G
Solar Wind Properties

By Section:
     1. Solar Wind Regimes
     2. Previous Data on Solar Wind Composition
         2.1 He/H Fractionation in Different Solar Wind Regimes
         2.2 "Heavy" Element Solar Wind Compositional Data
         2.3 Lunar Sample Data
         2.4 Solar Wind/Photosphere Elemental Fractionation
         2.5 Solar Wind/Photosphere Isotopic Fractionation
     3. Determination of Solar Wind Regime

1. Solar Wind Regimes

There are two types of solar wind flow: quasi-stationary and transient (see, for example, the review by Neugebauer, 1991). There are at least two sources of quasi-stationary solar wind: fast flows from coronal holes (CH) and slow "interstream" (IS) flows, some or all of which originate in or near visible structures called coronal streamers. Coronal holes are regions of the solar corona with open or unipolar magnetic fields and anomalously low density which appear dark in x-ray images of the Sun. Coronal streamers, on the other hand, are bright, dense structures extending well out into the corona. The "quiet corona" between the holes and the streamers may also lead to slow, interstream flow; alternatively, the entire quasi-stationary solar wind may originate only in coronal holes and streamers whose flows expand laterally to fill all solar latitudes and longitudes.

Transient flows of the solar wind are produced by solar eruptions referred to as coronal mass ejections (CME). CMEs are associated with the disruption of closed magnetic field lines above the solar surface. Depending on the energy released in the eruption, the solar wind from a CME can have either low or high speed. The frequency of occurrence of CMEs varies in concert with the solar activity cycle.

Research on the properties of the solar wind reveals that in the near-equatorial solar regions, the elemental abundances in the solar wind are different for the different types of flow. In the next section we review what is known about the abundance variations from one solar wind regime to the next and in the final section we discuss how the different solar wind regimes can be recognized from the monitor data.

2. Previous Data on Solar Wind Composition

2.1 He/H Fractionation in Different Solar Wind Regimes. The solar wind He/H is the best known solar wind compositional parameter, and it is also the most highly variable. It can be readily measured with spacecraft instruments, and the behavior of this ratio in different solar wind regimes is well documented, even if not well understood (e.g. Neugebauer, 1981; Ogilvie, et al., 1989). Variations of factors of 100 or larger are observed on time scales ranging from minutes to days. In solar cycle 20, long term averages (six months) showed a systematic trend with the solar cycle, with higher (about 0.05 atomic) ratios associated with solar maximum and lower (about 0.03) ratios with solar minimum. The solar cycle 21 data shown in Ogilvie et al. were consistent with cycle 20, but showed more structure. In general, although short term variations in the solar wind abundances are probably present for all elements, one might expect that long-term averages would converge to well-defined values, even if consistently different from the photosphere. This topic is discussed further in Section 2.2.

During the decline from solar activity maximum the solar wind is strongly influenced by high speed streams coming from equatorial extensions of large polar coronal holes. In those high speed streams the He/H ratio is relatively constant at about 0.05, and other solar wind parameters are relatively uniform as well. The uniformity led Bame et al. (1975) to propose that 0.05 is the photospheric He/H. Although that suggestion has not gained support, there is now clear evidence that coronal hole solar wind has the least elemental fractionation (Geiss, 1982; Reames et al., 1991; Geiss et al., 1994a; 1995).

Very low values of He/H are observed in the plasma sheet associated with flow from coronal streamers, commonly reaching a value of ~0.02 (Borrini et al., 1981). At the other extreme, CME plasma often contains patchy regions with extremely high helium enrichments, with He/H sometimes reaching 0.4, or greater (e.g., Neugebauer, 1981; Borrini et al., 1982). At least qualitatively, the long-term variability of the He/H ratio can probably be explained on the basis of the contributions of coronal holes, coronal streamers, and CMEs to the ecliptic plane solar wind in proportions that vary over the solar cycle. Independent compositional measurements of solar wind from coronal holes, coronal streamers, and CMEs are thus clearly important.

Helioseismology has provided an estimate of the solar He/H ratio of 0.07-0.08 by atom (Dziembowski et al., 1991) which, if correct, means that, on the average, the solar wind essentially always has a depletion of He relative to H, a conclusion previously adopted by several workers (e.g. Geiss, 1982, Meyer, 1985). The large He/H variations and their apparent failure to converge on the average solar value do not, in themselves, exclude the possibility that appropriate averages of the solar wind abundances of heavier elements give correct average solar system proportions. For two plausible mechanisms of fractionation, incomplete acceleration of heavy ions by proton Coulomb friction or incomplete ionization of some elements in the initial acceleration phase, He is expected to be the most fractionated element (see Neugebauer, 1981; Geiss, 1982, for general discussions).

In summary, the high degree of variability in He/H, the lack of a well-defined long term average, and the possibility of a systematic difference with the photospheric He/H are problems in interpreting solar wind composition in terms of solar abundances for heavier elements; however, it is quite likely that H and He are unique elements. It is clearly important to have independent compositional data on the different solar wind regimes, even in a mission focused on planetary science objectives.

2.2 "Heavy" Element Solar Wind Compositional Data. Good data are available for the light noble gas isotopes (³He, ⁴He, ^20,22Ne, and ³⁶Ar) from the Apollo Solar Wind Composition foils (Geiss et al., 1972). Aluminum and platinum foils were exposed to the solar wind for times ranging from 1 to 44 hours during five Apollo missions and returned for laboratory mass spectrometric analysis. On these short time scales, the absolute fluxes varied by about a factor of 4, as would be expected from spacecraft data; the ratios, however, showed less variation (total ranges: 26% in ⁴He/³He and 35% in ⁴He/²⁰Ne). These ratios are correlated in a way which can be interpreted as systematic mass fractionation, but there is no correlation between ratios and absolute fluxes, as might have been expected for some mass fractionation mechanisms. An alternative interpretation is that the above variations are in ⁴He relative to ³He and ²⁰Ne, thus the ³He/²⁰Ne ratio in the foil data is constant within errors. The ratio ²⁰Ne/³⁶Ar is also constant, although the errors are larger.

Although limited to a few nuclei, the Apollo foil data provide an important reference for the interpretation of results from our solar wind sample return instrument, most importantly in testing the significance of long term averages. In this context the He and Ne fluxes and the overall average He/Ne (470) from ISEE-3 data (Bochsler et al., 1986) are fairly close to the Apollo foil results (average He/Ne of 6 missions=530). Thus, in these ratios there appears to be some convergence in a longterm average for species relatively less "diverse" (e.g. in charge/mass) than H and He, supporting the above interpretation that He/H shows the largest compositional variation.

Electronic instruments have also provided data on elements other than H and He. For cycle 21, the ISEE3 instrument (Coplan et al, 1983), employed a combined velocity (Wien filter) and energy/charge measurement to obtain mass/charge data. This instrument provided long-term data for O, Ne, (Bochsler, et al., 1986), Fe (Schmid et al., 1988) and Si (Bochsler, 1989), but could not overcome the problems of exact or close mass/charge coincidences between different ions, e.g. ¹²C⁺⁶ and ⁴He⁺⁺ or ¹²C⁺⁵ and ²⁴Mg⁺¹⁰. Even with minimal overlaps of charge to mass for one charge state, uncertainties in the overall charge distribution produces errors in abundances. An additional difficulty with the ISEE-3 abundances is that they represent a biased sampling of solar wind because of the limitation to velocities <600 km/sec and low temperature conditions (Bochsler et al., 1986), although a large number of spectra are available (Ogilvie et al., 1989). Several reviews of the ISEE-3 data are available (e.g. Bochsler and Geiss, 1989 or Coplan et al., 1990).

The solar wind composition instrument on Ulysses (SWICS) involves three parameter measurements: velocity (by timeofflight), energy/charge (electrostatic analyzer), and energy (semiconductor detector). The Ulysses instrument has performed well, and good abundance and ionization state distribution data are becoming available for C, O, Ne, Mg, Si, S, and Fe for both low speed, coronal hole, and transient regimes (e.g., Geiss et al., 1995a, 1995b; von Steiger et al., 1995).

In the latest generation of in situ detectors, the principal new instruments are spectrometers that use a novel electrostatic mirror to determine the ion mass independent of its velocity or original charge state (Hamilton et al., 1990; Gloeckler et al., 1995). The first flight test of this type of instrument was on the WIND spacecraft launched in late 1994. Presently a similar instrument has been functioning on SOHO since its launch in 1995, with another due on the Advanced Composition Explorer (ACE) in 1997. This type of Solar Wind Ion Mass Spectrometer (SWIMS, as it is called on ACE, MASS on WIND, or MTOF, part of the CELIAS package on SOHO) is the first to separate ions by mass alone. It has a mass resolution M/(delta)M > 100, where (delta)M is the FWHM peak width, allowing relatively good separation of mass units up to about 60 amu. Transmission is between 2.5 and 10% and is element-specific, as it depends on the fraction of ions emerging from a thin foil in the singly ionized state. With an effective aperture of 0.013 cm², the anticipated counting rate of Fe, for example, is ~3,700 ions/minute. The in-flight performance of MASS was disappointingly limited by a background caused by solar wind helium which hampers analysis of ions with mass/charge near 2 amu/q. CELIAS on SOHO has done better so far, as the electronics are improved, it has a larger aperture, and the spacecraft is 3-axis stabilized, so that the instrument is continuously sun-pointing. Results so far (e.g., Bochsler et al., 1996) are consistent with expectations that SWIMS/MASS/MTOF will gather abundance data on elements more abundant than Ar, and on a few isotopic ratios such as ²⁰Ne/²²Ne or ²⁴Mg/²⁵Mg/²⁶Mg, ²⁸Si/²⁹Si/³⁰Si, and ⁵⁴Fe/⁵⁶Fe to a precision of ~5%. Elements heavier than ~Ni are not expected to be measured by SWIMS due to the low abundances and the relatively poor mass resolution in that range. Li, Be, and B will not be measurable either because of their expected low abundances. We therefore expect a solar wind sample return to be highly complementary to the present and planned in-situ solar wind missions.

In summary, electronic instruments and foil collectors provide very complementary kinds of information. In general electronic instruments have difficulty in providing the elemental and isotopic coverage of the collector foils, in covering the total range of solar wind conditions, and in providing accurate long term averages. Although our design provides independent collector arrays for 3 major solar wind regimes, these arrays cannot provide the detailed time resolution or the ionization state distribution information obtainable from electronic instruments. The time and charge data are vital for solar and heliospheric physics, but of lesser importance for planetary science. The level of abundance precision for planetary science applications is not especially important for solar physics. Close comparisons of foil and instrument data can be made through the relative elemental abundances of C, O, Ne, Mg, Si and Fe.

2.3 Lunar Sample Data. In principle, lunar samples are solar wind collector materials. In practice, although a variety of elements of solar wind origin have been recognized in lunar samples, it has proved very difficult to interpret the resulting concentrations in terms of fluxes or even relative elemental abundances. This is a result of perturbation of implanted solar wind ions by lunar processes: erosional/depositional modification of mineral grain surfaces by meteorite bombardment, thermal diffusion complicated by radiation damage, sputtering, etc. For example, essentially all of the noble gas atoms in lunar soils are of solar wind origin; however, a wide variety of noble gas abundance patterns have been observed. Only in a few highly selected ilmenite mineral separates, subjected to special analytical treatments, have elemental abundance patterns been obtained for noble gases which resemble what is expected from relative solar abundances (e.g., Frick et al., 1988; Becker and Pepin, 1989; Wieler et al., 1986, 1993; Nichols et al., 1994), and even in these the N/(noble gas) abundance ratio is roughly an order of magnitude higher than expected for the solar wind. Moreover there exist significant variations in the relative xenon abundance and the nitrogen isotopic ratio which appear to correlate with "indexes of antiquity" of the lunar soils (e.g., Kerridge, 1980; 1992).

It is now clear from the newer results and, in retrospect, from much of the earlier data dating back to the beginning of lunar sample studies, that the experimental task of extracting samples of pure solar wind gases from their shallow implantation zones in these grains is not trivial. There is no firm assurance that extracted gases have not been compositionally altered in situ by any of a number of post-implantation regolith processes acting to varying degrees over long periods of time. Significant disagreements still exist concerning isotopic compositions determined in this way for the contemporary solar wind, in particular for several crucial atmospheric processing tracers: C, N, and the heavy noble gases Ar, Kr and Xe. The SU data will be centrally important for evolutionary modeling in at least three ways: (1) they will provide an accurate and unambiguous measure of current solar wind compositions for these species; (2) they will validate -- or not, as the case may be -- the extensive modeling of solar wind composition based on ilmenite data; and (3) they will establish a "ground truth" data point for the contemporary wind in the search for possible secular changes in solar isotope ratios using data from old lunar and meteoritic samples. This latter investigation can be carried out only with fossil regolith grains, but in order to understand these possible variations in the ancient solar wind, it is imperative to characterize the present-day solar wind first. Item (3) was a major Apollo science objective which may require SU data for its realization. In summary, we regard the SU and regolith data as powerfully complementary rather than redundant.

2.4 Solar Wind/Photosphere Elemental Fractionations. There is now clear evidence for fractionation between the photosphere and the solar wind. Table G1 is a summary of the most recent data compiled by Geiss et al. (1994) and updated by von Steiger (1994). Abundances are presented relative to O. The column labeled "SEP-based Corona" presents the abundances in the corona deduced from the study of solar energetic particles. The solar wind Xe and Kr data are inferred from lunar sample analyses.

Table G1. Compilation of solar wind and photosphere abundances (von Steiger, 1994)
FIP Solar Wind in Ecliptic Solar Wind From Coronal Holes SEP-based Corona Photosphere

H 13.60 1900±400 ^a 824±80 ^d,1 1175 ^o

He 24.59 75±20 ^b 48.5±5 ^d,1 55±3 ^m 115 ^o

C 11.26 0.72±0.10 ^c 0.70±0.10 ^c 0.428±0.043 ⁿ 0.468 ^p

N 14.53 0.129±0.008 ^d,1 0.145±0.011 ^d,1 0.123±0.009 ⁿ 0.117 ^q

O 13.62 1 1 1 1

Ne 21.56 0.17±0.002 ^b 0.136±0.011 ^d,1 0.142±0.014 ⁿ 0.138 ^o

0.14±0.02 ^e,2

Mg 21.56 0.16±0.03 ^c 0.083±0.02 ^c 0.193±0.011 ⁿ 0.0447 ^o

Si 8.15 0.19±0.04 ^f 0.054±0.009 ^d,1 0.164±0.0099 ⁿ 0.0417 ^o

0.18±0.02 ^g

S 10.36 0.038±0.009 ^d,1 0.019±0.003 ^d,1 0.0377±0.0016 ⁿ 0.0191 ^o

0.05±0.02 ^e,2 0.022±0.008 ^h,2

Ar 15.76 0.004±0.001 ^i,2 0.0037±0.0006 ⁿ 0.00447 ^o

Fe 7.87 0.19±(0.10,0.07) ^j,2 0.057±0.007 ^d,1 0.172±0.023 ⁿ 0.0355 ^r

0.12±0.03 ^k

Kr ³ 14.00 3.4±0.4 ^l,2 1.89 ^o

Xe ³ 12.13 0.88±0.12 ^l,2 0.197 ^o

**Table G1. Compilation of solar wind and photosphere abundances (von Steiger, 1994)**
	FIP	Solar Wind in Ecliptic	Solar Wind From Coronal Holes	SEP-based Corona	Photosphere
H	13.60	1900±400 ^a	824±80 ^d,1		1175 ^o
He	24.59	75±20 ^b	48.5±5 ^d,1	55±3 ^m	115 ^o
C	11.26	0.72±0.10 ^c	0.70±0.10 ^c	0.428±0.043 ⁿ	0.468 ^p
N	14.53	0.129±0.008 ^d,1	0.145±0.011 ^d,1	0.123±0.009 ⁿ	0.117 ^q
O	13.62	1	1	1	1
Ne	21.56	0.17±0.002 ^b	0.136±0.011 ^d,1	0.142±0.014 ⁿ	0.138 ^o
		0.14±0.02 ^e,2
Mg	21.56	0.16±0.03 ^c	0.083±0.02 ^c	0.193±0.011 ⁿ	0.0447 ^o
Si	8.15	0.19±0.04 ^f	0.054±0.009 ^d,1	0.164±0.0099 ⁿ	0.0417 ^o
		0.18±0.02 ^g
S	10.36	0.038±0.009 ^d,1	0.019±0.003 ^d,1	0.0377±0.0016 ⁿ	0.0191 ^o
		0.05±0.02 ^e,2	0.022±0.008 ^h,2
Ar	15.76	0.004±0.001 ^i,2		0.0037±0.0006 ⁿ	0.00447 ^o
Fe	7.87	0.19±(0.10,0.07) ^j,2	0.057±0.007 ^d,1	0.172±0.023 ⁿ	0.0355 ^r
		0.12±0.03 ^k
Kr ³	14.00	3.4±0.4 ^l,2			1.89 ^o
Xe ³	12.13	0.88±0.12 ^l,2			0.197 ^o

Values determined in the Earth's magnetosheath
Original values not referred to O
Values x10⁶

a. Bame et al. 1975	g. Galvin et al. 1993	m. Reames 1993
b. Bochsler et al. 1986	h. Shafer et al. 1993	n. Garrard and Stone 1993
c. this work	i. Cerutti 1974	o. Anders and Grevesse 1989
d. Gloeckler et al. 1989	j. Schmid et al. 1988	p. Grevesse et al. 1992
e. Geiss et al. 1970b	k. Ipavich et al. 1992	q. Grevesse and Anders 1991
f. Bochsler 1989	l. Wieler et al. 1993	r. Hannaford et al. 1992

The data in Table G1 support the conclusion that the elements with low first ionization potentials (FIP) are over-abundant and that, collectively, the low-FIP elements Mg, Si, and Fe have a greater fractionation in the average in-ecliptic solar wind than they do in the wind from coronal holes. At the present limits of error, there is no evidence of fractionation among the low FIP elements Mg, Si, and Fe. These points are illustrated in Figure G1 which shows abundances in the solar wind, again relative to O, as a function of first ionization time (FIT). FIT takes into account all modes of ion excitation and ionization, and is therefore a more physically relevant parameter than FIP. Figure G1 shows a clear differentiation between the interstream and the coronal-hole regimes. (It should be noted that the Xe, Kr, and Ar points in Figure G1 were derived from lunar and meteorite data and assumed to be more representative of the interstream than of the coronal hole flow.)

Fig. G1. Solar wind-photosphere fractionation as a function of first ionization time (Geiss et al., 1995b)

Progress continues to be made in reducing the uncertainties in the systematics of elemental fractionation between the photosphere and the solar wind. New data from Ulysses, Wind, SOHO, and eventually ACE will provide ever smaller error bars in figures such as G1. At the same time, theoretical understanding of elemental fractionation is also continuing rapidly. It is clear that both theoretical understanding and the ability to correct observed solar wind abundances for the fractionation occurring between the photosphere and the wind is close at hand. But even today, we can say that to a fairly good approximation, the higher-FIP elements with first ionization times longer than ~10 s lie on or close to a power law curve with a slope of approximately -0.5, whereas for elements with FIP < ~10.2 eV (the energy of Ly-alpha photons), the fractionation is a constant depending on the solar-wind regime, implying that the relative solar wind abundances of low FIP (planet forming) elements are the same as in the photosphere.

The message to be derived from Fig. G1 is that, as time goes on, it will become increasingly easy and reliable to correct for elemental fractionation between the photosphere and the solar wind. This will be a bootstrapping procedure. If we know a few (e.g., C, N, O, Mg, Si, Fe) photospheric abundances well, comparison to solar wind data provides solar abundances for those elements poorly observed in the photosphere.

2.5 Solar Wind/Photosphere Isotopic Fractionation. What little data exist on the isotopic composition of the solar wind are consistent with there being little or no compositional difference between the different solar wind regimes. No significant variations in the ²⁰Ne/²²Ne ratios were found with the Apollo foil experiments (Geiss et al., 1972). Although the ISEE-3 instruments and the SWICS instrument on Ulysses observed large hour-to-hour and day-to-day variations in the ³He/⁴He ratio, there were apparently no systematic variations associated with solar wind speed or regime (Coplan et al., 1983; Ogilvie et al., 1989; Bodmer et al., 1995). First results from the WIND/MASS experiment show that the ²⁴Mg:²⁵Mg:²⁶Mg ratios are the same, within the measurement uncertainties, for low- and high-speed solar wind (Bochsler et al., 1996). The theoretical model of Marsch et al. (1995) predicts solar wind/photospheric fractionations of typically less than 1 ^o/_oo for the Mg isotopes. Finally, as discussed in Document II-A, it is very important to note that variations in heavy element isotopic abundances should be relatively insensitive to fractionations based on FIP or any other atomic property.

3. Determination of Solar Wind Regime

This section deals with how we plan to distinguish different solar wind regimes using real-time data from the monitors in order to expose different collector panels to the different types of solar wind.

As discussed above, the principal distinction between the quasi-stationary flows from coronal holes and that from streamers is that of their speeds, although there are other differences as well. Transient flows associated with CMEs have several characteristic features; the ones most relevant to the present discussion are:

High helium abundance (Hirshberg et al., 1972; Borrini et al., 1982)

Low ion and electron temperatures (Gosling et al., 1973; Montgomery et al., 1974)

Increased proton anisotropy, with large T_parallel/T_perp. (Zwickl et al., 1983; Gosling et al., 1987)

Counter-streaming superthermal electrons (Montgomery et al., 1974; Temnyi and Vaisberg, 1979; Bame et al., 1981; Gosling et al., 1987; Pilipp et al., 1987)

There are additional features involving the magnetic field and energetic particles, but those parameters are not planned to be measured on the Genesis mission. The low temperatures and high anisotropies are believed to be caused by the expansion of the CME material as it moves through the ambient, quasi-stationary plasma, while the counter-streaming hot electrons are probably caused by the plasma being situated on closed magnetic loops anchored in the Sun. The high helium abundance is unique to transient flows; if it is seen, the plasma is almost certainly transient. After removal of periods in which a spacecraft is magnetically connected to the Earth's' bow shock, the counterstreaming signature is similarly highly diagnostic of CMEs . The reverse is not true, however [Zwickl et al., 1983]; there are probably transient flows which exhibit neither high helium abundance nor counterstreaming superthermal electrons.

Another discriminator between quasi-stationary and transient high speed flows is that interplanetary shocks are often detected about half a day ahead of fast CMEs, whereas, near 1 AU, the fast flow from coronal holes is seldom preceded by a shock. Interplanetary shocks are easily identified as simultaneous, abrupt jumps in the speed, density, and temperature of both protons, alpha particles, and electrons.

The approach to the real-time identification of solar-wind regimes proposed for the Genesis mission is based on analysis of plasma parameters measured by the ISEE3 spacecraft between August, 1978 and February, 1980 and by the IMP 68 spacecraft in 19714 (Neugebauer, 1992). The first step in that study was to identify periods when the spacecraft were either (1) clearly in quasi-stationary flow associated with either coronal holes (CH), the plasma sheet (PS) surrounding interplanetary magnetic sector boundaries, or low-speed interstream (IS) flows or (2) clearly in transient flows marked by helium abundance enhancements (HAE) or counter-streaming superthermal electrons (also called bidirectional electron streaming events (BES)). (See Neugebauer and Alexander, 1991 or Neugebauer, 1992 for additional background.) For those selected intervals of presumably unambiguous flow type, the relations between different solar-wind parameters and the solar-wind speed were determined.

Some of the pertinent relations are shown in Fig. G2. In the three frames of this figure, intervals of quasi-stationary flow are indicated by open symbols and intervals of transient flow are indicated by +'s or x's. Panel (a) displays proton temperature (units of 10⁵ K) plotted versus proton speed. Above ~400 km/s, the quasi-stationary flow from coronal holes (squares) is well separated from the transient flow, as identified by bidirectionally streaming electrons (BES) or by helium abundance enhancements (HAE). At lower speeds, the two types of flow cannot be distinguished on the basis of proton temperature. Panel (b) is a similar plot showing the ratio of proton temperatures parallel and perpendicular to the interplanetary magnetic field. We do not plan to carry a magnetometer on the Genesis mission, but the direction of the interplanetary magnetic field can be determined (except for its sign) from the ion and electron distribution functions which are symmetric (gyrotropic) about the field direction. Panel (b) shows that at velocities >400 km/s, a high value of T_parallel/T_perp. is indicative of transient or CME flow. Once again, this distinction cannot be used at low speeds. (There is some indication that the method used to reduce the ISEE3 ion data exaggerated the values of T_parallel/T_perp. when the plasma was unusually cold, as in CMEs. In Phase B we will examine data from other instruments, such as those on Ulysses, to study this matter further.) Panel (c) shows the helium abundance n_alpha/n_p versus proton speed. For HAE events n_alpha/n_p > 0.08 by definition; some of the BES events had high values of n_alpha/n_p while others did not.

Fig. G2. Scatter diagrams of several parameters diagnostic of solar wind regime plotted versus solar wind speed. The squares are coronal hole data; the circles and triangles are interstream data, while the + and crosses are CME data.

In Fig. G2, lines have been drawn in each frame to indicate separation of a range of parameters that contains only intervals of transient flow from a range of parameters that contains a mixture of quasi-stationary and transient flows. On the basis of this separation of ranges together with previous studies of transient flows in the solar wind, we have defined three indices to use as possible discriminators of transient versus quasi-stationary solar-wind flow. (Recall that coronal hole and streamers are both quasi-stationary, but can be distinguished on speed alone.)

I₁ = (0.005 v_p - 1.75)*10⁵/T_p, where the proton speed v_p is in km/s and the proton temperature T_p is in Kelvin. This relation is calculated from the line shown in the top panel of Fig. G2.
I₂ = (T_parallel/T_perp.)/2, from Panel (b). We might have used a downward sloping line instead of a horizontal one; the effect of making such a choice will be explored in future work.
I₃ = (n_alpha/n_p)/0.08, where n_a and n_p are the alpha and proton number densities, respectively.

These indices were defined to be greater than unity for the transient-only ranges of parameters in Fig. G2 and less than unity for the mixed quasi-stationary and transient flows.

We have calculated and plotted these three indices for each five-minute interval of ISEE3 observations between August, 1978, and February, 1980, when the ion detector ceased operation. Fig. G3 displays the indices for the period Sept. 28 (Day 271) through Oct. 4 (Day 277), 1978, which included the detection of a well-documented, high-speed CME (Galvin et al., l987). A period of bidirectional electron (BES) streaming is indicated at the top of the figure. We conclude that transient flow was probably present from ~0600 UT on Day 272 through ~1500 on Day 274, with perhaps a brief return in the middle of Day 275.

Figure G3. Time series of 5-minute averages of the three indices of CME flows obtained through a CME event observed by the ISEE-3 spacecraft in 1978.

The study of the entire set of ISEE3 data agrees with other studies in that no single parameter is a foolproof indicator of CME flow because of the inhomogeneous and time-variable properties of the transient plasma. Although all three indices and an indication of bidirectional electron streaming occasionally increased in concert, more often each one increased and decreased during a transient event, with two or more indicators usually being greater than unity at any time. In quasi-stationary flow, all indices were generally well below unity. We found three cases, however, all occurring in very low speed flow, for which bidirectional electron streaming was reported but for which there was no other indication of transient flow; we will try to determine the nature of those events by examining other plasma and field parameters in greater detail.

In conclusion, we believe we can quite successfully distinguish between transient and quasi-stationary plasmas in the high speed (> ~400 km/s) solar wind using only ion and electron monitors. Although research is continuing, we are not yet able to do so with great confidence for the slower wind. We are therefore currently proposing three sets of collectors--for the high-speed CME, the high-speed coronal hole, and the low speed solar wind.

References

Bame S.J., Asbridge J.R., Feldman W.C., and Montgomery M.D. (1975) Solar wind heavy ion abundances. Solar Phys. 43, 463-473.

Bame S.J., Asbridge J.R., Feldman W.C., Gosling J.T., and Zwickl R.D. (1981) Bi-directional streaming of solar wind electrons > 80 eV: ISEE evidence for a closed-field structure within the driver gas of an interplanetary shock. Geophys. Res. Lett. 8, 173.

Becker R.H. and Pepin R.O. (1989) Long-term changes in solar wind elemental and isotopic ratios: A comparison of two lunar ilmenites of different antiquities. Geochim. Cosmochim. Acta 53, 1135-1146.

Bochsler P. (1989) Velocity and abundance of silicon ions in the solar wind. J. Geophys. Res. 94, 2365-2373.

Bochsler P. and Geiss J. (1989) Composition of the solar wind. In Solar System Plasma Physics, Geophysical Monograph 54 (J.H. Waite, Jr., J.L. Burch, and R.L. Moore, eds.), 133-141.

Bochsler, P., M. Gonin, R. B. Sheldon, T. Zurbuchen, G. Gloeckler, D. C. Hamilton, M. R. Collier, and D. Hovestadt, (1996) Abundances of solar wind magnesium isotopes determined with WIND/MASS, Proceedings of Solar Wind 8, edited by D. Winterhalter et al., in press.

Bochsler P., Geiss J., and Kunz S. (1986) Abundances of carbon, oxygen, and neon in the solar wind during the period from August 1978 to June 1982. Solar Phys. 103, 177201.

Bodmer, R.,, P. Bochsler, J. Geiss, and R. von Steiger (1995) Solar wind helium isotopic composition from SWICS/Ulysses, Space Sci. Rev., 72, 61-.

Borrini G., Gosling J.T., Bame S.J., Feldman W.C., and Wilcox J.M. (1981) Solar wind helium and hydrogen structure near the heliospheric current sheet: a signal of coronal streamers at 1 A.U. J. Geophys. Res. 86, 4565-4573 and 7370-7378.

Borrini G., Gosling J.T., Bame S.J., and Feldman W.C. (1982) Helium abundance enhancements in the solar wind. J. Geophys. Res. 87, 7370.

Coplan M.A., Ogilvie K.W., Bochsler P., and Geiss J. (1983) The relative abundance of 3He++ in the solar wind. In Solar Wind Five (M. Neugebauer, ed.), NASA Conf. Publ. 2280, 573-590.