Spectroscopy and Imaging of Radio Bursts on Acfve M Dwarfs

Image credit: NASA GSFC

Spectroscopy and imaging of radio bursts on acve M dwarfs

Jackie Villadsen NRAO Charloesville - [email protected] Radio Stars: from kHz to THz – Nov 2017 M dwarf magnec acvity shapes evoluon of planetary atmospheres

M dwarf planets have likely evolved under condions of strong stellar magnec acvity (e.g., West et al. 2008: M dwarf acvity lifeme ~ Gyr) 2 Image credit: Chuck Carter, Gregg Hallinan, and the Keck Instute for Space Studies M dwarf magnec acvity shapes evoluon of planetary atmospheres

3 Image credit: Chuck Carter, Gregg Hallinan, and the Keck Instute for Space Studies M dwarf magnec acvity shapes evoluon of planetary atmospheres

4 Image credit: Chuck Carter, Gregg Hallinan, and the Keck Instute for Space Studies 300 Radio

200 dynamic spectrum of a solar coronal harmonic mass ejecon 100 Frequency (MHz) 50 fundamental

Kouloumvakos et al. 2014 25 0 10 20 Time (minutes) 5 300

200 Sun

harmonic more dense 100

ν p = (9kHz) ne Frequency (MHz) 50 fundamental

Kouloumvakos et al. 2014 25 0 10 20 Time (minutes) 6 300

200 Sun

harmonic less 100 dense

ν p = (9kHz) ne Frequency (MHz) 50 fundamental

Kouloumvakos et al. 2014 25 0 10 20 Time (minutes) 7

to lower frequencies with time. tion and maximum bandwidth, 0.01 s and 487 MHz, respectively,

structures drift to higher frequencies with time, while others drift with our data set arise from a combination of the minimum dura-

shows that the sub-bursts appear to be fast-drift structures. Some The absolute value of the smallest inverse drift rates measurable

reaching a peak just below a frequency of 1500 MHz. Figure 2 with nonzero duration; the observed distribution is symmetric.

The bursts are conﬁned to frequencies above 1350 MHz, Figure 4 displays the inverse drift rate distribution for events

v 3.1.3. Start Frequencies and Bandwidths 3.1.4. In erse Drift Rates

¨ 1986; Bastian et al. 1990; Gudel et al. 1989; Stepanov et al. 2001). and 36%, respectively.

Á show quasi-periodic oscillations in some cases (Lang & Willson widths we are able to determine with our data are / 1%

for periodicities in the data, although previously reported bursts do sured bandwidths up to 15%. The minimum and maximum band-

Apowerspectrumanalysisoflightcurvesrevealsnoevidence fractional bandwidth is 5%, with individual bursts having mea-

behavior in two different bands are illustrated in Figures 2 and 3. Figure 4 details the distribution of bandwidths. The median

variations. (b)Dynamicspectrumofcircularpolarizationvariations.

Frequency coverage extends from 1120–1220 MHz, 1320–1589 MHz; gaps are due to RFI excision or poor bandpass response. (a)Dynamicspectrumoftotalintensity

Fig. 5.—Dynamic spectrum of slowly drifting behavior seen on AD Leo with the Arecibo Observatory. Frequency increases down and time increases to the right.

Acve M dwarfs oen produce coherent bursts at 1-5 GHz, many with frequency dri – are these space weather events?

6 Coherent radio burst on AD Leo (M3.5V) - LR ~ 10 x quiet Sun 1600

Frequency (MHz) v ~ 1200-12000 km/s

Osten & Basan 2006 1100 Arecibo

1 1020 Vol. 637 0 OSTEN & BASTIAN 2 Time (min) 8 Two coherent emission mechanisms produce solar radio bursts

Electron cyclotron Plasma emission maser (ECM) emission

Andris Vaivads hp://tempest.das.ucdavis.edu/pdg/ECE/

Solar and stellar flares, Solar and stellar flares, CME shock fronts, proton events aurorae of brown dwarfs & planets 9 Coronal density and magnec field strength correspond to low radio frequencies Electron cyclotron Plasma emission maser (ECM) emission

Andris Vaivads hp://tempest.das.ucdavis.edu/pdg/ECE/

10 -3 1/2 ωpe = (0.9 GHz) (ne/10 cm ) ωce = (2.8 GHz) BkG Observed up to ~1-2 GHz Observed up to ~10 GHz on Sun and stars on magnec brown dwarfs 10 VLA+VLBA survey of acve M dwarfs: detect coronal moon and image structure

Solar CME VLBA - UV Ce 8.4 GHz ﬂare

diameter of Kouloumvakos et al. 2014 photosphere Benz et al. 1998 VLA: Dynamic spectrum VLBA: Imaging 58 hours, 22 epochs, 5 stars 24 hours (AD Leo, UV Cet) 0.22-0.48, 1-4 GHz 8.3-8.5 GHz or 1-6 GHz Resoluon ~ stellar diameter 11 VLA: Diﬀerent frequencies probe diﬀerent distances from the star P band: 230-490 MHz 8-9 -3 L and S band: 1-4 GHz ne ~ 10 cm 10-11 -3 ne ~ 10 cm or B ~ 0.3-1.4 kG or B ~ 0.1 kG low, dense corona (~1-2 R*) height of a few R*

12 VLBA: Image large-scale structure in non-thermal radio corona

1000x brighter than Sun

Sun UV G0 Ce M6

13 4.6 GHz Sun. Stephen White/NRAO/AUI. 8.4 GHz VLBA UV Ce ﬂare. Benz et al. 1998. VLA survey: Coherent radio bursts in 13 of 23 epochs, occur on variety of mescales Long bursts (>~1 hour) Short bursts (sec - min) Requires ongoing electron acceleraon Powered by individual ﬂares? AD Leo (M3.5) UV Cet (M6) YZ CMi (M4.5) 4 Frequency (GHz)

1 0 220 0 220 0 10 Time (min)

Intense emission No emission 14 VLA survey: Coherent radio bursts in 13 of 23 epochs, occur on variety of mescales Long bursts (>~1 hour) Short bursts (sec - min) Requires ongoing electron acceleraon Powered by individual ﬂares? AD Leo (M3.5) UV Cet (M6) YZ CMi (M4.5) 4 Frequency (GHz)

1 0 220 0 220 0 10 Time (min)

Intense emission No emission 15 AD Leo (M3.5): A two week long coherent radio storm? 100% circular polarizaon

Benz 2004 Epoch 1: July 5, 2015 Epoch 2: July 19, 2015 2

1 Frequency (GHz) 0 220 0 220 Time (min) Time (min)

Frequency (GHz) 16 AD Leo: Long-duraon narrowband emission can be produced by emerging magnec ﬂux X-ray ﬂares and coherent radio emissions 17 Benz 2004

Epoch 1: July 5, 2015 Epoch 2: July 19, 2015 2

1 Frequency (GHz) 0 220 0 200 Time (min) Time (min)

Frequency (GHz) 17 Figure 2. Schematic drawings illustrating the presence of two simultaneous reconnection sites (labeled 1 and 2) in a flare. Loop system A is the main driver, and reconnection 1 is the main energy release site. Reconnected field lines are indicated by dashed curves. Left: Ahelmetconfigurationisdisturbedbytheemergingloop system A. Right: Three magnetic field systems (A, B, and C) are involved.

Multiple reconnection sites may explain the relative independence of some flare associated radio emissions such as type III bursts and possibly some decimetric emissions. Reconnection sites at high altitude are suggested to be more prolific in coherent radio emissions,suchas type III bursts, and reconnection sites at low altitude may dominate the total X-ray emission, but have radio beam signatures onlyinex- ceptional circumstances. We propose that in the ’standard pattern’ the meter and most decimeter wave emissions originate near reconnection site 2, the X-rays near reconnection site 1. In the ’type IIIm-only’ and ’afterglows’ patterns, reconnection may occur only at site 2, in the ’radio-quiet’ pattern only at site 1. In ’noise storm’ associated X-ray flares, the two reconnections sites seem to be without much physical connection.

5. Conclusions

Hard and soft X-rays indicate the dominant energy release in ﬂares. Radio emissions, on the other hand, have inspired ﬂare physics for the

RHESSI_Ph.tex; 2/02/2008; 13:57; p.17 VLBA imaging of AD Leo: Flare oﬀset from quiescent emission

8.4 GHz emission is 1 mas gyrosynchrotron from non-thermal electrons

First epoch of long-duraon 1.1-1.6 GHz coherent storm: Quiescent 8.4-GHz emission at high levels

18 VLBA imaging of AD Leo: Flare oﬀset from quiescent emission

Flare is separated from quiescent peak by 0.9 stellar diameters

0.7 mas 1.8 R* Dynamic spectrum does not show evidence of outwards source moon :(

19 VLA survey: Coherent radio bursts in 13 of 23 epochs, occur on variety of mescales Long bursts (>~1 hour) Short bursts (sec - min) Requires ongoing electron acceleraon Powered by individual ﬂares? AD Leo (M3.5) UV Cet (M6) YZ CMi (M4.5) 4 Frequency (GHz)

1 0 220 0 220 0 10 Time (min)

Intense emission No emission 20 UV Ce (M6): 90-minute bursts, similar over months

2013 May 2015 July 2015 July 2015 September 2015 80 4 mJy Frequency (GHz) 1 0.49 0.22 0 40 0 40 0 150 0 100 0 150 2 Time (min) mJy All bursts have strong right circular polarizaon à long-lasng stable magnec ﬁeld dominates in source region

Resolved imaging idenﬁes bursts with UV Cet, not BL Cet 21 Interpretaon: UV Ce has a periodic radio aurora, like brown dwarfs and planets

Frequency dri due to geometric modulaon!

22 Dipole image: Wikipedia Interpretaon: UV Ce has a periodic radio aurora, like brown dwarfs and planets

Barnes et al. 2016: inclinaon = 60 deg

Prot = 5.45 hours

UV Ce bursts also detected at 150 MHz in le and right polarizaon (Lynch et al. 2017)

- South magnec pole obscured at high frequencies? - Beaming angle eﬀect?

23 Dipole image: Wikipedia Interpretaon: UV Ce has a periodic radio aurora, like brown dwarfs and planets

Detected at 150 MHz to 8.5 GHz à B = 50 G to 3 kG if ﬁrst harmonic, 25 G to 1.5 kG if second harmonic

Kochukhov et al. 2017:

Bdip = 1.3 kG, = 6.7 kG, magnec north pole always visible

Bursts are from UV Cet, not BL Cet – is ability to sustain powerful radio aurora linked to strong large-scale B ﬁeld?

24 Dipole image: Wikipedia VLBA 8.4 GHz: Quiescent emission

RCP

1 mas

25 VLBA 8.4 GHz: Coherent burst

RCP

If this oﬀset is real:

Coherent emission comes from 3 kG 0.8 mas ﬁeld, near photosphere à 3.2R* quiescent (incoherent) emission originates oﬀ stellar limb

26 VLA survey: Coherent radio bursts in 13 of 23 epochs, occur on variety of mescales Long bursts (>~1 hour) Short bursts (sec - min) Requires ongoing electron acceleraon Powered by individual ﬂares? AD Leo (M3.5) UV Cet (M6) YZ CMi (M4.5) 4 Frequency (GHz)

1 0 220 0 220 0 10 Time (min)

Intense emission No emission 27 YZ CMi (M4.5): Short duraon burst with vercal spikes, implying rapid frequency dri

4 Peak ﬂux: 90 mJy

2 Frequency (GHz)

1 0 5 10

Time (min) 28 YZ CMi: Vercal spikes may be due to electron beams from magnec reconnecon site

4 Peak ﬂux: 90 mJy

Vercal spikes 3 imply source moon at v > 0.1c

2 Frequency (GHz) No detecon at 0.2-0.5 GHz à no evidence for parcles escaping along open ﬁeld lines

1 0 5 10

Time (min) 29 Coherent radio bursts trace electron acceleraon in corona Long bursts (>~1 hour) Short bursts (sec - min) Requires ongoing electron acceleraon Powered by individual ﬂares? AD Leo (M3.5) UV Cet (M6) YZ CMi (M4.5) 4 emerging radio magnec aurora ﬂux

Frequency (GHz) electron 1 0 220 0 220 0 beams 10 Time (min)

Intense emission No emission 30 Acve M dwarfs are eﬀecve parcle accelerators

At 1-1.4 GHz, acve M dwarfs spend 25% of me bursng at >~10x quiescent ﬂux

Signiﬁcant source of galacc radio transients

Bright bursts most frequent at 1-1.4 GHz; solar bright bursts most common at

<1 GHz (Nita et al. 2002) 31 The Astrophysical Journal, 818:105 (27pp), 2016 February 20 Mooley et al.

Transient density due to coherent radio bursts on acve M dwarfs

VLA survey of acve M dwarfs Mooley et al. 2016 CNSS galacc transients (Villadsen et al. in prep)

Figure 22. Top: the phase space of slow extragalactic transients. The panel shows the upper limits to the transient rates from previous radio surveys (colored wedges; 95% conﬁdence), the rates derived from radio transient detections (2σ error bars), and the expected transient rates. The transient detection labeled as “Le+02” represents an SN II having a peak radio luminosity of 3×1027 erg s−1 Hz−1 and an evolution timescale of ∼15 yr (Levinson32 et al. 2002; Gal-Yam et al. 2006). The one labeled “Ba+11” is a nuclear transient, SUMSS J060938–333508, with a peak radio luminosity of 6×1029 erg s−1 Hz−1 and an evolution timescale of <5 yr (Bannister et al. 2011a, 2011b, K. Bannister 2015, private communication). All observed quantities are color-coded according to the observing frequency. The solid gray line is the rate claimed by Bower et al. (2007), plotted for reference. The upper limit to the extragalactic transient rate from our pilot survey (this work) and the phase space probed by the full CNSS survey are shown as thick green wedges. The phase space probed by the VLA Sky Survey all-sky tier (VLASS) is also shown. The solid red line denotes the source counts from the FIRST survey, and the dashed red line denotes the approximate counts for strong variables at 1.4 GHz (1% of the persistent sources). Bottom: the Galactic transient phase space. Symbols have similar meanings as in the top panel. Black solid lines denote the source counts from the FIRST and the MAGPIS 1.4 GHz surveys. The source counts for variable Galactic sources approximated from Becker et al. (2010) are shown as a blue dashed line. The transient rate for active binaries resulting from our pilot survey is shown by the green error bar, and the upper limit for the rate of all other classes of Galactic transients is denoted by a thick green wedge. See Section 7.2 for more details.

24 Stokes V YZ CMi polarizaon of long-duraon >1 GHz bursts consistent EQ Peg across epochs

AD Leo RCP LCP

UV Cet

33 2128 P. Lang et al. The hidden magnetic ﬁeld of low-mass stars 2127

2128 P. Lang et al. Downloaded from Downloaded from http://mnras.oxfordjournals.org/ http://mnras.oxfordjournals.org/

2128 P. Lang et al. The hidden magnetic ﬁeld of low-mass stars 2127 at California Institute of Technology on December 15, 2016 Downloaded from at California Institute of Technology on December 15, 2016 Compare burst

YZ CMi http://mnras.oxfordjournals.org/ polarizaon to Downloaded from

large-scale B ﬁeld Downloaded from

(Zeeman Doppler http://mnras.oxfordjournals.org/ at California Institute of Technology on December 15, 2016 EQ Peg Imaging) http://mnras.oxfordjournals.org/ at California Institute of Technology on December 15, 2016

Figure 4 – continued Figure 4 – continued at California Institute of Technology on December 15, 2016

in Fig. 5, the rise and saturation of the X-ray emission measure is where B 6percentB , for the partly convective M dwarfs, ls = Total North rotation period, P, to the convective turnovernot as prominent time, τc;e.g.Pizzolato with the addition of 500saturation G of small-scale of the X-ray flux emission (red measureand B for the14per large-scale cent B field, for the fully convective M dwarfs AD Leo ls = Total et al. 2003a,b;Jeffriesetal.2011); insymbols) general, L andX/Lbol is noincreases longer and evident fordetected 1000 G through of small-scale ZDI. These flux resultsshown are shown in Fig. in6, Figs the rise5 and and6 saturationas of the X-ray emission with 3 then saturates (LX/Lbol 10− ; Delfosse(blue et symbols). al. 1998 This) with means increas- that addingblack the same symbols. surface distribution Rossby number is once again apparent. The magnitude of the X-ray ≈ ing rotation rate. This behaviour is attributedof small-scale to coronal field to saturation each star destroysWith the relationshipthe addition between of small-scaleemission field, the has correlation increased between by approximately five orders of magnitude (Vilhu & Walter 1987; Stauffer et al.magnetic1994) andflux occurs and Rossby at Ro number. Ifthe we X-ray now emission consider measure case (2) and Rossbydue to number the increase changes in flux depend- (Table 2). 0.1. For a sample of M dwarfs, Lang et al. (2012)reproducethe≈ ing on the amount of small-scale flux added. For case (1), shown South

MNRAS 439, 2122–2131 (2014) MNRAS 439, 2122–2131 (2014)

UV Cet Figure 4 – continued in Fig. 5, the rise and saturation of the X-ray emission measure is where B 6percentB , for the partly convective M dwarfs, (EQPeg B as ls = Total not as prominent with the addition of 500 G of small-scale flux (red and Bls 14 per cent BTotal , for the fully convective M dwarfs proxy) symbols) and is no longer evident for 1000 G of small-scale flux shown in= Fig. 6, the rise and saturation of the X-ray emission with (blue symbols). This means that adding the same surface distribution Rossby number is once again apparent. The magnitude of the X-ray 34 of small-scale field to each star destroys the relationship between emission has increased by approximately five orders of magnitude magnetic flux andZeeman Doppler Imaging: Morin+2008, Lang+2014; Rossby number. If we now consider case (2) due to the increase in flux (Table 2). Kochukov & Lavail 2017

Figure 4 – continued MNRAS 439, 2122–2131 (2014) Figure 4 – continued in Fig. 5, the rise and saturation of the X-ray emission measure is where B 6percentB , for the partly convective M dwarfs, ls = Total not as prominent with the addition of 500 G of small-scale flux (red and B 14 per cent B , for the fully convective M dwarfs ls = Total rotationsymbols) period, and isP, no tolonger the convective evident turnover for 1000 time, G ofτ small-scalec;e.g.Pizzolato flux shownsaturation in Fig. of the6, the X-ray rise emission and saturation measure of the for X-ray the large-scale emission with field (blue symbols). This means that adding the same surface distribution Rossby number is once again apparent. The magnitude of the X-ray et al. 2003a,b;Jeffriesetal.2011); in general, LX/Lbol increases and detected through ZDI. These results are shown in Figs 5 and 6 as of small-scale field to each star3 destroys the relationship between emission has increased by approximately five orders of magnitude then saturates (LX/Lbol 10− ; Delfosse et al. 1998) with increas- black symbols. ingmagnetic rotation flux rate. and This Rossby behaviour≈ number. is attributed If we now to coronal consider saturation case (2) dueWith to the the increase addition in fluxof small-scale (Table 2). field, the correlation between (Vilhu & Walter 1987; Stauffer et al. 1994) and occurs at Ro the X-ray emission measure and Rossby number changes depend- 0.1. For a sample of M dwarfs, Lang et al. (2012)reproducethe≈ ing on the amount of small-scale flux added. For case (1), shown MNRAS 439, 2122–2131 (2014)

MNRAS 439, 2122–2131 (2014) 2128 P. Lang et al. The hidden magnetic ﬁeld of low-mass stars 2127

2128 P. Lang et al. Downloaded from Downloaded from http://mnras.oxfordjournals.org/ http://mnras.oxfordjournals.org/

2128 P. Lang et al. The hidden magnetic ﬁeld of low-mass stars 2127 at California Institute of Technology on December 15, 2016 Downloaded from Polarizaon of at California Institute of Technology on December 15, 2016

YZ CMi long-duraon http://mnras.oxfordjournals.org/

>1 GHz bursts Downloaded from is determined Downloaded from http://mnras.oxfordjournals.org/ at California Institute of Technology on December 15, 2016

by large-scale http://mnras.oxfordjournals.org/ EQ Peg B ﬁeld at California Institute of Technology on December 15, 2016

Figure 4 – continued Figure 4 – continued at California Institute of Technology on December 15, 2016

in Fig. 5, the rise and saturation of the X-ray emission measure is where B 6percentB , for the partly convective M dwarfs, ls = Total RCP/North rotation period, P, to the convective turnovernot as prominent time, τc;e.g.Pizzolato with the addition of 500saturation G of small-scale of the X-ray flux emission (red measureand B for the14per large-scale cent B field, for the fully convective M dwarfs AD Leo ls = Total et al. 2003a,b;Jeffriesetal.2011); insymbols) general, L andX/Lbol is noincreases longer and evident fordetected 1000 G through of small-scale ZDI. These flux resultsshown are shown in Fig. in6, Figs the rise5 and and6 saturationas of the X-ray emission with 3 then saturates (LX/Lbol 10− ; Delfosse(blue et symbols). al. 1998 This) with means increas- that addingblack the same symbols. surface distribution Rossby number is once again apparent. The magnitude of the X-ray ≈ ing rotation rate. This behaviour is attributedof small-scale to coronal field to saturation each star destroysWith the relationshipthe addition between of small-scaleemission field, the has correlation increased between by approximately five orders of magnitude (Vilhu & Walter 1987; Stauffer et al.magnetic1994) andflux occurs and Rossby at Ro number. Ifthe we X-ray now emission consider measure case (2) and Rossbydue to number the increase changes in flux depend- (Table 2). 0.1. For a sample of M dwarfs, Lang et al. (2012)reproducethe≈ ing on the amount of small-scale flux added. For case (1), shown LCP/South

MNRAS 439, 2122–2131 (2014) MNRAS 439, 2122–2131 (2014)

UV Cet Figure 4 – continued in Fig. 5, the rise and saturation of the X-ray emission measure is where B 6percentB , for the partly convective M dwarfs, (EQPeg B as ls = Total not as prominent with the addition of 500 G of small-scale flux (red and Bls 14 per cent BTotal , for the fully convective M dwarfs proxy) symbols) and is no longer evident for 1000 G of small-scale flux shown in= Fig. 6, the rise and saturation of the X-ray emission with (blue symbols). This means that adding the same surface distribution Rossby number is once again apparent. The magnitude of the X-ray 35 of small-scale field to each star destroys the relationship between emission has increased by approximately five orders of magnitude magnetic flux andZeeman Doppler Imaging: Morin+2008, Lang+2014; Rossby number. If we now consider case (2) due to the increase in flux (Table 2). Kochukov & Lavail 2017

MNRAS 439, 2122–2131 (2014) 2128 P. Lang et al. The hidden magnetic ﬁeld of low-mass stars 2127

2128 P. Lang et al. Downloaded from Downloaded from http://mnras.oxfordjournals.org/ http://mnras.oxfordjournals.org/

YZ CMi long-duraon http://mnras.oxfordjournals.org/

>1 GHz bursts Downloaded from is determined Downloaded from http://mnras.oxfordjournals.org/ at California Institute of Technology on December 15, 2016

by large-scale http://mnras.oxfordjournals.org/ EQ Peg B ﬁeld at California Institute of Technology on December 15, 2016

Figure 4 – continued Figure 4 – continued at California Institute of Technology on December 15, 2016

MNRAS 439, 2122–2131 (2014) Long duraon coherent bursts from acve M dwarfs MNRAS 439, 2122–2131 (2014) are x-mode emission from electrons in the large-scale UV Cet Figure 4 – continued in Fig. 5, the rise and saturation of the X-ray emission measure is where B 6percentB , for the partly convective M dwarfs, (EQPeg B as ls = Total magnec field – same characteriscs as radio aurora, not as prominent with the addition of 500 G of small-scale flux (red and Bls 14 per cent BTotal , for the fully convective M dwarfs proxy) symbols) and is no longer evident for 1000 G of small-scale flux shown in= Fig. 6, the rise and saturation of the X-ray emission with (blue symbols). This means that adding the same surface distribution Rossby number is once again apparent. The magnitude of the X-ray 36 of small-scale field to each star destroys the relationship betweeneven if not periodic emission has increased by approximately five orders of magnitude magnetic flux andZeeman Doppler Imaging: Morin+2008, Lang+2014 Rossby number. If we now consider case (2) due to the increase in flux (Table 2).

MNRAS 439, 2122–2131 (2014) 000 7

(a) EV Lac (06): 0.32M ; Large-scale (b) Large- + small-scale field field VLA survey: Coherent radio bursts in 13 of 23 epochs, occur on variety of mescales Long bursts (>~1 hour) Short bursts (sec - min) Requires ongoing electron acceleraon Powered by individual flares? AD Leo (M3.5) UV Cet (M6) YZ CMi (M4.5) 4

1 0 220 0 220 0 10 Polarizaon consistent with star’s Time (min) Polarizaon variesà small- opcal Stokes V polarizaonà scale magnec structures large-scale magnec structures 37 (e) V374 Peg (05/06): 0.28M ; Large- (f) Large- + small-scale ﬁeld scale ﬁeld

(g) EQ Peg B (06): 0.25M ; Large-scale (h) Large- + small-scale ﬁeld ﬁeld Figure 4 – continued

c 2014 RAS, MNRAS 000, 1–11 CN Leo (Wolf 359): Radio & opcal monitoring

K2 VLA Swi GMRT ATCA

ﬂares ﬂares, coherent radio bursts (CMEs, aurorae)

38 VLA X band image: CN Leo ﬂare

Project lead: Alia Woﬀord, undergraduate at Elizabeth City State University

Please talk to me if you have any recommendaons for astrobiology PhD programs/ advisors!

temperature plot was made. Using the brightness temperature formula 2 7 dpc Sν T b = (2.04 × 10 K) 2 2 where Sv is the Stokes I flux in mJy, we calculated brightness νGHz R* temperature by assuming a source radius equal to the radius of the star R* (units of solar radii).

Results First look: Large radio and opcal ﬂare

K2: opcal (Tom Barclay, Elisa Quintana)

VLA: radio (Alia Woﬀord)

20 days 40

Figure 2: Time vs. Left (blue) and Right (green) circular polarization flux.

In this section we describe the results from the X-band data. The time vs. flux of left and right circular polarization showed CN Leo produced a bright multi-day radio flare. In Figure 2 at observations 1-7 the left circular polarization(blue) is slightly stronger than the right circular polarization(green). Images of observations 4-6 (Figure 1) show peak flare activity on CN Leo.

Wideband radio spectroscopy Very long baseline imaging reveals dynamic processes in pinpoints radio flares coronae of acve M dwarfs relave to quiescent corona emerging periodic e- beams AD Leo magnec flux radio aurora from flare AD Leo UV Cet YZ CMi Search for off-limb flares associated with outwards source moon seen 1.8 R* in coherent bursts

What mechanisms accelerate the electrons Analyzing: opcal + radio to producing long-duraon coherent bursts? determine relaonship to ﬂares

No source moon observed passing Wolf 359: K2, beyond minimum observed frequency Swi, VLA, à no CMEs, or CMEs with no emission? GMRT, ATCA

Collaborators and advisors: Gregg Hallinan, Stephen Bourke, Ryan Monroe, Tim Basan, Alia Woﬀord, Elisa Quintana, Tom Barclay, Beverly Thackeray, Starburst team, JPL DSN team Research funding: NSF GRFP, NSF ATI Program (Starburst project), Troesh Fellowship, PEO 41 Scholar Award, Jansky Postdoctoral Fellowship