fl037.doschek02 Posted: 17-Feb-92 Updated: 03-Sep-92, 06-Jun-93, 22-Aug-95 Events specified: N/A
Collaborators: George Doschek (11334::DOSCHEK), John Mariska, Uri Feldman, Alphonse Sterling, Andrzej Fludra, Leonard Culhane, PLUS ANY OTHER WORKERS WHO WOULD LIKE TO PARTICIPATE.
The BCS crystal spectrometers are much more sensitive than the crystal spectrometers flown on P78-1, SMM, and Hinotori. As a result, BCS obtains spectra of Fe XXV and lower stages of ionization much closer to flare onset than possible with the previous instruments. It is therefore possible to investigate the change in electron temperature as measured by the temperature diagnostic, Fe XXIV(j)/Fe XXV(w), as a function of time during the rise phase of flares. This information is important for deriving the emission measure of the Fe XXV producing region, which in turn is important for comparison to numerical simulations of chromospheric evaporation. In addition, the j/w temperature can be used to compute ion abundances for Fe XXII - Fe XXIV, and the results can be compared to predictions of numerical simulations assuming ionization equilibrium. The technique involves separating the rise-phase blueshifted iron spectrum from the stationary iron spectrum using spectral synthesis techniques. The results of this give the j/w ratio as well as empirical ion abundances for Fe XXII - Fe XXIV relative to Fe XXV. In an event like the 9 Nov. flare, the entire profile appears to be blueshifted and for the earliest stages only one spectral component might need to be considered. For other events, the blueshifted component is small and emission is dominated by the stationary component.
Update 22-Aug-95
A paper describing this work has been published in The Astrophysical Journal, 435, 898 (1994). The title, author list, and abstract for that paper are as follows:
Title: Fe XXV TEMPERATURES IN FLARES FROM THE YOHKOH BRAGG CRYSTAL SPECTROMETER
Authors: Sterling, A. C., Doschek, G. A., & Pike, C. D.
Abstract: Studies by Doschek et al. (1990) using P78-1 and SMM data have shown that the ratio of intensities of the Fe XXV and Ca XIX resonance lines can be expressed as a function of Fe XXV temperature. Using a more recent data set consisting of 13 flares observed by the Bragg crystal spectrometer (BCS) experiment on board Yohkoh, we find a nearly identical functional relationship between the same resonance line ratios and Fe XXV temperatures. We use this functional relationship to obtain resonance line ratio temperatures (T_{RLR}) for each flare in our data set, and compare them with temperatures resulting from application of a simple spectral fitting method (T_{SSF}) to individual Fe XXV spectra. We also use a more involved free-parameter spectral fitting method to deduce temperatures (T_{FSF}) from some of these spectra. On average, agreement between T_{RLR} and T_{SSF} improves as a flare progresses in time, with average agreements of 10.0% +- 5.2%, 6.4% +- 5.4%, and 5.0% +- 3.9% over the rise, peak, and decay phases, respectively. Deviations between T_{RLR} and T_{FSF} are about the same or smaller. Thus, for most analysis purposes, all three methods yield virtually identical temperatures in flares. The somewhat poorer agreement between T_{SSF} and T_{RLR} during the earlier phases may be partially a result of difficulties in obtaining precise values for temperatures from spectral fits when blueshifts and large non-thermal broadenings are present in the spectra. Because of the high sensitivity of the {\it Yohkoh} BCS compared to that of BCS experiments on earlier spacecraft, we can for the first time consistently observe the heating phase of flares in Fe XXV.
Update 6-Jun-93
Close investigations of our work since the previous update reveal that the spectral fitting code we use to derive Fe XXV temperatures (from the w-to-j line ratio) must be applied with caution during the early phases of flares. Using more sophisticated, two-component voigt codes, we have found that we previously often underestimated rise phase flare temperatures by about 1-3 MK.
The magnitude of the underestimate is related to the amount of blue shift in the event. We have now reworked our analysis to account for these effects, and are in the process of summarizing our results so far.
Update 3-Sep-92
We are continuing to analyze electron temperature behavior as a function of time in flares. So far we have studied about ten events for which we have good rise phase data. Flare temperature variations with time are well represented by examining the Fe XXV(j) to Ca XIX(j) resonance line intensity ratios. We compare temperatures estimated in this manner with temperatures measured by the Fe XXIV(j)/Fe XXV(w) ratio at different times in individual events. Because of the increased sensitivity of the Yohkoh BCS compared to the BCS experiments flown on earlier spacecraft, we are able to estimate temperatures earlier in flares than previously possible. We have found, for example, in one case (4 Jan 92, near 11:06 UT) that we are able to follow Fe XXV electron temperatures from the relatively cool value of approximately 12 - 13 MK early in the event's rise phase, to approximately 17 MK near the peak of the event. This latter temperature is a typical value quoted for flares in studies prior to Yohkoh.