AXAF/HETGS Spectroscopic Analysis Scenario: A Coronal Spectrum

AXAF/HETGS Spectroscopic Analysis Scenario:
A Coronal Spectrum

Contents:

  • Observational Program
  • Scientific Goals
  • Simple Analysis Scenario
  • Data Inspection and Preparation

    • Observational Program
    • Target: AR Lac with HETG/ ACIS-S (HETG GTO program)
    • Type: Stellar, Coronal, eclipsing binary.
    • Observations:
      • HETG/ACIS-S
        • primary eclipse 1 (40ks)
        • primary eclipse 2 (40ks)
        • quadrature 1 (5 ks)
        • quadrature 1 (5 ks)
        • quadrature 2 (5 ks)
        • quadrature 2 (5 ks)
    • P=1.998 days.
    • approx first order count-rate, 1.0 Hz.
    • approx flux f(0.1-10.0) = 40e-12 ergs/cm^2/s
    • Assume, get 10 3000 ct spectra through each eclipse => 20 spectra ; get 1 3000 ct spectra for each quadrature => 4 spectra => 24 spectra to analyze.
    • In addition to phase-selected time intervals, select based on presence of any flares: pre, high, decay (as count-rate allows).

    Scientific Goals

     This is a list of quantities to be derived from the spectra. They range from purely empirical diagnostics, to ballpark estimates which indicate the range or scale of the physical conditions, to detailed quantitative models (some of which are open-ended topics of fundamental scientific interest).

    Item Goals Difficulty Remarks Ref. Items
    a f(t), f(phi) easy Is there eclipse modulation? Modulation pre/post eclipse? Integrated flux (or counts) light-curve. 26
    b f(line,phi(t)) easy Does the modulation depend upon temperature? Line-strength vs phase light-curve. 27, 2..6
    c T(t), Ne(t) hard Model flares to determine density and temperature vs time (e.g., 2-ribbon models) 6, 11..13
    d DEM(T,t) low-fi easy How hot is it? Low-fidelity EM, as ratio rather than integral: EM = flux/ <G(T,Ne)*stuff > 6, 8, 10
    e Ne moderate What are approximate electron densities? (from line ratios) 11..13, 8, 9
    f Te moderate What are approximate electron temperatures? (from line ratios) 8, 9, 11..13
    g sigma(T,f) easy Are lines broadened (thermally or in flares)? Compare Gaussian fitted width to instrumental width for different lines. 6, 15
    h lambda(phi,f) easy Are lines shifted by flares? Orbital phase? 6, 9
    i fcontin(lambda,t) moderate What is continuum level and shape? Is it altered by flares? phase? 5, 6
    j DEM(T,t) hi-fi hard Detailed analysis: what is temperature structure (vs ion, species, abundance)? "Solve" integral over emissivity, emission measure by one of many possible methods. 6, 8, 9, 11..14
    k Abundance vs IP hard Detailed analysis: are there abundance anomalies? E.g., does abundance depend upon Ionization Potential? (coupled to detailed DEM analysis) 6, 8, 9, 11..14
    l f(lambda,T,t) hard Detailed line-shapes: is broadening instrumental? thermal? turbulent? flare-dependent? 6, 15
    m f(theta,phi,r) hard Use modulation, DEM(T,t) to geometrically map emitting regions. 6, 10, 11
    n Finput(E,t,A,h,B) very hard Determine the heating function as a function of everything, or at least, a heating function consistent with the observable radiative loss and mass motions. 6, 10, 11, ...

    Simple Analysis Scenario

    • Optionally non-interactive analysis:
      1. Select spectrum (grating, order).
      2. Detect features. Use default criteria on S/N, width, separation. Do in counts or flux. (Width and separation are saved w/ each feature.)
      3. Tentatively identify features (may require specification of temp. range). Use default line list with emissivities and temperature dependence.

    • Optionally interactive analysis (usually interactive first time, may be scripted for another time):
      1. Look at selected lines. Select interesting features.
      2. Specify type of continuum to use (none, explicitly marked, median value, minimum value, polynomial, model#response, other).
      3. Measure features: position, flux, width, w/ uncertainties; add to measurements table. Flag if wider than instrumental resolution. Options: instrumental width or free width parameter; fit method (Gaussian, Lorentzian, instrumental LSF (RMF).
        Table will have: n, wavelength, +-, width, +-, Tentative IDs, Continuum, +-
        Issue: emission file identification strings should be compatible with various plasma codes.
        Issue: Interface to customized databases should be possible (e.g. XSTAR, SPEX, NEI)
      4. Show instrumental features or response. Flag uncertain regions (add qualifier to table).

    • Emission modeling begins here:
      1. Browse feature characteristics, via plots or tables: Tmax, Ne sensitivity, ionization potential, A-values, relative emissivity, ionization fraction, excitation fraction, branching ratios and pairs (common upper-levels).
      2. Query on database keys: show lines of given species, ion, H- or He-like. Plot line positions; plot models. Plot residuals (e.g., measured wavelength - database wavelength); mark outliers, flag in table.
      3. Plot simple DEM of interesting features. (k*flux/<G(Tmax)> ); overplot k*flux/<G(T)> vs T.
      4. Define interesting line ratios.
      5. Plot database ratio vs Tmax (or Ne), and plot measured ratio point.
      6. Plot ratios vs other ratios as contours in Tmax (from DB). Overplot measured ratios point.
      7. Look for anomalies: normalize emissivity scale on one line flux; show expected strengths of other lines for some Te, Ne, abundance (or weighted sum over (Tmax)i).
      8. Analyze line profiles. Specify intrinsic broadening (thermal, turbulent). Plot predicted profile#instrument. Fit/compare profile to data.
      9. Write table of identified features and properties (wavelength, width, flux, other info).
    Steps can be saved, edited, and used as a script. (e.g., for application to another spectrum, or order).

    Issue: Do we require joint analysis; e.g., performing same actions on multiple spectra (e.g., HEG and MEG), in constrained way?

    Data Inspection and Preparation

     These are some of the operations one may wish to perform prior to the spectroscopic analysis in order to ensure that the spectra have been properly extracted. Some concerns may be to check the wavelength scale or order-sorting accuracy. There may be confusing sources in the field. If necessary, re-define parameters and re-extract. (NOTE: This is interactive Level 1.5 and Level 2 processing, for which there are no specs.)
    1. Plot MEG +1 order spectrum count-rate
    2. Over-plot MEG -1 order spectrum count-rate; do they align?
    3. Cross-correlate +1,-1, verify wavelength scale, or determine relative shift as wavelength correction (vs order)
    4. Ditto for HEG +1, -1
    5. View sky image, overlay grating mask regions; are masks proper angles? widths? adjust as necessary.
    6. Adjust zero-order centroid; save ZO, wavelength offsets.
    7. View spectral-spatial image (tg_R, tg_D), by HEG, MEG, +- parts.
    8. Overlay wavelength and energy scales, prominent line positions.
    9. View dispersion, pulse-height distribution, vs HEG, MEG, color-coded by order; Overlay pulse-height region.
    10. View light-curves (zero-order, or summed counts or flux in diffracted orders' regions); define temporal regions.
    11. Adjust regions, re-extract L1.5 events and binned spectra, if necessary.




    For further information, contact
    David Huenemoerder
    (617) 253-4283

    Last update: 06 January 1999


    This page is: http://space.mit.edu/ASC/analysis/Scenarios/Scenario_AR_Lac.html