The Starburst Program

We are developing the first facility dedicated to radio observation of coherent stellar bursts, recording spectra with ultra-wide bandwidth and fast cadence.  We will use dynamic spectra of stellar radio bursts and complementary observations to constrain properties of CMEs around active stars: rate, velocity, mass, dependence of direction on star's magnetic field configuration - properties needed to assess the role of coronal mass ejections in space weather around active stars (stellar mass loss and angular momentum loss, planetary habitability).

The Starburst program will observe a single-

baseline correlation with the two 27-meter

telescopes at the Owens Valley Radio

Observatory (OVRO).  The telescopes have new

cooled wideband receivers installed for the Owens

Valley Solar Array (OVSA), with an anticipated

system temperature of less than 25 K.  The

dedicated Starburst correlator will produce full

polarization spectra from 1 to 6 GHz with 1 MHz

channels and 100 ms cadence. An automated

real-time burst detection pipeline will identify

bursts as they occur, publishing them to the

Starburst web site and potentially triggering

VLBA observations.

The Starburst program will observe a sample of nearby active stars for hundreds of hours each.  The sample includes active M dwarfs (such as UV Ceti, AD Leo, EV Lac), close binaries (Algol), and stars with a high mass loss rate identified by Wood et al. 2005 (eps Eri, 70 Oph).

Imaging CMEs with VLBI

One goal of the Starburst program is to trigger observations with the Very Long Baseline Array (VLBA) in order to image the extended coronal structure of our targets after coronal mass ejections.  Benz et al. 1998 (see image below) observed UV Ceti during a flare and found that the coronal structure was resolved into two lobes aligned with the putative stellar rotation axis (and potentially with a dipolar stellar magnetic field), which raises the possibility that the magnetic morphologies of active stars may affect the direction distribution of mass loss.

Over time, VLBI observations triggered by the Starburst program will help determine if the direction of CMEs depends on the star’s magnetic field topology (as determined by Zeeman Doppler Imaging).  All of the active M dwarfs in the Starburst sample have been observed with spectropolarimetry, with published ZDI maps available for the majority.

Motivation

Left: Dynamic spectrum of a long-duration solar radio burst associated with the shock front of a coronal mass ejection (Green Bank Solar Radio Burst Spectrometer/Stephen White).  The shock front of a coronal mass ejection accelerates electrons, resulting in emission at the fundamental plasma frequency and the lowest harmonic.  Right: As the CME moves outwards, it moves into lower densities and thus the radio emission sweeps to lower frequencies.  The frequency structure of the burst can be used to measure the radial velocity of the CME.

There are two emission mechanisms that produce coherent solar radio bursts:

1. -Electron cyclotron maser emission: frequency determined by magnetic field strength
   

2. -Plasma emission: frequency determined by electron density
   

A few coherent radio bursts have been observed in other stars with dynamic spectroscopy (Güdel et al. 1989, Osten & Bastian 2006).  These are short-duration bursts, likely associated with magnetic reconnection but not necessarily with CMEs.

In order to search for long-duration bursts associated with CMEs, we have conducted VLA observations as a pathfinder for Starburst’s dedicated observing program.

JVLA Observations of UV Ceti Sweeping Radio Bursts

The figures below show a radio burst detected serendipitously during VLA observations of UV Ceti (an active M6 dwarf). The burst was bright enough to be detected on all baselines (one JVLA baseline has roughly the sensitivity that the Starburst system will have), with strong circular polarization.  The burst sweeps upwards in frequency over time from 2 to 3 GHz, potentially continuing to sweep up to 4.5 GHz.  The upwards sweep in frequency may imply bulk plasma motion downwards.  The emission mechanism must be coherent in order to produce the high degree of circular polarization, high brightness temperature (~1011 K at peak for a source the size of the stellar disk), and distinctive frequency structure.

Left: A coherent radio burst detected serendipitously during a VLA observation of UV Ceti (Hallinan et al., in prep).  The first 20 minutes were observed in S band, the latter 15 in C band.  Right: Band-averaged intensity and degree of circular polarization versus time.

After the initial, serendipitous detection of the burst shown above, we conducted follow-up observations with the JVLA with simultaneous 1-6 GHz observations of UV Ceti and AD Leo.  A preliminary reduction of the higher frequency data for UV Ceti revealed a second burst sweeping upwards in frequency (dynamic spectrum below).

Early results from JVLA observations of UV Ceti searching for sweeping radio bursts.  Like the prototype burst shown above, this burst shows strong circular polarization.  The burst sweeps upward in frequency from L band to C band with a duration of about 15 minutes; simultaneous data exists for 1 to 3 GHz that will reveal the origins of the burst at lower frequencies and earlier in time.

Both of the bursts shown above share many of the characteristics of the solar radio bursts associated with coronal mass ejections: they are produced by a coherent emission mechanism and they sweep in frequency on timescales of tens of minutes.  However, these two bursts detected on UV Ceti sweep upwards in frequency (rather than downwards) and occur at higher frequencies than the CME-associated solar radio bursts, making these UV Ceti bursts an unexpected phenomenon.  The image below illustrates two scenarios that may produce these bursts.

Two possible scenarios to explain the observed long-duration bursts that sweep upwards in frequency.  Left: bulk plasma motion downwards in the stellar corona. Right: rotationally-modulated auroral emission similar to that produced by Jupiter.  Further observations can distinguish between these scenarios by checking if the emission repeats with the stellar rotation period.  (Dipole image credit: Wikipedia.)

The 27-meter telescopes from the Owens Valley Solar Array, available for almost round-the-clock observing for the Starburst program.

VLBI image of UV Ceti’s post-flare radio corona (Benz et al. 1998).

Dipolar magnetic field structure of M4 dwarf V374 Peg, as measured by Zeeman Doppler Imaging (Moira Jardine/Donati et al. 2006).  UV Ceti likely has a similar magnetic topology; does this affect the direction of CMEs?

M dwarfs are an attractive target in the search for nearby habitable-zone planets.  However, most M dwarfs remain magnetically active for >1 Gyr (West et al. 2008). The strong flares observed in active stars are most likely accompanied by coronal mass ejections (CMEs), but these CMEs have yet to be directly observed.

The properties of CMEs on active stars are an important input to theoretical studies of stellar mass loss and angular momentum evolution and of the impact of space weather on exoplanets.  Due to lack of observational constraints on CMEs on active stars, current theoretical work scales up the properties of solar CMEs (e.g., Khodachenko et al. 2007, Lammer et al. 2007, Drake et al. 2013).

Artist’s representation of a CME interacting with Earth’s magnetosphere.  For active stars, CMEs may dominate stellar angular momentum loss and mass loss from planetary atmospheres.

Stellar Radio Bursts

The Sun produces a wide variety of radio flares.  One type of long-duration radio burst (which generally lasts minutes to tens of minutes) is due to plasma emission produced by electrons accelerated at a shock front.  Such long-duration bursts are frequently associated with coronal mass ejections.