Explosive Chromospheric Evaporation and Warm Rain in a C3.1 Flare Observed by EIS, IRIS, and RHESSI

Jeff Brosius


 






In order to understand the mechanisms that trigger and drive solar flares, as well as the processes that occur during flares, we need spectroscopic observations at cadences comparable to the timescales on which flares evolve. IHOP 241 was designed with this goal in mind. Here we present results from coordinated, rapid cadence IRIS and EIS stare spectra of a GOES C3.1 flare in AR 12002 (S21 W14) obtained through IHOP 241 on 2014 March 15, along with coordinated hard X-ray observations from RHESSI. We find that flare emission in chromospheric and transition region lines showed an earlier start than did emission from plasma at "flare temperatures" (~10 MK), and that explosively evaporated material observed by IRIS fell as warm rain and accumulated in a separate area observed by EIS. For a more complete description of the observations, analysis, results, and conclusions see Brosius & Inglis (2017).

In GOES soft X-rays, the flare started at 00:21:35 and peaked at 00:26:30 UT. The IRIS slit was pointed nicely within the flare, and the EIS slit, deliberately offset 35 arcsec west of the IRIS slit, observed areas remote from the flare. See Figures 1 & 2.

Figure 1: IRIS 1400 Å slit-jaw images at two different times during the flare, with the IRIS and EIS slits overplotted. The horizontal yellow lines across the IRIS slit indicate the 4-arcsec segment in which the Fe XXI 1354.1 Å line intensity became brightest during the flare, and the horizontal lines across the EIS slit indicate the two 8-arcsec segments in which weak activity became evident. Images and spectra show that flare activity began at the location of the IRIS slit several minutes before the 00:21:35 UT start of the flare in GOES observations.




Figure 2: Image from AIA's 131 Å channel at 00:20:08 UT, after the flare was observed to begin in IRIS slit spectra (00:16:56 UT) but before the start of the flare observed in the GOES soft X-ray light curves (00:21:35 UT). IRIS and EIS slits are overplotted as labeled, and the horizontal lines have the same meaning as in Figure 1. Contours of the HMI line-of-sight magnetogram obtained at 00:19:14 UT are overplotted, where red indicates inward-directed fields of -100 and -500 G, and white indicates outward-directed fields of +100 and +500 G.


With IRIS spectra, we focus on the time evolution of intensities and relative Doppler velocities in emission lines of C II 1334.5, C II 1335.7, Si IV 1402.8, and Fe XXI 1354.1Å. The exposure duration was 15 s for each of the 1000 IRIS stare spectral exposures, and the average cadence was 16.5 s. In order to reduce the noise, for every exposure we averaged the spectra into 1-arcsec segments, that is, we averaged over six slit spatial pixels (subsequently identified with a YPIX6 number). We obtained pre-flare reference wavelengths for C II and Si IV (and, indirectly, for Fe XXI) by averaging spectra from 130 exposures before the flare over a 27-arcsec segment of the IRIS slit in the active region. Because the C I wavelength in the pre-flare averaged spectrum does not differ significantly from its 1354.288 Å rest wavelength (Kelly 1987), we adopt the rest wavelength of Fe XXI (1354.0714 +/- 0.0108 Å) derived by Brosius & Daw (2015) as its reference wavelength. Because the C II and Si IV line profiles became complex during the flare (as well as at other times), and do not conform to a single Gaussian shape, we fit the profiles of these three lines with three Gaussians each. Figure 3 shows sample line profiles during the flare.

Figure 3: Sample line profiles of Fe XXI and Si IV at 00:25:43 UT, averaged over a 1-arcsec segment of the IRIS slit within the flare. The reference wavelengths are given in each frame and plotted as long vertical lines toward the bottom. The Fe XXI line is blueshifted, with a relative Doppler velocity of -42.5 +/- 4.1 km s-1, while C I is not significantly redshifted with a velocity of +5.5 +/- 6.4 km s-1 . The Si IV profile's redshift corresponds to +30.5 +/- 6.3 km -1, and the simultaneous, redshifted C II profiles at 1334.5 and 1335.7 Å yield velocities of +32.1 +/- 3.6 km s-1 and +35.4 +/- 4.3 km s-1. The combination of redshifted low-temperature emission and blueshifted Fe XXI emission indicates that this flare underwent explosive chromospheric evaporation. Single Gaussian profile fits were applied to the weak blending lines of Fe II and Si II near Fe XXI (e.g., Young et al. 2015, Brosius & Daw 2015), with the overall fit to all lines in the spectrum plotted in red.


About 4 minutes before the start of the flare was observed by GOES, the C II and Si IV line intensities observed by IRIS became (and remained) significantly greater than their pre-flare average values; this indicates that the flare had begun and that the chromosphere and transition region were involved. IRIS first detected significant, blueshifted Fe XXI emission at 00:22:42 UT, by which time the C II and Si IV line intensities had increased by factors around 100 and their profiles were significantly redshifted. This combination of simultaneous, cospatial blueshifted Fe XXI emission with redshifted C II and Si IV emission indicates explosive chromospheric evaporation. See Figure 4.



Figure 4: Light curves and relative Doppler velocities derived with IRIS spectra in a 1-arcsec segment of the IRIS slit (YPIX6=29). Curves are color-coded as indicated, and the Si IV intensity is reduced by a factor of 10 to enhance visibility. Uncertainties are overplotted for all but the Si IV velocity, where we avoid doing so to enhance visibility. The overplotted X's indicate measurements derived from spectra affected by saturated spectral pixels; for the sake of clarity we avoided overplotting the X's associated with the Si IV velocity measurements. The horizontal color-coded lines in frame (a) indicate the pre-flare quiescent average intensities plus their associated 1-sigma scatters. The vertical orange line in frame (a) indicates the 00:16:56 UT time of the first exposure (391) at which the C II line intensities continuously significantly exceeded their pre-flare quiescent averages. We interpret this to be the start of the flare at this location along the IRIS slit. The vertical orange line in frame (b) indicates the 00:21:03 UT time of the first exposure (406) after which the C II and Si IV lines exhibited continuous significant redshift during the flare. We infer gentle evaporation from 00:16:56 to 00:21:03 UT, followed by explosive evaporation once C II and Si IV were redshifted, and certainly by the time Fe XXI was observed and blueshifted (00:22:42 UT).


EUV spectra were obtained with the EIS study FLAREDOP_EIS, for which a detailed description is given by Brosius, Daw, & Inglis (2016). EIS began observing at 23:50:06 UT on 2014 March 14, and tracked the target until its observing run ended at 00:45:45 UT on 2014 March 15. The exposure duration was 10 s for each of the 300 EIS stare spectral exposures, and the average cadence was 11.2 s. In order to reduce the noise, for every exposure we averaged the spectra into 4-arcsec segments, that is, we averaged over four slit spatial pixels (subsequently identified with a YPIX4 number). We corrected for the well-known drift in the EIS wavelength scale using the procedure described by Kamio et al. (2010, 2011). With EIS spectra, we focus on the time evolution of intensities and relative Doppler velocities in emission of Fe XVI at 262.9, Fe XXIII at 263.8, Fe XIV at 264.8 and 274.2, and Si VII at 275.4 Å. EIS, like IRIS, contains no absolute wavelength scale, so flare velocity measurements derived with EIS spectra must be done relative to times and/or locations outside and away from flares. Here we obtain reference wavelengths for the above lines from exposures obtained between 00:00:09 and 00:18:02 UT on 2014 March 15 within the southernmost 64-arcsec segment of the slit. Unfortunately, at the location of the EIS slit, Fe XXIII emission is so weak and short-lived that no reference wavelength can be measured, so no reliable relative Doppler velocities can be derived.

At the location of the EIS slit, in a localized area of enhanced magnetic field strength seen in the HMI magnetogram that appears to be connected to the flare site by faint loops evident in AIA 131 Å emission (see Figure 2), EIS spectra reveal Fe XXIII emission that is too weak to measure velocities; intensity enhancements by factors up to about 1.7 in the Fe XIV and Fe XVI line emission are observed at this location. Lines from both of these coronal ions show redshifts about 9 km s-1 around 00:24:00 UT. See Figure 5.

Figure 5: Flare light curves (a) and relative Doppler velocities (b) derived from EIS spectra in YPIX4 = 0. Curves are color-coded as indicated, and some have been scaled or shifted to enhance visibility. In frame (a), the Fe XVI brightening starts at 00:23:59 UT, Fe XIV at 00:23:37 UT, and Fe XXIII at 00:23:37 UT; Si VII shows no significant brightening at this location, but its intensity becomes slilghtly enhanced around 00:25:06 UT. In frame (b) Fe XVI exhibits a small but significant redshift starting at 00:23:26 UT, and Fe XIV exhibits a small but significant redshift starting at 00:23:04 UT.


The density sensitive line intensity ratio of Fe XIV 264.8/274.2 observed by EIS reveals an increase of electron density from (1.03 +/- 0.20) X 109 before the flare to (3.58 +/- 0.68) X 109 cm-3 during the flare. This combination of redshifted coronal line emission and increased coronal electron density is consistent with explosively evaporated flare material observed by IRIS falling as warm rain and accumulating in the remote area observed by EIS.

A thermal/nonthermal fit to the hard X-ray spectrum observed by RHESSI yields a nonthermal energy injection rate of 4.9 X 1026 ergs s-1; combining this with an estimated injection area of (2.1 +/- 1.7) X 1017 cm2 based on the brightening seen in IRIS slit-jaw images during the rise of the flare, we estimate a HXR beam energy flux of (6.7 +/- 5.5) X 109 ergs cm-2 s-1, a wide range whose larger values are consistent with explosive evaporation.

The link to the paper is here:
Explosive Chromospheric Evaporation and Warm Rain in a C3.1 Flare Observed by IRIS, Hinode/EIS, and RHESSI.

References
Brosius, J. W., & Daw, A. N. 2015, ApJ, 810, 45
Brosius, J. W., Daw, A. N., & Inglis, A. R. 2016, ApJ, 830, 101
Brosius, J. W., & Inglis, A. R. 2017, ApJ, 848, 39
Kamio, S., Fredvik, T., & Young, P. R. 2011, EIS Software Note 5
Kamio, S., et al. 2010, Sol. Phys., 266, 209
Kelly, R. L. 1987, J. Phys. Chem. Ref. Data, 16, Supplement 1
Young, P. R., Tian, H., & Jaeggli, S. 2015, ApJ, 799, 218

 
For more details, please contact: Dr. Deb Baker.


Next EIS Nugget    »»  coming soon...

TBC




Last Revised: 27-Oct-2011

Feedback and comments: webmaster

Email
 
Print
    
EIS partner logo