The Solar and Heliospheric Observatory, SOHO, is an international cooperative project between ESA and NASA to study the Sun. This space-based Observatory is viewing and investigating the Sun from its deep core, through its outer atmosphere - the 'corona' - and the domain of the solar wind, out to a distance ten times beyond the Earth's orbit. The spacecraft provides a highly-stabilised platform for a complement of twelve sophisticated, state-of-the-art instruments, developed and furnished by twelve international consortia involving 39 institutes from fifteen countries (Belgium, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, Norway, Russia, Spain, Switzerland, the United Kingdom, and the United States).
Three helioseismology instruments are providing unique data for the study of the structure and dynamics of the solar interior, from the very deep core to the outermost layers of the convection zone. A set of five complementary remote-sensing instruments, consisting of extreme-ultraviolet (EUV), UV and visible-light imagers, spectrographs and coronagraphs, are currently giving us our first comprehensive view of the outer solar atmosphere and corona, leading to a better understanding of the enigmatic coronal heating and solar-wind acceleration processes. Finally, three experiments complement the remote- sensing observations by measuring the composition and energy of the solar wind and energetic particles at the spacecraft (Table 1).
Table 1. The SOHO scientific instruments* Investigation Principal Investigator GOLF Global Oscillations at Low Frequencies A. Gabriel, IAS, Orsay, France VIRGO Variability of Solar Irradiance and Gravity Oscillations C. Fröhlich, PMOD Davos, Switzerland MDI/SOI Michelson Doppler Imager/Solar Oscillations Investigation P. Scherrer, Stanford University, USA SUMER Solar Ultraviolet Measurements of Emitted Radiation K. Wilhelm, MPAe Lindau, Germany CDS Coronal Diagnostic Spectrometer R. Harrison, RAL, Chilton, UK EIT Extreme-Ultraviolet Imaging Telescope J.-P. Delaboudinière, IAS, Orsay, France UVCS UltraViolet Coronagraph Spectrometer J. Kohl, SAO, Cambridge, USA LASCO Large Angle Spectroscopic Coronagraph G. Brueckner, NRL, Washington, USA SWAN Solar Wind Anisotropies J.-L. Bertaux, SA, Verrières, France CELIAS Charge, Element and Isotope Analysis System P. Bochsler, Univ. Bern, Switzerland COSTEP Comprehensive Supra Thermal and Energetic Particle Analyser H. Kunow, Univ. Kiel, Germany ERNE Energetic and Relativistic Nuclei and Electron Experiment J. Torsti, Univ. Turku, Finland *IAS : Institut d'Astronomie SAO: SmithsonianAstrophysical Observatory PMOD: Physikalisch-Meteorologisches Observatorium Davos NRL: Naval Research Laboratory MPAe: Max-Planck-Institut für Aeronomie SA : Service d'Aeronomie RAL : Rutherford Appleton Laboratory
Just as seismology reveals the Earth's interior by studying earthquake waves, solar physicists are probing the solar interior by the use of helioseismology. Oscillations detectable at the Sun's visible surface are due to sound waves reverberating through its stellar interior. Using seismology techniques and wave measurements from SOHO's Michelson Doppler Imager (MDI), which records the vertical motion of the Sun's surface at a million different points once a minute, SOHO's investigators have already been able to generate the first maps of horizontal and vertical flow velocities, as well as sound- speed variations, in the convection zone just below the Sun's visible surface (Fig. 1). The convection zone lies directly beneath the photosphere, which forms the visible surface and effectively hides what is below. As a result, very little is known about the convection zone's internal structure, despite the fact that it is the source of sunspots, solar flares and most other forms of solar activity that affect the Earth.
Figure 1. A vertical cut through the outer 1% of the Sun, showing flows and temperature variations inferred by helioseismic tomography using measurements from SOHO's Michelson Doppler Imager (MDI). The arrows indicate directions and relative speeds of the vertical motions within the Sun. Colour shading indicates temperature changes.
MDI data have been used to calculate the time it takes for sound to travel between many different points on the solar surface. Because the paths of these sound waves loop down into the solar interior, one can use this information to map the temperature and flow patterns beneath the surface, similar to techniques used in computer-aided tomography to produce CAT scans.
Figure 1 provides a tantalising first view of how the convection zone is organised internally. For example, this map provides the first direct evidence for the depth of the features called 'granules', which cover the face of the Sun and are typically about 1500 km across. These granules are typically organised into larger domains called 'supergranules' that average about 25 000 km across. Theoretical calculations predicted that supergranule thicknesses should be between 25 and 30 percent of their width. The MDI mapping effort, however, suggests that they are shaped more like pancakes, with thicknesses only one-tenth of their width.
More significantly, the new map shows no evidence of the giant convection cells that had been predicted by a popular theory called the Global Circulation Model. It does, however, show evidence of narrow plumes of cooler gases streaming downwards towards the boundary with the radiative layer a feature consistent with the results of some numerical simulations of the region. Surprisingly, however, the plumes appear to originate from the middle of the supergranules, rather than at their edges as had been proposed. Additional observations at other times and locations are needed to determine whether the features that the map reveals are representative. Future observations will also allow the researchers to make a 'movie' of this part of the convection zone so that they can observe how its structure changes over time.
Random motions on the Sun, due to convection and other dynamic activity, lead to a continuum of frequencies (i.e. a noise spectrum) when the measured surface velocities are Fourier- analysed. This presumed solar background level imposes the eventual limit for the detection of weak g-mode oscillations.
It is possible to model this noise spectrum, by making assumptions about the physical properties of the various sizes of convection cells and their distribution. The result of a tentative model based upon the combined effects of granulation, mesogranulation, super-granulation and active regions, integrated over the complete solar disc, is given in the same format as the spectrum actually observed by SOHO's GOLF instrument, or with the ground-based networks IRIS (Installation d'un Reseau International de Sismologie Solaire) and BISON (Birmingham Solar Oscillations Network) which also measure global Sun velocity oscillations in the full-disc integrated light using a similar technique to GOLF (vapour resonance spectrophotometers).
Earlier comparisons of the noise model with ground-based data showed agreement to within a factor of 2 over the frequency range in which the so-called 'g-mode oscillations' (which, one hopes, will reveal the structure of the innermost part of the solar globe) are to be expected. This had been taken as confirmation of the model, as well as an indication that the Earth's atmospheric disturbances do not make a significant contribution to the observed data.
Now, for the first time, the frequency spectrum of surface velocity has been measured from space and these observations from SOHO have completely changed the picture. Figure 2 shows the Harvey model noise spectrum plotted together with preliminary results from GOLF and from Mark-I, a ground-based instrument in the BISON network at IAC Tenerife, regarded as one of the best of the ground-based observing stations. The data from GOLF and Mark-I are the result of an analysis made by Pere Pallé at the Instituto de Astrofisica de Canarias (IAC). The peak at 3 MHz is caused by the 5-minute p-mode (i.e. sound-wave-driven) oscillations. At lower frequencies, the signal level from GOLF is an order of magnitude lower than from Mark-I or the model. We conclude that atmospheric disturbances account for the major part of the noise received on the ground at these frequencies, and that the model greatly overestimates the solar background level. The consequences for SOHO are very positive: the strategy of making such observations from space is fully vindicated and the detection limit for low-level oscillations or g-modes is lower than anticipated, by a factor of ten or more.
Figure 2. Comparison of the background solar-noise model (as estimated by J. Harvey in 1988) with measured velocity power spectra from GOLF on board SOHO and the ground-based Mark-I instrument. Note the significantly reduced noise in the GOLF spectrum at low frequencies.
VIRGO's operation started with Sun Photometer (SPM) measurements of the spectral irradiance at 402, 500 and 862 nm and with the radiometers for total irradiance in mid-January 1996. The measurements with the Luminosity Oscillation Imager (LOI) started only at the end of March after the successful opening of its cover (which had earlier reverted to the closed position each time it was actuated with the open command!).
The power spectra calculated from the time series of the total and spectral observations are shown in Figure 3. Firstly, the 5-minute oscillations can be seen very clearly: on a widely expanded scale, the amplitudes of the individual p-modes have a signal-to-noise ratio never previously seen. In the range between 50 and 1000 µHz (0.5 5 h periods), which could not be observed before SOHO, the power spectra from the red and total to the blue flatten out at around 100 µHz to an increasingly stronger degree and, in the range below about 80 µHz, have equal power for all wavelengths. In the range in which the granulation is the dominant noise source, and where the p-mode oscillations are observed, the blue/green/red to total ratio is about 7:5:1 in terms of power. The g-modes are expected to be observed in the range of low solar power. Thus, what was said above in the context of the GOLF data is confirmed: the Sun seems to be quite cooperative in the search for g-modes. This is ultimately the main objective of VIRGO. At present, g-modes still seem to be hidden in the noise, but if they do indeed exist longer-duration observations will reveal them.
Figure 3. Power spectra from time series of total solar irradiance and of spectral irradiance at 402, 500 and 862 nm, obtained by VIRGO. The dominant noise sources, as well as the expected g-mode frequency range, are indicated.
At very low frequencies (corresponding to periods of days and weeks), the small activity features that have passed the visible disc have been observed with all of the VIRGO instruments. An example is a very small active region which passed the central meridian in mid-May. It could be followed by the LOI passing through its four pixels north of the equator, showing an increase in intensity close to the limb, a decrease near the central meridian, and again an increase before vanishing at the west limb. The simultaneous observations with the SPM and the radiometers are now allowing temporal and spatial variations to be disentangled, which in turn will greatly improve our understanding of solar irradiance variability. Although the passage of the spot around the central meridian decreased the total radiation by only about 0.02 percent, the whole effect can be investigated in detail. This demonstrates the high signal-to- noise ratio being achieved with the Sun in the quiet part of its activity cycle, but also, and perhaps mainly, due to the very quiet environment of SOHO.
Figure 4 is an image of the Sun's corona at about 1 500 000 K, taken on 13 March 1996 with SOHO's Extreme Ultraviolet Imaging Telescope (EIT). Every feature in this image traces magnetic-field structures. Because of the high quality of the instrument, we can see more subtle and detailed magnetic features than ever before. This is the first time we have been able to get such images except during five-minute rocket flights. SOHO's constant viewing allows movie loops to be made of the continuously changing, highly dynamic and complex outer atmosphere of the Sun.
Figure 4. Full-disc image of the Sun's corona at about 1 500 000 K, taken by the Extreme Ultraviolet Imaging Telescope (EIT) in the Fe XII emission line at 195 Å.
'Plumes' of outward-flowing hot gas in the Sun's atmosphere
may be one source of the solar wind of charged particles. Figure
5 shows (top) magnetic fields on the Sun's surface near the solar
south pole; an ultraviolet image (centre) of the
1 000 000 K plumes from the same region; and an ultra-violet image (bottom) of the 'quiet' solar atmosphere closer to the surface. The topmost image was taken by SOHO's Michelson Doppler Imager/ Solar Oscillations Investigation (MDI/SOI) instrument. The centre and bottom images were taken by EIT.
Figure 5. The south pole of the Sun: magnetic field (top), the 1 000 000 K corona as seen in the Fe IX emission line at 171 Å (middle), and the upper chromosphere as seen in the He II emission line at 304 Å (bottom).
These images represent the first opportunity scientists have had to see the detailed development over time of the polar areas, with their plume structures where the solar wind is accelerated. Again, SOHO's continuous viewing of the Sun from outside the Earth's atmosphere allows movies to be made, which will help us to understand the relationship between the magnetic field and the polar plumes in the wider context of the solar polar caps.
Figure 6 shows a sequence of images of the Sun in ultraviolet light taken by EIT on 11 February 1996. An 'eruptive prominence', or blob, of 60 000 K gas, over 130 000 km long, was ejected at a speed of more than 25 000 km/h. The gaseous blob is shown to the lower right in each image. These eruptions occur when a significant amount of cool dense plasma or ionised gas escapes from the normally closed, confining, low-level magnetic fields of the Sun's atmosphere to streak out into the interplanetary medium, or heliosphere. Eruptions of this sort can produce major disruptions in the near-Earth environment, affecting communications, navigation systems and even power grids. SOHO, with its comprehensive diagnostic instrumentation and its uninterrupted view of the Sun, can observe such events continually and is allowing us for the first time to acquire a better understanding of how such violent events occur.
Figure 6. Eruptive prominence recorded by EIT in the He II emission line at 304 Å (images, rotated here through 90°, taken approx. 20 minutes apart)
SUMER - Solar Ultraviolet Measurements of Emitted Radiation - observed its first light on 24 January 1996, and obtained a detailed spectrum in the wavelength range from 500 to 1490 Å: of a solar region near the north pole. Using the second detector of the instrument, this range was later extended to 1610 Å. Many more features and areas of the Sun have been observed since then, including coronal holes, polar plumes and active regions. Because of the technological advances employed in SUMER, we have been able to detect lines that are much fainter than previously observed. SUMER has already recorded over 2000 extreme- ultraviolet emission lines and many identifications have been made. The ions emitting this radiation persist at temperatures between 10 000 and 2 000 000 K and are thus ideally suited for investigations of the solar transition region where the increase occurs from chromospheric temperatures near 10 000 K to coronal conditions at several million Kelvin.
Among the many theoretical concepts of how the Sun heats its corona, reconnection processes of oppositely directed magnetic field lines play a prominent role. SUMER with its high spatial, spectral and temporal resolution is allowing us to investigate small-scale reconnection events. One example, in the transition region that lies between chromo-sphere and corona, has been observed near the Sun's centre in the O VI line (1031.91Å:), which is formed in plasma at temperatures of approximately 300 000 K.
Figure 7 shows a time series for the event in question with exposures every 6 seconds in the same format. Each spectrum strip depicts a 1 arcsec x 2 arcmin area on the Sun, corresponding to 700 x 77 000 km², which is about 1/100 000 th of its total surface area. The spectral range shown is 1.7 Å and the pronounced protrusions of the spectral-line images to shorter wavelengths, which developed with time, stems from Doppler- shifted radiation indicating a bulk velocity of the emitting ions of up to 120 km/s towards SOHO, which is superimposed on their thermal and turbulent velocities. The north south dimension of the event (along the slit) is more than 5000 km. As the axis of the protrusion is inclined with respect to the dispersion direction, it appears as if, in this particular case, the velocity vectors of the ions were at an angle to the line-of- sight. The distance of the apparent motion along the slit would be of the order of 4000 km. The growth phase for this event lasted about 30 s, which is quite typical and, consequently, a true (as opposed to projected) velocity of 130 km/s was deduced.
Figure 7. Temporal development of an explosive event as recorded by SUMER in the O VI line at 1031.91 Å. An exposure was taken every 6 s.
The Coronal Diagnostic Spectrometer (CDS) on board SOHO is a twin extreme-ultraviolet spectrometer that looks at the Sun in the wavelength range 150 - 780 Å. The Normal Incidence Spectrometer (NIS) provides stigmatic spectral images in two wavelength bands. The Grazing-Incidence Spectrometer (GIS) has a more complete and extended spectral coverage, particularly including the wavelength range 150 - 220 Å, which is particularly well-suited for plasma temperature and density diagnostics, yet has only been little explored so far. CDS is therefore being used to determine detailed properties of the solar atmosphere in order to understand its heating and other dynamic processes.
The CDS wavelength range covers lines and continua emitted
from the chromosphere at temperatures of approximately
20 000 K, to the hottest parts of active coronal loops at several million degrees. A large number of lines from a dozen chemical elements in several stages of ionisation are apparent. This allows simultaneous measurements of temperatures, densities, flows and elemental abundances within individual solar structures in the quiet Sun, coronal holes and active regions. By rastering the instrument slit, the spectrometer also builds up quasi- monochromatic images of solar features in a number of spectral lines. Each image contains detailed spectral information about the line and its immediate vicinity, and thus can be used to produce maps of total emission, velocities, plasma densities and temperatures over a wide temperature range.
By use of its normal- and grazing-incidence spectrometers, CDS has recorded spectral atlases for a variety of features: active regions, quiet Sun and coronal holes. These spectra represent a great improvement over earlier spectral mapping of the solar emission, particularly at the short end of the CDS spectral range. The reason for this is the combination of good angular and spectral resolutions, with a CDS spatial element size of 2 arcsec the spectral element in the range 0.07 - 0.2 Å, depending on the spectrometer type and spectral range. Thus, the short- wavelength region below 220 Å with many strong lines from different ionisation stages of iron, from Fe IX to Fe XIV, have provided good temperature information in a typical coronal plasma at 1 to 2 million Kelvin. A comparison of these spectra taken for different targets shows strong emission from Fe XIV lines in an active region. These lines become weaker in a quiet-Sun area, while the lower ionisation stages (Fe IX), representative of cooler plasma, completely dominate in the coronal hole. The evolution and structure of coronal holes, and their role in supplying the fast component of the solar wind, are currently being studied.
Images of an active region, recorded by the Normal Incidence Spectrometer (NIS) for two wavelengths and at different times on 22 March 1996, are displayed in Figure 8. It is clear that the emitting plasma is confined in magnetic loops. The two panels across the top of the image show the active region as observed in two ionisation states of magnesium, Mg IX and Mg X, emitting at temperatures only approximately 150 000 K apart around 1 000 000 K. The detailed loop structures are different and CDS is able to distinguish these differences on a much finer temperature scale than, for example, is possible from X-ray images taken by the Yohkoh satellite. The lower two panels show images from the same ions taken approximately one hour later: considerable evolution of the structures has taken place.
Figure 8. Images of an active region recorded by CDS with a one- hour interval in two ionisation states of magnesium, Mg IX and Mg X. The field-of-view is 4 x 4 arcmin² (4 arcmin corresponds to 1/8th of a solar diameter)
Analysis of line profiles recorded with the NIS has shown the presence of significant plasma flows in active-region loops. Up- flowing plasma reaching velocities of 100 km/s is seen in hot coronal lines like Fe XVI (335 Å). Other observations of an active region show down-flowing material with a velocity of 50 km/s. Similar flows are observed in lines emitted at transition- region temperatures, demonstrating that the transition region and corona are very dynamic in nature. From the present preliminary measurements it seems possible that CDS/NIS can measure relative line shifts corresponding to velocity differences as low as 20 km/s.
CDS has also observed its first strong high-velocity event (see upper panel of Fig. 8). This event, located in the leg of an active-region loop, was characterised by extremely wide emission lines corresponding to a velocity dispersion of approximately 300-450 km/s. The spatial extent is small, less than 4 arcsec. The event occurred in all lines from He I to Fe XVI, i.e. over a temperature range from 20 000 to 2 500 000 K. This is a new result which has not previously been reported. The fact that such events extend simultaneously over a wide temperature range is a challenge to theoretical models, and may cause a re-examination of the contribution from explosive events to coronal heating.
SOHO's Ultraviolet Coronagraph Spectrometer (UVCS) uses ultra- violet spectroscopy to obtain an empirical description of regions in the Sun's extended atmosphere, or corona, where the primary solar-wind acceleration takes place. This information is being used to address a broad range of scientific questions regarding the nature of the extended solar corona and the acceleration of the solar wind. UVCS has observed helmet streamers (Fig. 9), which are believed to be a source of normal-speed solar wind, and it has observed coronal holes, which are the known source of high-speed solar-wind streams. An understanding of the physical processes that control solar-wind acceleration will also contribute to our understanding of mass loss in other stars.
Figure 9. These images obtained by SOHO's Ultraviolet Coronagraph Spectrometer (UVCS) are the first of the extended corona in the ultraviolet. They are of atomic hydrogen (a) and highly charged oxygen (b), which flow out of the Sun along with other atomic particles to form the normal-speed solar wind. This material is shaped by the Sun's magnetic field into a giant nozzle called a 'helmet streamer', which extends over 3 000 000 km from the visible edge of the Sun. UVCS has determined that the particle velocities reach 100 km/s at the tips of these structures.
Since the start of its observations in late January of this year, UVCS has made the first ultraviolet images of the extended solar corona above two solar radii from the centre of the Sun. It has sensed the presence of a broad range of chemical elements in the extended corona, and it has actually measured the speed of coronal material as it is accelerated away from the Sun. UVCS has confirmed that protons and the more massive oxygen particles are hotter than the electrons in the outflowing coronal gas. This temperature difference may be the key to identifying the physical processes responsible for solar-wind acceleration and for controlling the composition and temperatures of solar-wind particles near the Earth. UVCS has made the first measurements of the speed of highly charged oxygen as it flows out of the tips of streamers (Fig. 9b) and has also made the first measurements of the supersonic outflow of highly charged oxygen from coronal holes. This information is being used to test theoretical explanations of how the solar wind is accelerated.
LASCO uses three coronagraphs to observe the outer solar atmosphere from near the solar limb to a distance of 21 000 000 km (i.e. about one seventh of the distance between the Sun and the Earth). Coronagraphs, which are special telescopes that can image the faint corona in the presence of the glaring light of the visible solar disc, can also observe the innermost corona from the ground. The extreme reach of LASCO can, however, only be achieved from space because of the scattering of sunlight in the Earth's atmosphere.
Coronal Mass Ejections (CMEs) are huge clouds of coronal plasma (10(exp 12) 10(exp 13) kg) ejected from the Sun at extremely high speeds (from several hundred to 2000 km/s). After acceleration at the Sun, they travel through interplanetary space and reach Earth in 2.5 to 5 days. When they reach the Earth, CMEs cause disturbances in the magnetosphere, which trigger auroras, make magnetic navigation at high latitudes difficult, and sometimes cause current spikes in high-voltage power lines, resulting in power outages and occasionally in the destruction of power equipment. They can also damage or destroy Earth- orbiting satellites.
Figure 10 shows a frame of a large CME as observed by the outer two LASCO coronagraphs (C-2 and C-3) on 3 February 1996. This event, because it was seen in the field of view of these two coronagraphs, could be followed from 1 100 000 km above the solar surface out to 15 000 000 km. Bright clouds are seen travelling outward in the equatorial plane with speeds ranging from 90 to 540 km/s, over both the east and west limbs of the Sun. The acceleration takes place over a distance of 15 000 000 km. The sectors above the solar limb in which the bright clouds are seen seem to extend over an angle of 120 deg in the equatorial plane. Although the instruments cannot see material moving towards or away from the Earth Sun line, it is safe to assume that the CME extended all the way around the Sun. A faint event above the Sun's south pole is also apparent in this picture.
Figure 10. Global coronal disturbances as seen by the LASCO C2 and C3 coronagraphs: instead of a single coronal mass ejection erupting in one direction, movies made with the LASCO coronagraph confirm that material is simultaneously ejected both eastwards (left) and westwards (right) from the Sun. The eastern ejecta are brighter in this particular case. The combined field of view of the images spans 25 solar radii. The inner edge is at 1.6 solar radii. The dark circle marks the boundary between the C2 and C3 fields of view. Bright areas are coronal clouds moving outwards. Dark areas resemble the blown-out streamer belt. At the inner edge (420 000 km above the solar surface), velocities are typically 90 km/s; at 23 solar radii (15.7 million km above surface), they are 530 km/s.
A thin magnetic current sheet forms around the Sun during a minimum in the solar cycle. The acceleration of the global coronal disturbances seems to occur in this sheet. The upper and lower boundary layers of this sheet have opposite magnetic polarities. Consequently, the two boundary layers attract and form a 'lid', thereby trapping hot coronal material in the corona's outer layers. The amount of material stored varies greatly with time. The upper panel of Figure 11 shows the current sheet during a quiet period, the lower panel the same area two days prior to the 3 February CME. The obvious increase in brightness prior to the CME indicates that more material is stored before a global coronal disturbance than during quiet coronal periods. It is assumed that the current sheet will blow open when the pressure inside exceeds a certain value. An instability will then accelerate and release the stored hot coronal material as a global coronal disturbance.
Figure 11. Streamer belt seen in the field of view of the LASCO C2 coronagraph. Upper panel: during a quiet period. Lower panel: during an active period prior to a coronal mass ejection. The field of view covers 1.6 to 6 solar radii (420 000 km above the solar limb to 3.6 million km).
Figure 12 shows the inner corona as seen by the innermost LASCO coronagraph, C-1, in the light of the green, so-called 'forbidden' (i.e. inherently very weak) coronal line of Fe XIV. Coronal structures can be seen as high as 1 000 000 km above the solar surface. The large-scale solar magnetic field is being traced by loop systems, which are forming all around the Sun in different latitude zones, as demonstrated by the appearance of the corona above both the east and west limbs. Three loop systems can be seen from high northern to high southern latitudes, bridging the solar equator. This magnetic configuration is known as a 'magnetic quadrupole', because it has four magnetic zones, each zone bordering another of opposite polarity in an inherently unstable configuration. This picture was taken two days before the coronal mass ejection shown in Figure 10 was observed.
Figure 12. The inner corona as seen by the LASCO C1 coronagraph in the light of the green forbidden coronal line of Fe XIV. Coronal structures can be seen as high as 1 000 000 km above the solar surface.
The solar wind is a hot gas of electrically charged atomic particles that streams out of the Sun at hundreds of kilometres per second and fills the surrounding space, i.e. the heliosphere, and is therefore also present throughout the Solar System. The solar wind causes the aurorae that occur sporadically at middle and high latitudes over both hemispheres of the Earth in the form of luminous bands, it causes comets to have tails, and it sometimes induces changes in the Earth's magnetic fields that can also disrupt communications and cause power-grid failures. Changes in the solar wind may also affect the properties of the Earth's lower atmosphere, including weather and climate.
The Charge, Element and Isotope Analysis System (CELIAS) investigation on SOHO is a multi-sensor experiment consisting of three detectors that measure the composition and energy spectra of plasma (solar wind) and energetic ions of solar, interplanetary, and interstellar origin. A fourth sensor monitors the absolute EUV (extreme ultraviolet) flux from the Sun. The in- situ particle measurements, when combined with optical coronal measurements and modelling efforts, yield information on the processes by which matter from the underlying solar atmosphere is fed into the solar corona. These processes lead to enrichments/depletions and other variations in elemental abundances, ionic charges and isotopic heavy-ion abundances in the corona and solar wind and solar energetic particle events which, in turn, provide clues as to the acceleration and heating mechanisms of solar-wind particles in the inner corona, as well as the temperature and density gradients in the latter. Other studies include the dynamics of pick-up ions (involving both particle and EUV measurements), and particle sources and acceleration in so-called 'Corotating Interaction Regions' (CIRs), which are located far out in the heliosphere.
The CELIAS solar-wind mass spectrometer (MTOF = Mass Time-of- Flight sensor) possesses unprecedented mass resolution for solar- wind composition studies, and has already measured rare elements and isotopes that were previously not resolvable from more abundant neighbouring species, or were not previously observable at all. For example, as can be seen in Figure 13, the elements of sulphur, argon and calcium are now easily distinguished from their neighbouring species silicon and iron, as is nitrogen from carbon and oxygen. The rare elements phosphorus, chlorine, potassium, titanium, chromium and nickel are being measured in the solar wind for the first time.
Figure 13. Element and isotope spectrum as obtained with the MTOF sensor of the CELIAS experiment. Many elements such as phosphorus, chlorine, potassium, titanium, chromium and nickel are being measured in the solar wind for the first time.
The determination of the elemental abundances of these rarer species allows us to fill in the 'blanks' in the tables comparing the abundances in the solar wind with those in the photosphere, i.e. on the solar surface. This is important for better analysis of the processes that feed and accelerate the solar wind in the chromosphere and inner corona. The solar-wind and coronal abundances indicate an ordering of relative abundance enhancement (or depletion) to photospheric values partially correlated with the First Ionisation Potential (FIP) of the element - the so- called 'FIP effect'. Since these newly observed elements have different properties (such as first ionisation potentials and times, charge-state equilibrium times, atomic mass, etc.), knowledge of their relative abundances serves as a further diagnostic tool. They help in determining conditions in the chromosphere/transition region, where ions which eventually become the solar wind are separated from neutrals.
The FIP effect is not the same for all types of solar wind. The temporal resolution of CELIAS means that abundance variations in different types of solar wind (e.g. coronal-hole-associated versus slow solar wind) may be better traced to the varying conditions in the solar-wind source regions.
The MTOF sensor - the first of its kind to be mounted on a Sun-pointing (rather than spinning) spacecraft - is routinely measuring isotopic abundance variations for several elements (neon, magnesium, silicon, sulphur, argon, calcium, iron and nickel), some of which have not previously been observed either in the solar wind, in solar energetic particles, or spectroscopically. Isotopes are also being measured with a much finer temporal resolution than previously available (of the order of minutes/hours instead of months/years).
Matter in the corona and solar wind is derived from the Outer Convective Zone (OCZ) of the Sun and isotopic abundances of the less volatile elements in the solar atmosphere are probably very similar to terrestrial, lunar and meteoritic abundances. For such elements, it is possible to infer the amount of isotopic fractionation under varying conditions in the solar-wind source region. For many species, the solar wind provides the only source of information on the isotopic composition of the OCZ. This is important for many cosmo-chemical and astrophysical applications: knowledge about the OCZ's isotopic composition will yield information on the early solar nebula and the history of the Solar System.
The COSTEP instruments measure electrons from 45 keV to 10 MeV and hydrogen and helium nuclei from 45 keV to 53 MeV. Given the current phase in the solar-activity cycle, which is close to solar minimum, the Sun was extremely quiet during the first phase of the mission. Figure 14 shows, from the top, counting rates for electrons (200 700 keV, EPHIN), protons (80 150 keV, LION), pro- tons (4.3 7.8 MeV, EPHIN) and nuclei (53 MeV/n, EPHIN) from switch-on until 25 April 1996. Only three small solar events were observed, on 11 and 24 December 1995 and on 22 April 1996.
Figure 14. Count rates of electrons, protons and nuclei as measured by COSTEP.
In addition to upstream events when SOHO is connected via interplanetary field lines to the Earth's bow shock (seen in the second panel of Fig. 14), two different types of recurrent events are being observed by COSTEP:
Of the solar-wind particle detectors on SOHO, ERNE covers the highest energies, from roughly 1 to 500 MeV/n. At these energies, the present solar-activity minimum enables ERNE to collect particles originating outside our Solar System, namely the galactic cosmic radiation. During the next five years, solar activity will increase and, accordingly, the particles collected will increasingly be of solar origin, coming from the flares and coronal mass ejections, for which the instrument was designed.
Since the activation of ERNE on 15 December 1995, only one small particle event of solar origin has been detected, lasting from 20 to 25 January 1996. During this event, the number of protons counted rose about tenfold over the background. Initial analysis indicates that the high-energy protons detected originated in the shock front arising from a fast solar wind running into a slower part of the wind, with the collision region located several Sun Earth radii in the heliosphere. The ERNE LED (Low-Energy Detector) is particularly well suited for the detection of such particles. Other possible solar acceleration mechanisms include the strong electric fields present in solar eruptions (impulsive acceleration, lasting perhaps an hour), and solar coronal mass ejections driving shock fronts that accelerate particles continuously as the shock approaches and passes Earth (lasting several days).
The galactic cosmic radiation that is currently observed consists of particles originating in the Milky Way, including atomic nuclei heavier than helium possibly coming from super- novae. Due to transport effects, the galactic cosmic rays are characterised by their rising spectrum. For this radiation, the ERNE HED (High-Energy Detector) has succeeded in identifying significant amounts of hydrogen, helium, boron, carbon, nitrogen, oxygen, neon, magnesium, silicon and iron.
An example of composition and energy resolution is shown in Figure 15. The two horizontal axes describe energy deposited by the incoming particle in two detector layers. The vertical axis gives the observed count rate. Carbon (the highest peak), nitrogen and oxygen are seen in this picture. To the right of the carbon ridge, some boron can be found, as well as some neon to the left of oxygen. In addition, anomalous cosmic-ray components (i.e. fake cosmic rays originating within the heliosphere) have been measured with ERNE LED for at least helium, nitrogen and oxygen.
Figure 15. Composition of the galactic cosmic radiation as measured by ERNE. One can clearly identify carbon, nitrogen and oxygen. Closer inspection of the data has helped to identify other elements such as hydrogen, helium, boron, neon, magnesium, silicon and iron.
One instrument on board SOHO avoids looking at the Sun because it would be dazzled. Instead, SWAN surveys the sky all around and sees an ultraviolet glow from (neutral) hydrogen atoms lit by the Sun (Fig. 16). These atoms come on a breeze from the stars that blows through the Solar System. However, the competing wind of charged particles from the Sun breaks up (ionises) the incoming atoms so that they can no longer emit at their characteristic wavelength. The result is a hole in the pattern of emissions downstream from the Sun. This allows us to determine the strength of the solar wind in different directions.
Figure 16. Full-sky Lyman-alpha map in ecliptic coordinates as recorded by the SWAN instrument. Note the asymmetry between the northern and southern hemispheres, which has also been detected by Ulysses. The U-shaped yellow band is the Milky Way.
The Earth is also visible in the maps, because a cloud of hydrogen gas called the 'geocorona' envelops it and glows in the ultraviolet. In fact, the goecorona would hamper observations of the interstellar glow by satellites close to the Earth. SOHO is thus able to monitor the three-dimensional structure of the solar wind. It has also confirmed the north south asymmetry of the solar wind that had already been observed by the Ulysses spacecraft.
With the present quiet state of the Sun, the SWAN sky maps clearly indicate a situation of increased solar wind around the Sun's equator in the ecliptic plane (Fig. 16). It will be interesting to see what happens when the Sun becomes more active. We will probably see important changes in the solar wind's impact on the interstellar gas revealed by changes in the sky maps. Meanwhile, SWAN uses alternate days for special investigations, such as observations of Comet Hyakutake which has recently swung- by the Sun.
During the first couple of months, after the commissioning of both the spacecraft and its experiments, all of the instruments have performed observations and measurements that demonstrate that SOHO is fully qualified to achieve the goals for which it was designed. In-depth analysis of the data is just starting, but quick analyses of the early measurements have already provided new observations in all the fields addressed by the SOHO instruments. The following are of particular relevance:
In the field of helioseismology:
In the solar atmosphere:
In the solar wind: