RHESSI Working Group 4, GSFC meeting June 21, 2002

Fletcher: introductory comments for joint discussion with WG 3 on the
  spectral mismatch catastrophe that's developing

White

  * Radio data for July 23, NoRP fixed frequencies => high peak
    frequency
  * Nobeyama factor of 6.5 at highest frequency - is there RSTN or maybe
    other data?
  * 35-80 GHz radio spectral index "wrong", explained in terms of
    optically-thin gyrosynchrotron emission: radio -2.5 or -3 energy
    index, as opposed to X-ray -4 or 4.5. If there were residual optical
    depth at 35 GHz; high harmonics give upper limit of 200 G for B.
    The only way to get such opacity would therefore be non-thermal
    particle density
  * Silva, Wang, Gary paper with many examples of this problem; flux 
    correlation looks normal (BFS-like). A bimodal distribution of
    spectral indices suggesting some agreements, but with others
    interpreted as broken power laws. Early paper by Benz and Gold
    recalled.
  * Comment (Hurford) about distributions of parameters (source
    inhomogeneity, filling factor). Need source maps to match at the
    two high frequencies used for spectral determination.
  * Model radio spectra. Discussion of bounce time scales, beaming, 
    how one compares the semi-trapped coronal population with the 
    thick-target rate. "No trapping" (White) may allow trapping on
    bounce time scale, which could in princple affect the spectral
    distribution (comments by many)
  * Need to compare absolute fluxes (Hurford comment) to learn about
    spectral modification during loss-cone replenishment. Suggestion
    of wave scattering with different spectral dependence from ordinary
    Coulomb or bremsstrahlung collisions (Hudson comment)
  * What measurements do we need? (Gary query) - FASR!
  * Standard bremsstrahlung ideas => alpha = -1.20 +0.90*delta
    for the radio flux spectral-index comparison, in isotropic
    distribution
  * Radio spectra: to get opacity at high frequencies, one really needs
    relativistic electrons.  Recall McTiernan thesis regarding
    transition to relativistic cross-section
  * Discussion of the nature of the spectral flattening observed in the
    gamma-ray continuum. 
  * Conclusion. We are still perplexed, but some more detailed modeling
    might really straighten things out.
  * July 23 footnote. The pre-flare non-thermal hard X-ray coronal
    source... has no counterpart at 35 GHz, at least on a linear plot.
    Need to look at a log plot!
  
Melnikov

  * Modeling of joint spectral variation hard X-ray and radio spectra.
    The comparison is highly spectrum-dependent.  Discussion: Is the 
    Dulk-Marsh formula so wrong?
  * Dependence on plasma density. Worry about Razin-Tsytovich effect
    at 6-10 GHz
  * Magun mm-wave observations (March 13, 1991, for example) temporal
    flattening of spectral index comparing 35-50 GHz on recovery from
    impulsive burst, as opposed to softening in SHS pattern
  * Melnikov, Gary, Nita (2003) OVRO data finds most bursts to show
    this flattening. Spectral evolution of trapped electrons is blamed
    for this behavior via Coulomb collisions.

Joint discussion with another WG

Phillips tutorial on "Fe feature"

  * Never see FeXXII or lower
  * May 21, 1980 famous flare at peak shows no evidence even at FeII
    (or other) K-line
  * Line "w" and other members of a huge Chianti line table in this
    spectral domain
  * K-alpha nitty-gritty: there is a pair of lines at 6.4 keV, in 1:2
    ratio, almost certainly formed via flourescence (radiative
    excitation of inner shell). One gets completely screened 2s-1p
    transitions.
  * Collisional excitation of K shell? Limb darkening expected from
    absorption in spicules or whatever... One also has a time-variation
    signature which is a clinching argument in favor of flourescence
  * What about the continuum? BCS spectra were not good for that.
  * Phillips-Neupert paper re direct collisional excitation (March 29
    1980 huge spike event)
  * "The whole world is going towards Chianti..." and it is very useful
    stuff. Note that it will be adapted to kevocentric users soon, we
    hope. Meanwhile use ch_ss.pro
  * Smeared Chianti spectra, Mazzotta ionization equilibria, show Fe-Ni
    mixture above 7.5 keV at v. high temperatures (e.g. > 20 MK). At
    100 MK see v. strong FeXXVI complex plus many 8-9 keV line groups
  * Discussion of whether or not the Mazzotta revision of the ionization
    balances is "revolutionary" or not. Nice plot compares
    Arnaud-Rothenflug with Mazzotta, showing main differences at lowest
    temperatures
  * Antonucci et al. 1985 discusses the temperature dependence
  * Hudson comment about relying on the observations, not the models,
    and a plea for help in understanding the RHESSI Fe observations,
    which seem to have no counterpart in the earlier spectroscopic data
  * Abundances? Meyer 1985; Feldman-Laming 2000; Fludra-Schmelz - FIP 
    effect vagaries.  Low FIP enhanced by 3-4 x photospheric in the corona?
    Mason-Young 1997 on spatial dependences 
  * NEAR spectrum at 0.7 keV resolution from a PIN Si detector seems to
    show clear DEM effects. Also there are data from RESIK, now alas
    suffering from a power-supply problem

Plenary session

  * Fletcher presentation

    - Recommendation for a "clearing house" Web repository for events
      and analysis types for each
    - Warnings about time stamps in various instruments' FITS files;
      other things too such as aspect information (e.g. MDI roll
      correction)