Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://star.arm.ac.uk/preprints/2013/636.pdf Äàòà èçìåíåíèÿ: Tue Jun 25 17:02:59 2013 Äàòà èíäåêñèðîâàíèÿ: Fri Feb 28 00:54:27 2014 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: m 101

Mon. Not. R. Astron. Soc. 000, 000­000 (0000)

Printed 25 June 2013

A (MN L TEX style file v2.2)

Short duration high amplitude flares detected on the M dwarf star KIC 5474065
Gavin Ramsay1, J. Gerry Doyle1, Pasi Hakala2, David Garcia-Alvarez3 Adam Brooks1,6 Thomas Barclay7,8 Martin Still7,8 1
2 3 4 5 6 7 8

,4 ,5

,

Armagh Observatory, Col lege Hil l, Armagh, BT61 9DG ¨ Finnish Centre for Astronomy with ESO (FINCA), University of Turku, V¨isal¨ntie 20, FI-21500 PIIKKIO, Finland a ¨a Instituto de Astrofsica de Canarias, E-38205 La Laguna, Tenerife Spain Dpto. de Astrofsica, Universidad de La Laguna, 38206 La Laguna, Tenerife Spain Grantecan CALP, 38712 Brea Baja, La Palma Spain Mul lard Space Science Laboratory, University Col lege London, Holmbury St. Mary, Dorking, Surrey, RH5 6NT NASA Ames Research Center, M/S 244-40, Moffett Field, CA 94035, USA Bay Area Environmental Research Institute, Inc., 560 Third St. West, Sonoma, CA 95476, USA

25 June 2013

ABSTRACT

Using data obtained during the RATS-Kepler pro ject we identified one short duration flare in a 1 hour sequence of ground based photometry of the dwarf star KIC 5474065. Observations made using GTC show it is a star with a M4 V spectral type. Kepler observations made using 1 min sampling show that KIC 5474065 exhibits large amplitude ( F/F>0.4) optical flares which have a duration as short as 10 mins. We compare the energy distribution of flares from KIC 5474065 with that of KIC 9726699, which has also been observed using 1 min sampling, and ground based observations of other M dwarf stars in the literature. We discuss the possible implications of these short duration, relatively low energy flares would have on the atmosphere of exoplanets orbiting in the habitable zone of these flare stars. Key words: Physical data and processes: magnetic reconnection ­ astrobiology ­ stars: activity ­ Stars: flares ­ stars: late-type ­ stars: individual: KIC 5474065, KIC 9726699

1

INTRODUCTION

Flares with duration of a few to tens of minutes and energies of 1028-35 ergs have b een observed on low mass dwarf stars for many decades (eg Bopp & Moffett 1973, Gershb erg & Shakhovskaia 1983). The origin of these flares is thought to b e similar to Solar flares in that they are produced during magnetic reconnection events (eg Haisch, Strong & Rodono 1991). Studying stellar flares from a wide range of stars can give imp ortant insight to how magnetic activity varies as a function of stellar mass and age. In more recent years, the affects of flares on the atmosphere of exo-planets around dwarf stars has b een the sub ject of much interest (eg Segura et al 2010). Historically the study of stellar flares was p erformed on known M dwarf stars. However, with the advent of large scale surveys such as SDSS it has b ecome p ossible to identify events from many previously unknown flare stars (eg Davenp ort et al 2012). Whilst this will no doubt prove a goldmine for stellar flare researchers, the issue of separating
c 0000 RAS

extra-galactic transient events and flares from M dwarfs will b ecome increasingly difficult in future surveys such as that made using LSST. One survey which allows the virtually uninterrupted observation of sources is NASA's Kepler mission which covers an area of 116 square degrees. The light curves extend over many months (or years) and have a precision of parts p er million and allows models of stellar structure to b e tested in a way not previously p ossible. A key p oint is that the actual targets which are observed using Kepler can b e updated every month. Walkowicz et al (2011) presented Kepler observations of flares seen in cool stars, while Balona (2012) rep orted observations of stars with A/F sp ectral typ e and Maehara et al (2012) presented some examples of `sup er' flares on Solar typ e stars. In June 2011 we started the RATS-Kepler1 pro ject

1

RApid Temporal Survey-Kepler


2
whose aim was to identify sources which showed flux variations on short (<30 min) timescales (Ramsay et al 2013). We do this by taking a series of short exp osures (20 sec) using wide field cameras on telescop es such as the Isaac Newton Telescop e on La Palma on sp ecific fields for one hour. Light curves of each ob ject are derived and variable sources identified. One variable source which we identified was KIC 5474065 which showed a short duration (<20 min) flare with an amplitude of 0.6 mag in the g band. We were successful in placing KIC 5474065 on the Kepler 1 min sampling target list. This pap er presents the results of these Kepler observations and a comparison of the energy distribution of the flares with other low mass flare stars.

2

KIC 5474065


Although KIC 5474065 (2000 =19h 53m 02.3s, 2000 =+40 40 34.6 ) is included in the Kepler Input Catalog (Brown et al 2011), it does not have a measured temp erature or surface gravity. It is, however, included in the Kepler-INT Survey (U=20.61, g=19.00, r=17.33, i=15.60; Greiss et al 2012a,b); the UBV survey of the Kepler field (B=18.79, V=18.07; Everett, Howell & Kinemuchi 2012) and also the 2MASS survey (J=14.015, H=13.397, K=13.215; Skrutskie et al 2006). The optical colours indicate a relatively late-typ e star. Since KIC 5474065 is variable (due to its rotational modulation and flare activity) some degree of caution is required when determining its colours unless it is known that multi-band observations are made simultaneously. However, L´pine & Gaidos e (2011) show the relationship b etween the colour (V - J ) and sp ectral typ e for late-typ e stars. For KIC 5474065, V - J = 4.06 implies a sp ectral typ e of M3 to M4.

Figure 1. The lower spectrum (solid line) shows the optical spectrum of KIC 5474065 obtained using GTC and Osiris. We indicate the wavelength of telluric absorption features as light grey vertical bands (taken from Kirkpatrick, Henry & McCarthy 1991) and we have not attempted to remove them. The upper spectrum (dashed line) shows the M4 template spectrum provided by Bochanski et al. (2007).

Table 1. The absolute magnitude and luminosity of stars with spectral types M3 V ­ M5 V based on the data in L´pine & Gaidos e (2011). Spectral Type M3 V M4 V M5 V M
V

L (erg s

-1

)

11.2 12.4 13.5

1.1 â 1031 3.6 â 1030 1.4 â 1030

3

GRAN TELESCOPIO CANARIAS SPECTROSCOPIC DATA

We carried out low-resolution sp ectroscopy with the Optical System for Imaging and Low Resolution Integrated Sp ectroscopy (OSIRIS) tunable imager and sp ectrograph (Cepa et al. 2003) at the 10.4 m Gran Telescopio Canarias (GTC), located at the Observatorio Roque de los Muchachos in La Palma, Canary Islands, Spain. The heart of OSIRIS is a mosaic of two 4k â 2k e2v CCD44­82 detectors that gives an unvignetted field of view of 7.8 â 7.8 arcmin2 with a plate scale of 0.127 arcsec pix-1 . However, to increase the signalto-noise ratio of our observations, we chose the standard op eration mode of the instrument, which is a 2 â 2-binning mode with a readout sp eed of 100 kHz. Two sp ectra each with an exp osure of 300 sec were obtained using the OSIRIS R1000R grism in service mode on 11 May 2013. They were made as part of a GTC filler programme which utilies p oor weather conditions. We used the 1.0 -width slit, oriented at the parallactic angle to minimise losses due to atmospheric disp ersion. The resulting resolution, measured on arc lines, was R 700 in the approximate 5250­9200 ° sp ectral range. The star Ross 640 was used to A remove the instrumental resp onse. The data were reduced using standard Figaro routines2 .
2

We show the optical sp ectrum of KIC 5474065 in Figure 1: it is clearly a late-typ e dwarf star. Examining Figure 1 of Bochanski et al (2007), KIC 5474065 is later than an M0V sp ectral typ e. Judging by the depth of the Na i (8190°) A A feature and the Ca i i triplet around 8500 ° it is most likely that KIC 5474065 has a M4V sp ectral typ e although M3V and M5V are also p ossible. In order to determine the energies of the flares, we must first estimate the intrinsic luminosity of KIC 5474065. L´pine & Gaidos (2011) include V J H K and parallax data e for late typ e stars. We were able to extract data as a function of sp ectral typ e and estimate the mean MV using relationship b etween (V - J ) and MV outlined in L´pine & Gaidos e (2011). We show in Table 1 the mean absolute V magnitude for sp ectral typ es M3V­M5V and we assume the Sun has MV = 4.83 and L = 3.8 â 1033 erg s-1 .

4

KEPLER DATA

http://starlink.jach.hawaii.edu

The detector on b oard Kepler is a shutterless photometer using 6 sec integrations and a 0.5 sec readout. There are two modes of observation: long cadence (LC), where 270 integrations are summed for an effective 28.4 min exp osure, and short cadence (SC), where 9 integrations are summed for an effective 58.8 sec exp osure. When an ob ject is observed in SC mode, LC data is also automatically recorded.
c 0000 RAS, MNRAS 000, 000­000


Flares detected on the M dwarf star KIC 5474065

3

Figure 2. The Kepler Short Cadence light curve of KIC 5474065 made in Q14. F/F is the ratio of the difference between the flux at any point and the mean flux. The rotational period of 2.47 days is clearly seen as are the short duration flares. In the left hand panel we zoom in on four flares.

Table 2. The start and end times of the Short Cadence and Long Cadence Kepler observations of KIC 5474065. Mode SC LC Start 2012-06-28 15:03:34 2012-06-28 15:17:47 End 2012-07-29 12:02:49 2012-10-03 19:40:10

dwarfs, e.g. YZ CMI has a rotational p eriod of 2.78 days, while V577 Mon has a rotational p eriod of 1.95 days. For comparison, we also extracted the light curve of KIC 9726699 (2000 =19h 51m 09.4s, 2000 =+46 29 01.2 ) which has also b een observed using Kepler in SC mode. Savanov & Dmitrienko (2011) presented an analysis of this data, but concentrated on determining the extent and duration and sp ots on its photosphere and did not discuss the flares themselves. Like KIC 5474065 it has a M4 V sp ectral typ e (Reid et al 2004), but it is more rapidly rotating (a rotation p eriod of 0.593 days) which makes it similar to V374 Peg (0.45 days). The Kepler Input Catlog (Brown et al 2011) gives g =13.9 for KIC 9726699 making it more than 5 mag brighter than KIC 5474065 and hence the Kepler data of this source has a much higher signal to noise than KIC 5474065. KIC 9726699 has b een observed using Kepler in SC mode in four quarters, but here we have restricted our analysis to data from Q6. We show a 4 day section of the light curve of KIC 9726699 in Figure 3. It shows a relatively small numb er of large amplitude flares, but its light curve is dominated by short duration, low intensity, flares. For comparison, we show the light curve of KIC 5474065 also covering 4 days and on the same flux scale in Figure 3. Given that KIC 9726699 is very much brighter compared to KIC 5474065, it is p ossible that short duration low intensity flares are likely to b e hidden in the noise in KIC 5474065.

KIC 5474065 was observed using Kepler in SC mode in Quarter 14: the start and end times of the observations are shown in Table 2. The on-time for SC mode was 24.1 days and 75 days for LC mode. After the data are corrected for bias, shutterless readout smear and sky background, light curves are extracted using simple ap erture photometry (SAP). Data which do not conform to `SAP QUALITY=0' were removed (for instance, time intervals of enhanced solar activity) and the data were corrected for systematic trends. The light curve of KIC 5474065, shown in Figure 2, exhibits two main features ­ one a clear quasi-sinusodial modulation with a p eriod of 2.47 days and a semi-amplitude of 2 p ercent, and the presence of short but intense flares: two flares have an intensity F /F 46 p ercent. For comparison, MOST observations of the dM3e star AD Leo (made using a 1 min cadence), show flares with F /F 28 p ercent (HuntWalker et al 2012), while `sup er-flares' with amplitude of 8 p ercent are seen on Solar-typ e stars (Maehara et al 2012). The rotational p eriod of KIC 5474065 is typical of M4-5
c 0000 RAS, MNRAS 000, 000­000


4

Figure 3. In the top panel we show a section of the Kepler Short Cadence light curve of KIC 9726699 made in Q6. F/F is the ratio of the difference between the flux at any point and the mean flux. For comparison we show in the lower panel a section of the light curve of KIC 5474065 on the same scales which is 5 mag fainter than KIC 9726699.

5

FLARE CHARACTERISTICS

To identify flares from KIC 5474065 in an automatic manner we first removed the effects of the rotational modulation. After some exp eriment, we identified the time interval when a flare occured when (fi - f )/ >3 where fi was the flux of the ith p oint, f was the overall mean of the light curve and was the standard deviation of the overall light curve. The resulting flare times were then manually insp ected and p oints which were clearly part of the same flare were edited to ensure no `double counting' of flares were made. (If we defined a lower threshold for flare detection the flare rate goes up but the false p ositive rate also goes up as it b ecomes difficult to distinguish b etween a genuine flare and noise in the data). This strategy found 27 flares in the SC light curve of KIC 5474065 ­ in otherwords, on average one flare was detected every 0.9 days, and on average there was one flare with an intensity F/F>0.2 every 8 days. There was no evidence of any pre-flare dips such as that seen in V1054 Oph (Ventura et al 1995). For KIC 9726699 it was more difficult to fully remove the effects of rotation and we therefore set the detection threshold as (fi - f )/ >8. However, we identified over 260 flares in the Q6 data of KIC 9726699. The fact that KIC 9726699 app ears to b e more active compared to KIC 5474065 is consistent with the well known correlation b etween rotation p eriod and stellar activity, (eg Noyes 1985), although we note that short duration, low energetic flares would not have b een detected in KIC 5474065 due to the much higher noise level. To derive the flare frequency rate, we use the following formulae where E = f is known as the flare equivalent duration, see Lacy et al. (1976) for further details; E=
f

Figure 4. The cumulative energy distribution of flares (in the Kepler band-pass) as seen in KIC 5474065 and KIC 9726699.

[(I

f +o

- Io )/Io ]T

(1)

where Io is the stellar intensity of the star in its quiescent

state, If +o the intensity during a flare and T the integration time. Further, we assumed the underlying luminosity was 3.6 â 1030 erg/s (Table 1) which is appropriate for a M4V sp ectral typ e. This gave for KIC 5474065 a range in flare energy, L = 1.1 - 7.3 â 1032 ergs, whilst for KIC 9726699 the range was L = 0.01 - 2.2 â 1032 ergs (these energies will vary by a factor of three for one sp ectral sub-class either side of M4V). Since the activity of flare stars is frequently measured in the U band, we have estimated the equivalent energy of KIC 5474065 in the U band. van den Oord et al (1996) found Eopt /EU = 2.4 where Eopt and EU is the energy emitted in the optical U B V and U band resp ectively. In comparison, Lacy, Moffett & Evans (1976) find Eopt /EW L = 2.1 where EW L is the energy emitted in 'white light'. For our purp oses we assume EK epler /EU = 2.4 and note there will b e a 10
c 0000 RAS, MNRAS 000, 000­000


Flares detected on the M dwarf star KIC 5474065

5

Figure 5. In the upper panels we show the mean light curve of KIC 5474065 (left) folded on To = BJD 2456106.97 + 2.47295492 E and KIC 9726699 (right) folded on To = BJD 2455371.310 + 0.593 E, where the light curves have been normalised so that the mean flux is unity and =0.0 is defined as photometric minimum and the light curves have been binned into 50 bins. In the lower panels we show the energy emitted in each flare as a function of rotational phase. Deviations from a smooth curve are a result of flares in the light curve.

p ercent uncertainty in our values of EU . For KIC 5474065 we therefore find the flares have a range b etween LU = 0.5 - 3 â 1032 ergs and LU = 0.01 - 0.9 â 1032 ergs for KIC 9726699. The energy and duration of these flares are at the lower energy of those detected on another dM4.5 flare star, AD Leo (Pettersen et al. 1986). We now show the cumulative flare-frequency distribution of the flares in Figure 4. This shows that on average KIC 5474065 (KIC 9726699) shows a flare with energy L = 1031 ergs every 0.2 days (0.6 days) and a flare with energy L = 1032 ergs every 8.7 days (117 days), compared to AD Leo which can produce a 1032 ergs flare every 1.5 days (see Table 3). We show in Figure 5 the light curve of b oth KIC 5474065 and KIC 9726699 folded on the stars rotational p eriod together with the energy of each flare as a function of the rotational phase. Since the modulation in the light curve is caused by the rotation of stellar sp ots into and out from view, the minimum flux corresp onds to the viewing phase when the fractional surface area covered by sp ots is at its greatest (since the sp ots are cooler than the surrounding photosphere). In the case of KIC 5474065, the three most energetic flares were seen in a very short phase interval, =0.74­0.79. Curiously, in KIC 9726699 the two most energetic flares were also seen at a similar phase (=0.84­0.85) and separated by 6 rotational cycles. This indicates that the most active regions on the star are preferentially located in stellar longitude and last for timescales of (at least) several weeks (a
c 0000 RAS, MNRAS 000, 000­000

conclusion also reached by Savanov & Dmitrienko 2010). In b oth sources, flares were seen at all rotational phases.

6

KIC 5474065 AND KIC 9726699 AS FLARE STARS

The monitoring of flares from M dwarf stars has b een ongoing for the past 50 years, most of it in the Johnson U band. Although all M dwarfs monitored over an extended timescale app ear to show flares, only a few dozen have a well established flare-rate. In Figure 6 we summarise the results from several thousand hours of photometric monitoring (the caption indicates the original source of the data). Compared to the other M dwarf stars (which range from dM0 to dM8) shown in Figure 6, KIC 5474065 and KIC 9726699 show flares which are relatively energetic but occur less frequently. KIC 9726699 also produces more frequent but less energetic flares than KIC 5474065, despite the fact that b oth have a dM4 sp ectral class. Generally sp eaking, stars of sp ectral class M4 and later are fully convective and therefore have a very different magnetic top ology compared to stars with earlier sp ectral typ e. However, Morin et al. (2010) has shown this is not always the case as age may play a role in addition to mass and rotation p eriod. V374 Peg has a similar rotation p eriod as KIC 9726699, yet it can produce extremely energetic flares, e.g. Batyrshinova & Ibragimov (2001) detected an 11 mag sup erflare with an energy in ex-


6
Table 3. Flare rates for KIC 5474065, KIC 9726699 compared to AD Leo where the energies are the equivalent energy in the U band. AD Leo flare rate (days) 0.09 0.29 1.5 KIC 5474065 flare rate (days) 0.2 8.7 KIC 9726699 flare rate (days) 0.3 0.6 117

8

CONCLUSIONS

flare enery (ergs) 1030 1031 1032

cess of 1035 erg. A more plausible explanation may b e the relative sp ot coverage on these two stars. For instance Notsu et al. (2013) found that the energy of sup erflares is related to the total coverage of starsp ots and therefore the amount of magnetic energy stored around starsp ots. For those stars where monitoring exists over a numb er of years, the observed seasonably variability can b e a factor of two, p erhaps indicating cycles similar to the Sun. This has b een determined in a numb er of ways but includes narrow band photometric filters centered on the Ca i i H & K lines (eg Baliunas et al 1995) and sp ectrop olarimetric observations (eg Donati et al 2008) spread over a considerable time interval. With the p ossibility of observing flare stars using Kepler with a cadence of 1 min for weeks at a time, it will b e practical to map the activity of many stars over a timescale of years. This will also provide good motivation to re-examine the effects that stars which show many flares have on the chemistry of atmosphere's of exo-planets in the stars habitable zone.

We present Kepler short cadence observations of the M4 V star KIC 5474065 which has a rotation p eriod of 2.47 days. It shows two high amplitude short duration flares (F /F >0.4 which have integrated energies of 7 â 1032 ergs. Additional flares energies as low as 1 â 1032 ergs are also seen. We compare the flare rate of a second M4 V star KIC 9726699 which is more than 5 mag brighter than KIC 5474065 and has a more rapid rotation p eriod of 0.60 days. Compared to KIC 5474065, KIC 9726699 does not show such high amplitude flares but since the Kepler data has a higher signal to noise, it allows us to detect many short duration, low energy flares reaching energies as low as 1030 ergs. Although the effect of flares with energies of 1034 ergs on the atmosphere's of exoplanets in the habitable zone have b een investigated, it is of great interest to determine what effect the presence of many numb ers of lower energy events will have on exoplanet atmospheres.

9

ACKNOWLEDGEMENTS

7

THE IMPACT OF FLARES IN THE IMMEDIATE STELLAR ENVIRONMENT

This pap er includes data collected by the Kepler mission. Funding for the Kepler mission is provided by the NASA Science Mission Directorate. Some of the data presented in this pap er were obtained from the Mikulski Archive for Space Telescop es (MAST). STScI is op erated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS5-26555. Supp ort for MAST for non-HST data is provided by the NASA Office of Space Science via grant NNX09AF08G and by other grants and contracts. Observations were also made with the Gran Telescopio Canarias (GTC), installed in the Spanish Observatorio del Roque de los Muchachos of the Instituto de Astrofsica de Canarias, in the island of La Palma. Armagh Observatory is supp orted by the Northern Ireland Government through the Dept Culture, Arts and Leisure. We thank the referee for a constructive and helpful rep ort.

The implications of stellar flares on the atmosphere of an exo-planet orbiting around a flare star are imp ortant for the development of life as energetic flares could have a p otentially hazardous influence on its habitability. For a M4 V dwarf star the habitable zone is roughly 0.04­0.14 AU (eg Kopparapu et al 2013). For instance Segura et al (2010) determined the p otential effect of a flare as seen on the dM3e star AD Leo in April 1985. This flare which had a duration of 4 hrs was found to have an energy of L 1034 ergs in the UV/Optical wave-band (Hawley & Petterson 1991), more than one order of magnitude than the largest flare seen on KIC 5474065. Segura et al (2010) determined that such a flare was not a direct hazard for life (as we know it) on an exo-planet 0.16 AU distant from AD Leo. The flares which we rep ort here are b oth less energetic but also of much shorter duration. For KIC 5474065 the total radiated U -band flare energy budget during the SC monitoring interval was 1027 erg s-1 . Assuming that the total radiated energy over all wavelengths is one order of magnitude greater (eg Doyle & Butler 1985) implies a value two orders of magnitude less radiated energy than the large AD Leo flare referenced ab ove which may suggest a minimum effect on any nearby planet. However, what effect a sequence of frequently occurring flares still needs to b e investigated.

REFERENCES
aliunas, S. L. et al. 1995, ApJ, 438, 269 alona, L., 2012, MNRAS, 423, 3420 atyrshinova, V. M., Ibragimov, M. A., 2001, AstL, 27, 29 ochanski, J. J., Munn, J. A., Hawley, S. L., West, A. A., Covey, K. R., Schneider, D. P., 2007, ApJ, 134, 2418 Bopp, B. W., Moffett, T. J., 1974, ApJ, 185, 239 Brown, T. M., Latham D. W., Everett M. E., Esquerdo G. A., 2011, AJ, 142, 112 Byrne, P.B., Doyle, J.G. Butler, C.J., Andrews, A.D., 1984, MNRAS 211, 607 Byrne, P.B., Doyle, J.G. Menzies, J.W., 1985, MNRAS 214, 119 Cepa, J., et al., 2003, RMxAC, 16, 13 Dal, H.A., Evren, S., 2010, AJ 140, 483 Dal, H.A., Evren, S., 2011, PASP 123, 659 Dal, H.A., 2011, PASA 28, 365 Dal, H.A., 2012, PASJ 64, 82 Davenport, J. R. A., Becker, A. C., Kowalski, A. F., Hawley, S. L., Schmidt, S. J., Hilton, E. J., Sesar, B., Cutri, R., 2012, ApJ, 748, 58 Donati, J. -F., et al. 2008, MNRAS, 385, 1179 Doyle, J.G., Butler, C.J., 1985, Nat 313, 378 Doyle, J.G., Byrne, P.B., 1986, A&A 154, 370 c 0000 RAS, MNRAS 000, 000­000 B B B B


Flares detected on the M dwarf star KIC 5474065

7

0 Flare frequency v. Flare U-band energy KIC 5474065 KIC 9726699 AD Leo CN Leo CR Dra DO Cep EQ Peg EV Lac G9-38 LU Vel UV Ceti V1005 Ori V1054 Oph V577 Mon V780 Tau Wolf 424 YY Gem YZ CMi

-2

Log N/T

-4

-6

-8

28

30

32 Log Eu (ergs)

34

36

Figure 6. The cumulative flare frequency (in seconds) versus U-band flare energy (in ergs) for a large group of M dwarfs KIC 5464065 and KIC 9726699. This data has been complied from work by Moffett (1974), Lacy et al. (1976), Byrne et al. Pettersen (1975a,b, 1981a,b, 1983, 1985a,b, 2006a,b), Pettersen et al. (1983, 1984, 1986), Pettersen & Sundland (1991), D (1986), Doyle et al. (1986, 1989, 1990), Hawley et al. (1989), Leto et al. (1997), Dal & Evren (2010, 2011) and Dal (2011,

plus that for (1984, 1985), oyle & Byrne 2012).

Doyle, J.G., Byrne, P.B., Butler, C.J., 1986, A&A 156, 283 Doyle, J.G., van Oord, G.H.J., Butler, C.J., 1989, A&A 208, 208 Doyle, J.G., Butler, C.J., van Oord, G.H.J., Kiang, T., 1990, A&A 232, 83 Everett M. E., Howell S. B., Kinemuchi K., 2012, PASP, 124, 316 Gershberg, R. E., Shakhovskaia, N. L., 1983, Ap&SS, 95, 235 Greiss S., et al., 2012a, AJ, 144, 24 Greiss S., et al., 2012b, arXiv:1212.3613 Haisch, B., Strong, K. T., Rodono, M., 1991, ARA&A, 29, 275 Hawley, S. L., Pettersen, B. R., 1991, ApJ, 378, 825 Hawley, S.L., Panov, K.P., Pettersen B.R., Sundland, S.R., 1989, A&A 220, 218 Hunt-Walker N. M., Hilton E. J., Kowalski A. F., Hawley S. L., Matthews J. M., 2012, PASP, 124, 545 Kirkpatrick, J. D., Henry, T. J., McCarthy, D. W., 1991, ApJS, 77, 417 Kopparapu, R. K., et al, 2013, ApJ, 765, 131 Lacy, C.H., Moffett, T.J., Evans, D.S., 1976, ApJ S 30, 85 L´pine, S., Gaidos, E., 2011, AJ, 142, 138 e Leto, G., Pagano, I., Buemi, C.S. & Rodono, M., 1997, A&A 327, 1114 Maehara, H., Shibayama, T., Notsu, S., Notsu, Y., Nagao, T., Kusaba, S., Honda, S., Nogami, D., Shibata, K., 2012, Nature, 485, 478 Moffett, T.J., 1974, ApJS 29, 1 Morin, J., Donati, J.-F., Petit, P., Delfosse, X., Forveille, T., Jardine, M. M., 2010, MNRAS, 407, 2269 Notsu, Y., Shibayama, T., Maehara, H., Notsu, S., Nagao, T., Honda, S., Ishii, T. T., Nogami, D., Shibata, K., 2013, accepted ApJ, (arXiv1304.7361) Noyes, R. W., 1985, SoPh, 100, 385 Pettersen, B.R., 1975a, A&A 41, 87 c 0000 RAS, MNRAS 000, 000­000

Pettersen, B.R., 1975b, A&A 41, 113 Pettersen, B.R., 1981a, A&A 95, 135 Pettersen, B.R., 1981a, A&A 97, 199 Pettersen, B.R., 1983, A&A 120, 192 Pettersen, B.R., 1985a, A&A 148, 147 Pettersen, B.R., 1985b, A&A 148, 151 Pettersen, B.R., 2006a, Obs 124, 397 Pettersen, B.R., 2006b, MNRAS 368, 1392 Pettersen B.R., Kern, G.A., Evans, D.S., 1983, A&A 123, 184 Pettersen, B.R., Coleman, L.A., Evans, D.S., 1984, ApJS 54, 375 Pettersen, B.R., Panov, K.P., Sandmann, W.H., Ivanova, M.S., 1986, A&A S 66, 235 Pettersen, B.R., Sundland, S.R., 1991, A&A S 87, 303 Ramsay, G., Brooks, A., Barclay, T., Hakala, P., Still, M., Greiss, S., G¨nsicke, B., Steeghs, D., Reynolds, M., 2013, submitted, a M N RA S Reid, I. N., et al, 2004, AJ, 128, 463 Savanov, I. S., Dmitrienko, E. S., 2011, ARep, 55, 890 Segura, A., Walkowicz, L. M., Meadows, V., Kasting, J., Hawley, S., 2010, AsBio, 10, 751 Skrutskie, M. F., et al., 2006, AJ, 131, 1163 van den Oord, G. H.J., et al., 1996, A&A, 310, 908 Ventura, R., Peres, G., Pagano, I., Rodono, M., 1995, A&A, 303, 509 Walkowicz, L. M., et al 2011, AJ, 141, 50