IRIS, Hinode, SDO, and RHESSI observations of a white light flare produced directly by non-thermal electrons

2.7 MB·

by  Kyoung-Sun Lee

IRIS, Hinode, SDO, and RHESSI observations of a white light flare
produced directly by non-thermal electrons
7 Kyoung-Sun Lee
1
Hinode Science Center, National Astronomical Observatory of Japan (NAOJ), 2-21-1, Osawa, Mitaka,
0
2 Tokyo 181-8588, Japan
[email protected]
n
a
J
Shinsuke Imada
3
2 Institute for Space-Earth Environmental Research (ISEE), Nagoya University, Furo-cho, Chikusa-ku,
Nagoya 466-8550, Japan
]
R
S
Kyoko Watanabe
.
h National Defense Academy of Japan, 1-10-20 Hashirimizu, Yokosuka 239-8686, Japan
p
-
o
Yumi Bamba
r
t
s Hinode team, ISAS/JAXA, 3-1-1 Yoshinodai, Chuo-ku, Sagamihara, Kanagawa 252-5210, Japan
a
and
[

1
v David H. Brooks 1
6
College of Science, George Mason University, 4400 University Drive, Fairfax, VA 22030, USA
8
2
6
0
.
1
ABSTRACT
0
7 An X1.6 flare occurred in AR 12192 on 2014 October 22 at 14:02 UT and was observed
1 by Hinode, IRIS, SDO, and RHESSI. We analyze a bright kernel which produces a white light
:
v (WL) flare with continuum enhancement and a hard X-ray (HXR) peak. Taking advantage of
i the spectroscopic observations of IRIS and Hinode/EIS, we measure the temporal variation of
X
the plasma properties in the bright kernel in the chromosphere and corona. We found that
r
a explosive evaporation was observed when the WL emission occurred, even though the intensity
enhancement in hotter lines is quite weak. The temporal correlation of the WL emission, HXR
peak,andevaporationflowsindicatethattheWLemissionwasproducedbyacceleratedelectrons.
To understand the white light emission process, we calculated the energy flux deposited by non-
thermal electrons (observed by RHESSI) and compared it to the dissipated energy estimated
from a chromospheric line (Mg II triplet) observed by IRIS. The deposited energy flux from
the non-thermal electrons is about 3∼ 7.7×1010erg cm−2 s−1 for a given low energy cut-off of
30∼40keV, assuming the thick target model. The energy flux estimated from the temperature
changes in the chromosphere measured using the Mg II subordinate line is about 4.6−6.7×
109erg cm−2 s−1: ∼6−22%ofthedepositedenergy. Thiscomparisonofestimatedenergyfluxes
implies that the continuum enhancement was directly produced by the non-thermal electrons.
Subject headings: Sun: activity — Sun: chromosphere — Sun: corona — Sun: flares — Sun: UV
radiation — techniques: spectroscopic
1
1. Introduction Hudson et al. 2006). The electron flux from these
low energy events is not enough to penetrate
Solar flares are one of the most energetic en-
and heat the photosphere directly. Therefore,
ergy release processes in the heliosphere. When a
other heating mechanisms have also been consid-
flare occurs we can observe the multi-wavelength
ered. For example, Zirin & Neidig (1981) pro-
response from microwaves to X-rays such as ra-
posed that high energy protons carry the energy,
dio bursts in the impulsive phase, Hα emission
and Machado et al. (1989) suggested that WL
along the flare ribbons, soft X-ray emission in the
emission is produced by electrons that heat the
postflarelooparcade,andhardX-rayemissionat
chromosphere directly and the photosphere indi-
the footpoints and looptop region (Fletcher et al.
rectly through radiative backwarming. However,
2011).
the true heating mechanism in the lower atmo-
Based on these multi-wavelength observa- sphere remains unclear. We believe that under-
tions, several solar flare models have been pro- standing the flare dynamics during the impulsive
posed. The standard solar flare model (CSHKP; phase, and the lower atmospheric response to the
Carmichael (1964); Sturrock (1968); Hirayama flare, are key to clarifying the heating and energy
(1974); Kopp & Pneuman (1976)) proposes that transport processes.
magnetic reconnection occurs at coronal heights
Frompreviousobservationalstudiesofchromo-
and the released magnetic energy is transported
spheric evaporation, strong blue shifted emission
to the lower atmospheric layers (e.g. the chro- (> 100 km s−1) in coronal lines (Antonucci et al.
mosphere) by thermal conduction (Nagai 1980;
1982;Brosius2003,2009,2013a,b;Milligan & Dennis
Yokoyama & Shibata2001),non-thermalparticles 2009) and red asymmetries (40 − 100 km s−1)
(Nagai & Emslie 1984; Fisher et al. 1985c,b,a), or
in chromospheric lines (Ichimoto & Kurokawa
anAlfv´enwavepoyntingflux(Fletcher & Hudson
1984; Kamio et al. 2005; del Zanna et al. 2006)
2008). The transferred energy heats the plasma
have been found. Some flares also show a
and generates and overpressure in the lower at-
redshift in coronal lines formed around a few
mosphere. Dense plasma is then evaporated to-
MK (Imada et al. 2008; Milligan & Dennis 2009;
ward the corona along the magnetic field (chro-
Watanabe et al. 2010b), and direct imaging ob-
mospheric evaporation) and we then observe post
servations of chromospheric evaporation upflows
flare loops emitting in the EUV and soft X-rays.
have been observed by Hinode/XRT (Nitta et al.
Sometimesstrongflaresproducecontinuumen- 2012). Hydrodynamic simulations can re-produce
hancementsasaphotosphericresponseduringthe aspects ofthe observationsandpredict twodiffer-
impulsive phase of the flare, and this is termed a ent types of evaporative flows depending on the
whitelightflare(WLF)(Carrington1859;Sˇvestka deposited energy flux: ”explosive” and ”gentle”
1966). Previous observations in visible wave- evaporation (Fisher et al. 1985a,b,c).
lengths and hard X-rays showed that the contin-
Recently, the Hinode/EUV Imaging Spectrom-
uumenhancementinWLFsiswellcorrelatedwith
eter(EIS,Culhane et al.(2007))andInterfaceRe-
hardX-rayemissionbothspatiallyandtemporally
gionImagingSpectrograph(IRIS,De Pontieu et al.
(Neidig 1989; Hudson et al. 1992; Metcalf et al.
(2014)) have provided us with high spatial and
2003; Watanabe et al. 2010a; Krucker et al. 2015;
temporal resolution spectroscopic observations in
Kuhar et al. 2016). As a result of this correla-
the EUV/UV (Li & Ding 2011; Tian et al. 2015;
tion,ithasbeenthoughtthatWLFsareproduced
Graham & Cauzzi 2015). The combined power
by the transported energy from accelerated parti-
of the instruments allows us to investigate flare
cles such as non-thermal electrons (Brown 1971;
properties and dynamics through the entire at-
Hudson 1972).
mospsherefromchromosphereto corona(Li et al.
With recent high spatial resolution observa- 2015; Polito et al. 2015, 2016).
tions, white light emission has been reported
Most relevant to this work are several stud-
even in C-class flares (Matthews et al. 2003;
ies comparing the deposited flare energy with ob-
served continuum enhancements. For example,
1Current address: Hinode Team, ISAS/JAXA, 3-1-1 Watanabe et al. (2010a) found that the energy of
Yoshinodai, Chuo-ku, Sagamihara, Kanagawa 252-5210,
thewhitelightemissionobservedbyHinode/Solar
Japan
2
OpticalTelescope(SOT)wasequivalenttotheen- 2. Observations and data analysis
ergy supplied by all the electrons accelerated to
2.1. X1.6 flare in AR 12192
above 40 keV, which suggests that highly acceler-
atedelectrons areresponsible for producing white
We investigateanX1.6 flarewhichoccurredon
light emission. Recently, Kleint et al. (2016) in-
2014 October 22 in AR 12192. AR 12192 was the
vestigatedtheradiatedenergyfromthecontinuum
largestactive regionin this solar cycle and it pro-
enhancement observed from the UV to IR during
duced 6 X-class flares and 31 M-class flares. Fig-
a flare using IRIS, Solar Dynamics Observatory ure 1 shows the GOES soft X-ray (0.5-4 ˚A and
(SDO),andFacilityInfraredSpectrometer(FIRS), 1-8 ˚A) light curve, its time derivative, and the
and also found that the deposited energy was suf-
RHESSI hard and soft X-ray light curves of the
ficient to produce the UV and visible continuum
flare event. The flare started at about 14:02 UT
emission in the flare.
andpeakedat14:28UT.Thegradualphaseofthe
In this study, we describe the temporal evolu- flare emission declined until an M-class flare oc-
tion of the spectral properties and quantitatively curredat15:54UT.Theverticaldashedlinesmark
estimate the energy flux of an X1.6 flare using the specific times at which we present the spec-
combined observations from IRIS, EIS, Reuven tral properties of the flare: (a: the beginning of
Ramaty High-Energy Solar Spectroscopic Imager the impulsive phase (14:06 UT), b: the rise phase
(RHESSI,Lin et al.(2002)),andSDO.Thewhole (14:09 UT), and c: the peak (14:24 UT); see sec-
flare evolution, from the beginning of the impul- tion 3). Positions a, b, and c also mark the peaks
sive phase to the gradual phase, is captured by ofthetimederivativeofthesoftX-raycurvewhich
these instruments simultaneously and the strong correspond to the hard X-ray peaks (bottom), as
flare produces white light emission. Previously, expected from the Neupert effect (Neupert 1968).
Li et al. (2015), Thalmann et al. (2015), and
Figure 2 shows context images of the flare ob-
Veronig & Polanec (2015) also investigated this
tained by SDO/Atmospheric Imaging Assembly
wellobservedflareusingSDO,RHESSIandIRIS. (AIA, Lemen et al.(2012))inthe 211˚A(a-c)and
Thalmann et al. (2015) and Veronig & Polanec 1700 ˚A (e-g) channels, and IRIS C II 1330 ˚A slit
(2015) mainly focused on the magnetic reconnec- jaw images (SJIs) overlaid with EIS 195 ˚A con-
tion rates and the RHESSI HXR profiles of this
tours (panels i-k) for different timings, before the
flare,andLi et al.(2015)investigatedtherelation- flare (∼13:34 UT), the first HXR peak (∼14:06
ship between Doppler velocity patterns derived UT) and the SXR peak (∼14:24 UT). Panels (d)
from IRIS Fe XXI and C I and the RHESSI HXR and (h) display the SDO/Helioseismic and Mag-
intensity. Those studies showed that the flare en-
netic Imager (HMI, Schou et al. (2012)) contin-
ergy is injected into high energy electrons, and
uum and a running difference image at the time
that they could driver the evaporationflow in the
the white light flare occurred, respectively. Panel
flare. We present the flare observations from each
(l) showsthe polarityinversionline fromthe HMI
instrument in Section 2, and the temporal evolu-
magnetogram contoured on an IRIS SJI to show
tion of the spectral properties (intensity, Doppler
the magnetic field configuration.
velocity, line width, density and temperature) of
When we look at the continuum and lower at-
the flare kernel during the impulsive phase with
mospheric (AIA 1700 ˚A) images, we can see that
the other continuum and X-ray observations in
the flare occurred at the boundary of the large
Section 3. We discuss the comparison between
umbraandthesatellitepenumbra. Whenthefirst
the deposited energy from the non-thermal elec-
hardX-raypeakoftheflarewasobserved(around
tronsandtheobservedspectroscopicpropertiesin
14:06 UT), we can see continuum enhancements
Section 4. A summary of our results is given in
(the white lightflare)at twobrightkernels(panel
Section 5.
(h))whichcorrespondtothefootpointsoftheflare
loop structure (panel (c)). Then, two ribbons ex-
tend in the east-west direction (panels (g) and
(k)). Before the flare, around 13:34 UT (panels
(a), (e), and (i)), we can see a small brigtening
near the east side of the flare kernel. This ker-
3
nel was observed simultaneously in the IRIS and and these are marked with vertical dashed lines.
EIS scanning rasters. We analyzed the evolution InFigure4,thegreensolidlineisthefittedspectra
of the plasma properties of this bright kernel and and the dashed lines indicate each component of
estimated the energy flux. the multiple Gaussian fits. The red dashed fitted
components are FeXXI (in the upper panels) and
2.2. Spectroscopic observations from Hin- the second component of Si IV (in the lower pan-
ode/EIS and IRIS els). The reference wavelengths for IRIS were de-
terminedbytakingthedifferencebetweenthethe-
EIScapturedtheflarewiththestudy“HH Flare raster v6”
oreticalwavelengthsand averagedobservedwave-
which ranfrom 13:01:56UT to 15:56:56UT. This
lengths ofO I 1355.60˚A andS I 1401.51˚Abefore
EIS study is designed for observing flares using a
′′ the flare. These are marked by vertical dashed
moderate cadence raster scan. The 2 slit scans
′′ lines in the Figure. The theoretical wavelength
20 positions with a coarse 3 step between each
position and the field of view is 59′′×152′′. The of Fe XXI is taken from the CHIANTI atomic
database(Dere et al. 1997;Landi et al. 2012)and
exposuretimeateachpositionis9secondsandthe
is denoted by the vertical dot-dashed line.
raster scan takes about 3.5 minutes. The study
has 12 spectral windows and we used 10 spectral
2.3. SDO and RHESSI observations
linescoveringthetemperaturerangefromlogT=
4.9-7.2. These are listed in Table 1. The spectral We also used AIA and HMI onboard SDO to
resolution of the EIS is about 0.022 ˚A. understandtheglobalstructureandmagneticfield
Atthesametime,IRISwasrunningaverylarge configuration of the flare. AIA provides multi-
coarse 8 step raster. It uses the 0.33′′ × 175′′ slit ple temperature images covering log T = 3.7-7.2
and with 2′′ steps and so covers a field of view of with a high time cadence ofabout 12 secondsand
about14′′ × 175′′ inaround130seconds. The ex- a spatial resolution of 1.2 arcsec. We used AIA
posure time at each position is 16 seconds. The 1700 ˚A and 211 ˚A filter images for context (Fig-
spectral and spatial resolution of IRIS is 0.025 ˚A ure 2), and took advantage of the high temporal
and 0.32′′, respectively. The IRIS slit direction resolutionof AIA to investigate the flareintensity
was rotated 45 degrees relative to its center for variations and compare them to what is observed
this observation. The observing programincludes by EIS and IRIS. The temporal variation of the
9 spectral windows in the FUV (1332-1358˚A and intensity in different AIA filters is plotted in Fig-
1389-1407 ˚A) and NUV (2783-2834 ˚A). In this ure 5. HMI provides full Sun line of sight (LOS)
study, we only analyze the spectral lines which magnetogramsandcontinuum images at a spatial
are close to optically thin, OI, SiIV, and FeXXI, resolution of ∼ 1′′ and a temporal cadence of 45
for measuring the Doppler velocity, and used the seconds. Using the continuum data,we confirmed
OIVandMgIIlinesforinvestigatingthechromo- that the white light flare kernel is coincident with
spheric response. The spectral lines we used are the same bright UV kernel we analyzed.
summarized in Table 1. We also investigated the flare hard and soft X-
To obtain the intensity, Doppler velocity, and ray emission using the RHESSI X-ray spectrome-
linewidthasafunctionoftime,wefittedthespec- ter. We plotted the light curve of the emission in
trallinesinTable1usingsingleandmultiplegaus- the 30-100 keV and 12-25 keV range in Figure 1.
sians. Figures 3 and 4 show examples of the line To obtain images of the HXR and SXR emission,
profilesfromEISandIRIS,respectively. InFigure we used the ”Clean” method with 300 iterations
3, the greensolidline is the fitted spectraand the and a temporal resolution of 2 minutes, which is
red dashed lines indicate each component of the similartothatoftheIRISraster. Theleftpanelin
multiple Gaussianfitting. The red and greendot- Figure 6 shows an HMI intensity difference image
ted vertical lines are the fitted line center of each withthecleanedRHESSIHXRandSXRintensity
spectralline componentfroma multiple Gaussian contours (50, 60, 70, 80, and 90%) overlaid.
fitting, andthe estimated velocities are written in
the Figure 3. To obtain a reference wavelength
for the EIS spectra, we measured the averageline
centersbeforethe flare(between13:01-13:51UT),
4
2.4. Co-alignmentofSDO,Hinode/EIS,and gests that the enhanced Fe XXI emission might
IRIS be caused by the density enhancement of the hot
plasma.
We co-aligned the flare observations from Hin-
ode, IRIS, RHESSI, and SDO as follows. First,
3.2. Temporal evolution of the spectral
we used SDO/AIA observations as the reference
properties of the bright kernel
imageandalignedtheAIA1600˚Aimagewiththe
IRIS SJI 1330 ˚A images. The IRIS slit was ro- 3.2.1. Intensity
tated 45 degrees from the north-south direction,
Figures 5, 8, and 10 (upper panel) show the
so we de-rotated the SJI images and then aligned
temporal variation of the intensities of the bright
them with the AIA images. Then, we aligned the
kernel at different wavelengths from SDO, EIS,
EISFeXII195.12˚ArasterimageswithAIA193˚A
and IRIS, respectively. The SDO/AIA intensities
filterimages. Forthisalignmentwecalculatedthe
in Figure 5 are normalized by their maximum in-
offset values using the procedure ‘align map.pro’
tensity during the period from 13:00 UT to 16:00
available in the Solar Software (SSW) package.
UT. Together EIS and IRIS provide the tempo-
The offset values vary within ∼2′′, and the EIS
ral variation of the intensity over a wide range of
and IRIS SJI images are overlaid in the bottom
temperature from log T=4.5 - 7.2.
panels of Figure 2.
From the intensity variation, we can see anal-
ogous behavior in certain temperature ranges,log
3. Results
T= 4.5 - 5.8 (cooler), log T= 5.8 - 6.4 (middle),
3.1. Temporalevolutionofthecontinuum, and log T= 6.4 - 7.2 (hotter). It seems that the
UV, and X-ray emission in the bright temperatureresponseoftheflareissimilarinthese
kernel temperature ranges and transitions at log T∼ 5.8
and log T∼ 6.4. In the cooler temperature emis-
We found that the white light flare signature sion, such as 1600 ˚A and 1700 ˚A in Figure 5 or
of the bright kernel in the HMI images and the
HeII or OV in Figure 8, the response of the flare
HXR emission observed by RHESSI were located
isseenasasharpintensityenhancementwhenthe
at the same position (left panel in Figure 6). To
flare starts (time (a) in Figure 1) with no sig-
check the temporal evolution we plotted the light
nificant enhancement during the gradual phase.
curve of the HMI continuum intensity in the bot-
For the hotter emission lines (e.g. Fe XXIV and
tompanelofFigure7,andtheHXRandSXRlight
FeXXIII), it seems that the intensities peak a lit-
curvesforthebrightkernellocationinthefirstand
tle later than in the cooler temperature lines and
second rows. We also plotted the light curves for
there is a significant intensity enhancement dur-
thechromosphericandflaringlineintensitiesfrom
ing the gradualphase. In the middle temperature
IRIS (OI, SiIV, FeXXI) for comparison.
range, e.g. Fe X - Fe XVI in Figure 8, there are
First,wenotethattheHXRpeakandenhance- several peaks during the gradual phase. It seems
ment ofthe chromosphericline intensity (OI) ap- that after the main flare, there are still bursts of
pear at the beginning of the flare (∼14:04 UT). intensityenhancementalbeitnotstrongenoughto
Second, the continuum enhancement starts to ap- produce white light or HXR emission.
pear around the same time and reaches its max-
imum within 2 minutes. Third, the SXR emis- 3.2.2. Doppler velocity
sion shows two borad humps, one near the HXR
The upper panel in Figure 9 and middle row
peaktime andanotherlowerpeakaround16min-
in Figure 10 show the temporal variation of the
utes later. This suggests that the heated plasma
Doppler velocity at different wavelengths. The
wasevaporatedandtheincreaseddensityandtem-
most significant velocity variation is seen at the
perature are observedas enhanced SXR emission.
impulsive phase of the flare around 14:06 UT. A
Fourth,theemissionfromFeXXIincreasesaround
strong blue shift is observed in the flaring tem-
14:24 UT. We note that even although the inten-
perature lines, Fe XV - Fe XXIV, and a red shift
sity of FeXXI is weak, emission is detected at the
beginning of the flare (14:04 ∼ 14:11 UT). It sug- is observed in the chromospheric lines, O V and
SiIV, which is consistent with the Doppler veloc-
5
ity pattern expected from explosive evaporation single gaussian, and the whole lines are blue or
(Figures 3 and 9). After the impulsive phase the redshifted withoutarestcomponent(seee.g. the
blue shift in the higher temperature lines quickly line profiles of Fe XXIII and Fe XXIV in Figure 3
changesto a weak redshift which lasts more than and Fe XXI in Figure 4). This implies that the
an hour; essentially until the hotter emission dis- enhanced line width at the beginning of the flare
appears (Figure 10). A strong red shift is also isrelatedtonon-thermalbroadening,notDoppler
seen in the cooler temperature lines during the velocity.
impulsive phase and weak red-shifted emission is The non-thermal broadening in UV and X-
observed during the gradual phase. ray emissions is usually regarded as a manifesta-
Weplottedthevelocityasafunctionoftheline tion of unresolved mass motions of the plasma,
formation temperature in Figure 11. Panels (a), such as, multiple flows, turbulence or waves
(b), and (c) show the velocity pattern before the (Alexander & MacKinnon 1993; Dere & Mason
flare, during the impulsive phase, and during the 1993;Chae et al.1998;Hara et al.2011;Kawate & Imada
gradual phase. The velocity pattern we discussed 2013). The significant enhancement of the line
is clearly seen in the Figure. Moreover,the veloc- width and the HXR peak at the beginning of the
ity become larger at higher temperatures, which flarearetemporallywellcorrelated. Ifthe coronal
is a similar behavior to what has been reported reconnection occurs at the HXR peak timing, or
previously by e.g. Milligan & Dennis (2009) and the HXR comes from the accelerated electrons,
Polito et al. (2016), and is consistent with theo- the non-thermal broadening possibly caused by
retical expectations (Nagai & Emslie 1984). turbulence motion or waves due to the magnetic
We note that a strong blue shift was only ob- reconnection or accelerated electrons.
served at the beginning of the flare even though Thereareseveralotherpeaksatdifferenttimes
the intensity enhancement in the hotter lines is in He II, Si IV and Fe XII that appear to be re-
quite weak. Figure 10 shows that the Fe XXI in- lated to small chromospheric brightenings; one of
tensity from IRIS is weak, but the Doppler ve- which may be a candidate for lower atmospheric
locity shows a strong outflow at the start time of reconnection that triggers the flare (Bamba et al.
the flare(∼ 14:06UT). The EISobservationsalso 2016).
showthatastrongblueshiftisobservedinthehot-
ter lines even when the intensity is weak (Figures 3.2.4. Density
3 and 8). So, it appears that higher temperature
We measured the density of the bright kernel
emission exists and we can observe the dynamics
in the chromosphere and corona during the flare.
even when there is no strong intensity signature.
Assuming the plasma is optically thin, thermal,
The intensity may be weak because the density is
andincollisionalionizationequilibrium,wederive
low in the high temperature plasma in the early
the density using the intensity ratio of emission
phase, which we tried to verify with density diag-
lines from the allowed and forbidden transitions
nostic measurements (see below).
which are sensitive to the density. The diagnostic
method is well described by Mariska (1992) and
3.2.3. Line width
Phillips et al. (2008).
We also checked the line width variation with Thegooddensitysensitivelinepairsobservedin
time, which is shownin the bottom panels of Fig- EIS and IRIS are reported in Young et al. (2007)
ures 9 and 10. The strongest line width enhance- andYoung(2015),respectively. WeusedtheOIV
ment was observed in the flaring emission lines line pair for measuring the density in the transi-
(Fe XXI and Fe XXIII) at ∼14:06 UT when the tion region plasma, and the Fe XIV line pair for
first hard X-ray peak and Doppler velocity peak thecoronalplasma. Figure12showsthetemporal
in the hot plasma appears. We note that if the variation of the density measured by IRIS (upper
emission has a bulk Doppler shifted velocity com- panel) and EIS (lower panel).
ponent, it will make the line broader due to the The O IV line pair (1399.77 ˚A and 1401.16 ˚A)
combination of the rest and moving component. shows density sensitivity in the Log N =10−13
e
However,atthebeginningoftheflare,theDoppler
range with an intensity ratio range of 0.17-0.43.
velocity pattern of the lines can be fitted with a
6
We plotted the intensity ratio with time rather the region we extracted for the spectral profile.
than the converted density because the inten- Thesolidlineshowsthelineprofileduringtheim-
sity ratio during the impulsive phase of the flare pulsive phase (∼14:07) and the dotted line shows
exceeds the maximum value of the theoretical the spectrum around 12:00 UT when there is no
calculation. Therefore, we cannot measure the specific brightening or X-ray response, as a ref-
density during the impulsive phase. Before and erence. One interesting point to note is that the
after the impulsive phase, the averaged density Mg II triplet lines emit strongly during the im-
is about 6.3 × 1010 cm−3. The enhanced in- pulsive phase compared to during the non-flaring
tensity ratio during the impulsive phase implies time.
densities in excess of 1013 cm−3, which could Leenaarts et al.(2013)andPereira et al.(2015)
be the result of compression from the explo- proposed that Mg II h & k, and its subordinate
sive evaporation. Alternatively it could mean lines(atriplet: 2791.60˚A. 2798.75˚A,and2798.82
that the plasma is not in ionization equilib- ˚A) can be used as diagnostic tools of the chromo-
rium (Kafatos & Tucker 1972; Imada et al. 2011; sphericplasma. Inparticular,Pereira et al.(2015)
Olluri et al. 2013; Mart´ınez-Sykoraet al. 2016) showed that the Mg II triplet blends at 2798.75
and therefore that the density measurements ˚A and 2798.82 ˚A will be seen in emission when
are invalid during that period. However, the chromospheric heating occurs, and the line core
ionization-relaxationtimeisonlyabout13seconds to wing intensity ratio has a linear relationship
for a plasma with the average density measured with the temperature increase.
before and after the impulsive phase, which is
We applied their quantitative method to inves-
much shorterthan the durationthat the intensity
tigatethetemperaturechangesinthisflarekernel.
ratio was enhanced.
We measured the line core (the average intensity
Another possibility is that the measured line between2798.66-2798.93˚A) to wing (takenatthe
intensities are blended with cool lines which was 2799.32˚A)intensityratioofthetwoblendedMgII
mentioned by Young (2015) and Polito et al. triplet lines, 2798.75 ˚A and 2798.82 ˚A. The left
(2016). For example, Fe II 1399.96 ˚A, SI 1401.51 columninFigure14showsthe variationofthe in-
˚A and unidentified lines at shorter wavelengths tensityratiowithtime. Weconvertedtheintensity
are close to O IV. Even if we perform multiple ratio to ∆T from the linear relationship between
Gaussian fitting to take these blended lines into them derived from the flare simulation conducted
consideration, the intensity ratio is still enhanced by Pereira (private communication). The varia-
during the impulsive phase. tionof∆T withtimeisshownintherightcolumn
We measured the coronal density with EIS us- of Figure 14. The estimated temperature changes
ing the Fe XIV 264.79 ˚A and 274.20 ˚A line pair. areabout3kKfortheflarekernelandthetempo-
Comparedto the resultsfromOIV,FeXIVshows ralvariationshowsthatthetemperaturesuddenly
that the density is slightly enhanced in the early increasedatthe beginning ofthe impulsive phase.
phase of the flare, then increases significantly in
the later phase and peaks atthe same time as the 3.3. Summary of results
SXR.Thetemporalvariationofthedensityissim- • The flare kernel is localized during the im-
ilar to the intensity variation, suggesting that the
pulsive phase and the HXR emission, chro-
low intensities observed when explosive evapora-
mospheric intensity, and white light contin-
tionoccursareduetothelowdensityofthehigher
uumemissioninthe kernelarespatiallyand
temperature plasma. After evaporation, intensity
temporally correlated. The bright kernel
enhancements can also be seen in the higher tem-
peaksfirstinintensityinthewhitelightcon-
perature lines.
tinuum and O I, and then HXR and SXR
emission peaks are observed consecutively.
3.2.5. Chromospheric temperature: Mg II triplet
lines • The Doppler velocity and line width are en-
hanced during the impulsive phase. The
We also checked the response of the chromo-
strongest line width enhancement appears
spheric Mg II line. Figure 13 shows the Mg II
during the first HXR peak, which is just af-
line profile and the green horizontal lines mark
7
ter the white light flare (within a minute). high, the simulation shows a blueshift in coronal
This may be a signature of turbulence from lines, ”chromospheric evaporation”, and a red-
reconnectionorheatingbynon-thermalelec- shiftinchromosphericlines,”chromospheric con-
trons. densation”. This is called explosive evaporation.
Conversely, if the injected energy is less than a
• Chromospheric(OIV)andcoronal(FeXVI) critical value, about F = 1010 ergs cm−2 s−1,
20
density diagnostics show a strong enhance-
gentle evaporation is observed, and most spectral
ment during the impulsive phase. The
lines are blue shifted.
temporal variation indicates that the den-
Figure 11 shows that explosive evaporation oc-
sity is first enhanced in the chromosphere
curs at this bright kernel in the impulsive phase.
and later in the corona. This is consistent
Toconfirmtherelationshipbetweenthedeposited
with compression of the chromosphere, the
energy from the accelerated electrons and the ob-
Doppler velocity pattern, and the chromo-
servedDoppler velocity pattern,we comparedthe
spheric evaporation process.
depositedenergyfluxmeasuredbyRHESSItothe
• The Mg II subordinate line blend is seen in criticalvalueoftheenergyfluxinthesimulations.
emission during the pre-flare and impulsive Assuming this white light flare is produced by
phasebrightenings. Theenhancementofthe accelerated non-thermal electrons, we calculated
MgIIcoretowingintensityratioimpliesthe the total power (P) in the non-thermal electrons
existenceofsteeptemperaturegradientsand above a given electron energy (low cut-off en-
heatingatthelowatmosphere(Pereira et al. ergy) under the thick target approximation using
2015). At the time the white light flare oc- theequation(Hudson et al.1978;Watanabe et al.
curs, the ratio becomes over 10, suggesting 2010a),
thatstrongheatingoccursintheloweratmo-
b(γ)
sphere. Moreover, the strong enhancement P(ǫ>ǫ )=4.3×1024 Aǫ −(γ−1) (erg s−1).
c γ−1 c
of the ratio is correlated to the hard X-ray
(1)
peak, which implies that the non-thermally
For this purpose, we fit the RHESSI hard X-ray
accelerated electron detected as the HXR
photon spectrum. The right panel of Figure 6
emissionmightbedirectlyrelatedtothelow
showsthefittedRHESSIspectrumwhenexplosive
atmospheric heating.
evaporationoccurs(14:06UT).ǫ isthelowcut-off
c
energy,γ isthespectralindex,andb(γ)istheaux-
4. Discussion
iliary function fromBrown(1971) for the relevant
range of γ. To measure the energy flux, we deter-
4.1. Spectroscopic results related to the
mined the size of the hard X-ray emitting region
evaporation flow and white light flare
where the HXR (30-100 keV) integrated intensity
The flare kernel we discuss in this paper pro- is greater than 60 % of the maximum intensity.
ducedawhitelightflarewhenthefirsthardX-ray The calculated energy fluxes in the non-thermal
peak appeared. The correlation of the HXR peak electrons at the HXR peaks during the impulsive
andDoppler velocity variationin this flare has al- phase (∼14:05 and 14:11 UT), assuming a cut-off
readybeenreportedbyLi et al.(2015)usingIRIS energyof30keV,isabout7.7×1010ergscm−2s−1
observations,andthey suggestedthat this implies and 6.1×1010 ergs cm−2 s−1, respectively (Fig-
thattheflareiselectrondrivenandthatenergyde- ure 15). The energy of 30 keV is the lowest en-
position from non-thermal electrons produces the ergy that contains a negligible amount of ther-
chromospheric evaporation flows and white light mal emission, and still contains large fluxes of
flare. non-thermal photons. We also estimated the de-
Fromelectronbeamheatingsimulations(Fisher et al. posited energy fluxes by non-thermal HXR elec-
1985c,b,a),itisexpectedthatgentleandexplosive trons above different threshold energies because
evaporationshouldbeobserved,dependingonthe we don’t know which energy electrons affected
injected energy flux, and that it can be detected to the WL emission. For example, non-thermal
by examining the velocity variation at different electrons in low energies cannot penetrate to the
temperatures. When the energy flux injected is photosphere which produces WL emssion. On
8
the other hand, the higher energy electrons may nection process high in the corona since there
not transport enough energy due to the low pho- is also a temporal correlation between the HXR
ton flux. Even though we assume the different emission and Doppler velocity peak. Most of the
threshold energies of 40 keV and 50 keV, the en- observations show that the electron beam heat-
ergyfluxes atthe impulsive phaseareabout3.0× ing model well explains this flare, and the accel-
1010 ergs cm−2 s−1 and 1.4×1010 ergs cm−2 s−1, erated electrons can produce the white light con-
respectively. Thisshowsthatthedepositedenergy tinuum emission. However, we cannot rule out
from the hardX-rays is strongenough to produce the possibility ofAlfv´en wavesas a heating mech-
explosive evaporation and is consistent with the anism (Fletcher & Hudson 2008) given the large
observed Doppler velocity pattern. non-thermalwidthinthe lowertemperaturelines,
Furthermore, the flows from explosive evapo- Fe XII, He II. Recently, Reep & Russell (2016)
ration are expected to have a reversal, where the showedthatAlfv´enwavedissipationproducessim-
flows change from downflow in the chromosphere ilar heating signatures to electron beam heating,
to upflowinthe corona,andourobsrvationsshow for example, explosive evaporation and a signif-
such a reversal in the temperature range 0.5-2 icant temperature enhancement in the chromo-
MK. The velocities in this range also appear to sphere. If Alfv´en waves transport the energy flux
befairlysteady. Interestingly,Somedownflows,of tothe lowatmosphere,linewidths canalsobe en-
5km s−1 inthesinglegaussianfitsand15km s−1 hanced by the waves.
when a double gaussian is used, are observed in
the coronal lines (T∼2MK), which is a surpris- 4.2. RHESSI and IRIS energy flux compar-
ison
ingly high temperature. A similarly high tem-
perature velocity reversal has previously been re-
One of the important issues for understanding
ported, however (Li & Ding 2011).
the white light flare mechanism is how the energy
To explain the downflows in the coronal lines, is transferred to the lower atmosphere to produce
Imada et al.(2015)investigatedthedependenceof the photospheric emission. Recently, it has been
the flowreversaltemperatureondifferentthermal reportedthat not only strong flaresbut also weak
conduction coefficients. It turns out that if the flares(Cclassflares)canproducewhitelightemis-
thermal conduction is strong, the energy is trans- sion (Matthews et al. 2003; Hudson et al. 2006).
portedquicklybythermalconductionandthissce- Soitisimportanttoknowhowmuchenergyisdis-
nario shows similar characteristics to the electron sipated in the chromosphere and whether the en-
beam driven case. If the thermal conduction is ergytransferredbytheelectronsisenoughtopro-
weak, however, the energy is mainly transported duceawhitelightflare. Untilrecently,ithasbeen
by the entalphy flux and advection. In this case, difficult to estimate the energy flux in the chro-
the enthalpy flux dominantcase, the flow reversal mosphere due to a lack of observations. After the
temperature is much hotter than in the thermal launch of IRIS, however, high resolution spectro-
conductioncase. Inourobservations,astrongred scopic observations of the chromosphere have be-
shift is seen at temperatures around 0.4MK, and come routine, and Pereira et al. (2015) suggested
the Doppler velocities in the 0.5-2 MK tempera- that the MgII triplet could be used as a diagnos-
ture range are mostly steady. This implies that tic tool for quantitatively measuring temperature
the flow reversal temperature in this flare is not changes in the chromosphere.
very high, and is similar to that expected in the
We have measured the temperature changes
electron beam model. However, the Doppler ve-
during the flare using the Mg II triplet intensity
locity around 2MK also shows small downflows,
ratio (section 3.1.5), and used the results to es-
suggesting that we cannot neglect the possibility
timate the energy flux deposited in the chromo-
ofenergy transportby directenthalpy flux, in ad-
sphere in response to the flare. The Mg II h & k
dition to thermal conduction.
components show similar peak intensities during
AtthesametimeastheDopplervelocitypeaks, the flare implying that they might be in emission
enhanced line widths are observed at flaring tem- even though the plasma is optically thick. Fur-
peratures. This may indicate the presence of tur- thermore, the densities measured using the O IV
bulence from the non-thermal electrons or recon- chromospheric line are strongly enhanced during
9
the impulsive phase (Figure 12), which suggests In this study, even though we have not per-
the plasma may be in local thermodynamic equi- formedanynumericalmodeling,wehavebeenable
librium (LTE). If we assume the plasma is op- to measure the temperature changes in the chro-
tically thick and in LTE, the energy flux can mosphere quantitatively using the Mg II triplet
be determined by Stefan-Boltzmann’s law, F = observed by IRIS, as suggested by Pereira et al.
σT 4. We estimate the temperature enhance- (2015). It is the first attempt to apply this di-
eff
mentduringtheimpulsive phasetobe 3∼3.3kK agnostic technique to flare observations, and the
(Figure 14), so the corresponding energy flux is results show that the temperature changes in the
4.6−6.7×109 ergs cm−2 s−1. Taking this energy chromosphere are about 3000 K, which is consis-
flux as an estimate of the amount of energy dissi- tentwith the resultsfromthe numericalmodeling
pated in the chromosphere, it is about 6 ∼ 22 % of the UV to IR spectra of Kleint et al. (2016).
of the deposited energy from the accelerated non- The results support their suggestion from numer-
thermal electrons measuredby RHESSI assuming ical modeling that quite a strong temperature en-
the cut-off energy of 30 ∼ 40 keV. This result hancement is needed in the chromosphere to pro-
implies that the majority of the energy from the duce noticeable continuum enhancement in the
non-thermalelectrons acceleratedin the coronais flare.
stillavailabletodirectlyproduceawhitelightflare
in this event. 5. Summary
Recently,Milligan et al.(2014)andKleint et al.
A bright kernel in an X1.6 flare on 2014 Octo-
(2016) investigated the continuum enhancement
ber 22 was observed by Hinode, IRIS, SDO, and
acrossthe EUV, UV, visible, and infra-redduring
RHESSI. The simultaneous observations covered
a flare, and compared it to the energy deposited
the whole duration of the flare and the bright
bynon-thermalelectronsobservedbyRHESSI.In
kernel produces a multi-wavelength intensity en-
their investigation, Milligan et al. (2014) showed
hancement from continuum to hard X-rays. We
that 15 % of the deposited energy is radiated by
investigatedthetemporalvariationofthespectral
line and continuum emission in the lower atmo-
properties of this kernelandestimated the energy
sphere and Kleint et al. (2016) found that 23%
flux at different wavelengths.
of the deposited energy is radiated by continuum
The multi-wavelength spectroscopic observa-
emission. From their investigations, more than
tions showthat the flare kernelis localizedduring
60 % of the energy is unaccounted for, and they
the impulsive phase and the HXR emission, chro-
suggested that it is dissipated by heating, plasma
mospheric intensity, and white light continuum
motions, or radiated away in other spectral re-
emission in the kernel are spatially and tempo-
gions or lines. Using IRIS and HMI continuum
rally correlated. We found that explosive evap-
observations, we also measured the energy flux
oration occurs and there are strong line width
from the UV and WL continuum. We converted
enhancements during the first peak of the hard
the observed Mg II DN to intensity using the
X-ray emission, which is also coincident with the
iris get response.pro routine in SolarSoft and we
timing of the white light flare. This may indi-
calibrated the HMI intensity by comparison with
cate that electron beam heating produces strong
the disk intensity reported in the atlas of Brault
evaporation flows and there is turbulence from
& Neckel. We then estimated the radiatedpower,
the reconnection or non-thermalelectron heating.
P = π I A ∆λ, using the intensities (I ) and
λ λ λ
band widths (∆λ) of the IRIS and HMI continua. Furthermore, the Mg II subordinate line blend
is in emission during the impulsive phase. The
The energy fluxes from the RHESSI HXR, IRIS
Mg II triplet, and HMI continuum are shown in strong enhancement of the Mg II core to wing in-
tensityratiolineiscorrelatedwiththehardX-ray
Figure 15 and the energy fluxes in the UV and
peak, imply the existence of a steep temperature
HMI continuum are only very small fractions of
gradient and heating. The correlated temporal
the power in the HXR emission. The estimated
variation of the hard X-ray, white light, explosive
energy flux from dissipation by heating in the
chromosphere may amount to ∼ 10% of the de- evaporationflows,the MgIIline responseandthe
comparison of the energy flux through the corona
posited energy.
10