Genesis will measure solar wind composition, but the desired data are the abundances for the solar outer surface layers (photosphere) which are expected to preserve the initial nebular composition. To the extent that differences exist, corrections must be applied. In situ measurements of heavy ions have demonstrated differences in some element ratios in the solar wind from those in the photosphere. The situation is different for elements and isotopes in that corrections will be required for at least some, but not necessarily all, elements. There is no observational evidence for isotope fractionations.
Figure H1. Solar wind-photosphere and SEP-photosphere
fractionation as a function of first ionization time (von Steiger et al.,
1995).
2. Elemental fractionation
Figure H-1 (from von Steiger et al., 1995)
summarizes available data from in-situ
instruments, plotted as a fractionation factor defined as a
double ratio, first of the element X to oxygen, then to an
adopted (X/O) ratio for the photosphere. The observed
fractionation factors are plotted against the theoretical
time required for a neutral atom of an element to be
ionized (FIT) in order to be accelerated into the solar
wind (Marsch et al., 1995). FIT is a function
of the atom's first ionization potential (FIP) and solar
physical conditions. Thus, high FIP elements, which are
difficult to ionize tend to be left behind. The data for Kr and
Xe should be ignored because these are inferred from lunar
samples as opposed to being directly measured in the
solar wind and also because direct observations of these
elements in the photosphere is not possible. Similarly, there
are no photospheric abundances for Ar, and the photospheric
Ne abundance must be inferred indirectly from impulsive
flare observations. The He point is legitimate in that a solar
He abundance can be derived from helioseimology which can
reasonably be assumed to apply to the photosphere (Dziembowski
et al, 1991). Relative to O, low FIP elements are
uniformly enhanced, thus, the primary fractionation
is a relative enhancement between low-FIT (easily ionized)
elements as a group and high-FIT elements. However, the
strength of the low-FIT enhancement depends on the type
of flow [interstream (IS) or coronal hole (CH) regimes
(Document G)], being
less pronounced in coronal hole flow than in interstream
flow. The fractionations of elements with long ionization
times tend to lie on a single curve, roughly proportional
to (FIT)1/2, somewhat surprisingly independent
of regime. The large amount of scatter and the large error
bars in Figure H1 arise from measurement uncertainties
in both the solar wind and the photospheric data; with
time, these uncertainties will certainly decrease.
Little is currently known about elemental fractionation
in the third (coronal mass ejections, CME) regime
sampled in our present instrument package design,
although good progress is expected as a result of the
Ulysses, WIND, SOHO, and ACE missions.
3. Accounting for FIT elemental fractionation
Two points should be emphasized: (1) The above example is for
today's best model. By 2005, the earliest possible time for
analysis of a returned solar wind sample, in-situ instrument data
will produce greatly improved models, yielding more
accurate corrections. In general, abundances of tracer elements
of different atomic properties (mass, ionization potential, etc.)
can establish constraints on theories of ionization and acceleration.
To the extent that different solar wind regimes represent
different emission mechanisms, differences in fractionation
patterns revealed by independent compositional measurements
of the regimes would be important in understanding these
mechanisms. It is quite possible that when all the data from
Ulysses, SOHO, and ACE are complete, enough elements with
different atomic properties will have been measured in a wide
variety of solar wind conditions to provide understanding
of the fractionation mechanisms independent of any sample
return data, and this understanding can be applied to
collector foil data on a wider range of elements as a correction
leading to improved solar system abundances. (2) Taking
present data (Figure H-1) at face value, little, possibly no,
correction will be necessary for relative abundances
among the group of low (< 9eV) FIP elements. This group
includes most of the periodic table, including all the non-volatile
planet-forming elements. The primary correction is among
the high FIT/FIP elements and between the low FIT elements
as a group and individual high FIT elements. Theoretical
models of FIT fractionation predict no fractionation for the
low FIP element group (Geiss et al., 1995; Marsch
et al., 1995).
4. Non-FIT Elemental Fractionations?
5. Isotopic fractionation
Independent of contingency planning for what might be observed,
a major point is that the only realistic source of isotopic data for
solar matter is by analysis of the solar wind. Important
data will be obtained by in-situ instruments; but it is also
likely that the required precision, for both solar and planetary
science purposes, can only be obtained via sample return.
Theoretically the most plausible mechanism of isotope fractionation
is "Coulomb drag" (Geiss et al., 1970), in which the solar
wind acceleration process for heavy ions is viewed as proceeding
through multiple collisions with the electromagnetically accelerated
protons. Inefficiencies in this process cause heavier species to be left
behind. The recent model of Bodmer and Bochsler (1996) predicts
that the fractionation relative to H of a species of mass number A
and atomic charge q should scale as
q-2*(2A-q-1)*[(1+A)/A]1/2 with
predicted fractionations up to 30% for He and 1.5%/amu for Mg
for IS (low speed) solar wind. Present data do not rule out these
predictions. However, less isotopic fractionation is predicted for
CH (high speed) solar wind, but there is no evidence for any
CH-IS differences in either the He or Mg data. It may be that
the theoretical amounts of isotopic fractionation are significantly
overestimated.
We have developed a boot-strap process, assuming that a few
photospheric abundance ratios, e.g. Na/S and Ca/Si, are
known accurately (nominally around 10%). Exactly
analogous to Figure H-1, these are the denominators in
calculated fractionaton factors with the same sample return
elemental abundance ratios being the numerators. These
provide a few, nominally accurate, measures of fractionation
within each solar wind regime. A minimum of three ratios
(4 elements, not necessarily O but spaced in FIT) are required
to calculate the model illustrated in Figure H-1. There are
roughly 10 photospheric abundances that have quoted errors
of around 10% which should give sufficient flexibility to produce
reliable corrections. Adopting theoretical FITs, the resulting model
fractionation curve will then be used to correct for the fractionation
of those elements whose photospheric abundances were not used
in calculating the model. The choice of adopted photospheric
normalization elements can be varied to get the most internally
consistent set of abundances.
The FIT fractionations are systematic effects which apply
to long term (years) average compositions. Although
there has been no systematic study reported in the
literature, the published time profiles of Ulysses data
suggest that there may be short term compositional fluctuations
in most elemental ratios. Similar variations
are better documented for
3He/4He
(Document A), but these
appear to converge, even on time scales of days, to a
well-defined long term average. The regular fractionation
systematics illustrated in Figure H-1, based on long term
averages, also argues for convergence to a well-defined set
of solar wind abundances for a given regime. For each regime,
sample return abundances are long term averages, and should
converge, after FIT corrections, to well-defined photospheric
values independent of regime, mission duration, when in the
solar cycle samples are taken, etc. Even if this analysis is overly
optimistic, convergence might occur for selected sets of elements
or for the data from a specific solar wind regime. Tests of
convergence are also possible. Independent compositional
measurements of the different solar wind regimes would again
play a major role. For example, if, despite large and variable
interregime variations in elements with high first ionization
potential, relative abundances for low FIP elements were the
same in all regimes, a good case could be made that average solar
system abundances were being measured for low FIP elements.
Alternatively, again for the subset of low FIP elements, close
agreement between solar wind and some CI
chondrite abundances would give confidence in the convergence
for many other elements. As a third example, if comparisons
of FIT-corrected solar wind and accurate photospheric
abundances showed a residual correlation with
charge to mass or similar variable, this trend could be used
to correct abundances for the large number of elements
with uncertain photospheric abundances.
Because FIT depends on atomic number, rather than mass, it is
an elemental rather than an isotopic variable. Isotopic
fractionation therefore will be much smaller than elemental
fractionation. No systematic long term, or inter regime
variations of He or Mg isotopic composition
are observed (Document A).
However, there are short term fluctuations in He isotopes (compare
Section 4, above) and the precision of the
Mg isotope ratios is around 5% at present. Thus, it is possible,
but not proved, that isotopic fractionation will be negligible even
for planetary science purposes. However, it will not be necessary
to assume this because, with our present instrument
package, isotopic data from different regimes can be compared.
Present data show that the amount of FIT elemental fractionation
varies with regime, thus it is very unlikely that the same mechanism
and/or amount of isotopic fractionation will occur in different
regimes. Thus, if isotopic fractionation exists, different solar wind
regimes would be expected to have different amounts of fractionation
leading to interregime isotope variations. The noble gases
will permit high precision comparisons of the regimes, to test
whether interregime isotopic differences exist. It is also
important to check isotope fractionations for elements which differ
greatly in mass, as the magnitude of any isotopic fractionation
will vary with mass. Thus, comparisons of isotopic abundances for
elements widely separated in mass are important; consequently,
measurements of C and Xe plus other noble gases, in different
regimes have a high priority (Measurement objectives 4 and 6,
proposal Table 1-2). Having data for three
regimes is important because a single anomalous regime can be
recognized. If all regimes show fractionations, this would be
significant information for solar physics, although a complication
for us. However, the increased theoretical understanding of the
isotope fractionation processes for different regimes would likely
generate the required corrections.
Please send comments on the html to wiens@gps.caltech.edu