 |
|
 |
A differential emission measure (DEM) analysis provides the temperature distribution of emitting material within a given solar feature. Treatment of the DEM has undergone substantial evolution since the initial treatment of the emission measure by Pottasch (1963; Withbroe 1975, Chapman 1981, Harrison & McWhirter 1991). The DEM of a given solar feature can be empirically derived as a function of the temperature f(T) by combining measured spectral line intensities with their corresponding emissivities per unit emission measure (generally known as the contribution functions). The numerical procedure used to evaluate the DEM in this work was developed in preparation for the SOHO mission (Monsignori-Fossi & Landini 1991, 1995). The method consists of an iterative procedure in which f(T) is described by a cubic spline function through a small number of n mesh points (f(T_i), i=1,...,n), and the f(T_i) are varied until the best fit with the observations is obtained. This method assures only positive solutions, and allows control of the smoothness of the solution through the number of mesh points. In order to obtain the best f(T) distribution it is important to select spectral lines which are relatively density-insensitive, and which cover a wide range in temperature. Large indeterminations may occur in temperature intervals where poor constraints are given by the available observations.
Results of the DEM analysis for the active and quiet regions observed by SERTS in 1991 and 1993 are shown in the following figures. The x-axis gives the log of the temperature in K, and the y-axis give the log of the DEM in cm^{-5} K^{-1}; the long error bars on the cool ends of the distributions indicate upper limits. All of the curves were derived by fitting available lines of Fe, Ne, Mg, Si, and Ni, assuming the elemental abundances of Feldman et al. (1992). The low temperature end of each DEM distribution was constrained by upper limits and/or relatively uncertain measurements of intensities of emission lines of O III, C IV, and Mg V. For the 1991 active region DEM, one Ca XVIII emission line was also used to help constrain the distribution at the high temperature end. Different symbols are used for the different elements in each figure. The x-coordinate of each symbol indicates the location of the temperature at which the product of the DEM and contribution function is maximized; this is the temperature of the maximum contribution to the line emission. The y-coordinate is the product of the DEM and the ratio of the observed to the predict intensities. The error bars on each symbol represent the uncertainties attributed to intensity measurements only.
The quiet Sun and the 1993 active region DEM curves show a peak between log T = 6.1 and 6.2; the 1993 active region DEM curve shows a second peak between log T = 6.6 and 6.7, and the 1991 active region DEM shows only one peak, between log T = 6.5 and 6.6. The presence of only the high-temperature peak in the latter DEM curve indicates a proportionately greater amount of hot plasma than in the 1993 active region. This may be a manifestation of enhanced activity in region 6615, as surmised from its greater measured photospheric magnetic fields or from its flaring activity.
Reference
"Measuring Active and Quiet Sun Coronal Plasma Properties with EUV Spectra from SERTS," J. W. Brosius, J. M. Davila, R. J. Thomas, and B. C. Monsignori-Fossi, Ap. J. Sup. 106, 143 (1996).
Last Revised: Wednesday, 29-Nov-2006 07:36:52 EST
Responsible NASA Official: joseph.davila@gsfc.nasa.gov
NASA iT Security Banner & Privacy Statement
Feedback and comments: Emilie Drobnes (curator)