The Sun and HeliosphereIn Three Dimensions

Report of the NASAScience Definition Teamfor the STEREO Mission

....... 1........... 4

........ 5

.........

...... 1

...... 15

....

.... 21..... 21... 21..... 21... 22.... 23

...... 25

............ 28..... 31

iii

Table of Contents

Executive Summary ....................................................................................................1. Introduction ............................................................................................................. 32. Scientific Objectives ...............................................................................................

Coronal Mass Ejections ....................................................................................................................... 4CME Onset ..................................................................................................................................... 4CME Geometry and Onset Signatures....................................................................................Reconnection .................................................................................................................................. 6Surface Evolution ........................................................................................................................... 6

The Heliosphere Between the Sun and Earth .............................................................................7CMEs in the Heliosphere ..................................................................................................................... 9Tracking Disturbances from the Sun to Earth .............................................................................0Particle Acceleration by CMEs .......................................................................................................... 11Magnetic Clouds ................................................................................................................................ 12Escape of Magnetic Flux and the Solar Dynamo ........................................................................Coronal Magnetic Fields ................................................................................................................... 15Coronal Loop Heating ....................................................................................................................... 15

Loop Cross Sections ..................................................................................................................... 17Scaling Laws................................................................................................................................. 18Axial Gradients ............................................................................................................................. 18Loop–Loop Interactions ............................................................................................................... 18

Solar Wind Origins ............................................................................................................................ 18Solar-B Collaboration ........................................................................................................................ 19Collateral Research ............................................................................................................................ 19

Helioseismology ........................................................................................................................... 19Solar Irradiance............................................................................................................................. 20X-ray and Gamma-ray Bursts ................................................................................................... 20Faint Objects ................................................................................................................................. 20

3. Making the Best Use of STEREO Images ..............................................................Determining the 3-D Structure and Dynamics of the Corona ......................................................

Resolving Line-of-Sight Ambiguities with Stereo Observations ..............................................Use of Triangulation to Determine the 3-D Coordinates of Coronal Features ........................Automatic Feature Tracking ........................................................................................................ 22Use of Magnetic Field Models in Conjunction with X-ray and EUV Observations ..................Magnetic-Field-Constrained Tomographic Reconstruction of the Corona ..............................

Visual Evaluation of Stereo Images ................................................................................................... 244. Space Weather .......................................................................................................

Solar Energetic Particles (SEPs) ........................................................................................................ 25The STEREO Beacon ........................................................................................................................ 27

Space Weather Forecast Data ................................................................................................... 27Maximizing the Science Return ............................................................................................... 28

5. Mission Overview ...................................................................................................6. Phases of the STEREO Mission ............................................................................

Phase 1: The 3-D Structure of the Corona (first 400 days, α ≤ 50°) ................................................. 31Phase 2: The Physics of CMEs (days 400 to 800, 50° ≤ α ≤ 110°) ................................................... 31

.... 31..... 3

.... 34

... 35.I-1. II-1III-1

Phase 3: Earth-Directed CMEs (days 800 to 1100, 110° ≤ α ≤ 180°) ............................................... 31Phase 4: Global Solar Evolution and Space Weather (after day 1100, α > 180°) ............................. 31

7. Observational Approach .........................................................................................Chromosphere and Low Corona Imager .......................................................................................2Coronagraph ...................................................................................................................................... 32Radio Burst Tracker ........................................................................................................................... 33Heliosphere Imager ............................................................................................................................ 33Solar Wind Plasma Analyzer ............................................................................................................. 33Magnetometer .................................................................................................................................... 34Solar Energetic Particle Detector ....................................................................................................... 34

8. Conclusions ............................................................................................................9. Bibliography ............................................................................................................Appendix I. Mission Requirements and Proof of Feasibility .........................................Appendix II. Data Compression ...................................................................................Appendix III. The Case for Magnetographs .................................................................

iv

d spacempson,view-

Patrice

v

James KlimchukNaval Research LaboratoryWashington, D. C.

Paulett LiewerJet Propulsion LaboratoryPasadena, California

Richard MewaldtCalifornia Institute of TechnologyPasadena, California

Marcia NeugebauerJet Propulsion LaboratoryPasadena, California

Victor PizzoNOAA/Space Environment LaboratoryBoulder, Colorado

Dennis SockerNaval Research LaboratoryWashington, D. C.

Keith StrongLockheed Martin Advanced Technology CenterPalo Alto, California

NASA Study ScientistGeorge WithbroeNASA HeadquartersWashington, D. C.

Goddard Study ManagerJames WatzinNASA/Goddard Space Flight CenterGreenbelt, Maryland

Acknowledgements

We are grateful for the encouragement and support of so many of our colleagues in the solar anphysics community. In particular, we thank Bernie Jackson, Allen Gary, Don Reames, Barbara ThoJohn Davis, Jeff Hall, and Eric De Jong for illustrations, and Ralph McNutt and Tom Potemra for reing a draft of this report. Jim Watzin led the Goddard study team. Murrie Burgan, Eric Benfer, and Zurvalec helped in editing the manuscript and preparing it for publication.

Science Definition Team Members

ChairmanDavid RustThe Johns Hopkins UniversityApplied Physics LaboratoryLaurel, Maryland

NASA Study ScientistJoseph DavilaNASA/Goddard Space Flight CenterGreenbelt, Maryland

MembersVolker BothmerUniversity of KielInstitut für Reine und Angewandte KernphysikKiel, Germany

Lennard CulhaneMullard LaboratoryUniversity CollegeLondon, United Kingdom

Richard FisherNASA/Goddard Space Flight CenterGreenbelt, Maryland

John GoslingLos Alamos National LaboratoryLos Alamos, New Mexico

Lika GuhathakurtaNASA/Goddard Space Flight CenterGreenbelt, Maryland

Hugh HudsonInstitute of Space and Astronautical ScienceKanagawa, Japan

Michael KaiserNASA/Goddard Space Flight CenterGreenbelt, Maryland

h f

e

h

v

r

fc

t

dd--

os-el-

e-p-ally theor-E

tountheis- ac-d ofnc-r-ldlyis-gh

outiclein-ar-ces

resedEsrth,ceshiss-ci-also

rthraftory

1. Introduction

NASA’s Sun-Earth Connections program aimsimprove mankind’s understanding of the originssolar variability, how that variability transforms tinterplanetary medium, how eruptive events onSun impact geospace, and how they might aclimate and weather.

STEREO is the third of five Solar-Terrestrial Probcalled for in NASA’s Space Science Enterprise Stegic Plan to accomplish the goals of the Sun-EConnections program. The first of these missiothe Thermosphere-Ionosphere-Mesosphere Eneics and Dynamics mission (TIMED) is scheduledlaunch in 2000. The other missions are

• Solar-B, which will obtain high-resolutionimages of the solar magnetic field to determhow it emerges, evolves, and dissipates its enat the solar surface. These processes aredrivers of the solar activity we observe.

• STEREO, which will obtain simultaneous imagof the Sun from two spacecraft and build a thrdimensional (3-D) picture of coronal maejections (CMEs) and the complex structuaround them. STEREO will also study tpropagation of disturbances through theliosphere and their effects at Earth orbit.

• Magnetospheric Multiscale, which will providea network of in situ measurements of Earthmagnetosphere that can be combined to gi3-D image of magnetic substorms and otactivity in geospace.

• Global Electrodynamics, which will probeEarth’s upper atmosphere to determine hvariations in particle flux and solar electrmagnetic radiation affect it.

While these missions individually will doubtless pduce exciting discoveries about the complex SEarth system, together they are a formidable that will greatly improve our ability to prediweather in space, enhance our knowledge of sinfluences on climate change, and give us freshsight into the origins and future of life on EarAlthough the Sun is much quieter today than indistant past —it was once a rapidly rotating, stronmagnetic, violently active star with a massive ste

to ofethefect

estra-arthns,rget-for

ineergy the

see-ssrese

he

’se a

her

owo-

o-un–leettolar in-h.theglyllar

wind—it is still capable of violent explosions ansubstantial variations in radiative output. Understaning present-day solar activity will help mankind understand the history of the Sun’s climate and its psible influence on Earth’s evolution and the devopment of life.

It is through CMEs that solar activity is most forcfully felt at Earth. CMEs are the most energetic erutions on the Sun. They are responsible for essentiall of the largest solar proton events, and they areprimary cause of major geomagnetic storms. Unftunately, no one can predict reliably when a CMwill occur or what its effects will be.

In order to make progress in this area, we needfollow CME-generated disturbances from the Sto at least the orbit of Earth and we need to know state of the ambient solar wind in front of these dturbances. Coronal observations need to providecurate measurements of the pre-CME corona anCME timing, size, geometry, mass, speed (as a fution of height), and direction at a minimum. Infomation about the strength and polarity of the fieembedded within the CMEs would also be highdesirable. We need to track the evolution of the dturbances optically or by radio emissions throuthe interplanetary medium to 1 AU. The in situ ob-servations need to provide accurate information abthe state of the ambient wind and energetic partpopulations ahead of the CMEs while also determing the plasma, magnetic field, and energetic pticle characteristics of the interplanetary disturbanas they pass.

Unfortunately, the CMEs that most affect Earth aalso the least likely to be detected with ground-baor Earth-orbiting telescopes. To understand CMbetter and to forecast their arrival and effects at Eaa totally new perspective on CMEs and their sourin the solar atmosphere is needed. Achieving tperspective will require moving away from our cutomary lookout point. This report describes the sentific progress that can be achieved toward the goof the Sun-Earth Connections program with twspacecraft at 1 AU, one drifting well ahead of Eaand one well behind. Together the two spaceccomprise the Solar-TErrestrial Relations Observat(STEREO).

3

s

st

c w

w

w

u

n

aresrg-

of

ss

f

l

,

2. Scientific Objectives

The principal science objectives to be addressedthe STEREO mission are as follows:

• Understand the origin and consequencesCMEs

• Determine the processes that control CMevolution in the heliosphere by tracking CMEdriven disturbances from the Sun to Earth’s or

• Discover the mechanisms and sites of soenergetic particle acceleration

• Determine the 3-D structure and dynamicscoronal and interplanetary plasmas and magnfields

• Probe the solar dynamo through its effects the corona and heliosphere

Coronal Mass Ejections

A primary scientific motivation for studying CMEstems from their enormous and difficult-to-explaspatial scales, masses, speeds, and energies. Cappear to be the means by which the corona evothrough the solar cycle. They may be the meanremoving dynamo-generated magnetic flux from Sun. They appear thus to be a crucial link to Eafrom the solar dynamo. Further, the striking effeof CMEs on planetary magnetospheres, comets,cosmic rays extend the interest in mass ejections beyond the traditional realm of solar physics, emphasized in the Sun-Earth Connection RoadmFinally, there may be astrophysical analogues of mejections, perhaps in accretion disks and activelactic nuclei, that will be better understood when understand CMEs.

Explaining the sudden expulsion of a highly coducting plasma from the magnetized Sun presenmajor challenge to space physics. The spectacnature of these large mass ejections is illustrateFigure 1 by a time sequence of images obtained the Large Angle Spectroscopic Coronagra(LASCO) flown on the SOHO mission. The brighloop-like feature contains more than 1015 g of plasma.The energy required to lift the material off the Smay be as high as 4 3 1032 ergs.

CMEs are complex and poorly understood. Notheless, it is now evident that many ejections invo

4

by

of

E-

bit

lar

ofetic

on

inMEs

lves of

herthtsandell

asap.assga-e

n-ts aulard in

ithpht,

n

e-lve

the eruption of coronal and chromospheric plasmfrom a region pervaded by magnetic fields that ainitially closed and possibly twisted. What triggerthe eruption? How is the energy built up and ovewhat scale? Which, if any, of the several competinmodels of CME origins is the correct physical description of what happens on the Sun?

CME Onset

Here are some of the many models for the origins CMEs that have been proposed:

• Magnetic shear by surface motions, causing loof equilibrium in the corona

• Magnetic helicity charging from beneath thesurface, causing a kink instability in the corona

• Emerging magnetic flux, causing loss oequilibrium in a coronal arcade

• Magnetic helicity charging of the corona byflares, causing loss of equilibrium in a coronaarcade

• Thermally driven blast wave from a large flareblowing the corona open

• Buoyancy, due to a low-density cavity in thecorona

Figure 1. Four images from the LASCO coronagraphon SOHO, showing a CME on 7 April 1997.

h

M

?

enrelyy to

tlyenlarndp-p-eri-rlynr-esr-heres

gro-and

gf-

Anhtow

One should be able to distinguish among the moby careful examination of the structure of the pCME corona. CMEs frequently follow several daof “swelling” of a coronal helmet streamer. Themay be corresponding changes in the low coromagnetic shear (by differential rotation’s effect emerged coronal loops) or magnetic helicity chaing (loop twisting by subsurface flows). Thredimensional reconstructions by triangulation on conal features should reveal the key signatures of tprocesses and even allow us to specify the dentemperature, and magnetic fields of the pre-evstructures.

As the list of models suggests, several fundamequestions must be answered if we are to understhe physical causes of CME eruption:

• Are CMEs driven primarily by magnetic ononmagnetic forces?

• What is the geometry and magnetic topologyCMEs?

• What key coronal phenomena accompany Conset?

• What initiates CMEs?

• What is the role of magnetic reconnection?

• What is the role of evolving surface features

These questions cannot be satisfactorily addreswith single-vantage-point observations of the tycurrently available. The corona is optically thin, botin the emissions seen by X-ray and UV imagers

5

Figure 2. SOHO-EIT observations of a wave expandare successive differences of images in Fe XII 195 Å

delsre-ysrena:onrg-e-ro-esesity,ent

ntaltand

r

of

E

sedpehand

in the Thompson-scattered photospheric light seby coronagraphs. Line-of-sight integration effects aa major source of ambiguity and confusion. OnSTEREO can provide the observations necessarsort out the overlapping 3-D structures.

CME Geometry and Onset Signatures

The geometry of erupting CME structures is currenunknown. Several basic configurations have beproposed: a simple dipolar arcade, a quadrupomulti-arcade system, a half-emerged flux rope, aa suspended flux rope. The different physical proerties of these configurations predict different erution scenarios, as discussed later. Detailed numcal models of CMEs have been developed, but neaall of them have made the simplifying assumptiothat the field is two dimensional (e.g., an infinite acade). In reality, of course, the erupting structurhave finite extent, and 3-D effects must be impotant. With the range of view angles accessible to tSTEREO telescopes, CMEs and coronal structucan be reconstructed in three dimensions.

A CME frequently starts with a sinuous brighteninin the low corona and an outward movement of conal structures on many scales. Chromospheric coronal plasma is accelerated to 300 km s–1 or more.Observations with the Extreme-ultraviolet ImaginTelescope (EIT) on SOHO show that CMEs are oten accompanied by a wave front in the corona. example is shown in Figure 2. It was once thougthat such waves are triggered only by flares, but n

ing from a CME initiation site on 12 May 1997. Images.

ht

t

i

c

a.

i

u

p

e

t

-n

no

E-heeds-n,d.g-r-

asls.is is

dt.

toldsdwestser-

tos-erygs, arephveedf

herg-er-

the whole issue must be reexamined. While waves, variously called Moreton waves or EIT wavdo not appear to be “thermal blast waves,” they sto be intimately involved with CMEs. What is trelationship of the wave to the CME? Which is trigger? In order to resolve these questions, SREO should be designed to provide images wimuch higher cadence than SOHO does.

It is well established that pre-eruption magnetic fconfigurations contain ample free energy to expthe gravitational and kinetic energies of CMEs. Hoever, it is not so obvious that enough of this energreleased as the field is opened during the erupThere can actually be an increase in the magnetiergy locally, even though the global magnetic enemay decrease. This has led some to argue that Cmust be buoyancy driven, so it is important to be to gauge the size and density of coronal cavities

Many CMEs have a three-part structure that inclua dark cavity, a bright frontal loop, and an embded core (probably erupting prominence materThere are, however, many examples of CMEs pre-event streamers with no obvious cavity. Ifcavity exists, then buoyancy can be ruled out, CMEs must be magnetically driven. Whether or cavities are present in these structures is difficuknow due to the possibility of overlapping brigfeatures in the foreground and background. OSTEREO can provide the observations needeunambiguously establish whether cavities are versal and, therefore, whether CMEs are magncally or buoyancy driven. Determination of crucphysical parameters such as the density and sure of the cavity will require simultaneous obsvations of emissions that are sensitive to den(white-light coronagraph) and to density squa(emission-line imager).

Reconnection

In many CME models, magnetic reconnectionnecessary for the eruption to begin or to procThe physical role of reconnection varies from moto model, however, and it is possible threconnection plays no active role whatsoever. Indipolar arcade and suspended flux rope modelsstretched fields reconnect beneath the CME a

6

thees,eemeheTE-h a

eldlainw-y istion. en-rgyMEsble

desed-al).andnoandnotlt tohtnlyd toni-eti-

ialres-er-sityred

ised.

delat the, the the

same time that the CME is lifting off. Withoutreconnection, a full eruption is not possible. In contrast, the quadrupolar model involves reconnectiohigh in the corona above the erupting arcade, andreconnection is necessary at low altitudes.

These different scenarios can be tested with STREO observations. For example, reconnection in tdipole arcade and flux rope models produces closmagnetic loops under the CME that should be viible at the time of the eruption. If no loops are seethen the models must be either rejected or modifieExisting observations suggest that the erupting manetic fields in many events remain open to the suface for a considerable time after the eruption hbegun. This would seem to contradict the modeHowever, the interpretation of the observations open to debate, since the orientation of the arcadenot known. An end-on view should reveal closeloops if they exist, but a sideways view might noOnly STEREO can resolve this ambiguity.

Surface Evolution

It is widely believed that CMEs are a response changes in the surface magnetic fields. These fieare constantly evolving, either by the twisting anshearing of existing flux or by the emergence of neflux from below the photosphere. These processstress the overlying coronal field, and the field erupwhenever the stresses become too great. An undstanding of how surface fields evolve leading up eruption is vital if we are to understand why CMEoccur. It might also prove valuable for CME prediction, a long-term goal of the National Space WeathProgram. At present, it is extremely difficult to studsurface evolution in the days immediately proceedina CME. Surface features, especially magnetic fieldare best observed near disk center, whereas CMEsbest observed near the limb. A STEREO coronagrawould be able to detect CMEs that originate abosurface locations that are at solar disk center as viewfrom Earth. This would allow us to study the role osurface evolution and, in particular, to address trecent suggestion that CMEs are triggered by emeing flux. The needed vector magnetograms will bavailable from Solar-B and the National Solar Obsevatory SOLIS magnetographs.

m

k

The Heliosphere Between the Sun andEarth

The heliosphere extending from 30 RSun to 1 AU (215Rsun), i.e., from the edge of the widest-field LASCcoronagraph to Earth orbit, contains nearly 400 tithe volume of the currently imaged region closethe Sun. This volume has remained unexploredcept during the Helios mission 20 years ago. two Helios spacecraft carried solar wind analyzand low-resolution photometers that mapped thelar wind density distribution in the heliosphere aproved that CMEs can be detected well bey30 RSun, even well into the heliosphere (see Fig3). No such observations are available now.

In general, dense heliospheric structure followslocation of the heliospheric current sheet. Howethis structure evolves and is segmented, with the dsegments generally being associated with regionhigh solar activity. Thus, the heliosphere is populanot so much by a continuous dense “ballerina sof slow-speed solar wind forming a wave aroundSun but, instead, by a set of spikes (see Figure 4) athe skirt that vary continually in strength.

Oes

to, ex-Theers so-nd

ondure

thever,enses oftedirt”

thelong

Figure 4. Perspective view of the corotating solar windpattern for Carrington Rotation 1653. The ellipseindicates Earth’s orbit. Solar wind densities, indicatedby shades of gray, were derived by fitting aheliospheric model to Helios photometer observationsof Thompson-scattering from solar wind electrons.

Figure 3. Polar plot of a CME recorded by the Helios B visible light photometer. Signal at 90° elongationcorresponds to a CME passing overhead. The spacecraft was at 0.484 AU from the Sun at the time.

7

ey at

n-e isyderell thently

The quiet solar wind flows out nearly radially frothe Sun in corotating patterns that evolve slowly wtime. The distribution of matter in the heliosphernever entirely certain because of a lack of timelD information. Figure 5 shows what has beencomplished with Helios data analysis, but that imwas built up under the assumption that heliospheric density distribution was unchanged

mith is 3-ac-gehe for

a whole solar rotation. Determination of the instataneous distribution of matter in the heliospheran important goal for STEREO. Solar wind densitand velocity measured in situ at 1 AU can be relateto 3-D reconstructions from STEREO heliosphimagers and coronagraphs and traced almost away down to the solar surface. This is not currepossible.

8

Visibility of Earth-Directed CMEs

The quantities observed by coronagraphs arethe polarized brightness pB and the brightnessB, and they are related to the electron densitythrough a line-of-sight integral over the light-scattering electrons. The (Thompson) scatter-ing cross section is quite small and the densityof the corona is very low. As a consequence,the white-light corona is very faint, and becausethe incident light is polarized by the scattering,one has to consider carefully how coronal vis-ibility is affected by the angle between the Sun,the scattering electrons, and the observer. For aspherically symmetric coronal density distribu-tion, roughly half of the total scattered lightcomes from within about ±20° of the plane ofthe sky. Contributions of structures that are 60°in front of or behind the solar disk are a fewpercent. Because of this limb-viewing bias ofcoronagraphs, most CMEs observed to date couldnot be related uniquely to other observed kindsof solar activity. Using simple mathematical

models, we investigated the pB properties of atypical CME as viewed in projection in theplane of the sky and at various angular dis-tances away from the plane of the sky. Thesimulated CME extended from 2.5 RSun to 8.0RSun. It was 35° wide in longitude, and the shellwas 1 RSun thick (these numbers are represen-tative of CME observations from LASCO C2and C3). Figure 5a presents a view with theCME in the plane of the sky, while Figures 5band 5c show the views when the CME is ro-tated 30° and 60° out of the plane of the sky,respectively. The CME appears fainter andsmaller when rotated 30° from the plane ofthe sky. It appears as a halo when rotated to90°. But our simulation shows that correct in-terpretation of the halo in terms of physicalsize or angular extent is almost impossible.Also, images from a single vantage point donot provide any information on the longitudeof the CME. With two vantage points, one canascertain its longitude and compare it directlyto other solar disk observations.

Figure 5. Three views of a simulated coronal mass ejection. (a) CME in the plane of the sky on the westlimb of the Sun, (b) CME rotated 30° from the plane of the sky, and (c) CME rotated 60° from the plane ofthe sky.

a

t

o

l

e

oi

s

s

s

l

eh,e

reyle,ed

is

e-d

-

o

re

CMEs in the Heliosphere

CME propagation and CME effects at Earth depenot only on the character of the coronal event also on the state of the heliosphere. Even as a Cemerges from the corona, its motion is influencby surrounding streamer belt structures and coroholes. A fast CME propagating into a slow ambiewind compresses the wind in its path; the interpletary magnetic field is intensified, its orientatiochanges as it is compressed, and it drapes arothe advancing structure. If the speed of the CMEhigh enough relative to the ambient solar windforward shock forms ahead of it. When a CME ocurs near the boundary between fast and sstreams, part of it is accelerated and part of islowed down. In general, a CME can be acceleradecelerated, deflected, distorted, attenuated, or plified, depending on the details of its interactiwith the ambient medium. Many of these interactioare of special interest because they produce occasstrong enhancement of geomagnetic effects.

To understand and predict the effects of CMEsEarth, it will be necessary to map the inner hesphere and reveal fast and slow streams, interacregions, and the interplanetary magnetic field. Thmaps will be generated from numerical modelsthe solar wind based on the 3-D observations of cnal holes and streamers together with observatof the magnetic field at the surface of the Sun.Insitu data obtained at STEREO spacecraft will be uto correct the heliospheric maps.

Current maps of the heliospheric magnetic fields based on Carrington Rotation maps built up froline-of-sight magnetograms taken from ground-baobservatories. The fidelity of the resulting modelssuspect because the determining characteristicspartly global in nature. Consequently, some partthe input data for the models are more than 3 weout of date. We know from Yohkoh and SOHO oservations that even the quiet Sun during solar mmum changes on timescales measured in hoursdays rather than weeks.

Maps derived from STEREO data will allow testsa new generation of advanced models of interpletary propagation of solar disturbances. The inrequirements are relatively straightforward but a

9

ndbutMEednalntn-nund is

, ac-low isted,am-nnsional

onio-tionse

ofro-ons ed

aremed is are ofeksb-ini- and

ofan-putso

impossible to obtain with present capabilities. Thcrucial elements are the time and location of launcthe initial direction, the speed, the spatial extent, thmagnetic configuration, and the mass.

Some tentative tests of CME propagation models abeing undertaken with existing data, but the efficacof such tests is severely constrained. For exampcoronagraphs are sensitive only to the portion of thdisturbance lying close to the plane of the sky anthe relation of that special slice to the entire CME generally ill defined.

Current models show that CME interaction with thheliosphere is likely to be very complex and to produce confusing results. An illustration is presentein Figure 6, which shows two slices through a 3-Dhydrodynamic simulation of a CME interacting witha tilted-dipole background flow. Although the resolution of this exploratory calculation is crude andthe model structures highly idealized, it suffices tshow how an initially compact, spheroidal CME(characterized as a modest velocity and pressu

Figure 6. Simulation of a model CME pulseinteracting with a tilted-dipole ambient solar windflow structure. The panels represent slices through thesolution in the ecliptic plane (left) and the centralmeridian plane (right) 10 days after CME launch fromthe Sun. The computational domain runs inheliocentric distance from 0.14 to 5.0 AU, with whitesemicircular gridlines every AU. The model CME wasinjected into the solar wind at the equator, at the baseof the streamer belt. Note how the CME is compressedat the interface between the slow streamer belt flowand fast high-latitude flow above the equator, whereasthe CME is drawn out and accelerated by therarefaction flow ahead of it below the equator.

d

t

t

r-ck)mag.

ec-hein- ise

iseltn-nt.i-

e-n-

ng

pulse injected at the base of the slow flow alongmodel coronal streamer belt) is subsequently torted by variations in the flow structures it encouters. Such models predict certain patterns in interplanetary evolution of CMEs that are direcrelated to the surrounding background flow structuSTEREO observations will allow us to make a reistic assessment of the models.

Tracking Disturbances from the Sun to Earth

Interplanetary disturbances will be detected remonot only by heliosphere imagers but also by ratelescopes. Only two types of radio emission generated far from the Sun in interplanetary spaThese are (1) the quite common flare-associated III bursts, and (2) the much rarer type II bursts asciated with CME-driven shocks. Type III events acaused by energetic electrons traveling at speed0.2 to 0.4 times the speed of light, so they move frthe vicinity of the Sun to 1 AU in 20 to 30 minuteFor the type II events, typical speeds are 300–1km/s, so the travel time from the Sun to 1 AU

10

theis-n-thelyre.al-

elydioarece.

typeso-res ofoms.000is

several days. For both types of emission, the distubance (that is, the energetic electrons or the shogenerates radio emissions at the local electron plasfrequency or its second harmonic as it moves alon

With one spacecraft, parameters such as the eltron density at the emission sites and the path of tdisturbance through interplanetary space can be ferred but they are model dependent because itnot possible to determine exactly where along thmeasured line of sight the radio source lies. Whatusually done in this case is to rely on a global modof interplanetary electron density, taking the poinwhere the line of sight intersects the appropriate desity (thus, plasma frequency) to be the emission poiAn example of this type of source location determnation can be seen in the top part of Figure 7.

As illustrated by Figure 7, reliance on a single spaccraft does not solve one of the outstanding and fudamental problems involved with predicting theterrestrial impact of CMEs, namely, determinatioof the propagation speed of the correspondin

d

Figure 7. Radio tracking of a CME shock from the Sun to beyond Earth. The Wind/WAVES instrument wasable to measure the azimuth and elevation of the type II emission associated with the spectacular January1997 CME, but the actual location of the source could be found only by intersecting the lines of sight with theassumed position of the shock front. Since the speed of the shock front was not known until the shock passeby Earth, the sketches at the top could only be constructed after the fact.

rc

p

d

te

icegl

u

h

f

a

se

rlj

fwro

ve

te

sd/

r-utllP

e-e byalentt

es

ndsre

aresere

nd andre-he

el-earofat-leill

disturbance through interplanetary space. Forample, in the low corona, where CMEs are obsein visible light, their speeds and accelerations be measured. However, these measured speedgenerally quite different from the speeds of progation through the interplanetary medium and ctherefore, lead to predictions of the arrival timethe coronal disturbance at Earth that can be in eby a day or more. Similarly, single spacecraft rameasurements like those of Figure 7 of the shfront can lead to large errors because the elecdensity in the solar wind can vary over a wide ranRadio telescopes on the two STEREO spacecrafmake reliance on models unnecessary becausradio source location is simply the intersectionthe two measured lines of sight.

The radio direction-finding capabilities, together wthe wide separation between the STEREO spacewill permit the type II radio source, at a givfrequency, to be located by triangulation. A sintriangulated source position is sufficient to estabthe density scale and, therefore, determine the Cshock speed through the interplanetary mediOnce the shock speed and density scale are obtawe can readily predict, to within about 2 hours, wEarth will encounter the disturbance. By triangulat-ing the type II radio source at many times and quencies, the CME shock can be precisely tracthrough interplanetary space and the predictedrival time at Earth can be refined.

With the stereoscopic observations, trajectoriekilometric type III radio bursts will be constructand studied in a systematic way for the first timThe type III radio burst trajectory can be construcfrom measurements made at a number of diffefrequencies. Stereoscopic observations will alinterplanetary densities along the radio burst tratory and electron exciter speeds to be remotely msured and the average interplanetary magnetic topology to be mapped. STEREO observations also allow intrinsic properties of the radio souregion, such as the brightness temperature, the ssize, and the source effective beam width, to berived and studied in an unambiguous manner.

1

ex-vedans area-

an,ofrrorio

ocktronge. will theof

thraft,nle

ishMEm.ined,en

re-kedr-

ofde.

tedentowec-ea-ieldill

ceurce

de-

Particle Acceleration by CMEs

Solar energetic particle studies with STEREO hatwo main objectives:

• To understand how and where CMEs acceleracharged particles

• To develop tools for greatly improved forecastof large solar energetic particle (SEP) events anor to warn of their onset

More than 95% of the largest solar energetic paticle events are associated with CMEs, but only aboone-third of CMEs produce shocks, and not ashocks result in large events. In the largest SEevents, the particle fluxes spread out over 180° inlongitude. Since the particle flux at a given spaccraft depends on how well it is connected to thshock, the objectives above are best addressedobservations of particles and fields at severheliospheric longitudes, as emphasized in the recNASA report “Foundations of Solar Particle EvenRisk Management Strategies.”

SEP events can be classified into two different typ(see Figure 8): impulsive events, which have minorincreases in particle flux, are rich in He3, heavy ions,and electrons, and last from minutes to hours; agradual events, which are major proton flux increaseon timescales of hours to days. Impulsive events aassociated with solar flares, and gradual events associated with fast CMEs that drive. In both typeof events the propagation and properties of thcharged particles depend crucially on the structuof the coronal and interplanetary magnetic fields aplasmas, so particle flux measurements will allownew kind of remote sensing of the acceleration apropagation regions, especially when the measuments are combined with stereoscopic images of tcorona and heliosphere.

Compression, plasma turbulence, and shock acceration of particles are expected to be strongest nthe western face of fast CMEs, and the duration the particle events should depend on the time ththe shock and CME affect the field lines that connect the observer with the Sun. Thus, the particexperiments on the two STEREO spacecraft w

1

sbe

Figure 8. Impulsive and gradual solar energetic particle events. Particles accelerated in impulsive solar flareare generally detected only by observers that are well connected to the flare site. These events tend to relatively small. They are rich in He3, heavy elements, and electrons. In gradual events, the particles areaccelerated by a CME-associated shock and are observed over a wide range of longitudes.

c a

tcc

n

a

Hrt

r

se

on theused. Ins,n-ting

i-rightds,

eticfit-fds. in

llyflux

provide stereoscopic observations of the large-sstructure of CMEs, their effects on the ambientterplanetary medium, and their evolution in interpletary space. The effects of shocks and CMEs casensed from CME onset up to and beyond the when the CMEs or shocks pass over the spaceModel calculations can then determine the large-sstructure and the position of the CME’s center inheliosphere. To test the models, it is important toable to sense the different regions around CMEsshocks on a large scale. This probing can be coued throughout the STEREO mission.

The particle detectors aboard the STEREO spcraft trailing Earth will probe the corona and its dnamics near disk center (as seen from Earth). is where the most geoeffective CMEs start. Figushows an example of how particle measuremendifferent energies provide information from the oset of the CME at the Sun up to its arrival at EaEarly in the event, the particle intensity peaks at Menergies, followed by increases at keV energieto and beyond the time the shock passes the obs

12

alein-n-

n beimeraft.ale

the beandtin-

ce-y-eree 9s atn-th.eV uprver.

The MeV electrons and protons provide information the dynamics of the corona and CME close toSun, and the keV particle measurements can be to track the shock and CME on the way to Earthaddition to helping profile CME-associated particlethe trailing spacecraft will detect steady interplaetary particle streams, such as those in co-rotainteraction regions, before they pass Earth.

Magnetic Clouds

Many CMEs are associated with erupting promnences (called filaments when seen against the bsolar disk). They often appear to be twisted stranlike ropes (see Figure 10). Interplanetary magnclouds are also flux ropes, as determined from ting their in situ fields to flux rope models. Many, inot all, CMEs are associated with magnetic clouThis may be a vital clue to their possible originshelicity charged features in the corona.

Magnetic fields in astrophysical settings are usuafilamentary and tend to concentrate as “magnetic

CME on 7 April 1997. EIT observations

13

Figure 9. Collected SOHO/EIT/LASCO/COSTEP EUV, white-light, electron, and proton observations for the show the extent of the wave in the corona. The shock/CME passed the Earth on 10/11 April.

ms-ute

n-ntstic fitties

Oag-ds inent

ropes.” It is expected that magnetic helicity is con-served in such flux ropes once they leave the Sun.Magnetic helicity, Hm, is defined as

H dVm

v

= ⌠⌡

⋅A B ,(1)

where A is the magnetic vector potential and B is themagnetic field vector. With suitable specification ofgauge and boundary conditions, Hm can be specifiedin practical terms. For example, the magnetic helicityof a twisted flux rope is TΦ2, where T is the totaltwist in radians and Φ is the magnetic flux in therope. The STEREO chromospheric and low coronaimagers should have enough spatial resolving powerto allow determination of T in eruptive prominences,

such as the one shown in Figure 10. Magnetograwill allow reasonable estimates of the flux. The imagers should also be able to test models that attribCME onset to a helical kink instability.

Helicity-conserving flux-rope models have been costructed under the assumption that twisted filameand their surrounding loops become the magneclouds seen in interplanetary space. The modelsthe average thermodynamic and magnetic properof magnetic clouds. If the helicity of eruptiveprominences can be determined with STEREobservations, one should be able to predict the mnetic field structure and strength of magnetic clouat 1 AU. This would represent a major advanceestimating the geomagnetic effects of the most potCMEs.

Figure 10. The Sun in He 304 Å emission as recorded on 26 August 1997 by theEIT telescope on SOHO. The prominence on the northwest limb demonstratesthe twisted structure characteristic of eruptive events.

14

ol

y

t

aeseib

mc

a

h or

n

or-

in-

en-s-ricof

utITst

earnns

th in

cta.,

des,etseet-theensce

ste

larn-

medes in

re-r-

thetic

Escape of Magnetic Flux and the SolarDynamo

Two mechanisms can facilitate net flux escape frthe Sun: helicity charging to push open the fiewith reconnection to close them off. Measuremeof the solar wind magnetic fields at 1 AU appearshow that 1024 Mx of azimuthal flux is ejected bthe Sun in each solar cycle. This rate is the samthe expected rate of toroidal flux generation by solar dynamo. This measured flux ejection ratealso consistent with estimates of flux escapingCMEs and prominence eruptions and with the parent rate of flux emergence at the solar surfacmeasured by ground-based magnetographappears that escaping toroids (idealized in Figuras resulting from the net effect of many CMEs whelical fields) remove at least 20%, and possi100%, of the emerging flux in each cycle. Flux es-cape can be checked with STEREO data, and it prove to be the key to understanding the cybehavior of the Sun.

Coronal Magnetic Fields

“Moreton waves,” once also known as flare blwaves, were discovered by Gale Moreton, an obseat the Lockheed Solar Observatory in the 1960s. Twaves propagate horizontally across the disk ofSun at velocities up to 1000 km/s. They are fast-mmagnetosonic waves associated with large flaMoreton waves are visible in the wings of the Ha line,and until the SOHO mission, they were known o

Figure 11. The escape of CME-associated helical fieldfrom the Sun, idealized as northern and southerntoroids. The net effect of the ejection of many CMEswith helical fields may be the removal of most of thedynamo-generated flux each solar cycle. The direction(arrows) of all the fields in the figure reverses at aboutthe time of solar cycle maximum.

1

mdsnts to

e ashe is inp-, as. It 11thly

aylic

strveresethedees.

ly

by their effect on the chromosphere and by their crelation with type II radio bursts.

As Figure 12 shows, there are distinctive wavescoronal emission line images from the EIT instrument on SOHO. They sweep across almost the tire corona, and it is probably true but yet to be etablished beyond doubt, that the chromosphemanifestation, the Moreton wave, is just the skirt the wave front (Figure 13). The waves in the EITimages (“EIT waves”) have been seen throughothe low corona. Considering that the cadence of Eimages is only four per hour (at most) and that moof the SOHO mission to date has taken place nsolar minimum, it is surprising that so many Moretowaves have been detected. The EIT observatiomake it clear that two coronal emission-line imagersoperating at higher cadence than is possible wiSOHO would be able to specify the wave frontsthree dimensions.

The motion and distortions of the wave fronts reflethe conditions for wave propagation in the coronAccording to Uchida’s theory of Moreton wavespropagation of the slow mode and the Alfvén mowave packets is confined to local magnetic field linebut the propagation of the fast mode wave packcan reveal the distribution of the field strength. Thfield strength distribution in the corona can binferred by entering a field distribution and compuing the paths of the wave packets, then adjusting field distribution until there is agreement with thobserved wave fronts. Thus, STEREO observatioof the wave fronts can achieve a dramatic advanin measuring the coronal magnetic field. This “seis-mology of the corona” may finally achieve what habeen impossible with older approaches: a complespecification of coronal magnetic field strength.

Coronal Loop HeatingAt a temperature of several million degrees, the socorona is 3 orders of magnitude hotter than the uderlying photosphere. The reason for these extreconditions has challenged solar physicists for decaand remains one of the great unsolved problemsspace science. What is the physical mechanismsponsible for heating the corona? A number of inteesting ideas have been proposed, including dissipation of electric currents in stressed magne

s

5

of

Figure 12. An EIT image of Fe XII 195 Å emission in the corona on 24 September 1997 showing the loops a major flare and the EIT wave. Both the wave and the flare were associated with a CME.

m

iu

tedf dis-tialng.eat-n ofrop-

fields and the dissipation of magnetohydrodynawaves generated in the photosphere, but none ofhas been demonstrated to be correct.

The solar corona is a highly structured mediuCoronal loops, which trace closed magnetic flines, are the primary structural elements. Althomagnetic fields fill the entire coronal volume, on

1

icthem

m.eldghly

those flux tubes that are filled with strongly heaplasma appear as bright loops. The existence otinct loops is, therefore, a direct indication of spainhomogeneities in the rate of coronal heatiClearly, if we are going to understand coronal hing, we must understand the nature and origicoronal loops. Why do they exist, what are their perties, and how do they evolve?

6

ntica.

Figure 13. Wave fronts at various times after a CME event on 25 May 1997. Distortions of the wave froreflect variations in fast-mode wave velocity along ray paths from the site of the eruption. Stereoscopobservations of such wave fronts can be used to derive the distribution of magnetic field strength in the coron

grbth

ticn-s.psarecher- the

Loop Cross Sections

One of the basic properties of coronal loops is thcross section. The shape of the cross section istermined by the spatial distribution of the enerrelease and, to a lesser extent, by the transpothis energy within the loop. It thus provides valuainformation about the spatial dependence of heating process. If energy is released within t

17

eir de-yt oflehein

layers, as might be expected for magnereconnection, for example, loops should be ribbolike structures with highly noncircular cross sectionIf energy is instead released axisymmetrically, looshould be more like the tubular structures that commonly assumed. It is impossible to know whiof these possibilities is correct in single-view obsvations, where loops are seen as projections onto

u

te

er

vte r

r t

ht

tn

eys

hl

t ofcon-theesedyint,are of

sp

lara isrees thatre

astalomayinde-w isder- ac-di-ind

s ofndetor-mo-elld to

in-ld.in-ndis-ofouto-aed

flat plane of the sky. Their depth is completely known. Only STEREO can provide the necessinformation to determine the all-important 3-D struture of coronal loops.

Scaling Laws

Scaling laws are very useful for studying the heaand equilibrium properties of coronal loops. Thlaws describe relationships among loop-averaquantities. One well-known scaling law relates product of the average temperature and pressurloop to the loop’s length. Another relates the avepressure and length to the average heating ratecomparing theoretically and observationally deriscaling laws, it is possible to determine if the heais steady and how it differs among loops of differsizes. This information provides vital clues aboutnature of the physical mechanisms involved. Curprogress is hampered, however, by an inability to maccurate measurements of the fundamental looprameters. Lengths can only be approximated becthere is little information about the extent of the strtures along the line of sight. Densities and pressare also highly uncertain, since they are derived femission measures with an assumption about theof-sight thickness of the emitting plasma. The sition will improve dramatically with the two-viewobservations from STEREO.

Axial Gradients

Many coronal loops observed in soft X-rays Yohkoh appear to have far more variation along taxes than is predicted by theory. If this observais correct, it has major implications for both the heing and transport of energy within the loops. Idifficult to know, however, whether the variatioare real or merely a consequence of misundersprojection effects. A bright loop section could beindication of locally enhanced densities and temptures or it could be a result of the loop geometrbend that increases the line-of-sight thickneSTEREO will resolve this ambiguity.

Loop–Loop Interactions

Both Yohkoh and ground-based coronagraphs observed what have come to be known as “loop–interactions.” These are transient events in whnearby loops concurrently brighten, with the grea

1

n-aryc-

ingse

gedthe in aage. Byedingnt

theentake pa-auseuc-uresomline-ua-

byeirionat- iss

toodanra- (as).

aveoopichtest

enhancement occurring at the presumed poincontact. It has been suggested that magnetic renection at the current sheet interface between loops is responsible for the energy release, so thevents may provide an excellent opportunity to stuthe details of the reconnection process. At this pohowever, we cannot be certain that the loops actually in contact, much less sort out the detailsthe overlapping structures. STEREO observationwill provide a definitive resolution of the loop–loointeraction issue.

Solar Wind Origins

Two of the most important current questions in soand heliospheric physics are how the solar coronheated to temperatures in excess of a million degand how the solar wind is accelerated to speedsrange from approximately 300 to 800 km/s. Theare at least three different types of solar wind: fwind from coronal holes, slow wind from coronstreamers, and transient wind of any speed frCMEs. The slow wind from coronal streamers malso be essentially transient in nature. Solar wmight also originate on open field lines in other rgions, such as the quiet, background corona. Hothe wind accelerated? We cannot be sure we unstand the processes responsible for solar windceleration until the theoretical models match contions both in the solar corona and in the solar wnear 1 AU for each type of wind.

To determine the conditions at the source regionthe solar wind, we rely on white-light images aspectral information in the UV, extreme ultraviol(EUV), and X-rays. These emissions contain infmation about the densities, temperatures, wave tions, and bulk flows of several ion species as was the electron density. Because the plasma is tiethe magnetic field lines, imaging also provides formation on the geometry of the magnetic fieCurrent models try to incorporate some of this formation, but the problem is that the images aspectra are all obtained by integration of the emsion of the optically thin plasma along the line sight. This leads to fundamental ambiguities abthe actual 3-D structure. What is needed to make slar wind acceleration models more realistic is model of the 3-D structure of the streamlines bason observations, rather than on assumptions.

8

vg

O

-th

e

l

n

Bnyealveestly

ue

ylyc-m-it,ry

licic

un.entderThey,d,

eveereb-

n-r

vetheestuf-er

esry.nt

c

The STEREO mission will play an important roleimproving our understanding of the accelerationthe solar wind and in testing models deriving frothat understanding. Stereo images of the coronahelp remove the ambiguities arising from obsertions along a single line of sight. When the anbetween the two STEREO spacecraft approac90°, it will be possible to obtain both white-lighcoronal data and interplanetary data on the sastreamline; currently, the in situ measurements musbe combined with plane-of-sky observations obtainat a 90° separation in longitude, so that the corretion between white-light coronal observations asolar wind measurements can be done only on atistical basis.

Solar-B Collaboration

The approved Japanese Solar-B mission (Tabledue for launch in 2004, will overlap with STEREand can provide sophisticated Earth-perspective ctext observations. The STEREO and Solar-B dsets will be highly complementary since Solaremphasizes high-resolution observations of photosphere, transition region, and low corona, wSTEREO provides an extended view of the coroand heliosphere. The Solar-B data will include vtor magnetograms at 0.1-arcsec pixel size as wesensitive EUV spectroheliography and whole-SX-ray imaging. The Solar-B vector magnetograwill provide the best possible sensitivity for studing the underlying magnetic fields and their evotion before, during, and after a CME-launchininstability, and it will be operating when STEREspacecraft are optimally positioned for recordithose same CMEs.

Table 1. Solar-B key parameters.

Launch date 2004Orbit Sun-synchronous Earth orbitInstruments White-light telescope (0.2-arcse

resolution)– filter photometry– spectroscopy (vector B)

UV stigmatic slit spectrographX-ray imager

1

in ofmwilla-le

hestmetedla-nd sta-

1),

on-ataBheilenac-

ll asunphy-u-g

Og

Similarly, STEREO will add a great deal to Solar-science by helping to define the 3-D structure of aparticular target of observation. Solar-B’s scienctheme is a “systems approach” to the physiccoupling of the photosphere and corona. To achiethis goal in a completely satisfactory manner requiradditional knowledge, because the corona is mosoptically thin and one cannot accurately infer the trgeometry from a single set of observations.

Collateral Research

The STEREO platforms offer opportunities for manunique kinds of observation in areas not directrelated to the solar activity that affects Earth. We reommend that such observations should be accomodated to the extent that resources permproviding that this does not compromise the primamission objectives.

Helioseismology

Helioseismology has captured scientific and pubattention because it actually provides views, graphas well as quantitative, of the structure inside the SMoreover, these views challenge some of our presconcepts of the physics of the Sun and of the broauniverse (e.g., the abundances of the elements). current methods of observation in helioseismologled by SOHO in space and GONG on the grounhave certain limitations. These include access to thlowest-frequency p-modes of low degree, which githe best information about the deepest interior, whwe hope to resolve the puzzling solar neutrino prolem. At low frequencies one finds high values of Qin the p-modes, with obvious benefits for mode idetification and application to learning about solastructure.

Thus far, the existing helioseismic observations hanot discovered g-modes, which propagate only in radiative core of the Sun and, thus, would be the bguide to its structure. The present tools may not sfice to show the g-modes, and STEREO might offthe key help needed.

The low-degree p-modes, and possibly the g-modas well, can be detected with simple photometFrom a single geometrical perspective, the differe

9

t

r

r

a

h

l

t

r

i

i

re-ily

r-eti-ts.

Eon-ughs ofre-he

snp-

leis-

hesryully

is

fic-n-mendyfortingn

es.

anthe

e

modes of oscillation present foreshortened projtions, which lead to crosstalk between modStereoscopic observation can reduce both crosstalk and noise in the measurement. A pair ofwell-separated instruments would allow removalthe incoherent convective motions (which represnoise) and isolation of the almost strict coherencethe seismic effects. Instruments on SOHO (MDIVIRGO, and GOLF) provide some heritage fhelioseismic observations from deep space. In STEREO mission, these measurements would pably be done photometrically with irradiance instrments. If a more complex helioseismology instment could be accommodated, velocity measuments could be made. These are superior to irradimeasurements for low-degree, low-frequency mod

Solar Irradiance

The solar irradiance variability discovered in t1980s provides a substantial challenge for solar pics. What are the mechanisms that produce such goutput fluctuations in what was once called “the soconstant?” We had long known about the variabiof spectral emissions such as the 10-cm or sX-ray fluxes, but these relate to the corona, nothe photosphere of the star itself, and represrelatively minor energy fluctuations.

Several known mechanisms exist to explain diffent components of the observed variability of tosolar irradiance but there remains a substantialexplained variance, which may include secular telinking one solar cycle with the next. One methoddisentangling these different components of variaity is to construct models that relate independobservations (for example, sunspot area) to the diance variations. The lack of data from abovearound the solar limbs reduces the accuracy of thmodels and, thus, limits the study of the mechanisof solar variability. Broad-band irradiance measuments can be carried out with minimal resource mands, and such measurements can also be useseismic observations if specialized instruments not available.

Stereoscopic observations represent an impornext step in solar irradiance measurements. Insments from SOHO and other deep-space missprovide some technical heritage. Total irradiance

20

ec-es.his

ofent of

,ortheob-u-u-re-ncees.

heys-reatlarityoft toent

er-talun-ms ofbil-entrra- oresemsre-de-d forare

tanttru-ons

measurements (and broad-band spectral measuments) require only minor resources and could easbe accommodated on a STEREO platform.

X-ray and Gamma-ray Bursts

Hard X-rays, γ-rays, and radio bursts may help to chaacterize coronal structure and may represent energcally important components of major eruptive evenSolar flares and probably also developments at CMonsets accelerate high-energy particles. CMEs ctinue to accelerate particles as they propagate throthe corona and the heliosphere. Thus, observationthe byproducts of these high-energy particles repsent an important channel of information about toverall process involved.

Solar nonthermal radiation, including hard X-rayand γ-rays (>10 keV) may have anisotropic emissiobecause they are nonthermal in origin. This proerty of the radiation provides a relatively simpremote-sensing tool that can help study particle dtribution functions near the acceleration site. Tdirectivity of hard X-rays from bremsstrahlung iclosely related to the polarization, which is a vedifficult measurement that has never been successfcarried off. Effectively, then, the only way to observethe directivity of the hard X-ray bremsstrahlung by stereoscopic viewing.

Precise timing of γ-ray fluxes also provides some othe best position information for the enigmat(nonsolar) γ-ray bursters, and this position information improves with the baseline separation for triagulation. Other techniques have recently becoavailable for burster source localization, however, ait is not clear that this is a high-priority item anlonger. There is considerable technical heritage small high-energy instruments in deep space, starwith the Vela program and currently on Ulysses. Aeffective hard X-ray and γ-ray spectrophotometer fora STEREO platform would require modest resourc

Faint Objects

Unique studies of faint sources in the sky other thheliospheric plasmas can be undertaken with STEREO coronagraphs and heliosphere imager:

• Zodiacal light. The imagers can help determinthe dust distribution in the inner heliosphere.

o

as

r

pa-ndf

of

ita-ed

es,a-g

andnd

es.s a ofntal

iss-nene

stheandthe. If

ea-in by)int

oon

• Asteroids. It is estimated that a heliospheimager will discover between 10 and 100 asteroper year with radii greater than 12 m.

• Comets. Images of comets and of the distributiof dust down to the level of the zodiacal clobrightness will provide fundamental informatioabout the dust replenishment of the zodiacal clo

• Stars. Stellar light curves with ~0.1% photometric precision and 1-day time resolution cbe obtained for the 103 brightest stars.

3. Making the Best Use of STEREOImages

Determining the 3-D Structure andDynamics of the Corona

The coronal plasma radiates strongly in X-rays EUV. These emissions are sensitive to both pladensity and temperature, making them a powediagnostic of the coronal plasma. Moreover, becathe plasma follows the magnetic field lines in tlow corona, imaging in X-rays and EUV directshows the structure of the magnetic field lines whot plasma. Thus, stereoscopic observations ofcorona in X-rays and EUV can be used to determthe 3-D structure and dynamics of the coronal plasand magnetic fields.

Resolving Line-of-Sight Ambiguities withStereo Observations

Coronal loops are not in general isolated. Other sttures often lie along the line of sight, either in froof or behind the structure of interest, causing a “baground” problem. Many Yohkoh images show looapparently interacting with adjacent loops. Withoa stereo view, however, it is not possible to resothe ambiguity of whether the brightenings of tloops are a result of summing intensities along line of sight or if the loops physically interact. some eruptive event scenarios, the energy releatriggered by the interaction of neighboring flux sytems, but a close neighbor in a 2-D view mayquite distant when the third dimension is consideStereoscopic observations of the X-ray/EUV corocan resolve ambiguities in the interpretation changes in the coronal structure.

2

reids

nudnud.

-an

ndma

rfulusehelyith theinema

ruc-ntck-psutlvehetheInse iss-beed.naof

The effective depth resolution depends on the seration between STEREO #1 and STEREO #2 athe 2-D resolution of the imager. In the first year othe mission, the separation will dwell at 50°, so ifthe imager has 2000-km resolution in the plane the sky, it will have 2600-km depth resolution.

Use of Triangulation to Determine the 3-DCoordinates of Coronal Features

It has not been generally appreciated that quanttive information on the 3-D magnetic fields can bfound by using triangulation on coronal loops another features. Using classic surveying techniquthe coordinates in three dimensions of a coronal feture can be determined from only two views as lonas (a) one knows the stereo separation angle spacecraft distances and directions to the Sun a(b) one can recognize the feature in both imagCondition (b) can be a serious limitation in studieof diffuse features or features in crowded fields, inlarge active region, for example. Over the coursea year of observations, however, the Sun will presemany opportunities to study a wide range of coronfeatures under near-ideal conditions.

The triangulation technique for a simple case shown schematically in Figure 14, where it is asumed that both views are from the equatorial plaof the Sun. In this figure, the coordinates in the plaof the sky of the two views with stereo angle α arerelated by the simple rotational transform

T( )cos sin

–sin cos .aa aa a=

00

0 0 0 (2)

The triangulation calculation from a pair of pointin the stereo images can be done by determining coordinate transformation between the telescopes solar coordinate systems so that the rays from points in the images can be traced back to the Sunthere were no errors, the two points on the same fture in two images would map to a single point solar coordinates. However, errors are introducedthe manual tiepointing (joint feature identificationitself. Therefore, what is actually computed is the poof closest approach (in solar coordinates) of the twrays traced back toward the Sun from the points

1

eirar-tic-ed.e,dy.ti-n-s.i-e-

e

sll

-erelerong-e

the image pair. The location of the feature is then tato be midway between the rays at closest approa

The triangulation technique was tested on simulaSTEREO observations of well-defined loops. A ste-reo image pair was created by viewing the loops ftwo angles separated by 15° (see Figure 15). Alsoshown in Figure 15 are the (x,y,z) location of points(crosses) determined by triangulation from the stepair plotted over the input test loops (solid curveThe agreement between the input loops and thferred loops is excellent. When the same featurebe located in a time sequence of stereo pairs, onalso determine the 3-D velocity of the feature andthis way, obtain information on loop expansion ra

Automatic Feature Tracking

We recommend that studies be undertaken of sanalysis techniques used in other fields, such as E

x=y’-ycosαsinα

α

x

z

y

x’

y’

x=x’cosα+y’sinα

y=y’cosα-x’sinα

z=z’

Coordinates of two views related by simplerotational transform

Given y,y’ , Solve for x,x’

Coronal loop viewed from two angles separated by α

Figure 14. Determination of 3-D loop geometry fromtwo views via triangulation. Since both views of theloop are from the solar equatorial plane, thecoordinates are related by a simple rotationaltransform that can be inverted to give the 3-D solarcoordinates of the loop as shown.

2

kench.

ted

rom

reos).

e in- cane can, in

tes.

tereoarth

Figure 15. Test of determination of 3-D loopgeometry: known loops. The test stereo image pair wascreated by viewing the known loops from two anglesseparated by 15°. The (x,y,z) location of points (crosses)determined by triangulation from the stereo pair areplotted over the known test loops (solid curves). Theagreement is excellent.

and planetary surface imaging. Several fields havelong had the benefit of stereo data, and some of thdeveloped analysis techniques can probably be cried over to space physics. One example is automafeature tracking in which patterns within many sections, or “patches,” in one of the images are searchfor and identified automatically in the other imageIn contrast to the manual method described abovthe relative offsets of matching points are computewith cross correlations, usually to subpixel accuracThen, using ray intersection techniques, a sophiscated algorithm determines the coordinates for cojugate points in the two images in three dimensionWhile this is a standard technique in producing digtal terrain models, much development remains bfore we will understand its full potential and limita-tions in interpreting the optically thin features of thcorona.

Use of Magnetic Field Models in Conjunctionwith X-ray and EUV Observations

The use of magnetic field models with magnetogramsupplying the photospheric boundary conditions wigreatly enhance the information obtained fromSTEREO X-ray and EUV observations. Simultaneous stereo observations will allow a much bettidentification between features in the magnetic modand features in the observations. Loops and othfeatures that have been determined by triangulatican also be compared to features in the 3-D manetic field model. If a correspondence between th

2

a

o

fao

g-n-ilt.d-

hessem-

-

Dn-gl-7o-

a

model and observed features can be made, the netic field model can be tested.

Once a magnetic field model has been validated, lines from the model form a skeleton to which emting plasma can be attached to create a 3-D modthe corona. For whatever plasma model is chothe integrated line-of-sight emission calculated frthe 3-D coronal model (field plus plasma) must agwith the X-ray and EUV observations from boviewpoints. Figure 16 shows results from such D corona model using an iterative technique totermine the spatial distribution of emissivity. In tfigure, both the original Yohkoh/SXT view of a loocomplex and an image rendered from the 3-D mare shown. The model is viewed from the same aas the original SXT image.

The comparison in Figure 16 shows clearly thatmagnetic models need to be improved and that vemagnetic field measurements will probably needed to achieve convincing representations oservations. These enhancements will be availfrom Solar-B and the National Solar ObservatSOLIS magnetographs.

.he

Figure 16. Yohkoh/SXT image and rendered image from 3-D magnetic-field-based model of an active regionThe model assumes a potential magnetic field in the region. The 3-D model was rendered into the image on tright by computing the integrated line-of-sight emissivity.

2

mag-

fieldit-el ofsen,omreeth 3-de-hepdel

ngle

thectorbe ob-blery

From a time series of STEREO images and manetic models, a complete 4-D (three spatial dimesion plus time) model of coronal features can be buIn this way, the STEREO observations can be useto determine the magnetic field evolution that accompanies solar eruptions.

Magnetic-Field-Constrained TomographicReconstruction of the Corona

Tomography can be used to directly determine t3-D structure of the optically thin corona if one hamany viewing angles. STEREO will provide imagefrom only two angles. However, it is possible to maka tomographic-like reconstruction of the corona froonly two views by assuming a magnetic field configuration a priori. In this approach, the spatial distribution of coronal emissivity is determined byconstraining the stereo reconstruction with a 3-magnetic field model. The technique is a modificatioof the multiplicative algebraic reconstruction technique. In it, the constraint is applied by assuminthat emitting plasma only exists within a loose voume defined by the magnetic field model. Figure 1illustrates the technique and shows results of a tomgraphic reconstruction both with and without

3

r

-

e

--

-dst

of

ek-

d-

etos

b-a-

/ts

2

magnetic field constraint. The top panel of Figureshows the original test loops on the left and the mnetic field constraint to be applied on the right (viewfrom a second angle). The middle panel shows

Figure 17. Magnetic-field-constrained tomographicreconstruction. (Top left) Original simulated X-rayloops. (Top right) The envelope is the magnetic fieldconstraint applied. The view here is orthogonal to thaton the left. (Middle left and right) Two views of thetomographic reconstruction of the simulated loopsfrom a simulated pair of stereo X-ray images (28˚separation angle) with no magnetic field constraint;the white arrow head points to the more badly smearedloop. (Bottom left and right) Reconstruction from thesame image pair but with the added constraint thatthe loops are within the loose magnetic envelope showin the top right frame. The magnetic-field-constrainedtomographic reconstruction has reproduced theoriginal loops with little smearing (considerably lessthan the range of the envelope), illustrating theimportance of using a priori knowledge of the magneticfield.

4

17ag-edtwo

Figure 18. Simulated stereo pair of soft X-ray imagesof the corona. The Yohkoh/SXT image on the left wastaken on 27 April 1992 at 23:16 and the other about 6hours earlier. Some features can be seen in both imageswhile others are unrecognizable due to temporalchanges. These images may be viewed as a stereo paiby relaxing your eye focus or using a stereo viewer.

views of the result from a tomographic reconstruction of the test loops obtained without applying themagnetic field constraint. The bottom panel shows thresult from the tomographic reconstruction with themagnetic field constraint applied. The magnetic-field-constrained tomographic reconstruction has reproduced the original loops with very little smearing compared to the unconstrained reconstruction.

Visual Evaluation of Stereo Images

Human beings are equipped with an exquisite computer that quickly evaluates stereo image pairs andevelops an intense image in three dimensions. Juviewing stereo image pairs and time sequences stereo pairs will provide valuable insights on thestructure and dynamics of the phenomena we seto understand. Examination of image pairs with stereo viewers may be enough to eliminate some moels. For example, models of CME initiation involv-ing buoyancy require that there be a cavity, but thabsence of a cavity in a single image may be due a line-of-sight effect. However, if stereo observationshow some CME initiations with no cavity, buoy-ancy models can be eliminated.

To gain an impression of what can be gained fromdirect examination of stereo images, we have sustituted sequential images for true angular separtion of viewpoints. Figure 18 shows a simulated softX-ray stereo image pair created from two YokhohSXT images taken 6 hours apart. Solar rotation shif

n

d

i

l

o

o

t

h

S

a

s

heres.st toer,O

isbutity-ceehe

uld-

ndbynts

le.

it isic

the effective viewpoint 13° per day. The simulateangular separation in Figure 18 is then 3.3°, so theviewer has a leverage on 3-D structures on the similar to that achieved by examining somethabout 130 cm away.

Attempts to use such rotational synthesis to bui3-D picture of the active corona are defeated by constantly changing active regions. Even larger-scaleand longer-lived structures such as polar streamand quiet-Sun arcades are impossible to deconvbecause of slow evolution in brightness, shape,size. Only simultaneous images can give an acrate impression of coronal structure.

On the back cover of this report is an anaglyph image in relief) constructed from two EIT imageTo obtain the 3-D effect, the reader should viewwith a red filter over the left eye and a blue one othe right eye. Examination of the back cover imabrings out the dark veins in the corona, and one gan especially clear impression of the extent of a cnal-emission-absorbing prominence that is nearnorthwest limb. Figure 18 suggests that the pCME coronal arcade in the upper right quadranthe Sun is extraordinarily high—higher than aother feature. Whether any physical insight cangained from such simulated stereo observationsing solar rotation depends on the features being sbut it is clear that substantial insight can be gainefrom visual examination of the true stereo pairs tSTEREO telescopes will produce.

4. Space WeatherBesides the scientific reasons for studying the and heliosphere, there is a practical reason. Sactivity influences our lives. In our era of heavy sputilization, many more solar-terrestrial-related prolems occur than are commonly publicized or admted, especially in telecommunications and defesatellites. Enterprises known to be affected are

Cellular telephone serviceWeather satellite operationFusion and carbon dating experimentsGlobal Positioning System (GPS)Ozone measurement programCommercial airlines

2

Sunng

d athe

ersolveandcu-

(ans. it

vergeainsro-

thest-

t ofny be us-atic,dat

unolarceb-it-

nse

Commercial TV relaysCommunication satellite systemsSatellite reconnaissance and remote-sensing

systemsGeophysical exploration and pipeline operationSubmarine detectionPower distributionLong-line telephone systemsManned space programInterplanetary satellite experimentsVLF navigation systems (OMEGA, loran, etc.)Over-the-horizon radarSolar-terrestrial research and applications

satellitesSatellite orbit predictionBalloon and rocket experimentsIonospheric rocket experimentsShort-wave radio propagation

NASA should not ignore the needs of space weatusers. Of course, STEREO will greatly acceleratthe development of reliable forecast techniqueSTEREO data can be helpful almost from the firday of the mission. However, if STEREO data arebe used operationally in forecasting space weaththey must be available in real time and the STEREspacecraft must be monitored continuously. Thisoutside the scope of a NASA research mission, there could be a clear delineation of responsibilbetween NOAA and NASA, with NASA being responsible for collecting and transmitting the sciendata and NOAA being responsible for real-timtracking and forecasting, much as is done with treal-time solar wind data from ACE.

Although it would be impractical to transmit full-upSTEREO science data continuously, STEREO cowarn Earth of either coronal or interplanetary conditions indicative of impending disturbances. A net-work of modest antennas could detect the alert aeven trigger real-time tracking of the spacecraft the Deep Space Net to gain additional informatioand possibly lead to continuous monitoring of evenduring critical manned space activities, for examp

Solar Energetic Particles (SEPs)

In this era when man is always present in space, vital to improve our ability to forecast the energet

5

k

t

rt

e

n

ns

tedar-e

d-s ines

fcee,h

proton events that can present a radiation risastronauts. In the immediate future, there is a nfor warnings and short-term forecasts that mighused in scheduling astronaut extravehicular activ

The energetic particle flux at Earth depends ccally on how Earth is connected to the accelerasite at the shock. The degree of connection change quickly as the shock moves outward. Fig19 shows typical time-intensity profiles for observviewing a large CME-driven shock from threlongitudes. The observer seeing a western evewell connected early and sees a rapid rise andcline, while the observer seeing an eastern evepoorly connected until after the local shock pas

26

toeed beity.

iti-ioncanurerset is

de-t ises.

This represents an example where the well-connecobserver can provide several days’ warning of a pticle event that will eventually affect observers to thwest.

In the long term, as NASA considers mannemissions to the Moon and/or Mars, it will be important to forecast SEP events as much as 2 weekadvance. Such forecasts will require new approachwith improved accuracy. The observations to becarried out on STEREO will provide the first test owhat can be achieved with a future network of spaweather stations, and they will provide the databasexperience, and insight on which planning for succapabilities can be based.

lyketut

Figure 19. Longitudinal dependence of particles from CME-driven shocks. Typical intensity-time profiles forprotons of three different energies as seen by observers viewing a large CME-driven shock from three differentlongitudes. The observer seeing a “western” event (left panel) is well-connected to the nose of the shock earon and sees a rapid rise and decline. The observer near central meridian is well-connected until the shocpasses, and thus sees a flat profile. The observer viewing an “eastern” event is poorly connected until after thshock passes; it is not until then that he is connected to the nose of the shock. With a network of spacecraft asuch locations it is possible to study the accelerated particle spectra and composition as the shock moves ointo the heliosphere, to measure the plasma and magnetic field properties of the shock in situ, and to developthe necessary databases, understanding, and tools that can ultimately lead to a predictive capability.

axi-

lyv-al-tsthersoe,tal

re-ent

l-dor

leuldlert

ss,re

helt-tulds.

ee.

lueyh.

beass

e--

c-dby (tovalna

CMEs and Space Weather Forecasts

Virtually all transient shock wave disturbancesin the solar wind observed near Earth are driveby CMEs. In addition, the largest SEP eventsthe so-called gradual events, appear to be a cosequence of particle acceleration occurring inthe vicinity of strong CME-driven shocks.Finally, all of the largest (Kp > 7–) nonrecur-rent geomagnetic storms are caused by CMEdriven solar wind disturbances impactingEarth’s magnetosphere. On the other handabout two-thirds of all CMEs do not produceshock disturbances in the ambient wind nea1 AU, an even larger fraction do not producegradual particle events in interplanetary spaceand about five out of six CMEs directed Earth-ward do not produce large geomagnetic storm(Kp > 7–). Present evidence suggests that CMEspeed, in particular speed relative to the ambient wind ahead, is a key factor in a CME’sability to produce these phenomena. Howeverother factors, such as the mass and size of thCME, the strength and orientation of the mag-netic field within both the CME and the ambi-ent wind, the location of the observer (or Earth)relative to the disturbance center, and the avaiability of various particle seed populations foracceleration, enter into the disturbance equation as well. It is not currently known whichattributes of the CME/ambient wind combina-tion are most influential in producing large shockwave, energetic particle, and geomagnetic disturbances. In addition, we do not yet know howto translate observations of the ambient windand CME characteristics close to the Sun intoaccurate predictions of effects at Earth.

The STEREO Beacon

Space Weather Forecast Data

Successful integration of STEREO into the natiospace weather forecast effort hinges on the immentation of simple but robust onboard processschemes to automatically identify events of interebroadcast an alert, and trigger the transmission pre-stored, high-cadence image and ancillary d

2

n,n-

-

,

r

,

s

-

,e

l-

-

-

nalple-ingst,

of aata

stream necessary to sharpen the warning and mmize its utility.

A “beacon” mode of operation can be particularuseful in warning of dangerous solar particle actiity. For example, a microprocessor could make retime classifications of gradual or impulsive evenbased on the measured particle composition and ocharacteristics. The microprocessor would aldetermine the maximum particle flux, rate of risproton and helium energy spectra, and elemencomposition. If these parameters exceed pdetermined threshold levels, an alert could be sto Earth.

If suitably designed, STEREO will provide a reatime capability for warning of Earthward-directeCMEs. For example, if an onboard microprocessidentifies a coronal transient, significant particfluxes, or a strong interplanetary shock, an alert cobe sent to Earth at a low bit rate. The immediate awill need to provide positive identification of CMElaunch time and direction. Estimates of speed, maand relation to structures in the lower atmosphe(to provide an idea of the magnetic content of tCME) would be desirable but perhaps too difficuto include in a simple algorithm. Most likely, preliminary values will have to be derived from the firsfew images sent down, and more accurate ones wofollow from analysis of the full series of event image

The immediate alert algorithm will presumably bbased on some form of image differencing schemDetection of change beyond some threshold vawill be required but, in addition, it will be necessarto judge whether the motion is, in fact, toward EartThe fastest CMEs travel approximately 5 RSun perhour, so the underlying image cadence needs toquick enough to catch these events before they pentirely out of the field of view.

It is important that the false alarm rate of the immdiate alerts be kept low lest their utility be compromised. This will be no trivial task, since the detetion scheme will have to be run in real time anautonomously. These difficulties are compounded the desire to track as far from the Sun as possibleget the best estimate of CME properties and arritime), out where accurate subtraction of the F-coro

7

a

ap

l

u

toost

assil-ca-er-n-r

e.orgyk

ofr-

e-es’sr oril

Ese.

tedand

iste-yill

es.

othoneep-

on

becomes an issue. In addition, although the fasand most energetic CMEs are generally associwith the most dramatic geomagnetic disturbancsome CMEs that start out slow and unimpressivealso be geoeffective. For these reasons, an effectiveforecast scheme will have to rely heavily on thestruments that can track CMEs to 1 AU.

STEREO could improve upon predictions of SEfrom X-ray observations provided directional infeences can be drawn in real time. That is, if STERin the beacon mode can distinguish the locationthe parent X-ray flare, it could at the very least sout sources too far removed from the Earth–Sun to be effective (thereby reducing false alarms), it might be able in a statistical sense to narrow probabilities of the prediction by virtue of the moaccurate locations.

The value of the STEREO mission in pioneering adeveloping the use of deep space monitors at langles to the Sun–Earth line cannot be overemsized. The work here is truly exploratory, since athough we now have some idea of what is invoin gathering observations relevant to space weatapplications, the full scope of what is required conly be determined by direct experience.

Maximizing the Science Return

In addition to modern data compression strateg(see Appendix II), the STEREO mission can usunique new strategy for maximizing the data retby taking advantage of the beacon mode and omultaneous observations from Earth-orbiting aground-based solar observatories. The concept store much more imaging data on board than cadownlinked; data from periods of interest are thselectively downlinked. Using this strategy, vehigh-cadence data on the initiation and explosphases of CMEs or other eruptive events canobtained. This data strategy requires that missoperations include scientists monitoring the data pvided by other observatories and by the STERbeacon.

To implement this strategy on a STEREO misswith a downlink of about 1 Gbit per day, the onboa

2

testatedes,

can

in-

Psr-EO ofortlinend

there

ndrgeha-l-vedheran

iese arn

f si-ndis ton beenryive beionro-

EO

ionrd

data storage capacity would be sized at about 1020 Gbits. This does not cause a large mass or cpenalty because 10- to 20-Gbit erasable disk mmemory devices weighing about 5 kg are now avaable. Data would be recorded at a much higher dence than 1 Gbit per day. Scientists would detmine which portions of the data should be dowlinked and which portions should be marked fodeletion. This information would be uplinked to thspacecraft during the daily uplink/downlink periodSince the data could remain on the recorder fseveral days before being downlinked, this stratecan be implemented within a low-cost 40-hour/weemission operations schedule.

5. Mission Overview

STEREO must lead to a depth of understandingsolar activity that is incisive enough to predict solaeruptions and their effects throughout the heliosphere. To accomplish this, each STEREO spaccraft must carry a cluster of state-of-the-art telescopand environmental sensors. Images from STEREOsolar telescopes will be combined with solamagnetograms and other data from ground-basedEarth-orbiting observatories to document in detaboth the buildup of magnetic energy and CMliftoffs. Other STEREO telescopes will track CMEand their shocks through interplanetary spacOnboard sensors will sample particles acceleraby the shocks as well as the disturbed plasmas magnetic fields themselves.

We recommend that the STEREO mission consof two identically instrumented Sun-pointed spaccraft at 1 AU. The spacecraft should slowly drift awafrom Earth, so that after 2 years, STEREO #1 wlead Earth by 45° and STEREO #2 will lag by 60°.Each spacecraft will generate at least 250 imagper day plus in situ magnetic field and particle dataThe solar images should be simultaneous ±1 s. Sci-ence data should be transmitted once a day, and bspacecraft should provide real-time alerts (beacmode). When needed, a quick response by the DSpace Network (DSN) to an alert of especially important or dangerous events could provide detailsEarth-bound CMEs.

8

c

aw

l

t

r

hsa

udye- a

ibedonsis-nchtra-

tionnd andn to

n 2ectigh tobutlarvele 2nce

ed that

As described in Appendix I, the two spacecraft be launched in 2003 either separately by Taurockets or together by a Delta rocket. Solar interplanetary instruments on and near Earth provide a third vantage point from which to stuthe Sun and heliosphere together with the STERspacecraft. Solar-B will be launched in 2004. Thwill be improved ground-based telescopes, andNOAA GOES satellites will carry solar X-ray imagers. It is possible that the Solar and HeliosphObservatory (SOHO), WIND, and the AdvancComposition Explorer (ACE) will be operating stialthough those programs are expected to endfore 2003. The Yohkoh satellite, with its outstaning X-ray telescopes, will reenter the atmosphin 2002.

We have defined a STEREO mission that will demine the origins and propagation of solar activthat affects Earth. We assumed that only a netwof ground-based observatories and the Solar Tetrial Probes will be available to provide synergisdata. We particularly considered the possibility tSOHO could serve as one of the STEREO eyespace. We decided against relying on SOHO becif it should fail during the years leading up to launor shortly after, reliance on SOHO would result

2

anrusndill

dyEO

ere the-ericedl, be-d-ere

er-ityorkres-ticat onuse

ch in

the loss of the stereoscopic mission. As our stmade clear, the scientific value of another singlviewpoint mission is dramatically lower than formission with stereoscopic capability.

The mission is divided into four phases, as descrin Section 6 of this report. Primary science operatiwill occupy the first 2 years. The goal for total msion lifetime is 5 years. The schedule, with a lauin 2003, is based on the Solar Terrestrial Probe stegic plan developed for the Sun-Earth ConnecRoadmap. The scientific program does not depeon the phase of the solar cycle because CMEsthe other phenomena to be studied are commoall phases of the cycle.

To achieve the scientific goals outlined in Sectioof this report, the STEREO instruments must reflstate-of-the-art technology and achieve quite hspatial and temporal resolution. The technologyachieve the STEREO goals is available now, implementing it within the cost guidelines for SoTerrestrial Probes will be a challenge. We beliethe measurement objectives summarized in Tabare necessary and sufficient to achieve the sciegoals. A preliminary cost study (Appendix I) carriout at the Goddard Space Flight Center indicates

t

Table 2. STEREO measurement objectives.

Feature Size ResolvedPhenomenon (and/or Timestep) Physical Properties

CMEs near Sun 40,000 km = Density, velocity, internal structure, exten2 3 10–4 AU (6 min)

Flares 2,000 km Position, density, structureMoreton waves 5,000 km Wave front shape, velocity, underlying

(1 min) magnetic fieldCoronal loops 2,000 km Temperature, density, structure, deflection

by wavesCoronal streamers 40,000 km Distortion by CMEs, extentCoronal holes 2,000 km Footprint, spreadingSEPs 2 min 3-D distribution functionCMEs near Earth 0.01 AU (images) Magnetic field, density, velocity, shape,

(plasma) 1 min extent, temperatureInterplanetary shocks 0.02 AU (5 s) Extent, velocity, strength

9

t

nydhe

e

n

the measurement objectives can be achieved wthose cost guidelines.

As detailed in Appendix I, the STEREO mission cbe implemented with a mission cost (phase C/Dthe $120M (FY97 dollar) cap for Solar TerrestrProbes. Costs can be minimized by a deft execuof phase C/D, which, according to the study, laonly 32 months, and by early selection of an insment team.

3

ithin

an) atialtionstsru-

As detailed near the end of Section 2, there are mainteresting and important investigations, beyonthose baselined, that can be carried out from tSTEREO platforms. The principal restriction onadded investigations is the Solar Terrestrial Probcost cap. Hence, instruments provided by non-NASA-supported institutions may be included to strengthethe overall science program.

-

.

Sun

EarthE W

Figure 20. Position of the STEREO spacecraft afterabout 1 year. STEREO #1, leading Earth, can seearound the west limb; STEREO #2, lagging Earth,can see around the east limb.

What is the Optimum AngularSpacing Between the Two Spacecraft?

There is no single angular spacing that is bestfor all instruments and science goals. The coro-nagraphs effectively detect only the coronawithin ±60° of the plane of the sky. This im-plies that for triangulation on CMEs aimed atEarth, the spacecraft should be at least 60° apart.Other CMEs will be detectable by both coro-nagraphs for spacecraft separations rangingbetween 0° and 120°. On the other hand, it isbest to have the high-resolution chromosphereand low corona imagers separated by only 15°–60° so that features can be identified in the im-ages from both spacecraft. Triangulation onshock fronts with the radio receivers is likely tobe most accurate when the spacecraft are sepa-rated by ~60°. If ACE or WIND or other near-Earth spacecraft are not available, then a STE-REO spacecraft near Earth would be desirableto monitor the fields and particles input to themagnetosphere. The Science Definition Team’ssolution is the four-phase plan, which focuseson different mission objectives at differenttimes. Thus, we recommend that the two space-craft be launched into slightly elliptical orbitsat 1 AU, one leading Earth and one lagging (see

Figure 20), so that the angles between the spacecraft and the Sun–Earth line increase graduallywith dwells at selected angles (see AppendixI). STEREO #1, leading Earth, will dwell near20° between 200 and 400 days into the mis-sion, and near 45° between 600 and 800 days.STEREO #2, lagging Earth, will dwell near 30°and 60°, respectively. After this period, the twospacecraft will move to larger angles and focuson support of other Solar Terrestrial Probe missions

0

i

i

E

i

r

eyricill

e-

-ngmn

-ar-

aft0les

iversm-c-n-es to

a onI,le-n-

e

6. Phases of the STEREO MissionThe studied STEREO mission will have four dtinct phases corresponding to different scientific apractical applications of the data and to the anglαseparating the two spacecraft.

Phase 1: The 3-D Structure of the Corona(first 400 days, a ≤ 50°)While the angular separation α is small and the satellites are close to Earth, telemetry is hardly stricted, and the STEREO satellite configurationoptimum for making rapid-cadence high-resoluti3-D images of coronal structures. The coronal agers will be able, for the first time, to unambigously determine the important physical propertiescoronal loops and to determine whether coronal lointeractions include reconnection. Stereoscopic age pairs and sequences will capture the 3-D stture of the corona before, during, and after CMThey will also allow us to delineate the subtle sweing and the sigmoid features that often foreshadCME onset. Solar-B will be able to show thcorresponding magnetic developments in the phosphere. The period when the STEREO spacecrafclose together will also be used to intercalibrate instruments.

Phase 2: The Physics of CMEs (days 400 to800, 50° ≤ a ≤ 110°)As the two spacecraft drift farther apart, they becoideally placed to triangulate on CMEs to determtheir true dimensions and trajectory. These will breakthrough measurements. Further, each spacewill be able to image CMEs directed toward the othDetectors on each spacecraft will measure the mnetic field and plasma properties of CMEs trackby the other spacecraft, thereby linking the charteristics of a CME (composition, magnetic field oentation, density, and velocity at 1 AU) with its launand propagation parameters (size, velocity, asource region characteristics).

Phase 3: Earth-Directed CMEs (days 800 to1100, 110° ≤ a ≤ 180°)In Phase 3, the viewing angles become ideal forserving CMEs aimed at Earth. The coronagrapheliosphere imagers, and radio receivers will tra

3

s-nde

-re- isonm-u- ofop

im-ruc-s.ll-oweto-

t arethe

menebecrafter.ag-edac-i-chnd

ob-hs,ck

the development of CMEs and their shocks as thpropagate to Earth, where the MagnetospheMultiscale and Global Electrodynamics missions wmeasure their geoeffectiveness.

At this phase of the STEREO mission, the spaccraft will have nearly a 360° view of the Sun, allow-ing the longitudinal extent of CMEs and other activity to be determined. There have been tantalizisuggestions from Yohkoh soft X-ray images and frothe SOHO/LASCO experiment that CMEs castretch over more than 180° of longitude. STEREOwill not only test this suggestion but will also provide global maps of the coronal structures that pticipate in the activity.

Phase 4: Global Solar Evolution and SpaceWeather (after day 1100, a > 180°)When the separation of each STEREO spacecrfrom the Sun–Earth line becomes greater than 9°,events on the far side of the Sun that launch partictoward Earth will be visible for the first time. Activeregions can be tracked and studied for their eruptpotential from their emergence, wherever it occuon the Sun. The results will have a tremendous ipact on our ability to anticipate changes in solar ativity and to predict changes in space weather coditions. Such a predictive capability is vital if we arto build permanent lunar bases or send astronautMars.

7. Observational ApproachBased on our study of the scientific potential of STEREO mission, as described in Section 2, andthe practical limitations, as described in Appendix Iwe recommend that the baseline instrument compment for each of the two STEREO spacecraft cosist of seven instruments as summarized below.

• Chromosphere and low corona imager: anextreme ultraviolet (EUV) and/or X-ray telescopthat images 1 RSun to 1.5 RSun

• Coronagraph: a white-light coronagraph thatimages 1.5 RSun to 30 RSun

• Radio burst tracker: a radio receiver that tracksshocks from the outer corona to beyond Earth

• Heliosphere imager: a visible-light telescope thatimages 30 RSun to beyond Earth

1

a

f

h

to

h

d

o

lrn

r

tss

-nge be

ofEs,o-

0alra-

of

lctor-

andro-

liaronnd-

do-pa-tial

ter

no-

co-veky-

eoate

• Solar wind analyzer: a plasma analyzer thsamples CME and ambient plasmas at 1 AU

• Magnetometer: a sensor that detects magnefields inside and outside CMEs

• Solar energetic particle detector: detectors oprompt and delayed electrons and ions fromto 50 MeV

All the instruments needed for accomplishing STEREO objectives can be built with available tenology. In some cases, instruments essentially idtical to previously flown instruments will meet tobjectives and mission constraints. In other casome customizing will be needed. The instrumdescriptions given below are intended to demonsthat there is at least one well-established apprto each of the baseline instruments. The actual SREO instruments will be selected through a copetitive review process, and the instrument descriptions here are not intended in any way to restrictpossible approaches, nor do we intend by our lispreclude consideration of other instruments, suca magnetograph (see Appendix III). We believe, however, that the baseline instrument complement meet the mission science objectives.

Chromosphere and Low Corona Imager

This telescope should be able to obtain images least one coronal and one chromospheric emisline. The EIT multilayer normal-incidence extremultraviolet telescope on SOHO, for example, provithe kind of images needed. However, STEREfocus on solar activity will require a much highcadence of observations than EIT provides. Theages should show solar prominences and corloops and other coronal structures from the basthe corona to 1.5 RSun.

The capabilities of the chromosphere and low rona emission-line imager should include tunabiso that Doppler shifts can be measured. One appstudied at the NASA Goddard Space Flight Cewould include two mirrors, multilayer-coated to haa peak reflectivity chosen to observe He II 304 Å301 and 307 Å. Each of the two mirrors would foimages of the Sun on the same area of the deteA movable shutter could switch between the multilayer passbands. The difference in intenbetween the two images would provide a mea

3

t

tic

0.1

thech-en-e

ses,entrateachTE-m--thet to as-will

in atsione-es

O’serim-nal

e of

co-ity,oachterve atmctor.woityure

of the Doppler velocity for each pixel. The instrument should be able to measure velocities in the raof 10 to 3000 km/s. The spatial resolution shouldat least 3 arcsec, or 2000 km on the Sun.

Coronagraph

The white-light coronagraph should be capableobserving the current sheet, streamers, CMstreamer blow offs, and the acceleration of inhomgeneities in the solar wind from ~1.5 RSun to ~5 RSun.A substantially larger field of view, to at least 3RSun, is important for studying the three-dimensionstructure of streamers, the evolution and acceletion of streamer blow offs, and the accelerationinhomogeneities in the solar wind.

Coverage of the corona from about 1.5 RSun to 30RSun will likely require two channels if conventionaexternally occulted designs are used. The deteformat should be 1024 3 1024 or better and the dynamic range must be better than 104 in order to trackthe two orders of magnitude change in the signal background while detecting the ~1% contrast of conal features against the background.

Certain aspects of the coronagraph design pecuto the specific objectives of the STEREO missimust be carefully considered. The usual backgroulight rejection limitations affecting the determination of the field of view of an externally occultewhite-light coronagraph designed for the inner crona is complicated by the orbit eccentricity, the seration angle of the two spacecraft, and the sparesolution function in the inner field of view. Whilethe orbital eccentricity affects the apparent diameof the Sun, and hence the occulter inner cutoff R1,by a relatively small (~10%) amount, its effect othe background light rejection can be very prnounced.

The spacecraft separation significantly affects the observed field of view for separation angles aboabout 30°. A feature observed in the plane of the sand an altitude h by one coronagraph will be co-observed by the second coronagraph only if h > R1 seca, where a is the spacecraft separation or sterangle. The science objectives of STEREO indicR1 ~ 1.5 RSun is desirable. The co-observed h is 1.55RSun, 1.73 RSun, 2.12 Rsun, and 3.00 RSun for 15°, 30°,

2

n

/

e

i f

te

wsw

ct

seil

n

t

aiei

rlyun as-d itslikefivean- itsre-e-er

tor-hhe-ht

thee-ec-forhe

r isaton

si-ith

wothe

m- a

ewx ofry.ionf--in

le,the a

45°, and 60°, respectively. The effects at the oufield of view cutoff, R2, scale similarly. Using stadard design methods, a coronagraph with R1 = 1.5RSun could be expected to have R2 ~ 5–6 RSun. Thecoronagraph spatial resolution function, due tovarying obstruction of the entrance aperture byexternal occulter with changing altitude, is asymmric and deteriorates rapidly near R1. The SOHOLASCO/C2 coronagraph with an R1 of about 1.5 RSunhas a nominal spatial resolution of about 8 arcs6 RSun, but only about 110 arcsec by 2 RSun. The ef-fects that the low and radially varying resolutfunction will have on three-dimensional imageconstruction in the inner corona must be careexamined.

Radio Burst Tracker

The STEREO spacecraft should carry two idenradio receivers so that triangulation of solar evcan become routine rather than fortuitous asWind-Ulysses. A simple receiver like that to be floon Cassini connected to a triaxial antenna syalso similar to (but simpler than) that being floon Cassini can track solar radio disturbances to w±1° from 1 to 2 RSun to 1 AU. The correspondinradio frequency range is ~15 MHz to ~30 kHzkey scientific objective for the radio burst tracketo triangulate on radio emission from shock-acerated particles, 20 MHz to 30 kHz sweep, wifew seconds’ time resolution. This should effectivallow tracking of the location of particle acceletion sites through interplanetary medium

Heliosphere Imager

The heliosphere imagers should have 100 timespatial resolution of those on Helios and a cadof about one image pair per hour. These capabwill be adequate to map the solar wind and CMEheliospheric distances between 30 RSun and 215 RSun.Resolutions of ~1° in heliospheric latitude and logitude are feasible with current technology, so SREO should achieve the goal of approxima1-hour time resolution in CME tracking.

One design that has been studied has an opticager to view a hemisphere of sky starting withfew degrees of the solar disk and roughly centon the spacecraft-to-Earth line. Strictly speak

3

ter-

the theet-

c at

onre-ully

icalnts

forn

temn

ithing. Ar isel-

h aelyra-

thenceitiess at

-TE-ely

l im-n ared

ng,

this is a “half-sky” camera, although it covers neaall the path traveled by material going from the Sto beyond Earth. A hemispherical imager withmulti-element light baffle, a wide-angle optical sytem, and a CCD camera has been designed ancomponents have been tested. The baffle works a coronagraphic external occulter and consists of knife-edge walls spaced about 1 cm from one other, with each wall top placed in the shadow ofnext outer neighbor. The optical system further duces background-light contamination, down to blow the equivalent of one 10th magnitude star psquare degree. This optical system consists of aoidal mirror enclosing a simple thick lens, whicmaps the sky onto the CCD photometer; it is tequivalent of a “fish-eye lens,” but without a protruding glass element that can intercept stray ligcrossing over the edge of the baffle.

Solar Wind Plasma Analyzer

The solar wind plasma analyzer should measuredistribution functions (to provide density, vector vlocity, temperature, and anisotropy) of ions and eltrons over the energy ranges of 300–8000 eV (positive ions) and 1–1000 eV (for electrons). Trequired time resolution is a few minutes.

One approach to the solar wind plasma analyzean ion-electron spectrometer comprising two top-htoroidal electrostatic analyzers that share a commcollimator and steering lens. This design allows multaneous measurement of electrons and ions wan overall reduction of mass and volume over tdiscrete instruments. With no potential placed on steering lenses, the analyzer provides up to 360° fieldof view in the plane perpendicular to the axis of symetry through the entrance aperture. By placingpotential across the steering lens, the field of viof the instrument is changed to a cone, the apewhich is located on the analyzer’s axis of symmetBy sweeping the steering lens voltage, elevatangles through ±40° can be observed during an efective field of view of 2.6 p steradians. The elevation can be servo-controlled to follow deviations the solar wind direction.

If sufficient mass, power, and funding are availabthe plasma analyzer could be enhanced by addition of a time-of-flight section capable, at

3

a

iu

and (10o-dor

ht berslede-ith0ro-ed.

ingsoure

ne,-lnd

ci-ur-ive in

rrygere

minimum, of determining the ionization temperatuof the plasma through measurement of the ratioO6+ to O7+ ions. A second objective would be to dtermine the relative abundances of ions with hand low first-ionization potentials, such as the Mgratio. The ionization state and the composition pvide clues to the coronal sources of the plasma. energy-per-charge filtered ions would be accelerathrough a carbon foil floating at a high negative ptential. Secondary electrons emitted from the fwould provide a start pulse, while accelerated iowould be detected after passing through the timof-flight region to provide a stop pulse.

Magnetometer

A candidate magnetometer for STEREO is a sinminiature triaxial fluxgate magnetometer using rincore magnetic sensing elements. The magnetomlow-noise ring core sensors are derived from the stechnology used in the Voyager, Magsat, GiotCLUSTER, GGS, AMPTE, and MGS. A dynamrange of ±65,536 nT can be achieved with a resotion of 0.125 nT in one channel, and ±655 nT with aresolution of 0.00125 nT in a second channel. Tvector magnetic field can be obtained at a rate ovectors per second, but such high temporal restion is not necessary for the STEREO mission. If magnetometer sensor is mounted on a boom, theentation of the boom must be chosen to avoid inference with the field of view of other instrument

Solar Energetic Particle Detector

Only rather modest instrumentation based on proapproaches is required to address the mission obtives for energetic particles. The instrumentatshould be able to distinguish impulsive from gradevents based on the characteristics in Table 3.

Table 3. Characteristics of impulsive andgradual particle events.

Impulsive Gradual

Enriched in He3 Normal isotopiccomposition

Electron-rich Proton-richEnriched in Fe & Coronal abundances

other heavy ionsRapid rise & decay More extended

34

re ofe-igh:Oro-Thetedo-oilnse-

gleg-eterme

to,iclu-

hef 20olu-the ori-ter-s.

venjec-onal

It should be able to measure the absolute intensity energy spectra of energetic nuclei at the energiesto 100 MeV/nuc) that pose a potential risk to astrnauts. The following list gives the observational anscientific objectives of the energetic particle detectand representative energy ranges to be covered:

Observational ScientificObjectives Objective

Sample in situ, with Understand the1-min time resolution mechanism for the

• Energetic electrons: production of~0.1 to 3 MeV CME shock-

• Protons: ~0.1 to ~100 MeV accelerated• Helium: ~1 to 100 MeV/nuc particles• Heavy ions (6 < Z < 28):

~2 to 30 MeV/nucHe3 identification

Using modern approaches to low-power, lightweiginstrumentation, the required measurements canprovided with a package of several small detectothat would require ~3 kg and ~2 W. These partictelescopes could be based on silicon solid-state vices, which provide precise measurements wgood long-term stability. An average bit rate of ~20bits per second should be adequate if onboard pcessing and data compression techniques are us

8. ConclusionsWe have reviewed recent progress in understandCMEs and identified the major scientific questionto be answered. The key questions and many of conclusions are highlighted by italics throughout thtext. We concluded that two spacecraft at 1 AU, odrifting well ahead of Earth and one well behindwill serve the objectives of NASA’s Sun-Earth Connection Initiative by (1) enabling fundamentaresearch on the three-dimensional structure adynamical processes of CMEs, (2) providing the sence base for greatly improved forecasts of distbances at Earth, and (3) providing comprehensmeasurements of the interplanetary environmentsupport of follow-on Solar Terrestrial Probes.

We recommend that the STEREO spacecraft caidentical complements of instruments, includinchromosphere and coronal imagers, a heliosph

w

h

e

r

l

h

c

i

t

byial

rate

es.

yhate

x-cee

:

.d

re6,

.d

se-

am

ng

e:p/

e-nk-er

rch

-94,

imager, a radio telescope, and sensors of interpetary particles and magnetic fields. We believe tthe recommended complement of instruments accomplish the goals of the STEREO mission.

Together with the Goddard Space Flight Center,studied orbits, vehicles, programmatic requiremeand funding needed to carry out a 2-year sciemission with a 3-year extension for support of otSolar-Terrestrial Probes. We concluded that needed technologies are available now and thamission can be launched in mid-2003 within the crestrictions of the Solar-Terrestrial Probe line of msions. We also considered how existing or planspace assets, such as ACE and Solar-B, might athe scientific potential of the mission. We particlarly considered the possibility that SOHO couserve as one of the STEREO eyes on space. Wcided against relying on SOHO because if it shofail during the years leading up to launch or shoafter, then reliance on SOHO would result in the lof the stereoscopic mission. As our study made cthe scientific value of another single-viewpoint msion is dramatically lower than for a mission wistereoscopic capability.

In order to maximize the scientific return from tunique opportunity provided by STEREO, furthstudies should be conducted to maximize information that can be extracted from sterobservations. Such studies, which will include simlated observations of prescribed structures (eCMEs, streamers, loops), will help assure the omum design and selection of STEREO instrumention. We also recommend that studies of various tscopes, including magnetographs, be pursvigorously to minimize eventual costs and maximcapabilities.

As a result of this study, the Science Definition Teconcludes that:

1. Two suitably instrumented spacecraft in elliptisolar orbits, leading and lagging Earth at 1 Awill provide the measurements needed to sothe fundamental scientific issues surroundcoronal mass ejections.

2. The technology for the STEREO missionready, and NASA should act promptly

3

lan-hatill

wents,nceer

thet theostis-neddd tou-ld de-

uldtly

ossear,is-th

eertheeou-.g.,pti-ta-

ele-uedize

am

alU,lveng

iso

implement it. The spacecraft can be launched2003 within the cost cap of the Solar-TerrestrProbe program.

3. NASA should act, in cooperation with otheagencies, to implement a “beacon mode” thwould enable STEREO to provide near-real-timwarnings of impending geomagnetic disturbanc

9. BibliographyThe scientific literature dealing with solar activitand its effects is extensive. We have drawn on tliterature in discussing the scientific goals of thSTEREO mission, but for brevity we have not eplicitly cited the works. Several recent conferenproceedings are particularly helpful in explaining thscientific issues motivating the STEREO mission

Coronal Mass Ejections, Geophys. Monogr. Ser., vol99, edited by N. Crooker, J. A. Joselyn, anJ. Feynman, 1997, AGU, Washington, D. C.

Magnetic Reconnection in the Solar Atmosphe,edited by R. D. Bentley and J. T. Mariska, 199Astron. Soc. Pacific Conf. Ser., Vol. 111, San Fran-cisco, Calif.

Solar Wind Eight, edited by D. Winterhalter, J. TGosling, S. R. Habbal, W. S. Kurth, anM. Neugebauer, 1996, AIP Conf. Proc., Vol. 382,AIP Press, Woodbury, N. Y.

Solar Dynamic Phenomena and Solar Wind Conquences, Proc. Third SOHO Workshop, 1994, ESASP-373.

The recommended Sun-Earth Connections progris available on the Web:

Sun-Earth Connection Roadmap. Strategic Plannifor the Years 2000–2020, Roadmap IntegrationTeam (J. L. Burch, chairman), 1997. Websitht tp: / /espsun.space.swri .edu/~roadmaindex.html.

In the bibliography below, we offer a sample of rcent papers that are representative of current thiing in the field and that will lead the interested readto earlier works and to other works on the reseafrontier.

“Acceleration of energetic particles which accompany coronal mass ejections,” Reames, D. V., 19

5

i

a

sn

n

M

,

eh

.

un

e

a

n

“ D.

“ ndic

T.K.

“

“ g-n,ndC.

“ in

o,

“ a-d

“ ei,g,

L ee,the0,

“ h5,

“ ,”.

ic,

“ icnd

in Proc. Third SOHO Workshop—Solar DynamPhenomena and Solar Wind Consequences, ESASP-373, pp. 107–116.

“Coronal mass ejections and large geomagnestorms,” Gosling, J. T., S. J. Bame, D. J. McComand J. L. Phillips, 1990, Geophys. Res. Lett., 17,901–904.

“Coronal mass ejections and magnetic flux ropesinterplanetary space,” Gosling, J. T., 1990, in Phys-ics of Magnetic Flux Ropes, Geophys. Monogr.Ser., vol. 58, edited by C. T. Russell, E. R. Prieand L.-C. Lee, pp. 343–364, AGU, WashingtoD. C.

“Coronal mass ejections: A summary of SMM obsevations from 1980 and 1984-1989,” HundhauseA. J., 1997, in The Many Faces of the Sun: Scietific Highlights of the Solar Maximum Mission,edited by K. T. Strong, J. L. R. Saba, and B. Haisch, Springer-Verlag, New York, in press.

“Coronal mass ejections: The key to major interplaetary and geomagnetic disturbances,” WebD. F., Rev. Geophys., Supplement, pp. 577–583.

“Disruption of coronal magnetic field arcadesMikic, Z., and J. A. Linker, 1994, Astrophys. J.,430, 898.

“EIT: Extreme-Ultraviolet Imaging Telescope for thSOHO Mission,” Delaboudinière, J.-P. and oters, 1995, Solar Phys., 162, 291–312.

“The ejection of helical field structures through thouter corona,” House, L. L., and Berger, M. A1987, Astrophys. J., 323, 406-413.

“Eruptive prominences as sources of magnetic cloin the solar wind,” Bothmer, V., and R. Schwen1994, Space Sci. Rev., 70, 215.

“The escape of magnetic flux from the Sun,” BiebeJ. W., and D. M. Rust, Astrophys. J., 1995, 453,911.

Foundations of Solar Particle Event Risk Managment Strategies, Findings of the Risk ManagemenWorkshop for Solar Particle Events, M. AcunChair, July 1966.

“Geomagnetic activity associated with Earth passaof interplanetary shock disturbances and coromass ejections,” Gosling, J. T., D. J. McComaJ. L. Phillips, and S. J. Bame, 1991, J. Geophys.Res., 96, 7831–7839.

36

c

tics,

in

t,,

r-n,-

.

n-b,

”

-

e,

ds,

r,

-t,

geals,

Great geomagnetic storms,” Tsurutani, B. T., W. Gonzalez, F. Tang, and Y. T. Lee, 1992, Geophys.Res. Lett., 19, 73–76.

Heliospheric observations of solar disturbances atheir potential role in the origin of geomagnetstorms,” Jackson, B. V., 1997, in Magnetic Storms,Geophys. Monogr. Ser., vol. 98, edited by B. Tsurutani, W. D. Gonzalez, Y. Kamide, and J. Arballo, pp. 59–76, AGU, Washington, D. C.

The implications of 3D for solar MHD modelling,”Antiochos, S. K., and R. B. Dahlburg, 1997, SolarPhys., 174, 5–19.

The initiation of coronal mass ejections by manetic shear,” Mikic, Z., and J. A. Linker, 1997, iCoronal Mass Ejections, Geophys. Monogr. Ser.vol. 99, edited by N. Crooker, J. A. Joselyn, aJ. Feynman, pp. 57–65, AGU, Washington, D.

In situ observations of coronal mass ejectionsinterplanetary space,” Gosling, J. T., 1992, in Erup-tive Solar Flares, Lecture Notes in Physics, editedby Z. Svestka, B. V. Jackson, and M. E. Machadvol. 399, pp. 258–267, Springer, Berlin.

Interplanetary magnetic clouds, helicity conservtion and intrinsic-scale flux ropes,” Kumar, A., anD. M. Rust, 1996, J. Geophys. Res., 101, 15,667.

Interplanetary origin of geomagnetic activity in thdeclining phase of the solar cycle,” TsurutanB. T., W. D. Gonzales, A. L. C. Gonzales, F. TanJ. K. Arballo, and M. Okada, 1995, J. Geophys.Res., 100, 21,717–21,733.

agrange: The Sun and Inner Heliosphere in ThrDimensions, Schmidt, W. K. H., and others, 1996a proposal to the European Space Agency for next medium size mission (M3), ESA SP-118116.

The Large Angle Spectroscopic Coronagrap(LASCO),” Brueckner, G. E., and others, 199Solar Phys., 162, 357–402.

Magnetic chirality and coronal reconnectionCanfield, R. C., A. A. Pevtsov, and A. NMcClymont, 1996, in Magnetic Reconnection inthe Solar Atmosphere, edited by R. D. Bentley andJ. T. Mariska, pp. 341–346, Astron. Soc. PacifSan Francisco.

Magnetic field structure of interplanetary magnetclouds at 1 AU,” Lepping, R. P., J. A. Jones, a

e

n

a

u

s

9

6

in

y

of.,

yD.

n,

-

s,

,”

.8,

L. F. Burlaga, 1990, J. Geophys. Res., 95, 11,957–11,965.

“Problems and progress in computing three-dimsional coronal active region magnetic fields froboundary data,” McClymont, A. N., L. Jiao, aZ. Mikic, 1997, Solar Phys., 174, 191–218.

“Quasi-stereoscopic imaging of the solar X-rcorona,” Batchelor, D.,1994, Solar Phys. 155,57–61.

“Rendering three-dimensional solar coronal strtures,” G. Allen Gary, 1997, Solar Phys., 174, 241–263.

“Scaling of heating rates in solar coronal loopKlimchuk, J. A., and Porter, L. J., 1995, Nature,377, 131.

“The sizes and locations of coronal mass ejectioSMM observations from 1980 and 1984-198Hundhausen, A. J., 1993, J. Geophys. Res., 98,13,177–13,200.

“Solar activity and the corona,” Low, B. C., 199Solar Phys., 167, 217.

“The solar flare myth,” Gosling, J. T., 1993, J.Geophys. Res., 98, 18,949.

“Solar flares—An overview,” Rust, D. M., 1992, Life Sciences and Space Research XXIV/2/ Ration Biology; Proc. Topical Meet. Interdisciplinar

3

n-md

y

c-

,”

ns:,”

,

dia-

Scientific Commission F/Meetings F3, 12, 289–301.

“Solar tomography,” Davila, J. M., 1994, Astrophys.J., 423, 871.

“Spatial and temporal invariance in the spectra energetic particles in gradual events,” D. VReames, S. W. Kahler, and C. K. Ng, 1997Astrophys. J., in press.

“The spatial distribution of particles accelerated bcoronal mass ejection-driven shocks,” Reames, V., L. M. Barbier, and C. K. Ng, 1996, Astrophys.J., 466, 473.

“Spawning and shedding helical magnetic fields ithe solar atmosphere,” Rust, D. M., 1994Geophys. Res. Lett., 21, 241–244.

“Stereoscopic viewing of solar coronal and interplanetary activity,” Schmidt, W. K. H., and V. Bothmer,1996, Adv. Space Res., 17, 369–376.

“Three-dimensional reconstruction of coronal masejections,” Jackson, B. V., and H. R. Froehling1995, Astron. Astrophys., 299, 885–892.

“Tomographic inversion of coronagraph imagesZidowitz, S., B. Inhester, and A. Epple, 1996, inSolar Wind Eight, edited by D. Winterhalter, J. T.Gosling, S. R. Habbal, W. S. Kurth, and MNeugebauer, pp. 165–166, AIP Conf. Proc. 32Woodbury, N. Y.

7

STEREO Mission Concept November 14, 1997

APPENDIX I

Mission Requirements and Proof of Feasibility

I-1

STEREO Mission Concept November 14, 1997

I. Observatory Concept

A. OverviewThe uniqueness of the Solar STEREO mission concept lies simply in the geometry of theobservations and consequently the drifting heliocentric orbit requirement. The scienceinstrumentation utilizes proven, relatively straightforward designs. The spacecraft requiresmodest three-axis stabilization and a reasonably high performance communication system.It must be rather lightweight, radiation hard, and inexpensive. All of these requirements canbe met using current designs such as the newly developed SMEX•Lite architecture1.Though admittedly state-of-the-art today, this type of small spacecraft performance willundoubtedly be readily available by the 2003 timeframe in which STEREO will fly.Consequently, the SMEX•Lite architecture is cited in this study as an ample demonstration(a case study) of the ease in which this mission could be assembled.

B. Spacecraft BackgroundThe SMEX•Lite architecture is currently being developed by the SMEX Project atNASA/GSFC under the context of the NASA Explorer Program Technology InfusionProgram. This design will be built to protoflight standards, qualified for flight, andperformance demonstrated by early 1998. Prototype integration began in October 1997.This new spacecraft architecture has been optimized for versatility, ease of change, and lowcost. A three-axis stabilized version of this design is no more than one-foot tall, 38 inchesin diameter, and is anticipated to cost approximately $10M per mission to obtain. Thisstudy makes no presumption as to who or where the spacecraft are produced, but onlypresumes ready availability of this class of technology and the acceptance of aggressiveproject management and systems engineering techniques. Full cost accounting techniqueshave been utilized in estimating mission costs. The current development activity isproceeding very well, meeting nearly all of its cost and performance objectives. Thisprovides the confidence to cite this architecture as a proof of feasibility concept for theSTEREO mission concept definition study.

C. Mission OrbitThe STEREO mission requires two spacecraft to make their observations separated withinthe ecliptic plane by approximately 60 degrees from each other in their respectiveviewpoints of the Sun. This is an optimum viewing geometry for the selected scienceinstrumentation package, not an absolute geometry meaning that good science can beobtained from smaller as well as larger angles, but the best observations will occur as thisangle approaches approximately 60 degrees. In fact, there are reasons to argue that a varietyof viewing geometries will yield more information on CME structures and behavior than afixed viewing geometry. When you couple the science viewing requirements with therecognition that it takes a great deal of propulsive energy and resulting spacecraft weightand complexity to position a spacecraft in a fixed position in deep space such as a librationpoint, this study recommends a slowly drifting heliocentric orbit in which the spacecraftcan be directly inserted by the launch vehicle as the optimal low cost solution for thismission. This approach has the added benefit of only requiring two spacecraftconfigurations (or operating conditions) during the mission the launch configuration andthe science observation configuration. With no intermediate orbit transfer or parkingconfiguration and no propulsion system requirements to be met, the spacecraft designbecomes very straightforward.

I-2

STEREO Mission Concept November 14, 1997

The heliocentric orbit, which was chosen, has an energy requirement of approximately 0.78km2/s2. A C3=1.0 km2/s2 was used for launch vehicle performance analysis. One spacecraftwould be placed in a leading trajectory ahead of the Earth in its orbit and the other would beplaced in a lagging trajectory following the Earth in its orbit. This combination oftrajectories will yield a spacecraft separation angle of ~60 degrees that persists from aboutday 210 to day 460 of the mission (see Figure 1, Figure 2, Figure 3, and Figure 4).

Figure 1. Reference Orbit Selection

Figure 2. 9/8 Lagging Transfer C3=1.2

I-3

STEREO Mission Concept November 14, 1997

Figure 3. Ten-Day Time Ticks for 9/8 Trailing Transfer

Figure 4. Parameters for the STEREO Orbit C3=0.78

I-4

STEREO Mission Concept November 14, 1997

These trajectories are very straightforward to obtain and relatively insensitive to insertionerrors (see Figure 5).

Figure 5. Effects of Launch Time Variations on Heliocentric Transfer

The resulting orbital geometry has the added advantage of minimizing the distance fromthe Earth (simpler RF communications) and of putting all other Earth orbiting as well asground based solar observatories in a good position to provide collaborative data.

D. Launch Vehicle SelectionLaunch vehicle options ranging from the Pegasus to the DELTA were evaluated for thismission. Vehicle technical capabilities and cost were considered. Two options were studiedin detail launching both spacecraft together on a DELTA or launching each spacecraftseparately on a TAURUS.

The DELTA vehicle does not have the capability to provide separate 3rd stages to multiplepayloads. Consequently, at least one, if not both of the spacecraft would require their ownkick motors. The DELTA 7326 could lift a 600 kg combined payload mass into a transferorbit, leaving one spacecraft there, and subsequently boost the other into its heliocentricorbit (C3=1.0) using a STAR 37 upper stage. The other spacecraft would need to provideits own propulsive element to boost to the heliocentric orbit.

A single TAURUS vehicle could directly insert a single 350 kg payload into the requiredheliocentric (C3=1.0) orbit for approximately half the cost of the DELTA. The TAURUSconfiguration would utilize the 63 inch diameter fairing (54 inch useable payload diameter),the Star 37FM upper stage, and the standard 3712 payload adapter fixture/separationsystem (see Figure 6). The upper stage is spin stabilized (~50 rpm) and includes an integraldespin system. Launch would be from the Kennedy Space Center (KSC) range. It isassumed that all launch support facilities would already be in place. Two separate launcheswould be used for STEREO.

I-5

STEREO Mission Concept November 14, 1997

The TAURUS option was selected as the most practical for its simplicity in missionoperations, least impact on the spacecraft design, and for spreading the risk of catastrophicfailure by utilizing separate launches.

Star 37FMKick Motor

Figure 6. TAURUS Payload Envelope

I-6

STEREO Mission Concept November 14, 1997

E. Mission Lifetime and ReliabilityThe STEREO mission concept is optimized for the 60-degree separation angle scienceobservations between days 210 and 460 and the 110-degree separation between days 600and 800. Observations after this angle expands outward are considered an extendedmission opportunity. Any additional science data gathered in the extended mission isconsidered a bonus and will be relayed back to Earth at diminished data rates as thespacecraft gradually drifts further and further away from Earth. This approach eliminatesthe need for a propulsion system on-board the spacecraft. It also greatly simplifies the datasystem requirements, both on-board the spacecraft as well as on the ground. It yields adesign life of approximately 2 years implying that a single string observatory is areasonable design approach.

F. Spacecraft ConfigurationThe STEREO observatory was configured for maximum simplicity. The sunward facingplatform was balanced so as to minimize the separation of the Center of Pressure (CP) andCenter of Gravity (CG) in order to keep secular momentum build-up as small as possible.The spacecraft must be balanced in its launch configuration due to the spin stabilized upperstage. Balance mass has been allocated to simplify the implementation of this requirement.No attempt was made to provide a balanced torque couple configuration of the spacecraftthrusters since there are no stringent trajectory maintenance requirements. The sciencemagnetometer was placed in the spacecraft shadow in order to minimize thermal distortionof its boom. The instrument electric dipole antennas were placed to prevent interferencewith not only the sunpointed instruments, but also the high gain antenna and the startracker. The Heliosphere Imager is deployed on-orbit to a slightly outward and aft positionin order to clear its expansive FOV from shadowing by the high gain antenna.

The spacecraft assembly was decoupled as much as possible from the instrument modulein order to provide for the use of an accelerated development schedule (see Section VI) thatminimizes mission cost. The instrument module is a separate, fully integrated sub-assembly. The spacecraft components are housed in or attached to the one-piece integralspacecraft structure (see Figure 7, Figure 8, Figure 9, Figure 10).

I-7

STEREO Mission Concept November 14, 1997

Figure 7. STEREO Launch Configuration

Figure 8. STEREO On-Orbit Configuration Overview

I-8

STEREO Mission Concept November 14, 1997

Figure 9. STEREO On-Orbit Configuration Close-up

Figure 10. STEREO Configuration Details

EnergeticParticleDetector

SolarCoronalImagingPackage

Solar WindPlasma

Analyzer(2)

RadioBurst

Detector(3)

Booms

InstrumentModule

AuxiliaryInstrumentController

StarTracker

High GainAntenna

PulsePlasma

Thrusters(3)

SMEX•LiteSpacecraft

CMEInterplanetary

Imager

I-9

STEREO Mission Concept November 14, 1997

The internal architecture of the STEREO observatory (see Figure 11) utilizes the standardinterfaces and basic performance of the SMEX•Lite architecture2. Command and control,as well as housekeeping and low rate telemetry are managed on the MIL-STD-1553 DataBus. A single RSC6000 radiation hard, 32-bit microprocessor controls all observatoryfunctions. A single serial port is used to transfer the high data volume from the SCIPinstrument. The remaining instruments are interfaced by a single Auxiliary InstrumentController (AIC) to the spacecraft.

Computation Hub

3 Integrated Reaction

Wheel Assemblies

Communication Hub

TAURUS Telemetry

Test Port

X-Band Transponder

RSC 6000 Processor

Communications Interface

1 Gbit Bulk Memory

PCI

BACKPLANE

•Rad-Hard •Software Safehold

RS 422

Separation Loops

Sun Sensors

MIL-STD 1553 Data Bus

Thermistors

PPT

Utility Hub

12 Amp-Hour Commercial

NiCad Batteries

Modular Solar Array

(10 modules)

Power Node

28V Spacecraft Bus

28V Instrument Bus

(4 Amp-Hr packs)

X

Y

Z

High Speed Serial Ports

Solar Coronal Imaging Package

Instrument PCI Plug-In

3-Axis Gyro

Assembly

CCD Star

Tracker

STEREO Observatory

Block Diagram

11/14/97

DSN 34 Meter System

CCSDS

Auxiliary Instrument Controller

Magneto- meter

Energetic Particle Detector

Radio Burst

Tracker

Solar Wind Plasma Analyzer

CME Interplantary

Imager

Figure 11. STEREO Observatory Block Diagram

G. Attitude Control System (ACS)The STEREO mission requires a three-axis stabilized pointer that maintains the instrumentline of sight viewing the Sun with a ±30 arc-sec, 3-sigma accuracy. Jitter must be limitedto within ±5 arc-sec. The coronograph provides a pitch/yaw pointing error signal accurateto approximately 0.1 arc-sec that is used by the ACS for its solar reference. Roll about thesunline is unimportant to the science data collection, but is important for post flight dataanalysis. The solar roll angle must be known to 0.1 degrees. The spacecraft roll angle mustbe controlled such that the high gain antenna is pointed towards the Earth ±0.10 degrees.

All of these pointing requirements can easily be accommodated by an ACS configurationsimilar to that used on the NASA/SMEX Transition Region and Coronal Explorer(TRACE)3 mission. Three orthogonal reaction wheels are used to adjust the attitude of the

I-10

STEREO Mission Concept November 14, 1997

spacecraft and to provide a bias momentum vector oriented towards the Sun. This biasmomentum provides gyroscopic stability to the system, resisting disturbance torqueperturbations and providing short-term stability in the case of system upsets.

Three orthogonal Pulsed Plasma Thrusters (PPT) are utilized to manage secularmomentum build-up. These devices function as illustrated in Figure 12. They have beensuccessfully flown on the US Navy NOVA satellites. Further background data is availablein Reference 2. The PPT was chosen over the more traditional liquid or cold gas propulsionsystem because of its simplicity and small size. Once out of Earth orbit the spacecraftdisturbance torques are very small, dominated by solar pressure forces and weakgravitational pulls. The estimated disturbance torques were increased by a factor of 3 forthe purposes of this study. Designing the sunward face of the spacecraft to have a smallCG/CP displacement (≤1.0 in) further minimizes the effects of the solar pressure forces;however, the ACS design is not dependent upon this assumption. Increasing the CG/CPseparation to 1.0-foot approaches the practical limitation of the PPT, increasing the averagepower and Teflon propellant mass by a factor of twelve. However, it is rather straight-forward to balance the spacecraft projected area to within a few inches of the CG, so this isof little concern. This low torque environment is the ideal place to use PPT’s since thedevices themselves produce relatively small impulses. The PPT for STEREO were sizedusing the assumptions outlined in Table 1. The resulting calculations specified a verymodest PPT size that is well within the experience base of these devices. The NASAEO-1 spacecraft of the New Millennium Program will fly similar devices in 1999.

Figure 12. Pulsed Plasma Thruster System

I-11

STEREO Mission Concept November 14, 1997

Table 1. PPT Sizing Assumptions

Assumption CommentEnv

Solar radiation pressure 1.3 x 10–4 N Solar flux = 1358w/m2

ir

Solar radiation pressure Torque 3.3 x 10–4 N-m Spacecraft CP/CGoffset = 1.0 in

on

1 min. accumulated momentum 1.98 x 10–4 N-m-s

m Spacecraft surface area 6.0 m2 Sun faceent

Spacecraft surface reflectance 0.6 Sun face q

C h aT rh ar c 1.5 amp source 29.4 W 70% efficiencyu t Max. capacitor charge 29.4 Js e Time for max. charge 1.0 sect r Min. capacitor charge 5.0 Je i Time for min. charge 0.1701 secr s t i cC Pa e Required thruster impulse 3.97 x 10–4 N-sl r Required mass per firing 4.05 x 10–8 kg Isp = 1000 secondsc f Required thruster force 3.97 x 10–4 N at 1 Hz firingsu o Required capacitor charge 22.62 Jl r Required capacitor charge time 0.7694 sec at power levela m Average power over 1 hour 0.377 Wt a Firings over 2 years 1,051,200 0.0167 Hz frequencye n Teflon mass over 2 years 0.0426 kgd c Length of Teflon cylinder 0.0389 m 1 in. diameter e

Sun acquisition attitude signals are provided by a moderate Field of View (FOV) fine Sunsensor further aided by 2π steradian coverage coarse Sun sensors. Fine Sun pointing errorsignals come from the coronagraph.

Roll attitude information is provided by a star tracker oriented perpendicular to thespacecraft Sun axis, opposite the high gain antenna. The ecliptic plane is richly populatedwith stars, providing an easily recognizable roll reference throughout the mission. Theorientation of the Earth will be computed on-board utilizing this roll reference and anorbital ephemeris updated by mission controllers. The ACS will also compute the requiredhigh gain antenna pitch angle necessary to maintain an effective communication link to theEarth.

I-12

STEREO Mission Concept November 14, 1997

The ACS will also provide safe pointing at the start of the mission. Once released from thelaunch vehicle, the spacecraft will autonomously deploy the solar arrays and enable theACS to seek and point towards the Sun and orient the high gain antenna axis towards theEarth. This initial Sun acquisition utilizes all three reaction wheels and should beaccomplished in a few minutes time. The wheels must be sized to have sufficientmomentum storage capability to absorb the full tip-off momentum from launch vehicleseparation. Once stabilized in position the spacecraft will await ground command toinitialize the science instrumentation. This autonomous acquisition approach has beensuccessfully utilized on all SMEX missions.

Attitude control algorithms are executed within the spacecraft processor. The requiredsystem performance is well within prior SMEX mission capabilities and is not a driver forthis mission.

H. Data and Communication SystemThe STEREO data system must be capable of collecting one Gbit of science data per dayand relaying it to the ground. This can easily be handled with two hours per day of 150Kbps downlink transmission to the Deep Space Network (DSN) 34 meter ground systemduring the prime mission period (i.e., ~60 degree spacecraft separation angle). Later in themission the data rate must be reduced in order to maintain an adequate link margin (seeFigure 13) as the spacecraft-Earth separation distance increases. Reduced data volumemust be incorporated to balance the data flow. Downlink time can also be increased as longas a positive energy balance is maintained within the spacecraft power system. A thirdoption of maintaining data volume during the extended mission would be to switch to theDSN 70 meter ground system; however, system availability may make this impractical.The spacecraft will have 8 selectable data rates in order to support this flexibility.

-15.00

-10.00

-5.00

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

6.25E+061.25E+071.88E+072.50E+073.13E+073.75E+074.38E+075.00E+075.63E+076.25E+076.88E+077.50E+078.13E+078.75E+079.38E+071.00E+081.06E+081.13E+081.19E+081.25E+081.31E+081.38E+081.44E+081.50E+08Distance from Earth (km)

10k 25k

50k 100k

150k 200k

250k 500k

DSN 34m

DSN 70m

AssumptionsS/C EIRP 77dbm

Primary Science

Figure 13. STEREO Link Margins

A full day’s data set can be stored in a relatively modest solid state recorder. Spacecraftcommand loads can be received concurrent with the downlink.

In order to support these data flow requirements as well as to maintain the theme ofspacecraft simplicity, a 1.3 meter rigid antenna with 39 dB gain and a Field of View (FOV)

I-13

STEREO Mission Concept November 14, 1997

of ±0.25 degrees was chosen for the high gain downlink and beacon mode broadcast. Thisis the largest fixed antenna that could be easily packaged in the payload volume withoututilizing an elaborate deployable mechanism or antenna system. The antenna is stowedabove the spacecraft for launch and rotated into the operational position after separationfrom the launch vehicle. The spacecraft will roll in order to orient the antenna towards theSun-Earth line. The antenna will be oriented along the Sun-Earth line to point at the Earthby a small stepper motor that will pivot the antenna elevation axis relative to the spacecraftbody. The elevation angle (see Figure 14 and Figure 15) changes slowly throughout themission and consequently will only require periodic adjustment, rather than active pointing.The antenna elevation angle will therefore be controlled to ±0.1 degree. If the mission isextended much beyond two years the spacecraft will need to offset point from the Sun inorder to place the Earth within the antenna FOV since elevation angle adjustment is limited.Two low gain omni-directional antennas will provide coverage for health and safetycontingencies as well as initial Launch and Early Orbit (L&EO) activities.

Figure 14. Earth-Sun-Spacecraft Angle Variation (C3=2.01)

I-14

STEREO Mission Concept November 14, 1997

Figure 15. Typical Sun-Spacecraft-Earth Angle Variation (C3=2.01)

The data system will utilize a 20 watt X-band transponder for RF communications. X-bandallows for a smaller size antenna and avoids the frequency access issues which arebecoming more pronounced in the crowded S-band arena. DSN is already equipped tohandle X-band communications. This configuration meets the command and telemetryrequirements as well as satisfying the orbit tracking requirements for two-way Doppler andranging data. Orbit knowledge can easily be derived for this trajectory using this approachto better than ±200 km along track and ±100 km across track, more than sufficient forscience data analysis and ground station antenna pointing.

The data management functions as well as the ACS functions can be easily handled by thesingle spacecraft R6000 processor and interfaces integral to the SMEX•Lite architecture.The Command and Data Handling (C&DH), ACS and spacecraft health and safety flightsoftware requirements can easily be supported by existing SMEX software with anestimated reuse factor of 85%. This code is highly structured and utilizes C+/C++ highlevel language modules. The processor, running at 33 MHz, will only be ~15% utilized tosupport the spacecraft requirements of this mission.

I. Power SystemThe STEREO power system is only required to provide ~150 watts of 28 volt power tothe observatory. This requirement can be met with 1.08 square meters (12 SMEX•Litesolar array platelets) of GaAs solar array. A 12 Amp-hr Nickel Cadmium battery isincorporated to provide a power reserve for transmitter operation as well as to act as theprimary power source during L&EO prior to solar array deployment and Sun acquisition.

I-15

STEREO Mission Concept November 14, 1997

J. Power and Weight Budgets

Item Mass(kg)

AveragePower(watt)

Comments/Heritage

Primary Structure 24.0 – SMEX•LiteBalance Weights 10.0 –Solar Array 7.2 – SMEX•LiteBattery 12.0 – Sanyo D-cellTransponder 5.0 15.0 20 watt RF output

(daily average)Antenna Support Structure 11.0 –Harness 7.2 – SMEXHigh Gain Antenna 5.5 – TRMMLow Gain Antennas 0.8 – SMEXComputation Hub 5.5 18.5 SMEX•LiteUtility Hub 2.3 5.0 SMEX•LitePower Node 5.4 9.7 92% efficiencyReaction Wheels 12.0 12.0 SMEX•LitePulsed Plasma Thrusters 10.5 3.5 Factor of 3 power

marginSun Sensors 1.0 0.7 Adcole DSS, CSSGyros 5.0 12.0 SMEX IncosymStar Tracker 8.0 7.6 Lockheed ASTThermal 2.0 10.0 SMEXSpacecraft Subtotal 134.4 94.0Combined Emission-LineImager and Coronagraph

20.0 20.0 LASCO, EIT

Magnetometer 2.0 2.0 Existing versionsSolar Wind Analyzer 3.0 2.0 Existing versionsEnergetic Particle Detector 3.0 2.0 Existing versionsRadio Burst Telescope 5.0 2.0 Existing versionsHeliosphere Imager 15.0 20.0 New developmentAux. Instrument Controller 5.0 10.0Instrument Support Structure 10.0 –CME Structure DeploymentMechanisms

8.0

Instrument Harness 6.0Instrument Subtotal 69.0 58.0TOTAL 211.4 152.0

Mission Capability 350.0 156.0

I-16

STEREO Mission Concept November 14, 1997

II. Mission Operations Concept

The SMEX Program has been actively pursuing and promoting the concepts of automatedspacecraft operations. By combining the safing capabilities of the spacecraft and theadvancements in ground control software, the TRACE and Wide-Field Infrared Explorer(WIRE) missions are setting precedence for unattended spacecraft operations.

The location of the control center is somewhat arbitrary when considering theseadvancements in telecommunications. The SMEX control center is centrally located atGSFC with access to ground support resources as well as access to other solar physicsobservatories for coordinated science observations. Communications services between JetPropulsion Laboratory (JPL), the primary source of data, and GSFC are already in placeand would simply require bandwidth allocation. GSFC Flight Dynamics Facility (FDF)would provide tracking and navigation services which would allow STEREO to takeadvantage of the PC based FDF systems being developed for TRACE and WIRE. Thesesystems will process and distribute required FDF products.

The STEREO control center could easily be modeled after the TRACE and WIREsystems. The primary ground control software requirements can be met by the IntegratedTest and Operations System (ITOS). ITOS has 7 years of SMEX heritage and takesadvantage of cost effective automation features. The automation features allow forunattended health and safety monitoring, data processing, and anomaly notification.Anomaly notification is accomplished with automated configuration monitoring and limitchecking which activate pre-programmed paging systems. Spacecraft safing sequencesensure that all mission critical responses to anomalies are issued by the onboard computer.

Armed with the confidence that mission critical responses are satisfied and that thespacecraft will be in a stable condition at the next ground contact, the operations will bereduced primarily to mission planning and post-pass data evaluation. This will enableoperations support to be constrained to a standard 40 hour week. The size of the team willdepend primarily on the expertise of the staff and the spacecraft real-time requirements. Askilled team of four could reasonably support this mission as long as the instrumentoperations can be simplified to a consistent data acquisition mode. Operations of bothSTEREO satellites will be kept as identical as possible to minimize the operation to simplestation contact differences. A baseline of 1 Gbit per day at the primary science distance of.5 AU will require DSN station support for 2 hours @ 150 Kbps per day. This could besupported in multiple passes or one single pass. Additional pass coverage could bescheduled to increase science data volume. Much of this depends on station availability;however, either approach can be accommodated. The spacecraft will have selectable datarates to accommodate link margin restrictions. This will give the operations personnelconsiderable flexibility in mission planning. The DSN 34 meter system is the baseline forthe STEREO communication system; however, in times of conflict the 70 meter systemcan be used to dump science data faster and to conserve spacecraft power.

Command uploads will cover a one week period which once again will reduce operations.This upload should consist primarily of station contacts, science data dump commands andorbital ephemeris updates. The Command Management System (CMS), modeled after theTRACE system, will take ground station contact schedules and science command inputs to

I-17

STEREO Mission Concept November 14, 1997

produce spacecraft command loads. If routine operations are necessary these actions willbe done with the on-board Absolute Time Sequence (ATS) and Relative Time Sequences(RTS).

On-board science data analysis could identify data as a priority for near real-timetransmission and set off the “Beacon Mode” that would attempt to call for ground support.This mode would initiate a request for ground support by sending a low rate beacon toscheduled listening stations. Once the signal is detected the station would activate a pagingrequest to the operations personnel. The flight operations team would then schedule a DSNstation contact and issue the data collection commands. This concept does have drawbacksin that it would require constant ground monitoring. Ground stations with reasonableantenna gains of 52 dB (5.5 meter) could support a beacon mode surveillance concept.There are numerous commercial companies that are currently developing the groundstation infrastructure that would be needed.

I-18

STEREO Mission Concept November 14, 1997

III. Mission Cost Estimate

The STEREO mission cost estimate is based partially on actual SMEX mission historicaldata4, and partially on comparative evaluations of similar type instrumentation. SMEX-Litemission costs are derived from the ongoing technology development expenditure history.The technical maturity of most elements of this mission make this a realistic method toassess mission viability.

The STEREO mission cost estimate is based on an accelerated development cycle (seeFigure 16) beginning in October 1999 and ending with a launch of the first spacecraft inFebruary 2003, and the second spacecraft in April 2003.

FY00 FY01 FY02 FY03

Oct ‘99

Jun ‘00

Nov ‘00

Jun ‘02

Instrument Development (24 months)

Mission I&T

(6 mos)

Launch Ops 1 (2 mos)

L&EO (1 mo)

Feb ‘03

Launch 1

Spacecraft Development (12 months)

S/C Integ. (3 mos)

Spacecraft Environmental Test (2 months)

Spacecraft Performance

Test (1 month)

32 Month Mission

Development

18 Month Spacecraft

Development

Instrument Delivery

Apr ‘03

Launch 2

FY99

Oct ‘98

20 Month Instrument Definition

Period

Pre-Mission Instrument Maturation

(12 months)

Mission Confirmation

Review

L&EO (1 mo)

Reserve (2 mos)

Launch Ops 2 (2 mos)

Mission Design

(8 months)

Figure 16. Development Cycle

Postponing the spacecraft development is a valid technique for reducing system cost giventhe straightforward requirements for this mission. It is assumed that an existing spacecraftdesign could be incorporated with very minor mission unique modifications. The missiondesign period and the pre-mission development instrument definition period will focus ondefining the instrument configuration and interface specifications.

I-19

STEREO Mission Concept November 14, 1997

The STEREO mission total cost is summarized in Table 2 and Table 3.

Table 2. STEREO Mission Cost Summary (FY97 $M)

Phase A/B &Technology& Ground

System

Phase C/D Non-Flight

Instrument DefinitionMaturation 3.0Definition 5.0

Instrument Development 60.0Reserve (20%) 12.0Subtotal 72.0

Spacecraft Development 5.2 26.5Reserve (20%) 1.1 5.3Subtotal 6.3 31.8

Mission Integration 1.6 11.6Reserve (10%) 0.2 1.2Subtotal 1.8 12.8

Launch Segment 56.0Ground System 8.0

Reserve (20%) 2.0MO&DA 20.0Plan Subtotal 22.8 96.1 76.0Reserve Subtotal 3.3 20.5 0.0Subtotal 26.1 116.6 76.0GRAND TOTAL 218.7

Less Than$30M CostConstraint

Less Than$120M CostConstraint

I-20

STEREO Mission Concept November 14, 1997

Table 3. STEREO Mission NOA Phasing Estimate (FY97 $M)

Mission Segment(including reserve)

FY99 FY00 FY01 FY02 FY03 FY04 FY05

Instrument Definition 3.0 5.0

InstrumentDevelopment

14.4 26.4 21.6 9.6

SpacecraftDevelopment

6.3 17.2 12.7 1.9

Mission Integration 1.8 4.4 5.8 3.9

Launch Segment 23.0 24.6 8.4

Ground Segment 2.0 5.0 3.0

MO&DA 8.0 11.0 11.0

TOTAL 3.0 27.5 73.0 69.7 34.8 11.0 11.0

A. Launch Segment CostsThe TAURUS vehicle configured for STEREO is estimated by the NASA/GSFC OrbitalLaunch Services (OLS) Project to cost ~$28M. This includes payload support andintegration costs. Consequently the STEREO mission launch cost is $56M. The vehicle isordered 30 months prior to launch with a payment schedule as outlined in Table 4.

Table 4. Payment Schedule

MilestoneEvent No.

Amount(% of Price)

Estimated Date ofCompletion(Months)

1 10% L-302 5% L-273 8% L-244 9% L-215 9% L-186 9% L-157 12% L-128 12% L-99 11% L-610 5% L-311 5% Launch12 5% L+3

I-21

STEREO Mission Concept November 14, 1997

B. Instrument Costs

1. Instrument DefinitionInstrument definition costs include all instrument expenditures associated with Pre-PhaseA, Phase A, and Phase B activities leading up to the Mission Confirmation Review.

The projected instrument costs are summarized in Table 5, along with some indication ofthe heritage of the instrument design.

*All costs given are for a single instrument. The total mission instrument cost can be obtained bysimply doubling these amounts since two instruments are required.

The cost for the SCIP is based on the cost for comparable but more complex instrumentsfrom the TRACE and SOHO missions. TRACE is a much more sophisticated, much

Table 5. Summary of the Estimated Instrument Costs for the STEREOInstrument complement.

Instrument Heritage Estimated Cost ($M) forEach Instrument*

Solar CoronalImaging Package(SCIP)Coronagraph andemission-lineimager combined

Coronagraph is similarto the SPARTAN 201coronagraph with offsetocculter

Coronal and Chromo-spheric Imager is asimplified version of themulti-layer imaging tech-nology of TRACE andEIT

18

Energetic ParticleDetector

ACE 2

Radio Burst Tracker Similar to the instru-ments on Ulysses andWIND

2

Magnetometer Similar to instrument onGIOTTO, CLUSTER,Mars Global Surveyor

1

Solar Wind PlasmaAnalyzer

WIND 2

Heliosphere Imager SMEI (Air Force prog.)NEAR

5

Total Cost — 30

I-22

STEREO Mission Concept November 14, 1997

larger instrument than the SCIP with sub-arcsecond pointing, active focusing, multiplebandpass optics. Additional basis for this cost is obtained by comparing the SCIP with theSolar X-ray Telescope (SXT) on Yohkoh. But again, the instrument is much morechallenging than the SCIP with very expensive grazing incidence optics and a filter wheel.In addition, SXT had one of the first CCD cameras built for a solar physics mission, andsignificant development cost were incurred. Since that time, several groups have developedworking, inexpensive CCD cameras, which could be used for the SCIP.

2. Instrument Development and IntegrationInstrument development and integration costs include all costs incurred during Phase C/Dassociated with the instruments and the preparation of the Science Operations Center(SOC). This includes support to observatory Integration and Test (I&T) through launch +30 days.

C. Spacecraft CostsSpacecraft costs (see Table 6) begin with supporting the mission design. They include allcosts associated with spacecraft design, build, and qualification testing as an assembledsingle component up to the time of instrument integration. Spacecraft hardware is obtainedin a “protoflight” status, meaning that Engineering Test Unit (ETU) and spare units aregenerally not procured. This approach obviously presents some schedule risk, but isconsistent with the low cost approach that is recommended for the STEREO mission.Significant reserve is carried on the spacecraft, not because it is a risky development, butsimply because the aerospace industry has only infrequently demonstrated costperformance similar to that of the SMEX Program. Better performance is expected overthe next few years as the newer technology subsystems and interface standardizationsbecome the norm.

I-23

STEREO Mission Concept November 14, 1997

Table 6. Spacecraft Cost Itemization

FY00 FY01 FY02 FY03 Mission Total

Mission Element $M $M $M $M $M

Mechanical 0.111 0.810 0.280 0.105 1.306

Power Node 0.017 0.190 0.020 0.010 0.237

Battery 0.000 0.120 0.020 0.010 0.150

Solar Array 0.187 0.320 0.210 0.000 0.717

Harness 0.000 0.130 0.040 0.000 0.170

Thermal 0.153 0.160 0.150 0.040 0.503

Contamination 0.017 0.020 0.080 0.020 0.137

Utility Node 0.000 0.110 0.010 0.010 0.130

ACS Analysis 0.391 0.363 0.262 0.175 1.191

ACS Hardware 0.170 1.500 1.190 0.010 2.870

ACS Software 0.136 0.288 0.257 0.050 0.731

C&DH Software 0.204 0.400 0.300 0.070 0.974

C&DH Hardware 0.034 0.670 0.310 0.000 1.014

Communications 0.051 0.740 0.340 0.020 1.151

I&T GSE 0.357 0.440 0.340 0.200 1.337

EEE Parts 0.578 0.170 0.010 0.000 0.758

Spacecraft I&T 0.000 0.200 1.100 0.000 1.300

SUB-TOTAL 2.406 6.631 4.919 0.720 14.676

8% G&A 0.192 0.530 0.394 0.058 1.174

TOTAL $2.598 $7.161 $5.313 $0.778 $15.850

Phase B (Design)

Phase C/D (Development)

D. Mission Integration CostsMission integration costs (see Table 7) begin with managing the mission design activitiesand include all costs associated with integration of the instruments to the spacecraft and allsubsequent observatory testing and engineering support up through the initial 30 days onorbit. It also includes all project management costs including administrative, projectsupport, financial, flight assurance, Reliability and Quality Assurance (R&QA), inventorymanagement, preparation of the flight operations team for mission operations, missionsystems engineering, and launch operations support.

I-24

STEREO Mission Concept November 14, 1997

The STEREO mission integration costs are based on actual SMEX mission5 historicaldata. Though the STEREO mission requires the construction of identical observatories,the mission integration costs assume an activity level consistent with prior SMEXmissions that were unique and distinctly separate.

Table 7. Mission Integration Cost Itemization

FY01 FY02 FY03 FY04 Mission Total

Mission Element $ M $ M $ M $ M $ M

Scheduling 0.017 0.025 0.045 0.010 0.097

Configuration Management

0.068 0.100 0.100 0.030 0.298

Management, Systems Engineering and Administrative

0.306 0.430 0.510 0.410 1.656

Travel 0.020 0.048 0.059 0.150 0.277

Mission I&T 0.000 0.000 0.580 0.590 1.170

Field Operations 0.000 0.000 0.050 0.100 0.150

Quality Assurance 0.068 0.130 0.230 0.050 0.478

Misc. Flight Assurance 0.000 0.080 0.020 0.000 0.100

Reliability 0.000 0.040 0.080 0.030 0.150

EEE Parts Support 0.170 0.200 0.060 0.000 0.430

Flight Operations Preparation

0.085 0.540 0.460 0.130 1.215

Miscellaneous 0.000 0.050 0.050 0.000 0.100

SUB-TOTAL 0.734 1.643 2.244 1.500 6.121

8% G&A 0.059 0.131 0.180 0.120 0.490

TOTAL 0.793 1.774 2.424 1.620 6.611

Phase B (Design)

Phase C/D (Development)

I-25

STEREO Mission Concept November 14, 1997

E. Mission Operations and Data Analysis CostsThe STEREO MO&DA activities are very straightforward. Except for the routine use ofthe DSN 34 meter system, the mission operations activities are very similar to thoseplanned for the upcoming TRACE mission. These include not only the conduct of theSTEREO mission and analysis and distribution of its data products but also coordinationwith other orbiting and ground based solar observatories. For the purposes of this study theMission Operations and Data Analysis (MO&DA) requirements were assumed to beapproximately twice the scale of those of the TRACE mission. This is a very conservativeassumption.

I-26

STEREO Mission Concept November 14, 1997

IV. Study Conclusion and Recommendations

This study finds the STEREO mission to be very doable with today’s progressivetechnologies. The spacecraft and instruments can be developed with minimal technical risk.Cost and technical (power, weight, volume, and performance) margins are robust.

The STEREO mission is an excellent starter mission candidate for the new SEC initiative.The three-dimensional measurements of coronal structures will provide new and uniqueresearch tools to investigate the physics and evolution of the Sun. These will havetremendous public appeal, helping to communicate the excitement of this science tomembers of other scientific disciplines and to the public at large.

1 James G. Watzin. “SMEX•Lite–NASA’s Next Generation Small Explorer”. 10th Annual AIAA/USUConference on Small Satellites, 16-19 September 1996.2 Same as number 1.3 Darrell Zimbelman. “The Attitude Control System Design for the Transition Region and CoronalExplorer Mission”. 9th Annual AIAA/USU Conference on Small Satellites, 18-21 September 1995.4 Same as number 1.5 Same at number 1.

I-27

o f

s s

n

u

-

s

l

e

a

w

o

-s

tohece-

ur

1,dfteo-O/ad10 tom-

Othesywre

blyhh

e--

ay-ege by

Appendix II. Data Compression

The STEREO mission science objectives requtransmission of at least 250 images per day freach spacecraft. Since the spacecraft distancesEarth range up to ~1 AU during the prime scienphases, the capacity of the available resourcehandle the demand must be carefully examined.studied the issue using reasonable assumptionthe onboard computer, compression algorithmtransmitter power, high-gain antenna dimensioand use of the NASA DSN. The results of the stuare presented quantitatively in tabular form as unitimages received per DSN contact as a functionspacecraft distance from Earth and qualitatively asactual compressed SOHO/LASCO and images.

The onboard resource assumptions for the stwere as follows. The nominal imaging detector wassumed to be a 1024 3 1024 pixel format chargecoupled device (CCD) operated with 16-bit analoto-digital conversion. A single full-resolution image acquired with such a detector is treated as aimage. Other image rates with subframe imagelarger CCD pixel formats can be obtained by scing the tabular unit image rates as appropriate.

The onboard computer was assumed to be abhandle the computational compression demawithout significantly affecting the telemetry ratThis is a reasonable assumption based on eastudies. These had indicated that, when the spcraft are very near Earth, the telemetry rate ishigh that either the instruments themselves or onboard storage tends to limit the images per ctact, while at large spacecraft distances from Eatelemetry limits the rate. The compression algrithms assumed and used for the sample images Rice and the H-transform. Onboard storage wassumed to be adequate to hold up to 3 days of cpressed images.

II

irem

romce to

We fors,s,

dy

of

dyas

g--unit oral-

e tond.

rlierce-so

theon-rtho-ere

as-m-

A 50-W X-band transmitter coupled to a 1.3-mdiameter parabolic high-gain (38 dB) antenna waassumed for the study. The power level appearsbe practical for the anticipated launch date, and tantenna diameter is reasonable for the overall spacraft design.

Use of the DSN assumed a 34-m dish, an 8-hocontact, and a 6-dB margin, all standard values.

The results of the study are presented in Tablewhich shows the number of unit images transmitteper 8-hour DSN contact as a function of spacecralead (lag) angle from Earth and as a function of thapplied compression factor (CF). The CFs were chsen on the basis of actual experience with SOHLASCO. A CF of 1.0 indicates no compression; CF of 2.4 is typical of the compression achievewith the lossless Rice compression; and CFs of and 28 are for H-compression and were chosenmatch the compression applied to the sample copressed images shown in Figure 1.

The images shown in Figure 1 are SOHO/LASCC2 and C3 coronagraph images compressed with lossless Rice compression (CF = 2.4) and the losH-compression (CF = 10 and 30). The images shothat the lossy compression values cited in Table 1 aacceptable for the coronagraph images and probaalso for the heliosphere imager. Similar tests witSOHO/EIT images indicate that H-compression wita CF of 10 produces acceptable image quality.

We conclude that one 8-hour DSN contact per spaccraft every 3 days will provide ~200 to 600 stereographic image pairs per day around a 60° lead angleand ~100 to 300 stereographic image pairs per daround the 90° lead angle. This quantity of stereographic image pairs will be sufficient to meet thSTEREO science objectives. The onboard storarequirement for 800 images per day compresseda factor of 28 is ~1 Gbit.

-1

tor

Table 1. Unit images per DSN contact as a function of spacecraftlead angle and compression factor (CF).

Angle Telemetry(deg) (kbits/s) CF = 1 CF = 2.4 CF = 10 CF = 28

30 142 243 585 2437 682360 38 65 156 653 182790 19 32 78 326 913

Figure 1 Comparison of coronagraph images compressed by various factors. Even with a compression facof 30, it is difficult to detect any image degradation.

II-2

k

tro

y

l

t

ub

v

n-heedSva-s-are ofeksOlarursleld0elsld.-forlar

asicla-r

ow

eldn-r-

ag-ar

thetheedofer

(asy).-by-

flect

Appendix III. The Case forMagnetographs

Should the STEREO spacecraft carry a magngraph? If so, what type of magnetograph—vecor line-of-light (LOS)? In this appendix, we mathe case that it is important for STEREO to carrmagnetograph of some form on both spacecraf

Vector versus LOS Magnetographs

A simple, low-resolution longitudinal magnetograwould add immensely to the scientific return of STEREO mission, but is there a strong case focluding a vector magnetograph? It is a more cplex and expensive instrument and would take mtelemetry resources so would have a lower cadethan a longitudinal instrument.

A vector magnetograph would enable us to seetransverse component of the photospheric magnfields as well as measure the line-of-sight magnfield. It could help in understanding the role helicity in the processes that form magnetic strtures in the corona and their relative instability.

However, we need to investigate what sensitivitpossible on an instrument compatible with the smass, telemetry, and budget restrictions of the SREO mission. If it can only measure the transvecomponent in strong-field regions, i.e., active gions, it will not add greatly to the mission. Equaif the spatial resolution is too low, we run the dager of adding too much of the fine structure togeand ending up with a low-fidelity reconstructionthe coronal fields.

LOS magnetographs would provide valuable infmation on the general structure of the photosphfield, and having three views (including grounbased magnetographs) would provide some inmation on the transverse component, at least inplane of the ecliptic. If these data can be combiwith the 3-D data on coronal magnetic field strture from the coronal imager, it might be possito model the structure of the field from the phosphere to the corona. However, this could be a computationally challenging project, and how wconditioned the problem is would have to be stud

II

eto-torey at.

phhe in-m-orence

theeticeticofuc-

isize,TE-rsere-ly,n-herof

or-ericd-for- thenedc-le

to-eryellied.

Scientific Objectives of Magnetic FieldMeasurements from STEREO

Vital to our understanding of the propagation of iterplanetary disturbances is the ability to model tsolar wind flow. Such models currently are bason Carrington Rotation maps built up from LOmagnetograms taken from ground-based obsertories. The fidelity of the resulting models are supect because the determining characteristics partly global in nature. Consequently, some partsthe input data for the models are more than 3 weout of date. We know from Yohkoh and SOHobservations that even the quiet Sun during sominimum changes on timescales measured in hoand days rather than weeks. Having the multipviews of the Sun that STEREO spacecraft wouprovide, especially when they are more than 9°apart, none of the data that the solar wind modwere based on would be more than a few days oThis would provide significantly more reliable models of the solar wind and, hence, a firmer basis understanding the propagation of CMEs in the sowind.

These types of measurements impose some brequirements on the magnetograph. It can be retively low resolution (>5 arcsec) over the full soladisk, but it does need to have high accuracy at lfield strengths (≤ 100 G). The observations wouldbe required every ~6 hours.

The global measurement of the solar magnetic fiwould enable us, for the first time, to see its cotinuous evolution. From a single, Earth-bound pespective we only see about 30% of the Sun’s mnetic evolution due to foreshortening effects nethe limb, and we are blind to what happens on far side of the Sun. Hence, our understanding of emergence and evolution of magnetic fields is bason a statistical montage of the partial evolution many regions. We do not know, for example, wheththere is simultaneous global emergence of field implied by outbreaks of X-ray bright points seen bYohkoh and sympathetic flaring observed by SMMThis would imply a large-scale (global) interconnectivity of the field. We have been surprised the Yohkoh images showing how widely interconnected active regions are, but do these images re

I-1

nim

ge

oio

oe

ari

en

ilHdse

beuch

wies.isatps

fo-io-n-hatlarenue

he

udens.ges

nd

ithire-

a global phenomenon? STEREO, with a combition of a magnetograph and coronal imager, especin Phase 4, would be able to address this proble

The evolution of large-scale structures, althouslower than active regions, still cannot be followwith a sufficient temporal coverage to understaor predict their course. The appearance and evtion of coronal holes, particularly transequatorcoronal holes, is hard to understand without ctinuous coverage. They can last for several rotatioand some seem to be sheared by differential rtion while others do not. Are they a structure in thown right or are they formed and controlled by othglobal forces? Again, only STEREO, with a magntograph and coronal imager, can answer this qution by providing continuous spatial coverage of the

Polar plumes, polar-crown arcades, filaments, extended neutral lines are further examples of lascale structures that can only be effectively studusing STEREO, but it is vital that the two spaccraft carry magnetographs to address the problassociated with these globally related phenome

Magnetographs can also make helioseismolomeasurements, either in Doppler mode or by tunto the local continuum. Data taken from multipspacecraft and the Earth–Sun line (e.g., from SOor ground-based observatories) can be combinethat the entire Sun is in view; this will allow precifrequency measurements without crosstalk betw

III-

a-ally.

hd

ndlu-

aln-ns,ta-ir

ere-es-m.

ndge-ede-msa.

gyngeO so

een

the modes. In the low-l, global-mode regime, themode frequencies and rotational splittings that prothe energy-generating core may be measured mmore precisely than is currently possible.

If it were possible to obtain time series over a femonths, we could make advances in g-mode studAnother potentially interesting application of thtype of observation would be to obtain somewhhigher-resolution velocity measurements, perhain the range of l = 50–100. This is the direction osystematic, long-lived flows in and below the photsphere by the technique of time-distance helseismology. Observations with the MichelsoDoppler Imager on SOHO have demonstrated tthis approach is feasible but is limited by the sorotation, which removes any given portion of thSun in less than 2 weeks. By following a regioaround the Sun for a rotation or more, the techniqcould discover the long-sought giant cells in tconvection zone beneath the photosphere.

Conclusion

The baseline instrument complement does not incla magnetograph simply because of cost limitatioHowever, we have explored the potential advantaof stereoscopic magnetography, and we recommendthat studies of lightweight vector magnetographs aline-of-sight magnetographs be initiated. If a mag-netograph can be developed that is compatible wthe STEREO mission restrictions and science requments, it should be seriously considered.

2