[Previous] | [Session 2] | [Next]
J.D. Allen, J.T. Schmelz (University of Memphis)
When attempting to compare models and observations of the solar corona, an accurate determination of the multi-thermal distribution of the plasma along the line of sight is clearly superior to the single value of temperature obtained from an isothermal approximation. But methods of determining the Differential Emission Measure (DEM) are fraught with pitfalls. Forward Folding is subjective and time consuming: an initial model DEM curve is convolved with the emissivity function of the spectral lines. This produces a set of predicted intensities that are compared with the observed values. The emission measure distribution is then adjusted manually to improve the agreement between the observed and predicted intensities while keeping the curve as smooth a function of temperature as possible. The process is repeated until, ideally, the predicted and observed intensities agree to within approximately 1-2 sigma. Automatic inversion techniques are mathematically ill posed, so small changed in the observed spectral line intensities can result in large differences in the DEM solution. The programs are too often used as a 'black box,' and the physical relevance of these solutions has (rightfully so) been questioned. We attempt to minimize these pitfalls by comparing and contrasting the temperature distributions obtained from forward folding with two different automatic inversion techniques. We used three different data sets: (1) SERTS iron line intensities of an active region from the 1993 Aug 17 rocket flight; (2) Solar Maximum Mission Flat Crystal Spectrometer data for a flare on 1980 Aug 30; and (3) XSST spectral data of a flare from the 1982 Jul 13 rocket flight. Solar physics research at the University of Memphis is supported by NASA grant NAG5-9783.
[Previous] | [Session 2] | [Next]
Bulletin of the American Astronomical Society, 34
© 2002. The American Astronomical Soceity.