL Technology Readiness LevelTRS Technology Reference StudyVEX Venus EXplorer

Image CreditsCover Page top to bottom

- http://www.jhu.edu/news info/news/home04/oct04/kepler.html

- NASA Marshall Space Flight Center- Axel Mellinger, Univ. of Potsdam, Germany

Engineering diagrams EADS Astrium Ltd, UK

Other figures as cited in the text and/or by mem-bers of the Cross-Scale team

Cross-Scale CommunityThis proposal is submitted by the following 372 scientists from 23 countries.

Austria

Helfried K. BiernatKunihiro KeikaStefan KiehasWerner MagnesRumi NakamuraAndrei RunovKlaus TorkarMartin VolwerkTielong ZhangAustrian Academy of Sciences

Manfred P. LeubnerUniversity of Innbruck

Belgium

Fabien DarrouzetMarius EchimJohan De KeyserHerve LamyViviane PierrardMichel RothBelgian Institute for Space Aeronomy

Giovanni LapentaKatholieke Universitei Leuven

Bulgaria

Emiliya YordanovaSpace Research Institute

Canada

Farbod FahimiRobert FedosejevsSean KirkwoodIan R. MannJonathan RaeRobert RankinIllia SilinRichard SydoraClare WattUniversity of Alberta

Eric DonovanAndrew YauUniversity of Calgary

William LiuCSA

Abdelhaq M. HamzaKarim MezianeUniversity of New-Brunswick

China

Suiyan FuZuyin PuQiugang ZongPeking University

Jinbin CaoChao ShenState Key Laboratory for SpaceWeather/CSSAR

Xiaohua DengWuhan University

Czech Republic

Ondrej SantolikZoltan VorosInstitute of Atmospheric Physics

Dana Jankovicova-SaxonbergovaCzech Academy of Science

Finland

Olaf AmmNatalia GanushkinaTuija PulkkinenFinnish Meteorological Institute

Hannu KoskinenUniversity of Helsinki

Anita AikioUniversity of Oulu

France

Gabriel FruitVincent Genot

Philippe LouarnChristian MazelleHenri RemeCESR

Matthieu BerthomierPatrick CanuGerard M. ChanteurThomas ChustOlivier Le ContelDominique DelcourtDominique FontaineBertrand LembegeRaymond PotteletteLaurence RezeauAlain F RouxFouad SahraouiCETP

Olga AlexandrovaMilan MaksimovicLESIA

Thierry Dudok de WitVladimir KrasnoselskikhAurelie MarchaudonJean-Louis PinconLPCE/CNRS

28 June 2007 34


Germany

Joachim SaurUniversity of Cologne

Edita GeorgescuStein Haaland*Berndt KleckerMPE-Garching*also University of Bergen, Norway

Karl-Heinz GlaßmeierUwe MotschmannTechnische Universitat Braunschweig

Joachim VogtJacobs University Bremen

Jorg BuchnerPatrick W. DalyElena KronbergMPI fur Sonnensystemforschung

Greece

Anastasios AnastasiadisOlga MalandrakiNational Observatory of Athens

Hungary

Peter KovacsEotvos Lorand Geophysical Institute

Gabor FacskoKFKI RMKI

Italy

Patrizia FranciaMassimo VellanteUniversita di L’Aquila

Vincenzo CarboneAntonella GrecoGaetano ZimbardoUniversity of Calabria

Luca Sorriso-ValvoINFM/CNR

Raffaella D’AmicisRoberto BrunoMaria Bice CattaneoIgino CocoGiuseppe ConsoliniMaria Federica MarcucciGiuseppe PallocchiaIstituto di Fisica dello SpazioInterplanetario - INAF

Thomas PenzNational Institute for Astrophysics

Japan

Koji KondohDaisuke MatsuokaKen T. MurataTohru ShimizuMasayuki Ugai

Kazunori YamamotoEhime University

Takashima AsaiTsutomu NagatsumaInstitute of Information andCommunications Technology

Kazushi AsamuraMasaki FujimotoHiroshi HasegawaHajime HayakawaSatoshi KasaharaKiyoshi MaezawaAyako MatsuokaTakefumi MitaniYukinaga MiyashitaToshifumi MukaiMasato NakamuraTakuma NakamuraMasaki NishinoMiho SaitoYoshifumi SaitoYoshitaka SekiNobue ShimadaIku ShinoharaTaku TakadaTakeshi TakashimaKentaro TanakaTakaaki TanakaAtsushi YamazakiSho-Ichiro YokotaInstitute of Space and AstronauticalScience

Ryoichi HigashiHisato ShiraiIshikawa National College of Technology

Tooru SugiyamaJAMSTEC

Yoshitaka GotoMitsuru HikishimaTomohiko ImachiYoshiya KasaharaIsamu NaganoSatoshi YagitaniKanazawa University

Kozo HashimotoToshihiko IyemoriKaori KanedaHirotsugu KojimaShinobu MachidaHiroshi MatsumotoTomohiko MitaniYohei MiyakeDaisuke NagataMichi NishiokaNaoto NishizukaChoosakul NithiwatthnMasahito NoseMitsuo Oka

Yoshiharu OmuraAkinori SaitoKazunari ShibataTadahiro ShimodaKoichi ShinNaoki ShinoharaMariko TeramotoYoshikatsu UedaSatoru UenoHideyuki UsuiHiroshi YamakawaDaiki YoshidaKyoto University

Shuji AbeAkiko FujimotoTohru HadaAkihiro IkedaYasushi IkedaYoshihiro KakinamiHideaki KawanoShuichi MatsukiyoAoi NakamizoYasuhiro NariyukiTakashi TanakaTeiji UozumiAkimasa YoshikawaKiyohumi YumotoKyushu University

Sachiko ArveliusYusuke EbiharaRyoichi FujiiAkimasa IedaYoshiko KoizumiJunichi KuriharaYosuke MatsumotoYoshizumi MiyoshTetsuo MotobaAkimitsu NakajimaNozomu NishitaniSatonori NozawaKaori SakaguchiKanako SekiKazuo ShiokawaShin TanakaTakuo TsudaTaka UmedaTatsuhiro YokoyamaNagoya University

Shinsuke ImadaNational Astronomical Observatory ofJapan

Akira KadokuraYasunobu OgawaMasaki OkadaYoshimasa Tanaka

Akira YukimatsuNational Institute of Polar Research

Masao NakamuraOsaka Prefecture University

Hitoshi FujiwaraYasumasa KasabaAtsushi KumamotoHiroaki MisawaYukitoshi NishimuraTakayuki OnoTakeshi SakanoiHiroyasu TadokoroFuminori TsuchiyaManabu YamadaTohoku University

Masahiro HoshinoIchiro YoshikawaKazuo YoshiokaUniversity of Tokyo

Yoshihiro AsanoTsugunbu NagaiToshio TerasawaTokyo Institute of Technology

Keigo IshisakaToyama Prefectural University

Mexico

Xochitl Blanco-CanoOlivia L. Enrıquez RiveraUniversidad Nacional Autonoma deMexico

Norway

Nikolai ØstgaardKristian SnekvikEija TanskanenUniversity of Bergen

Jan A. HoltetBjørn LybekkJøran MoenEllen OsmundsenYvonne RinneUniversity of Oslo

Rico BehlkeUniversity of Tromsoe

Poland

Jan BleckiBarbara PopielawskaRoman SchreiberMarek StrumikSpace Research Center of the PolishAcademy of Sciences

Kris MurawskiUMCS

Romania

Adrian BlagauOctav MarghituInstitute for Space Sciences

28 June 2007 35


Russia

Grigory KoinashAnatoli A. PetrukovichSergey SavinOleg VaisbergLev ZelenyiIKI - Space Research Institute

Sergey ApatenkovVictor A. SergeevSt. Petersburg State University

Slovakia

Jozef MasarikComenius University

Karel KudelaJan RybakSlovak Academy of Sciences

Sweden

Lars BlombergTommy JohanssonTomas KarlssonPer-Arne LindqvistGoran MarklundUwe MotschmannRoyal Institute of Technology (KTH)

Mats AndreJan E. S. BergmanStephan BuchertJonas EkebergAnders ErikssonMaria Hamrin*Yuri KhotyaintsevTomas LindstedtRickard LundinHans NilssonAlessandro RetinoLisa RosenqvistIngrid SandahlK. StasiewiczGabriella Stenberg*Andris VaivadsMasatoshi YamauchiSwedish Institute of Space Physics*also Umeå University

Taiwan

Motoharu NowadaNational Central University

Chio-Zong (Frank)ChengNational Cheng Kung University

United Kingdom

William P. WilkinsonUniversity of Brighton

Mervyn FreemanRichard B. HorneAlan RodgerBritish Antarctic Survey

Francois RinconCambridge University

Chandrasekhar ReddyAnekalluYulia V. BogdanovaAndrew CoatesAndrew FazakerleyClaire FoullonPaul HendersonD. O. KatariaAndrew LahiffBranislav MihaljcicChristopher J. OwenI. RozumYasir SoobiahAndrew WalshUniversity College London

Leah-Nani S. AlconcelCesar BertucciLaurence BillinghamChris CarrRobert ForsythEdmund HenleyTim HorburyBertrand LefebvreElizabeth LucekAdam MastersAlexander SchekochihinSteve Schwartz

Anders TjulinImperial College London

Mike KoschPatrick DaumMichael H. DentonAndrew J. KavanaghJames WildLancaster University

Sarah V BadmanStanley W H CowleyRobert FearColin ForsythCarlos GaneAdrian GrocottMark LesterSteve MilanDarren WrightTim YeomanUniversity of Leicester

Emma E WoodfieldUniversity of Liverpool

David BurgessQueen Mary, University of London

Ruth BamfordRobert BinghamJacqueline A. DaviesMalcolm W DunlopMike HapgoodChris PerryRutherford Appleton Laboratory

Misha BalikhinYasuhide HobaraSimon N. WalkerUniversity of Sheffield

Mina AshrafiBetty LanchesterMike Lockwood*Southampton University*also STFC/Rutherford AppletonLaboratory

Andrew BuckleyPaul GoughUniversity of Sussex

Manuel GrandeUniversity of Wales, Aberystwyth

Sandra C ChapmanUniversity of Warwick

USA

Stuart D. BaleChris ChastonJonathan EastwoodJames McFaddenMarit OierosetTai PhanDavid SundkvistUniversity of California - Berkeley

Vassilis AngelopoulosUniversity of California - Los Angeles

Stefan ErikssonUniversity of Colorado at Boulder

Katariina NykyriEmbry-Riddle Aeronautical University

Melvyn GoldsteinMichael HesseGuan LeThomas E MooreAntti PulkkinenDavid SibeckJim SlavinTimothy J StubbsGoddard Space Flight Center

Ken-Ichi NishikawaNSSTC - Huntsville

Karlheinz J. TrattnerLockheed Martin ATC

Vania K. JordanovaBenoit LavraudGeoff ReevesMichelle F. ThomsenRob WilsonLos Alamos National Laboratory

Terry ForbesHarald KucharekJoachim RaederRoy TorbertUniversity of New Hampshire

James L BurchJerry GoldsteinJorg-Micha JahnDave McComasCraig J. PollockSouth West Research Institute

28 June 2007 36