1. INTRODUCTION

Many activities of the sun that affect the terrestrial environment and human society lie in the solar corona. Such activities include the solar wind, solar flares, and coronal mass ejections (CMEs). Tousey and Friedman at the Naval Research Laboratory (NRL) made the initial discoveries of ultraviolet and x-ray emission from the solar corona shortly after World War 2 with the aid of captured German V2 rockets [1]. In 1958 Friedman demonstrated that solar x-ray emission is extended beyond the visible solar disk and that the solar corona is structured across the disk. This was achieved through observation during a solar eclipse with a series of Nike-Asp flights carrying nonimaging x-ray sensors. This use of the moon as an optical element from a space platform was reprised in 1964 with Freidman’s experiment to measure the angular size of the x-ray source in the Crab nebula.

Koutchmy [2] gives an account of the early history of the development of the technique of white-light coronagraphy by groups at NRL and other institutions where outside of solar eclipses, an artificial occulter is used to block out light from the solar disk allowing the outer corona to be studied. Coronal mass ejections were discovered from space in 1971. NRL instruments on Skylab in 1973/74 revolutionized our view of the solar upper atmosphere, and the SolWind Coronagraph observed the first Earth directed halo coronal mass ejection (CME) in 1979. Our view of not just the solar atmosphere but the extended corona was further transformed with the launch of the Solar and Heliospheric Observatory [3] (SoHO) in 1995, and the Solar Terrestrial Relations Observatory [4] (STEREO) in 2006. Instruments on SoHO studied CME eruptions from about 1.5–30 solar radii (${\mathrm{R}}_{\odot }$), while STEREO provided a 3D view of these events from $1.5{\mathrm{R}}_{\odot }$ to the orbit of Earth.

This paper describes new instrumentation designs required to attack two specific problems arising out of the SoHO observations: the role of wave-particle interaction in the acceleration of the solar wind and the initiation of solar energetic particle (SEP) events by shock waves driven by CMEs. Section 2 describes observations of the solar corona in a broader context with the ultimate goal of extending the initial SoHO results on solar wind acceleration. Section 3 describes the particular problem of particle injection into the CME shock acceleration process while Section 4 describes the observables that should be associated with it. Finally, in Section 5 we discuss the instrumentation required both for a scientific validation of our hypothesis and for an experiment designed to yield real-time monitoring.

2. SPECTROSCOPY OF THE CORONA AND SOLAR WIND SOURCES

The extended solar corona far off the solar limb is the site where the solar wind is accelerated and coronal plasma is nonthermally heated to millions of degrees. It is also the region where propagating CME shocks are formed (e.g., see [5]) and subsequently produce SEPs [6,7]. Observations of the extended corona are thus essential for understanding the formation and evolution of these activities. Especially valuable are observations by spectroscopic means. The very same technique has long been used to understand our universe. The power of spectroscopy is that it allows us to probe the local physical state of the corona and other remote locations in the universe where direct in situ sampling is still not possible.

The solar corona presents dozens of ultraviolet and extreme ultraviolet (UV/EUV) emission lines that allow the plasma properties (e.g., densities, temperatures, outflow velocities, and abundances) in the corona to be determined from remote sensing diagnostics. In fact, NRL’s Space Science Division has a rich history of solar spectroscopy of the disk and limb (see Doscheck et al., this issue). However, to make such measurements at heights beyond about $1.5{\mathrm{R}}_{\odot }$ (from sun-center) requires that the solar disk be occulted in order to keep the intense brightness of the disk from overwhelming the faint UV and EUV emissions in the extended corona.

The observational problem is shown in Fig. 1. It shows coronal-hole intensities, relative to the disk, for a few of the brightest emission lines observable by the ultraviolet coronagraph spectrometer (UVCS) on SoHO [8] (plus He II 30.4 nm, which is not observed with UVCS). The He II intensity for all heights and the intensities of the weaker lines (Mg X and Si XII) above $2.5{\mathrm{R}}_{\odot }$ are extrapolated based on the density variation with height. (Intensities for coronal streamer and CMEs can be 3–10 times higher.) As can be seen in the figure, radiation from the disk is orders of magnitude greater than the emission lines we wish to observe. The solution is to use a combined spectrograph and coronagraph [9] to make the required observations of emission lines far off-limb in the extended corona. As described in Section 5, the coronagraph reduces the disk radiation to acceptable levels with occultation, while the spectrograph is used to reject off-band radiation from both nearby lines and scattered disk radiation within the instrument. The selection of optical coatings also plays a role in the rejection of off-band radiation.

 

Fig. 1. Coronal hole intensities, relative to the disk, for a few bright UV/EUV lines. The relative intensities of all lines, except He II, are based on UVCS observations.

Download Full Size | PDF

The UVCS mission on the SoHO was the first to use these techniques to make groundbreaking advances in our understanding how the fast solar wind is accelerated. Cranmer et al. [10] present a summary of these and other physical insights that were gained from the UVCS observations. The most important finding from UVCS has been the discovery of broadened emission line profiles of O VI (the spectrum of five-times ionized oxygen). These broad line widths suggest that the ${\mathrm{O}}^{5+}$ ions are heated by transverse Alfvén waves, which are in resonance at the ion cyclotron frequency for the ion. While there is no direct detection of the waves themselves, over the years a number of authors (e.g., [11] and references therein) have developed models to show how these waves are generated by motions at the coronal base. Another key UVCS observation that supports the presence of Alfvén waves is the fact that the oxygen ions in fast solar wind appear to flow out of coronal holes faster than the protons above $\sim 2.5{\mathrm{R}}_{\odot }$ [12]. This difference in speed is preserved in the distant solar wind as measured, for example, by the Advanced Composition Explorer at 1 AU, which shows that, in general, ions have outflow speeds above the protons by an amount on the order of the local Alfvén speed [13].

In addition to the investigations of the fast solar wind, more recent work involves understanding the sources of the slow-speed wind. The energy requirements are comparable for the slow wind but the large variability in many parameters, including wind speed [14], density, elemental abundances, and the first ionization potential (FIP) effect (see below), makes this problem difficult. Theoretical ideas involving the magnetic field foot-point interchange reconnection [15], expansion factor (e.g., [16]), and unique field-line topologies (e.g., S-Web, [17]) have been proposed to explain various properties of the slow wind. A surprising outcome of this work is that it is now becoming clear that the wind speed alone is no longer sufficient to identify slow solar wind from fast wind. Elemental and charge state abundances may be a better discriminant for determining the type of solar wind and its coronal source regions [18]. Off-limb coronal abundance determinations (e.g., [19–22]) have proved useful in identifying sources of slow wind in streamers and other localized regions of the corona. Elemental abundance measurements on coronal loops have provided constraints on ion-neutral fractionation processes occurring at the loop chromospheric foot points that point to the important role of magnetohydrodynamics (MHD) waves [23]. This leads to the FIP effect, where elements with low FIP that are ionized in the chromosphere (e.g., Fe, Si, and Mg) are enhanced in abundance in the corona with respect to high FIP elements (e.g., O, Ne, and Ar) that remain neutral. New observations with a dedicated coronal spectroscopy mission are needed to determine the properties of these waves and to understand their role in FIP fractionation and the solar-wind acceleration processes.

Doppler shift and linewidth measurements have contributed to our understanding of the 3D dynamic structure of CMEs and the CME shock properties, both of which can only be obtained unambiguously from spectroscopic means. Doppler shifts, when combined with the plane of sky velocities, have revealed helical motions of CME plasmas, which provide information on the 3D magnetic-field structure. Spectroscopic determinations of ion temperatures from linewidths and charge states have also provided valuable tools for understanding CMEs. Studies of CME energy budgets [24] show that additional heating beyond the kinetic energy of the mass is required to explain the observed spectral lines. These CME studies have relevance beyond providing a better understanding of physical processes in the solar atmosphere. They are also important because CMEs are a major driver of space weather [25–27], which has the potential of causing severe damage to space-based systems, power grids, and oil pipelines.

While SoHO UVCS provided critical clues about the processes that take place in the corona, it was limited by the effective area of its light gathering optics. As an example of the limitations, significant nonthermal heating above $\sim 1.5{\mathrm{R}}_{\odot }$ in coronal holes could only be determined for two heavy ions, ${\mathrm{O}}^{5+}$ and ${\mathrm{Mg}}^{9+}$. In order to study the charge-to-mass dependence of the nonthermal heating, the measurements of line profiles for many more ions are desired. For another example, UVCS was not able to determine if there exists a supra-thermal population of protons or other ions in the corona. The presence of sufficient quantity of these particles could provide the necessary seed particles for shock acceleration in the corona. (See [28] and Sections 3 and 4 for more details.) Direct measurement of the usually faint CME shock front and its interaction with the ambient corona would also benefit from enhanced light gathering power of a more advanced instrument.

3. INITIATION OF SEP EVENTS

SEPs are high-energy charged particles produced in eruptive processes such as CMEs and flares low in the solar corona. Their energy reaches from a few keV to several GeV. SEPs streaming to Earth can damage satellites, disrupt radio communication, and global satellite positioning. Gradual SEPs are thought to be accelerated primarily by the CME-driven shocks (a schematic picture is shown in Fig. 2.), as there is a correlation between peak proton intensity of an SEP event at 1 AU and associated CME speed. However, for any given CME speed, peak intensities may vary by three to four orders of magnitude. This indicates that there are other important physical factors affecting this variability beside CME shock speed.

 

Fig. 2. Schematic picture representing acceleration of SEPs at a CME shock.

Download Full Size | PDF

Several recent observations of SEPs supported by theoretical studies presented clear evidence for existence of supra-thermal seed particles in the corona as a prerequisite for intense SEP events. Kahler et al. [29] considered various factors other than CME speed that may be important for production of SEPs such as CME width and location, its origin in the corona, and ambient SEP intensity. They analyzed 17 fast CME events in a relatively narrow speed range of 650–850 km/s with associated SEP intensities spanning more than four orders of magnitude. They concluded that the background intensity of the SEPs prior to the CME eruption is the most significant factor for the CME-associated SEP intensity. Other factors were shown not to be important. Gopalswamy et al. [30] found that the high-intensity SEP events occurred whenever a CME is preceded by another wide CME from the same active region. Their interpretation was that a preceding CME disturbs the interplanetary medium and may produce seed particles for the acceleration at the following CME shock. Cliver [31] considered two exceptional ground-level enhancement (GLE) events associated with relatively weak flares and slow CMEs as opposed to GLEs normally produced in intense flares and fast CMEs. He suggested that enhanced-background proton-intensity enabled the acceleration of protons to GeV energies in these weak solar events.

Element abundance variations in SEPs have also been interpreted in terms of seed particles produced as supra-thermal ions in impulsive flares [32,33]. In large SEP events, at the highest energies, the abundance ratios Fe/O and sometimes ${}^3\mathrm{He}/{}^4\mathrm{He}$ are elevated over their usual coronal values. This suggests that close to the sun, where what ultimately become the highest energy particles are injected into the acceleration process, these particles have a different provenance than the rest of the SEP distribution, i.e., they are likely to be supra-thermals from an impulsive flare, where such abundance anomalies are well-known. All the above observations suggest that to explain the SEP intensity variations, it is crucial to understand the characteristics of the background plasma where they are produced and, in particular, the processes of generation of supra-thermal particles in the corona.

Theoretical studies point out the necessity of high-energy particles to initiate diffusive shock acceleration and support the impact of pre-existing supra-thermal particle distributions on the variability of shock accelerated SEPs. Injection of particles into the shock acceleration process likely occurs when the CME shock is within a few solar radii of the solar surface [32,34,35]. Close to the sun, CME driven shocks are often expected to be quasi-perpendicular in geometry, requiring a higher energy or harder spectrum of seed particles to initiate SEP acceleration than would be the case for quasi-parallel shocks. Zank et al. [36] showed that a particle has to have sufficient initial energy to be accelerated by the diffusive shock acceleration mechanism, and this energy threshold is higher at a quasi-perpendicular shock than at a quasi-parallel shock. Laming et al. [28] revisited this seeking a prescription in terms of the supra-thermal particle distribution function rather than the energy. Their conclusion was that quasi-perpendicular shocks require a harder spectrum of supra-thermals than do quasi-parallel to initiate diffusive shock acceleration. Their calculations demonstrated that a stronger non-thermal tail of particle distribution function is required for shocks closer to the sun with low Alfvén Mach number $M_A$ compared to those further out from the sun with higher Mach numbers (see Fig. 3). A simulation of proton shock acceleration [28] with higher and lower densities of seed particles (one order of magnitude difference) showed that the resulting high-energy proton intensity can vary by several orders of magnitude. The results of Zank et al. [36] and Laming et al. [28] support the inference of Tylka and Lee [32].

 

Fig. 3. Contours of upstream growth rate from Laming et al. [27] in $\kappa -{\theta }_{\mathrm{BN}}$ space. ($\kappa$ is the index of the kappa distribution function of the supra-thermal ions, and ${\theta }_{\mathrm{BN}}$ is the shock obliquity.) The region of growth is to the lower left of the dashed lines, which represent zero growth. The top panel is for shock Alfvén Mach number $M_A=2$ (at the distance $2{\mathrm{R}}_{\odot }$), the bottom for $M_A=4$ (at the distance $2.6{\mathrm{R}}_{\odot }$). At lower $M_A$, a lower $\kappa$ (i.e., harder supra-thermal distribution) is required to grow waves, which lead to SEP production.

Download Full Size | PDF

A number of theoretical studies demonstrated that particles can be accelerated in the process of magnetic reconnection, which ubiquitously occurs in the strongly magnetized solar corona. Magnetic reconnection is the process of reconfiguration and annihilation of the magnetic field over a small-scale volume. In this process the energy stored in the magnetic field is converted to the kinetic and thermal energy of the plasma and energy of accelerated nonthermal particles. Simple analytical studies [37–39] suggest that ions can be accelerated by the first-order Fermi mechanism in the interaction with convergent magnetized flows associated with magnetic reconnection. In this process, ions with mean free paths larger than the thickness of the current sheet cross the reconnection layer and gain energy from each crossing similar to the process of diffusive shock acceleration.

Other possible ion acceleration processes include Fermi acceleration of ions trapped in plasma islands as they pass through the fast termination shock in reconnection outflows [40] and acceleration of pickup ions, with charge-to-mass ratios above a certain critical value, by outflows in reconnection regions [41,42]. Reconnection could produce a suitably hard spectrum of seed particles, if sufficiently high-density compression can be achieved in the reconnection region. Further studies are needed to confirm this.

4. OBSERVATIONAL SIGNATURES OF SEP SUPRA-THERMAL SEED PARTICLES

Out to a distance of about $3.5{\mathrm{R}}_{\odot }$ heliocentric distance, protons and neutral hydrogen are effectively coupled by charge exchange reactions so that the proton distribution function may be revealed by spectroscopic observations of emission lines from neutral H [43,44]. The Lyman $\alpha$ transition ($\mathrm{Ly}\alpha$) is by far the strongest emission line available. It is mainly excited in the outer corona by absorption of $\mathrm{Ly}\alpha$ photons emitted from the solar disk followed by re-radiation. Thus the prediction of the observed line profile requires a treatment of the scattering of $\mathrm{Ly}\alpha$ by H atoms with a supra-thermal ion distribution and moving radially with respect to the sun as the solar wind begins to accelerate.

Photon redistribution functions in the corona are given by Cranmer [45] in the form of $\kappa$ distributions. Integrating over initial photon direction (azimuthal and polar angles), frequency, and along the observer’s line of sight, each point on the line profile is given as a four-dimensional nested integral. Such calculations are carried out by Laming et al. [28] for a variety of off limb positions, with the $\kappa$ distribution in each case chosen at the threshold for exciting waves ahead of a $2000\mathrm{km}{\mathrm{s}}^{-1}$ shock with obliquity $45°$ motivated by a study of the 20 January 2005 CME.

Figure 4 shows the results of such calculations meant to match the cases shown in Fig. 3. In each case, the thin histogram shows the effect of a Maxwellian H atom distribution. The thick dashed histogram shows the effect of a $\kappa$ distribution ($\kappa =2$ and 4.5, respectively), while the solid line shows a possibly more realistic case of 10% $\kappa$, 90% Maxwellian distributions. These simulations assume a ${10}^3\mathrm{s}$ integration time with an instrument of effective area $1{\mathrm{cm}}^2$, and give photon counts in 0.1 Å bins in a spatial pixel of $1{\mathrm{arcmin}}^2$. Under such observing conditions, the supra-thermal distribution appears to be distinguishable from a Maxwellian out to about $2.5{\mathrm{R}}_{\odot }$. The intensity decreases by about an order of magnitude between 1.8 and $2.5{\mathrm{R}}_{\odot }$. due to the decreasing density of scattering H atoms and the dilution of the incident radiation.

 

Fig. 4. Profiles of H I $\mathrm{Ly}\alpha$ predicted at distance $2{\mathrm{R}}_{\odot }$ ($M_A=2$; top) and at distance $2.6{\mathrm{R}}_{\odot }$ ($M_A=4$; bottom). The thin histogram shows the effect of a Maxwellian. The thick dashed histogram shows the effect of a $\kappa$ distribution ($\kappa =2$ and 4.5, respectively). The solid line shows 10% $\kappa$, 90% Maxwellian.

Download Full Size | PDF

5. NEW INSTRUMENTATION

The comprehensive summary of Kohl et al. [8], written after almost a decade of SoHO operations, outlines the technical and scientific progress made with the SoHO UVCS investigation. A comparison of the SoHO UVCS capabilities with the requirements of the investigations described in Sections 2–4 above leads to the identification of two technological areas in which improvements are necessary to achieve the new science outlined above.

The NRL SSD Solar and Heliospheric Physics Branch has undertaken a broad-based investigation of instrument technologies over the past 20 years directly applicable to improvements in these two areas. To address the first issue, achieving the requisite improvement in grating scatter, a specific technological development task is needed—primarily encompassing incremental improvements in optics fabrication, coating technology, and contamination control. In contrast, to address the second issue of increasing the light-gathering power of a space-based instrument, a systems approach is required. If nothing else, launch costs would prohibit a simple $100\times$ scaling of the SoHO UVCS instrument. A solution to this problem requires an approach combining development of each optical element and detector, a novel coronagraphic design, spacecraft design, and mission design.

A. Grating Scatter

Progress in the development of technology for the production of low-scatter optics of all kinds and of gratings in particular has been extremely rapid over the 20 years since the SoHO instruments were developed [46]. While these technologies were primarily driven by requirements from outside of the field of helio-physics, the happy consequence is the availability of gratings with scattering that can be characterized as an order of magnitude lower effective surface roughness over that of the SoHO UVCS gratings. A review of the relative contributions to the instrument-induced noise indicates this improvement is essential for the detection of supra-thermal seed particles within a timescale commensurate with their residence time in the corona.

In Laming et al. [28] we identified the 1–3 Å band from line-center as the most sensitive range for determining departures from Maxwellians. In this range the dominant contributions are the off-band scatter from the diffraction grating and the stray disk light from the coronagraph occulting system. These components are shown in Fig. 5 as dot-dashed and dotted lines, respectively. The grating scatter is shown assuming two different rms surface roughness models: $\sigma =20\AA$ (red curve) similar to the SoHO/UVCS gratings and $\sigma =5\AA$ (blue curve), which is more typical of modern gratings. The residual stray disk light is representative of a coronagraph with stray-light rejection of $B/B_{\odot }\le {10}^{-8}$ (typical of UVCS). The disk-line profile shows extended wings from radiation transfer effects making characterization/validation of the coronagraph stray-light performance an important issue for data reduction. Another contribution to the off-band profile is the broad electron-scattered $\mathrm{Ly}\alpha$ profile (triple-dot, dashed line in Fig. 5) produced by free electrons in the corona. This profile can be modeled and subtracted from the measurement with the help of knowledge about the electron density. The narrow geo-coronal $\mathrm{Ly}\alpha$ contribution (not shown) has little effect in the wavelength region where the $\kappa$ profile makes a departure from a pure Maxwellian. The expected resonantly scattered $\mathrm{Ly}\alpha$ profile is shown as the black dashed line. All curves are on the same vertical scale which is in counts/bin. The bin size ($\text{radial}\times \text{transverse}$) is 0.17′ by 1.1′.

 

Fig. 5. Coronal streamer observation at $1.8{\mathrm{R}}_{\odot }$ showing the different components of the observed $\mathrm{Ly}\alpha$ profile (black solid line). See the text for a description of the various components. $\kappa$ is as defined in [45].

Download Full Size | PDF

In order to reduce the grating scatter for an instrument optimized for seed-particle detection, it is desirable to use a holographic grating, which has been shown by Fineschi et al. [47] to have a factor of 4 lower scattering, compared to conventionally ruled gratings for the UV. Fineschi et al. [48] compared scattered light measurements of the UVCS $\mathrm{Ly}\alpha$ holographic grating to an analytical model of the grating scatter. From this they computed an effective rms roughness ($\sigma$) and correlation length ($\gamma$) for the grating surface and determined their values to be 20 Å and 6.0 μm, respectively. The UVCS gratings were manufactured in the mid-1990’s but more recent holographic gratings have been produced with better surface properties. Ion etched, holographic gratings for two recent NRL programs, the Joint Astrophysical Plasma Dynamics Experiment (JPEX) sounding rocket and the Extreme-Ultraviolet Spectrograph (EIS) on Hinode, were each measured to have an rms surface roughness $\sigma <3\AA$ [49,50]. Using only a value of $\sigma =5\AA$ for the surface roughness, the modeled grating scatter shows a dramatic improvement in the estimated scatter (see Fig. 5), greater than the factor of 4 difference between holographic and mechanically ruled gratings found in [42].

The residual radiation from the solar disk that is scattered by the occulting system and interior surfaces has a smaller effect on the coronal profile in the near wings. This stray-light background can be estimated before launch by laboratory measurements or calibrated in-flight by determining if there are detector counts in the cooler spectral lines that are not expected to be produced in the extended corona.

In principle, accurate measurements of the coronal line shape can still be made, even with backgrounds higher than the coronal signal in the wings of the line. This is achieved by carefully performing the optical characterization of the instrument and then subtracting out the background components from the observed line profile. However, even after these corrections, residual noise on these backgrounds will still remain.

B. Light Gathering Power

Various possibilities for improving the light gathering throughput are summarized in Table 1. One approach to achieving increased light gathering power is to consider optical designs with increased collecting area. However, simply scaling the UVCS design to achieve larger aperture quickly results in an impractical instrument size for spaceflight. The difficulty with externally occulted solar coronagraphs is, in general, the requisite occulter to objective distance increases linearly with the unvignetted aperture available for observations at any given coronal elongation. Thus, direct scaling of UVCS to the required collecting area would result in an instrument greater than 10 m length.

The resources to launch a monolithic spaceflight instrument of $>10$ m length will not be available for quite some time. The first solution to be considered for this problem was to employ an extendable boom to support the occulter at the correct distance from the coronagraph objective in flight, while compressing down to a very small volume for launch [51]. Considered risky at the time it was first proposed, the success of the Nuclear Spectroscopic Telescope Array (NuSTAR) small Explorer mission (see e.g., [52]) demonstrates a 10 m extendable boom can reliably perform with greater precision than required for a coronagraph. The flight dynamics of a boom-based coronagraph of $\ge 10\mathrm{m}$ demand a dedicated spacecraft with a specific attitude control system and thus limit the opportunities for flight.

Another approach to increasing the occulter to objective distance is to locate the occulter and coronagraph telescopes on separate, formation-flying spacecraft. Analogous to the NuSTAR use of an extendable boom, formation flight has been extensively studied for various forms of astronomical instrumentation. Use of this approach for a solar coronagraph has been proposed for the first two European Space Agency (ESA) M-class missions (e.g., [53]). Further, this approach is in development for the ESA Proba-3 formation-flight technology demonstration mission as ASPIICS (the French acronym meaning “Association de Satellites Pour l’Imagerie et l’Interférométrie de la Couronne Solaire”). ASPIICS is not the prime mission objective, but serves as a means to gauge the success of the formation flight technology. ASPIICS in its final form is a visible light coronagraph observing the K-corona and prominences in the He I 587.5 nm (D3) line [54]. Although the ASPIICS aperture is not the maximum allowed by the 144 m separation between the occulter and objective, the separation results in negligible vignetting of the telescope aperture at very low coronal altitudes ($1.08{\mathrm{R}}_{\odot }$ inner field of view)–similar to an eclipse of the sun by the moon as viewed from Earth ($3.5\times {10}^8\mathrm{m}$ separation between occulter to objective).

In addition to the added cost of implementing two spacecraft for a formation flight coronagraph, the lifetime of the mission is severely constrained by high propulsion consumption to achieve the correct formation in all but a few orbital configurations. Wood and Breakwell [55] showed that continuous formation flight in a low Earth orbit (LEO) would require the occulter spacecraft propulsion system to supply sufficient thrust to achieve a $\mathrm{\Delta }\mathbf{v}=8\mathrm{km}/\mathrm{s}/\mathrm{year}$ (compared to Earth escape velocity of 11 km/s). In order to achieve an extended formation flight mission, Proba-3 uses a highly elliptical orbit ($\mathrm{60,530}\mathrm{km}\times 600\mathrm{km}$ altitude, with an eccentricity of 0.81). The ASPIICS formation flight is limited to the approximately 6 h apogee arc, thus providing a total duty cycle of $\sim 30\%$. With this strategy, the required $\mathrm{\Delta }\mathbf{v}$ per orbit is less than 10 cm/s, which is manageable with a normal spacecraft propulsion system. For continuous observations, inner and outer Lagrangian point orbits have been considered (e.g., [53]). Moses and Fineschi [56] explored several variations on the Wood and Breakwell paradigm in LEO, including the use of a series of disposable nanosat occulters and the use of alternative means of supplying the requisite $\mathrm{\Delta }\mathbf{v}$ such as a photonic laser thruster [57] or a conductive tether [58] operating between coronagraph and occulter.

An alternative to gaining light-gathering power by increasing the separation of occulter to objective can be achieved by following the path taken in increasing the light-gathering power of grazing incidence x-ray telescopes, which are nesting multiple optical channels of the same focal length and the same focal plane. As in a nested-grazing incidence telescope, a nested coronagraph consists of a series of stacked confocal occulter-objective pairs in order to increase the net collecting area of the instrument. In contrast to nested-grazing incidence telescopes, one does not actually need to employ a physically separate annular optical segment for the objective elements of the individual nested channels of the class of nested coronagraphs considered in this paper. The difficult opto-mechanical challenge of multiple objectives can be circumvented in the case of a normal incidence instrument with a single objective and a segmented objective baffle. Figure 6 illustrates a design, which in cross-section is equivalent to four UVCS-like coronagraphs, with the same focal length but proportionally larger off-axis optical figures in order to maintain a common focal plane. As can be seen in the figure, the same result is achieved simply and accurately by masking a single off-axis mirror with an objective baffle with four appropriate apertures.

 

Fig. 6. Optical diagram of a nested-coronagraph design. For clarity, only four nested segments are shown in this illustration. Each of the four segments is roughly equivalent to UVCS. All direct solar-disk light entering the instrument through the forward external occulter (left) is prevented from reaching the objective by the heat rejection mirrors (middle). The objective baffles (right) serve as internal occulters to prevent any diffracted light from reaching the spectrograph slit.

Download Full Size | PDF

After the initial nested coronagraph design work was developed at NRL [59], we discovered the concept was not entirely new due to investigation by Beckers and Argo [60] for the same application of H $\mathrm{Ly}\alpha$ coronal spectroscopy. Although the stray-light performance of the Beckers design was successfully characterized in ground testing, and the performance of the nested-coronagraph approach was verified, the instrument was never used in a spaceflight environment for coronal observations. With the selection of UVCS for the SPARTAN program and then SoHO, development of the Beckers’ instrument stopped and this technique was largely forgotten.

The design goal at NRL for a nested coronagraph was specifically directed toward the objective of making a definitive test for the existence of supra-thermal seed particles considered essential in many acceleration models for the production of high-energy SEPs. As demonstrated above, this objective can be achieved with high-contrast spectroscopy of the resonantly scattered H $\mathrm{Ly}\alpha$ line at 120 nm if the instrument can achieve light-gathering power sufficient to collect ${10}^6$ photons per spatial bin (e.g., 15 arcsec) per temporal bin (1000 s). The field of view of the instrument must span the region of the corona where a fast CME could first exceed the local Alfvén speed and become a shock: $1.8–3{\mathrm{R}}_{\odot }$. We discovered an instrument dedicated to this specific measurement that could be designed to fit within the approximately $1{\mathrm{m}}^3$ volume allotted for standard International Space Station attached payloads.

In order to achieve this goal (and in contrast to the Beckers’ design), a linear occulter system was employed with the direction of the baffle aligned parallel to the orientation of the spectrometer entrance slits (as in UVCS). In order to observe the range of coronal heights over which a shock is formed in a point-and-stare mission operations approach, a spectro-coronagraph design employing two slits at two coronal heights is necessary, as described by Fineschi et al. [61]. Light is focused on separate spectrometer entrance slits for the two observational heights. The dispersion of the spectrometer is adjusted so that the off-band radiation from the each slit can be baffled to prevent overlap of the two spectra that are imaged by the grating onto the detector with spatial separation in the direction of dispersion. A more detailed consideration of the coronagraphic portion of such an instrument is presented in Fig. 7. Radiation from coronal heights of 1.8 and $3.0{\mathrm{R}}_{\odot }$ propagates through the slots in the A0 baffle aperture onto the primary mirror, which focuses the light on two spectrograph slits in the focal plane. (Note the solar illumination comes from the right side in Fig. 7 while it comes from the left in Fig. 6.)

 

Fig. 7. Schematic of a nested linear occulter coronagraph designed to obtain coronal spectra at $1.8{\mathrm{R}}_{\odot }$ and $3.0{\mathrm{R}}_{\odot }$. This combination of linear occulters (with light trap), internal occulters, and slits prevent any diffracted radiation from entering the spectrometer. Scatter from the mirror surface roughness and contamination is the only significant source of stray light.

Download Full Size | PDF

In Fig. 7 the path of coronal light from the two coronal heights is shown for three of the channels of the system, illustrating for this particular design how the light from each channel does not overlap the light from the next in the plane of the objective mirror. Thus, each channel can be considered separately. Figures 7(b) and 7(c) show the path of the direct illumination, Fig. 7(b), and the scattered illumination, Fig. 7(c), through a single channel of the front baffle and primary mirror assemblies. Note that in Fig. 7(b) of this design, light from the two different coronal heights do not overlap in the plane of the objective. The direct radiation from the disk and diffracted radiation from solar-illuminated edges is trapped by a set of mechanical baffles as shown in Figs. 7(b) and 7(c). In this example an extended occulter baffle, which absorbs the unwanted direct and scattered light, is used instead of the heat-rejection mirrors in the design presented in Fig. 6. Note in Fig. 7(c) that only two rays (illustrated in dark yellow) of all possible light diffracted from the front of the occulter baffle at point $\alpha$ can be focused onto either of the two spectrograph slits. This is the only singly scattered disk light that could potentially reach the slits. The objective baffles labeled (1) and (2) in Fig. 7(c) function as internal occulters and absorb those two rays. The same is true of the two rays illustrated in dark blue generated from diffraction at the rear of the external occulter assembly (point $\beta$), which are absorbed by the internal occulters labeled (2) and (3). These rays include the only doubly scattered disk light and singly scattered light from lower coronal heights that could potentially reach the slits. Note that no direct disk ray can hit the rear baffle (point $\beta$), so the light illuminating this edge is greatly reduced from that illuminating the front edge (point $\alpha$). The mirror has a set of internal occulters for each stack pair.

The objective baffle edges diffract any light illuminating them but with even minimal over-occultation this illumination is small. Complementary baffles can be located in front of the grating to address the objective baffle diffraction but a ray trace indicates this is not necessary for the design illustrated. The remaining scatter from the objective mirror roughness and from particulate contamination are minimized with super-polished mirrors, which are now available, and by carefully avoiding contamination during instrument development.

It is interesting to note that although the number of baffle edges in the system is much greater than that of UVCS, the ratio of stray light to signal from each channel (slot) is the same. Thus the total stray light to signal level ratio does not change for this new design. A modest over occultation is all that is required to achieve the necessary stray-light suppression ($<1\times {10}^{-7}\mathrm{B}/{\mathrm{B}}_{\odot }$ at $1.8{\mathrm{R}}_{\odot }$ and $<1\times {10}^{-8}\mathrm{B}/{\mathrm{B}}_{\odot }$ at $3.0{\mathrm{R}}_{\odot })$.

All of the above designs are externally occulted coronagraphs because this approach is usually necessary to achieve sufficient stray-light suppression of the disk intensity at the coronal heights of interest to both solar wind and SEP investigations (i.e., $\ge 3{\mathrm{R}}_{\odot }$). However, Kohl et al. [62], demonstrated the residual stray light of an internally occulted UV/EUV coronagraph at important lines useful for solar wind research can be removed analytically at heights up to $3{\mathrm{R}}_{\odot }$. The utility of this approach for the spectroscopically more demanding SEP seed-particle measurements is still under investigation. It is likely that the best approach to achieve both objectives at the present time is an externally occulting coronagraph with current state-of-the-art gratings, mirrors, and other optical components.

Table 1. Summary of Coronagraphic Design Approaches to Increase Light Gathering Power

View Table

6. SUMMARY

UV spectroscopy of the extended corona has proven to be an effective means for diagnosing the source regions of the solar wind and shows great promise for doing the same for solar energetic particles. Measurements of line intensities, Doppler widths, charge states, and abundances can reveal details about the plasma conditions of the bulk solar wind. These properties can be compared directly with models of coronal heating and acceleration.

Newer diagnostics for measuring the line shape in the far wings are being studied for investigating the properties of supra-thermal seed particles, which are believed to be necessary for SEP production. The theoretical work and advances in technologies for low-scatter gratings, high-efficiency coatings, and coronagraph designs have matured sufficiently for making the required SEP measurements.