Constraints on Solar Coronal Abundances from MESSENGER X-ray Solar Monitor Data

The chemical composition of the solar corona is fractionated from that of the photosphere, with elements having low first-ionization-potential (FIP <~10 eV) enriched relative to higher-FIP elements. However, the absolute normalization of coronal abundances relative to photospheric ones as well as possible time variation in coronal abundances (e.g., during flares) is a matter of longstanding controversy. Moreover, the effects of a recent downward revision in photospheric C, N, and especially O abundances on coronal composition have not yet been extensively studied. The shape of the soft X-ray spectrum emitted from the coronal plasma depends strongly on chemical composition both through emission lines and free-bound emission contributions to the continuum. The MESSENGER spacecraft, en route to orbit Mercury, includes a Si-PIN detector to monitor the solar X-ray spectrum (~1.5 to 8 keV) as part of an experiment to determine planetary surface composition via X-ray fluorescence. A pinhole and thin Be window in front of the PIN attenuate much of the flux below 2 keV, providing a high dynamic range in measuring the highly variable solar spectrum. The energy resolution of the solar monitor (~600 eV) is not sufficient to resolve individual solar lines but does allow line complexes of Ca and Fe to be distinguished from continuum during flares. In preparation for analysis of X-ray data from Mercury’s surface, we have begun a systematic effort to fit theoretical solar spectra to MESSENGER solar monitor data, using the CHIANTI 5.2 code and assuming isothermal plasma. The key fitting parameters are the plasma temperature, emission measure, and level of fractionation for low-FIP elements. Preliminary fitting of some 1400 individual spectra (300-450 s integration) from ~200 B-level and above solar flares during June-August 2010 reveals two interesting results: (1) The best fits are obtained for plasma with low-FIP elements enriched by a factor of ~2 relative to photospheric. This factor is comparable to some but not all previous flare X-ray analyses obtained by other spacecraft. (2) Significantly worse fits are found if photospheric CNO abundances are lowered from older abundance compilations. Since these high-FIP elements are not expected to be fractionated in the corona, this result does not apparently support the downward revision in photospheric abundances.