Transcript 404.06 Stephen Drake

2014 August 21st, HEAD Meeting
Very Bright, Very Hot and Very Long: Swift
Observations of the DG CVn "Superflare" of April
23rd, 2014
Steve Drake
HEASARC/GSFC
&
USRA/CRESST
Collaborators
•
•
•
•
•
•
•
•
•
Rachel Osten (STScI)
Adam Kowalski (ORAU/GSFC)
Kim Page (U Leicester)
Jamie Kennea (Penn State)
Samantha Oates (IAA-CSIC)
Neil Gehrels (NASA/GSFC)
Hans Krimm (USRA/GSFC)
Mathew Page (UCL)
Kosmas Gazeas (U Athens)
References
Osten et al. (2010, ApJ, 721, 785) for the 2008 EV Lac flare
Drake et al. (2014, Atel # 6121)
Kowalski et al. (2014, in preparation)
+ Thanks to Rob Fender for permission to show the 15-GHz radio data, and
to the Swift Project for their support for these observations
Structure of Presentation
•
•
•
•
•
Coronae and flares 101
The EV Lac ‘Superflare’ of April 25th 2008
The Properties of the M4 V ‘Star’ DG CVn
The DG CVn ‘Superflare’ of April 23rd 2014
Future Directions
Our Friend:
The Sun
SOHO/EIT: Fe XII
line @ 195 Å
T ~ 1.5 MK
Solar Corona
Lx ~ 1026.5-27.3 erg/s
T ~ 1 – 5 MK
Solar Flares
E ~ 1024-33 erg
T ~ 10 – 30 MK
RHESSI spectrum of the
impulsive phase of a solar
flare between 6 – 250 keV
Benz & Guedel 2010
ARAA, 48, 241
Standard solar and
stellar flare geometry
B
A
A & C loop footpoints
B loop-top source
C
Geometry of EV Lac flare (Osten et al. 2010)
Hudson et al. (1994, ApJ, 422, 25) Yohkoh image of X1.5 solar flare
So How Do you Observe Stellar X-Ray Flares?
 Inactive stars similar to the Sun flare for only a small fraction of the time
=> need either to observe 1 star for v. long time or multiple stars
simultaneously as in Kepler field
 Active stars can flare for a much larger fraction of the time (up to 20%),
but most flares are small (N(E) ~ E-α, where α ~ 1.5 – 2.5) => to catch a big
flare you still need to observe for pretty long time or be lucky
 You can use `all-sky monitor’ X-ray observations, e.g., Swift BAT, MAXI,
etc., => given present sensitivity levels ~10-9 erg cm-2 s-1, only the largest
flares will be found: e.g., Swift detections of stellar flares from the active
binary systems Algol, II Peg & HR 1099 and the M4.5V star EV Lac and MAXI
detections of ~12 flares
 You can ask for long pointed observations of your favorite targets with
Chandra, XMM or Suzaku => hard to estimate flare rates for many stars,
so no guarantee of success and hard to convince TACs to grant long
enough observations
X-Ray Flare from Very Nearby (5 pc) dM4.5 star EV Lac Caught by
Konus/Wind & Swift on 2008 April 25
XRT
UVOT
Osten et al. 2010
XRT and BAT Spectrum of Early Decay Phase of EV Lac Flare
No significant
non-thermal
hard X-ray cpt
Dominant ~100
MK thermal cpt
He-like Fe K 6.7
keV line
Cool Fe K 6.4 keV
line behavior is
affected by charge
trapping
Osten et al.2010
The April 2008 Superflare of EV Lac
•
Proved that some (~ several) apparent GRBs are actually stellar flares
•
Peak X-ray temperature ~ 100 MK drops to ~30 MK after 45 minutes
•
Evidence for Fe K-α 6.4 keV in first few minutes after trigger: later detections likely
spurious
•
No evidence for non-thermal power-law emission in BAT spectrum
•
Significant white-light flare (Delta V ~ 2.9 magnitudes over quiescence) seen by UVOT
•
Flare loop size of 0.1-0.4 R* inferred flare loop modeling & Fe K-α
•
Peak X-ray luminosity was 3 x Lbol = 104 x normal level: in fact for 500 s after trigger, Lx
> Lbol
•
Total summed flare energy in soft X-ray band >~7 x 1034 erg/s (but observations
stopped after half a day when Lx was still 10 x normal level)
The M4V Star(s) DG CVn = G 165-08 AB
•
DG CVn is a rapidly rotating (period < 1 day) pair of M4V stars (~ 0.17“
separation) that is fairly nearby (18 pc) but poorly studied
•
DG CVn is one of the most active stars in the solar neighborhood based on its
levels of radio continuum, X-ray emission, and Hα emission
•
These properties explained by its very young age ~30 Myr (Riedel et al. 2014,
AJ, 147, 85)
•
Its normal Lx ~ 1.5 x 1029 erg/s ~ 0.001 Lbol => DG CVn star(s) at so-called
‘coronal saturation’ level
•
Thus, DG CVn is a plausible candidate for producing flares bright enough for
Swift/BAT detections, albeit rather distant
X-Ray Flare from DG CVn Caught by Swift BAT
on 2014 April 23
XRT and BAT Spectrum of Early Decay Phase of DG CVn Flare
No nonthermal
hard X-ray cpt
Dominant ~220
MK thermal cpt
No Fe K 6.4 keV
line
Swift/XRT Light Curve of DG CVn
F2
BFF
F2
R
BFF
Swift/UVOT &
Ground-based Rband Light Curve
of DG CVn
XRT
V
B
U
UVW1
UVM/W2
AMI 15-GHz Observations of DG CVn
Fender et al. (2014)
BFF
F2
Soft X-Ray Energy Budget of DG CVn Flare Series
log (N/T)
0
-2
-8
-6
-4
Big Flares are Very Rare!
M dwarfs
32
solar maximum
solar minimum
30
G dwarfs
Flare frequency/sec v. Flare Energy (ergs)
28
34
Log E
36
AD Leo
V577 Mon
Ei Cnc
V780 Tau
CN Leo
UV Ceti
FL Vir
YZ CMi
EQ Peg
EV Lac
YY Gem
LU Vel
DO Cep
V1005 Ori
V1054 Oph
V1285 Aql
KIC5474065
KIC9726699
AF Psc
KIC10422252
KIC12354328
KIC10120296
KIC10220756
Sun (min)
Sun (max)
38
40
BFF
Courtesy of Gerry Doyle
The April 2014 Superflare of DG CVn
•
Big first flare (‘BFF’) followed by >6 secondary flares: total time to drop back to
usual soft X-ray level ~ 1 - 2 Ms
•
Peak X-ray temperature ~ 220 MK drops to ~43 MK after 90 minutes
•
No strong evidence for Fe K-α 6.4 keV in flare peak emission
•
No strong evidence for non-thermal power-law emission in BAT spectrum
•
Significant UV/optical emission from flare (Delta V ~ 2.5 & Delta UV >~ 5
magnitudes over quiescence) seen by ground-based obs. & UVOT
• Peak X-ray luminosity was 2 x 1032 erg/s (1.5 x system Lbol): in fact for several
minutes after trigger, Lx > Lbol
•
Total summed flare energy in soft X-ray band 2 x 1036 erg
•
Similar (1036 erg) summed flare energy in ‘white light’ band coming from 0.5-5%
of the stellar surface (cf. <0.05% for solar flares)
Future Directions and Questions About DG CVn
• Complete a full X-ray and optical analysis to infer geometry and physical
properties of flaring plasma (Kowalski et al. 2014)
• Continued Swift monitoring to build up flare vs energy statistics might be
useful: suspect this star must have frequent smaller flares
• New study of optical photometry: what is the cause of the periodic
variation and what is the ‘true’ period?
• Magnetic field measurement would be very desirable
• Is the (wide) binarity of this system a factor in flare properties?
Questions about Stellar Flares

Where is the Non-thermal Hard X-Ray Emission?
- Hidden by the thermal emission: Current and previous instruments with sensitivity above 10 keV range
have observed many big stellar flares, but still only 1 plausible non-thermal detection: We need more
sensitive hard X-ray instruments or must use simultaneous radio observations as proxy

How Hot Can Large Flares from Main-Sequence Stars Reach?
-
Peak X-ray temperatures can reach 10-20 keV (120 – 240 MK)

How Long Can Large Flares or ‘Flare States’ from Main-Sequence Stars Last?
-
Durations can be much longer than realized: weeks rather than hours to 1 day

How Energetic Are Flares from Main-Sequence Stars?
-
Peak X-ray flare luminosities in some stars can reach >~ 1032 erg/s, and can exceed Lbol (and these stars
can stay active for ~ 108-9 years) => significant impact on habitable zones for extended time range
-
X-ray and optical-band flare energies can reach ~ 1036 erg