PDS_VERSION_ID = PDS3 RECORD_TYPE = STREAM OBJECT = TEXT INTERCHANGE_FORMAT = ASCII PUBLICATION_DATE = 2001-09-01 NOTE = "N/A" END_OBJECT = TEXT END This writeup is was developed as part of the NEAR XGRS submission to NASA's Planetary Data System (PDS). The contents are specific to the NEAR XRS Level 3 partition of the dataset and are from the Nittler, 2001 MAPS paper. Tables 1-4 are listed at the end of the first section. XRS Results ----------- Elemental ratios for Eros have been determined for areas sampled by the XRS during five solar flares and two extended integrations during quiescent solar conditions. Some basic characteristics for these data are provided in Table 1, including the date and UT of each flare, the total integration time used in spectral sums, the average latitude and longitude of the instrument footprint on the asteroid surface and the average temperature of the solar plasma, determined from the Geostationary Operational Environmental Satellite (GOES-8). GOES measures the solar x-ray flux in two wavelength bands every three seconds. Isothermal plasma temperatures are determined from the flux measurements by convolving theoretical solar spectra with instrument response functions to generate calibration curves. The estimated uncertainty in derived temperature is 10-20%. Background-subtracted pulse-height spectra for the five flares and two quiet Sun integration periods are given in Figures 2 and 3. For each flare spectrum, the background is the sum of all spectra acquired the same day during periods when the instrument field of view did not include solar-illuminated asteroid and scaled to the same integration time as the flare spectral sum. Background for quiet Sun spectra are sums of all off-pointing spectra acquired over the same time period as the summed fluorescence spectrum. Errors on indivudal channel counts are based on counting statistics. Each asteroid-pointing spectrum was fitted using a non-linear least-squares technique to extract count rates of characteristic photons from Mg, Al,Si, S, Ca and Fe incident on the XRS. The best spectral fits are included in Figures 2 and 3. From the fits, photon ratios Mg/Si, Al/Si,S/Si, Ca/Si and S/Si were derived, along with statistical errors. These are provided in Table 3. Conversion of x-ray photon ratios to elemental abundance ratios requires knowledge of the incident solar x-ray spectrum. For the Eros flare spectra, solar spectra acquired by the solar monitor detector during the same integration period were fitted to derive plasma temperatures. This fitting procedure involved convolving theoretical solar spectra of different temperatures through the response function of the solar monitor filter and minimizing the difference between synthetic and observed spectra. For each flare, two fits were performed: F1 was a two temperature fit to the full solar monitor spectrum; F2 is a single temperature fit to the high energy portion (5-10 keV) of the spectrum. For more discussion of solar modeling see Figure 4 and GOES, Solar Modeling discussion in appendix of this paper. Given an observation geometry (determined from a shape model of Eros and spacecraft pointing data) and a solar spectrum, theoretical fluorescence and coherent backscatter models are used to generate calibration curves relating photon ratios to elemental ratios. Standard analytical models of x-ray scatter and fluorescence were performed for a wide range of compositions, typical of different meteorite types, and solar temperatures. Figure 6 gives example calibration curves for Mg/Si and Fe/Si. The x-ray models assume a homogeneous composition of the observed sample, but in real rocks different elements are contained in different minerals. To account for this, a correction is applied to elemental ratios derived from the model calibration curves. The correction factor calculated for four meteorite types is given in Table 2; for the results included here the average chondrite correction factor (last column) was applied. Elemental ratios derived for each flare and quiet sun spectrum are given in Table 3. Flare results for solar monitor fits F1 and F2 are given separately. S/Si, Ca/Si and Fe/Si ratios are only derived for the brighter flares for which sufficient signal was obtained. The average element ratios of Eros, determined from the individual analyses, is provided in Table 4. The F1 results for the May 4 and June 15 flares and the F2 result for the December 27 flare are believed to be in error, due to errors in solar modeling, and are not included in the average. The standard deviation of individual S/Si results is within 1 sigma of zero and a two-sigma upper limit is provided for this ratio. For the other ratios, three error estimates are provided. Sigma_stat is the statistical error bar, based on the standard error of the mean of the individual results. Sigma_var is the standard deviation of individual results; this includes both the effects of any km-scale heterogeneity on the asteroids' surface and a systematic error due to incorrect modeling of solar monitor spectra. Sigma_sys is the total systematic error. This was calculated in quadrature from three sources of error: the error in the mineral mixing correction factor (Table 2), a 10% error due to uncertain elemental abundance variations in the solar corona, and sigma_var to account for errors in solar modeling. Table 1: Basic data for analyzed solar flares and quiet Sun integrations ______________________________________________________________________________________ Spectrum UT Total Lat Long Average GOES Integration (deg) (deg) Temperature (MK) time (s) May 4, 2000 Flare 4:40 950 17 146 14.5 Jun 15, 2000 Flare 19:49 2250 -22 153 11.9 Jul 10, 2000 Flare 22:00 3750 6 266 10.6 Dec 27, 2000 Flare 15:40 1000 -5 175 14.2 Jan 2, 2001 Flare 7:52 1700 -27 106 12.7 Quiet Sun, May 2 - Aug 27, 2000 --- 402800 --- --- 4.23 +/- 0.4 Quiet Sun, Dec 12, 2000 - Feb 10, 2001 --- 357500 --- --- 3.89 +/- 0.4 Addendum to Table1: Mission Event Time Range of Flare(MET): May 4: MET=132826285-132827185 June 15:MET=136508736-136510936 Jul 10 MET=138677136-138680836 Dec 27: MET=153341842-153342792 Jan 2: MET=153832142-153833792 Table 2: Effect of assumed hetergeneous (mineral mixing) versus homogeneous compositions of predicted XRF photon ratios. ______________________________________________________________ Ratio Ratio of homogeneous model to mineral mixing modeli R Chon. LL Chon. H Chon. Eucrite Chondrite Average Mg/Si 1.11 1.05 0.95 1.23 1.04 +/- 0.08 Al/Si 0.7 0.76 0.8 0.71 0.76 +/- 0.05 S/Si 1.28 1.29 1.45 1.08 1.34 +/- 0.10 Ca/Si 1.05 1.08 1.25 0.95 1.12 +/- 0.10 Fe/Si 1.18 1.29 1.78 1.2 1.42 +/- 0.32 Table 3: Derived photon and elemental abudnance ratio data for Eros(*): ___________________________________________________________________________________________________________________ Spectrum Ratio Mg/Si Al/Si S/Si Ca/Si Fe/Si 4-May-00 Photon 0.93 +/- 0.03 0.08 +/- 0.02 0.055 +/- 0.007 0.052 +/- 0.004 0.101 +/- 0.004 Flare Element: F1 0.64 +/- 0.02 0.04 +/- 0.02 0.049 +/- 0.015 0.201 +/- 0.026 4.32 +/- 0.22 Element: F2 0.88 +/- 0.03 0.04 +/- 0.02 0.026 +/- 0.008 0.081 +/- 0.010 2.03 +/- 0.10 15-Jun-00 Photon 0.96 +/- 0.070 0.12 +/- 0.05 0.023 +/- 0.017 0.047 +/- 0.012 0.077 +/- 0.009 Flare Element: F1 0.70 +/- 0.05 0.081 +/- 0.045 0.004 +/- 0.011 0.152 +/- 0.066 3.12 +/- 0.46 Element: F2 0.90 +/- 0.06 0.083 +/- 0.045 0.001 +/- 0.009 0.079 +/- 0.034 1.75 +/- 0.24 10-Jul-00 Photon 1.16 +/- 0.05 0.10 +/- 0.04 0.009 +/- 0.014 --- --- Flare Element: F1 0.86 +/- 0.04 0.06 +/- 0.03 0.0004 +/- 0.0004 --- --- Element: F2 1.07 +/- 0.04 0.07 +/- 0.03 <0.004 --- --- 27-Dec-00 Photon 0.97 +/- 0.04 0.08 +/- 0.03 0.064 +/- 0.008 0.054 +/- 0.005 0.105 +/- 0.005 Flare Element: F1 0.81 +/- 0.03 0.05 +/- 0.03 0.047 +/- 0.013 0.082 +/- 0.012 1.32 +/- 0.08 Element: F2 0.97 +/- 0.04 0.05 +/- 0.03 0.032 +/- 0.009 0.043 +/- 0.007 0.75 +/- 0.04 02-Jan-01 Photon 0.93 +/- 0.04 0.14 +/- 0.03 0.036 +/- 0.010 0.045 +/- 0.007 0.080 +/- 0.005 Flare Element: F1 0.84 +/- 0.03 0.10 +/- 0.03 0.011 +/- 0.009 0.077 +/- 0.020 1.69 +/- 0.12 Element: F2 0.89 +/- 0.04 0.10 +/- 0.03 0.010 +/- 0.008 0.067 +/- 0.017 1.48 +/- 0.11 Quiet Sun Photon 2.19 +/- 0.08 0.16 +/- 0.04 --- --- --- Summer Element 0.72 +/- 0.10 0.05 +/- 0.03 --- --- --- Quiet Sun Photon 2.44 +/- 0.11 0.19 +/- 0.05 --- --- --- Winter Element 0.70 +/- 0.11 0.05 +/- 0.03 --- --- --- M ____________________________________________________________________________________________________________________ * For solar flares, element ratios were derived for two theoretical solar spectra: F1: two-temperature fit to solar monitor, F2: single temperature fit to high energy portion of solar monitor (see text). Element ratios for quiet Sun were derived using GOES temperatures (Table 1). Table 4: Average surface composition of 433 Eros Mg/Si Al/Si S/Si Ca/Si Fe/Si ratio 0.85 0.068 0.05 0.077 1.65 sigma_stat 0.04 0.007 0.003 0.12 sigma_var 0.11 0.022 0.006 0.27 sigma_sys 0.15 0.023 0.012 0.49 Appendix: --------- GOES Processing Issues: ----------------------- From the ratio of the two channels the temperature is computed either from the parameterized fit in Thomas et al (Solar Physics v95, 1983) or by interpolating from a lookup table computed by folding the GOES transfer functions with the MEWE emission line and continuum spectrum calculated by MEWE_SPEC.PRO. The responses of the GOES ionization chamber were obtained from Donnelly et al 1977. The responses for GOES8+ were obtained by a private communication from Howard Garcia to Hugh Hudson subsequently passed along to Richard Schwartz and implemented in July 1996. Tables for GOES6 and GOES7 were communicated by Garcia to Schwartz on 10 Oct 1996. The most recent version, Nov 22 1996, corrects the Thomas parameterization by including the change in definition for the transmission-averaged short wavelength flux which occurred from GOES4 onward. These are reported in the GBAR_TABLE produced by GOES_TRANSFER.PRO While there are real changes in the response between versions of GOES, this was a simple change of divisor from (4-0.5) to (3-0.5) and not in any real difference between detectors such as a change in thickness of the beryllium window. This change of definition results in the higher values found in Table 1 of Garcia, Sol Phys, v154, p275. This issue is disussed on page 284 in paragraph 2 of section 3.1 of Garcia. An alternative interpolation table was created by Howard Garcia and reported by private communication and partially published in Solar Physics v154, p275. The tables for GOES2, GOES6, GOES7, and GOES8 have been renormalized to 1e49 cm-3 and the ratio taken the ratio of the long current divided by the short current as given by Garcia. References for Solar Modeling and Temperature --------------------------------------------- Extensive calculations of solar emission over a wide range of temperatures (4 to 30 million Kelvins) include both continuum and line emissions from less than 1 to greater than 10 KeV. Continuum is derived from computations of Landini and Fossi (1970) for coronal continuum production (free-free+free-bound radiation) with a correction (Karzas and Latter, 1961) for free-free emission. Calculation of the extensive spectral lines in the soft X-ray region for major elements C, N, O, Ne, Na, Mg, Al, Si, S, Ar, K, Ca, Ti, Cr, Mn, Fe, and Ni, that result from both resonant and non-resonant transitions, are based on the work of Landini and Fossi (1970) and Mewe (1972), with improved excitation cross-sections provided by Mewe (1972). (See Mewe, 1972, Calculated solar x-radiation from 1 to 60 Angstroms, Solar Physics 22, 459-491.)