Impact of Electron Precipitation on Brown Dwarf Atmospheres and the Missing Auroral ${{\rm{H}}}_{3}^{+}$ Emission

The detection of auroral radio emission in brown dwarf atmospheres indicates that auroral processes may be creating significant quantities of ${{\rm{H}}}_{3}^{+}$ in their upper atmospheric regions. Using the relative energy scaling observed in the Jovian atmosphere between the different multiwavelength auroral emissions, we can estimate the amount of energy in ${{\rm{H}}}_{3}^{+}$ K-band emission features we expect to see from UCDs. In Jupiter, ∼10% of the total auroral energy emerges in ${{\rm{H}}}_{3}^{+}$ emission lines, with the total emission around 4 μm composing ∼85%–90% of the ${{\rm{H}}}_{3}^{+}$ energy (Bhardwaj & Gladstone 2000). However, compared to the energy in Hα (≲1% of total) and radio (≲0.1% of total), the ${{\rm{H}}}_{3}^{+}$ features carry considerably more energy.

If we assume that all of the visible auroral energy on Jupiter (∼100 GW) emerges as Hα and that only 10% of the ${{\rm{H}}}_{3}^{+}$ energy emerges in K band (∼600 GW), then the expected energy in the overtone band is at least several factors (∼6) greater than the Hα energy. However, the Hα emission is only a portion of the visible aurorae; additional features of H₂ and broadband variations likely contribute to the total visible flux (Ingersoll et al. 1998). Thus, the K-band ${{\rm{H}}}_{3}^{+}$ lines likely carry at least as much energy as Hα and are possibly several factors stronger. The clear Hα detections of our target objects suggest that the ${{\rm{H}}}_{3}^{+}$ emission features could be strong enough to detect despite the thermal photospheric emission in K band.

We show an example of ${{\rm{H}}}_{3}^{+}$ emission features with energy equivalent to that of Hα, L_Hα /L_bol ∼ 10^−5.5, in Figure 1, sitting on top of the K-band photospheric emission represented by PHOENIX BT-Settl model spectra (Baraffe et al. 2015). The figure shows two representative spectra normalized by the mean flux between 2.1 and 2.2 μm and offset by a constant, the top line having T_eff = 2300 K and log g = 5 and the bottom line having T_eff = 1200 K and log g = 5. Because of the bright photospheric flux in the K band, the emission features are harder to distinguish for the warmer object than they are for the cooler object, in which the emission stands out clearly. Similar calculations can be done using the radio emission; however, any auroral Hα is physically associated with the atmospheric conditions producing the ${{\rm{H}}}_{3}^{+}$ and is more likely to yield an accurate estimate.

Zoom In Zoom Out Reset image size

Although UCDs have been observed extensively at IR wavelengths in the past, these observations have predominantly been at very low resolution (∼ 500), or have not focused on potentially auroral brown dwarf targets with deep K-band observations, precluding the successful identification of these features (e.g., McLean et al. 2003; Rayner et al. 2009; Blake et al. 2010). Furthermore, the ${{\rm{H}}}_{3}^{+}$ emission features are likely to be very narrow, requiring highly sensitive and at least medium-resolution spectrographs. The deep observations are essential, especially for late M dwarfs and early L dwarfs, in order to distinguish from the photospheric emission. Although this is not an issue for the cooler objects, the faint K-band magnitudes of the T dwarfs makes high-resolution observations difficult. The development of highly sensitive IR instruments provides the opportunity to look for these features in brown dwarf atmospheres.

In order to search for ${{\rm{H}}}_{3}^{+}$ emission features, we focused on the 2ν₂ overtone band around 2 μm in the K band, where we expect potentially strong ${{\rm{H}}}_{3}^{+}$ emission features. We observed our target sample (see Table 1) using the Multi-Object Spectrometer For Infrared Exploration (MOSFIRE) on the Keck Telescopes on Maunakea (McLean et al. 2010, 2012). MOSFIRE is a multiobject spectrograph with a configurable slit-mask unit (CSU) with movable bars that can be repositioned as desired during the observations. MOSFIRE uses a highly sensitive teledyne H2RG HgCdTe detector with 2k × 2k pixels within a 61 × 61 field of view and is designed to provide near-IR (NIR) spectra of faint objects (usually galaxies), but it has been used for careful spectrophotometric work during exoplanet transits (e.g., Crossfield et al. 2013). MOSFIRE provides a resolution of R ∼ 3600 in the K band with a nominal 07-wide slit, spanning 1.95–2.4 μm. The resolution in each of our observations varied according to the slit width we used, from ∼2500 to ∼8400 (see Table 1 and below for discussion).

Our observations took place over the course of 2013 and 2014 (see Table 1), usually only one or two objects on a given night with total exposure times ranging from 3 minutes, for the two bright M dwarfs in our sample, up to 50 minutes for one of our targets, but usually only about 25 minutes. Our program was clouded out on dedicated observing nights; however, we were able to take advantage of cooperation with companion programs looking for photometric variability in brown dwarfs and/or exoplanet transmission spectroscopy. During periods between transits and/or when specific targets were not yet sufficiently high in the sky, we were able to collect deep K-band spectra of our sample target list. Consequently, the time available on each target varied considerably, and we were not always able to acquire telluric calibrators to correct the K-band spectra. However, the prominent ${{\rm{H}}}_{3}^{+}$ emission features avoid the regions of intense telluric absorption in the K band around 2.02 and 2.07 μm. We present these spectra in Section 3.

We configured the CSU to contain additional reference targets to provide a check on any astrophysical variations we observed and included both narrow-slit and wide-slit sections of the slit mask. These can been seen in the raw exposure example, shown in Figure 2. The bright bands in the image correspond to the sky emission lines seen through the wide open slit (width ∼10'') and can be traced vertically to the sections with narrow skylines from the narrow slit width (∼07). During the observations, we were able to dither along the slit direction on the mask to change between one mode and the other. This proved more expedient than moving the CSU bars, because it was quicker and slight differences in the repositioning of the CSU bars can introduce systematics into the observations from frame to frame (Crossfield et al. 2013). Although there can be differences introduced from the dithering due to slight shifts in the pointing potentially shifting the target slightly within the wide slits, these proved inconsequential for the search for ${{\rm{H}}}_{3}^{+}$ emission.

Zoom In Zoom Out Reset image size

The use of both slit widths allowed us to provide precise control of the observed spectral resolution with the narrow slits while using the wide slits for spectrophotometry and to help distinguish between skylines and narrow ${{\rm{H}}}_{3}^{+}$ emission features. Although the the wide slits increase the background noise by spreading out the bright night skylines, it eliminates any issues dealing with residual skyline subtraction and confusing a poorly subtracted narrow skyline with potential astrophysical emission features. The sky emission varies on the order of 1–2 minutes, and we found that for careful skyline subtraction we required exposures of ≤30 s (see Table 1). Nevertheless, we were able to use our wide-slit exposures as a verification when assessing any narrow features in the observed spectra. Additionally, the most prominent ${{\rm{H}}}_{3}^{+}$ features avoid the major OH skylines. We illustrate this in Figure 3, showing the error spectrum from our observations of DENIS 1058–15 with an ${{\rm{H}}}_{3}^{+}$ spectrum overlaid. The spikes in the error spectrum correspond to the locations of sky emission lines, with the strongest ${{\rm{H}}}_{3}^{+}$ lines usually located between sky features. In Section 3, we present our best coadded spectra (based on signal-to-noise ratio (S/N) and resolution) for each object, usually from narrow-slit observations. Our observations of J1254–01 were conducted under good seeing conditions (see Table 1), and we thus present the wide-slit observations of that target because of their high resolution and high S/N.

Zoom In Zoom Out Reset image size

For each target we used an ABBA nodding pattern, to allow clean background subtraction between adjacent frames with typical nodding sizes of at least 4''. The IR detector was read out in the multiple correlated double sampling (MCDS) mode, 4, 8, or 16, depending on the exposure time, available observing time, and brightness of the object. While additional readouts help limit the read noise in each exposure, they generate additional overhead, which may limit the total S/N that is ultimately achieved over the course of many exposures within a given amount of time. For the brighter objects, we typically used MCDS 4 and MCDS 16 for the fainter objects with longer exposures.

The data were reduced using the MOSFIRE Data Reduction Pipeline (DRP) ⁵ to perform the flat-fielding, background subtraction through frame differencing, cosmic-ray rejection, coaddition, and wavelength calibration. We took flat fields and any necessary arc lamps at the beginning and end of the night with the CSU configured to our science masks, and when possible we prevented changes to the CSU between calibration and science exposures to limit systematic effects. The flats in K band require the use of dome lamp exposures and additional "lamp-off" exposures to assess the thermal background contribution to the "lamp-on" signal. Additionally, because the lamps were too bright to properly calibrate the wide slits, we took many additional (∼50) lamp-off exposures to construct accurate flat fields with the wide slits. ⁶ In the K band, the OH skylines are used for the primary wavelength calibration; however, they become scarce at the long-wavelength end of the band and are thus supplemented by arc lamp spectra, with the solutions being merged and corrected for the lamp optical path differing from the astrophysical signal path. The DRP produces 2D spectra, from which we use optimal extraction to create our reduced science spectra.

To address our first question, we adapt the work of Hiraki & Tao (2008), which parameterizes Monte Carlo simulations for Jovian electron beam precipitation. They determined the rate of ionization from energetic electron impacts in a molecular-hydrogen-dominated atmosphere, including energy degradation, to define the generation of ${{\rm{H}}}_{2}^{+}$ as a function of height in the atmosphere from a vertically propagating electron beam. Their simulation includes 10 different collisional processes (e.g., electronic, vibrational, and rotational excitation; ionization) that can directionally scatter and/or remove energy from the electron beam. The fundamental simplifying assumptions are that the electron path can be traced through the bulk motion of the cyclotron guiding center, while ignoring collisions with any species besides molecular hydrogen. The latter assumption has only a ∼10% effect on the height of peak ionization, and hence the localized production rate of ${{\rm{H}}}_{3}^{+}$. The cyclotron motion may influence the collision rate of the lowest-energy electrons in their simulations (1 eV), which continue to elastically scatter until they lose energy through rotational/vibrational excitation of H₂. We do not expect the guiding center assumption to alter the ionization profile, as the low-energy electrons are generally below the ionization threshold, and higher-energy electron motion is still captured well by this simplification.

The results of Hiraki & Tao (2008) are verified with Jovian conditions, and their analytic parameterization provides a useful means of applying their results toward any molecular-hydrogen-dominated atmosphere, like those of brown dwarfs. The function q_ion(ε, Z) (their Equation (5)) gives the ionization atmospheric profile for a given electron energy, ε. We can then integrate over an assumed electron beam energy flux, F_ε , to determine the atmospheric ionization rate, Q_ion, throughout the atmosphere,

$$\begin{eqnarray}&&{Q}_{\mathrm{ion}}(Z)=\int {q}_{\mathrm{ion}}(\varepsilon ,Z){F}_{\varepsilon }(\varepsilon )\ d\varepsilon ,\end{eqnarray} \tag{ 2 }$$

where Q_ion is in units of per cubed centimeter per second and the height, Z, is relative to the 1 bar level. Because ${{\rm{H}}}_{2}^{+}$ should quickly react to form ${{\rm{H}}}_{3}^{+}$, this rate effectively gives the production rate of ${{\rm{H}}}_{3}^{+}$.

The ion production rate depends on the atmosphere and the assumed electron spectrum. We show in Figure 8 how the same electron beam energy spectrum produces different ion production rates between Jupiter (density profile as in Hiraki & Tao 2008) and brown dwarfs, as a function of height above the 1 bar pressure level. For the brown dwarf atmospheres, we used the cloudless Sonora–Bobcat models in radiative–convective equilibrium (Marley et al. 2021), with $\mathrm{log}g=5$ and effective temperatures of 900 and 2300 K. We had to extend the model temperature–pressure profiles isothermally in hydrostatic equilibrium in order to determine the atmospheric conditions at low pressures (<10⁻⁴ bars). This has the effect of fixing the volume mixing ratios of each species to the level at 10⁻⁴ bars for lower pressures. The result is consistent with the expectation that atmospheric mixing is more efficient at low pressures (Mukherjee et al. 2022). Although it might not be totally accurate for isolated brown dwarfs, and other effects may influence the profile, this provides a reasonable first approximation to understand the effects of electron precipitation. Moreover, our ionization model depends almost exclusively on the H₂ density structure, dictated predominantly by the dwarf gravity and effective temperature.

Zoom In Zoom Out Reset image size

Despite the changes in height at peak ionization for the different atmospheres (Figure 8), the profile peaks actually take place at similar pressure levels between the two example brown dwarfs. In general, the physical extent of the ionization is determined by the atmospheric scale height, which is much smaller for a higher-gravity brown dwarf relative to Jupiter, and smallest for the cooler T dwarf. The corresponding height of peak ion production above the 1 bar level is thus lower in the brown dwarf regime. For Figure 8, we employed the triple Maxwellian electron energy distribution utilized by Grodent et al. (2001), as representative of the Jovian electron beam spectrum. This function combines three distributions of the form

$$\begin{eqnarray}&&{F}_{\varepsilon }(\varepsilon ;{\varepsilon }_{0})=A\displaystyle \frac{\varepsilon }{{\varepsilon }_{0}}{e}^{-\varepsilon /{\varepsilon }_{0}},\end{eqnarray} \tag{ 3 }$$

where A sets the overall beam current density in units of e⁻ s⁻¹ cm⁻² eV⁻¹ and the characteristic energy ε₀ defines the mode of the distribution. We demonstrate the impact of using different electron beam spectra on the ion production in Figure 9, including comparisons against a power-law distribution, and a single high-energy beam with distribution according to Equation (3). The normalization of the electron flux scales the total ion production (Q_ion) but is generally unknown. As such, the absolute values of Q_ion may not be physically accurate, but the relative profile with height in the atmospheres (or pressure) is. In Figure 9, we show that with broader electron spectra we would expect a broader profile of ionization, with higher-energy electrons penetrating more deeply into the atmosphere.

Zoom In Zoom Out Reset image size

We illustrate this clearly in Figure 10, with electron beam spectra utilizing a range of characteristic energies (Equation (3)). Higher-energy electrons generally also produce more ionization because after their secondary, tertiary, etc., impacts, the electrons still retain enough energy to continue to collisionally ionize H₂. We further illustrate the dependence of the pressure location (height in the atmosphere) of the peak ion production rate with characteristic electron beam energy (ε₀, using spectra as Equation (3)) in Figure 11. The location of the ion production steadily declines as the stopping height goes lower into the atmosphere until it appears to level off for the highest energies (≳1 MeV).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

As a function of pressure, the location of the ionization peak is determined by the atmospheric gravity: similar gravity means similar pressures at peak ionization. Within the formalism of Hiraki & Tao (2008), the ionization rate is directly proportional to the mass density profile, but the profile shape is dictated by the depth to which individual electrons penetrate into the atmosphere. The latter is controlled by the column density (inversely proportional to gravity at fixed pressure). This is because, along the path traversed by the energetic electrons, the product of the column density with the interaction cross section sets the probability of ionizing collisions.

Hiraki & Tao (2008) did not test their parameterization at the highest energies—they only went as high as ∼200 keV—but we illustrate the general idea here that higher-energy beams go deeper into the atmosphere to some limit. Their results likely extend toward 1 MeV energies, but likely not higher. Furthermore, the atomic data at the foundation of the Hiraki & Tao (2008) simulations have not been experimentally verified at high energies (>1 keV), although more recent collision cross-section studies remain consistent with the data used by those authors (e.g., Karwasz et al. 2022). Nevertheless, their results are consistent with observational studies of the Jovian ionization profiles, lending confidence to their work and our extension of it. These results for pressure at peak ionization are also consistent with the work of Helling & Rimmer (2019), who modeled electron beam interactions with brown dwarf atmospheres, including transport effects, but with limited electron energies. Full electron ion impact models on brown dwarfs need to verify the beam stopping heights toward the highest electron energies. Our work and similar studies would benefit greatly from experiments and theory verifying electron impact cross sections in higher-energy collisions.

Higher-energy electron beams could be expected as a consequence of magnetospheric–ionospheric coupling in brown dwarfs with stronger magnetic fields and more rapid rotation, relative to Jupiter. Strong effective field-aligned voltages (equivalent to ∼0.5 MeV) are required to support the current generated in the brown dwarf electrodynamic engines (Turnpenney et al. 2017). These electron beam energies in the range from 5 keV to 1 MeV are still permissible for ECM generation (Dulk 1985). Additionally, high-energy electrons may also be needed to model the radio beam properties of the detected gigahertz emissions (Lynch et al. 2015). We do not expect auroral electron beams to exist beyond 1 MeV energies. The ECMI is limited to only mildly relativistic particles (higher energies lead to gyrosynchrotron and synchrotron emissions; Dulk 1985; Treumann 2006). Electron beam spectra much more energetic than the Jovian case, reaching energies of several hundred keV, are consistent with our available observational constraints on auroral brown dwarfs.

Detections of atmospheric auroral emissions and the determination of where that energy is being deposited should provide a constraint on the typical electron beam energies. Auroral energy deposition at deep levels (10⁻² bars) may indicate energetic beams of up to 1 MeV, with deposition predominantly at higher altitudes giving indications of weaker electron beam energies.

In order for ${{\rm{H}}}_{3}^{+}$ to be detected as molecular emission, it must survive long enough in the atmosphere, once created, in order to radiatively de-excite. The results of the previous subsection underscore how the sustained presence of ${{\rm{H}}}_{3}^{+}$ depends on the typical electron beam energies, because deeper in the brown dwarf atmosphere the molecular ion is increasingly likely to react with molecules that would destroy ${{\rm{H}}}_{3}^{+}$—an idea first presented in Pineda et al. (2017).

H${}_{3}^{+}$ is very reactive and will be destroyed in brown dwarf atmospheres through reactions of the form

$$\begin{eqnarray}&&{{\rm{H}}}_{3}^{+}+{\rm{X}}\to {\mathrm{HX}}^{+}+{{\rm{H}}}_{2}.\end{eqnarray} \tag{ 4 }$$

Most molecular species can take on the role of X in brown dwarf atmospheres, including CH₄, CO, and H₂O. For each such reaction we can define a timescale associated with the rate at which a particular species is destroying ${{\rm{H}}}_{3}^{+}$,

$$\begin{eqnarray}&&{\tau }_{a}={[{k}_{a}{n}_{a}]}^{-1},\end{eqnarray} \tag{ 5 }$$

where k_a is the reaction rate, typically in units of per cubed centimeter per second. For faster rates (greater k_a ) and a greater number density of the particular species (n_a ), the timescale is smaller, i.e., quicker. These timescales do not depend on the amount of ${{\rm{H}}}_{3}^{+}$, but instead on the atmospheric properties wherever there is any ${{\rm{H}}}_{3}^{+}$.

Comparing the timescales of the dominant reactions driving the destruction of ${{\rm{H}}}_{3}^{+}$ with the radiative timescale for IR emissions lets us determine whether there is enough time for the molecule to appreciably emit before being destroyed. Since multiple species allow for multiple pathways to decrease the population of ${{\rm{H}}}_{3}^{+}$, we define the effective destruction timescale as

$$\begin{eqnarray}&&{\tau }_{\mathrm{eff}}\equiv {\left[\sum {\tau }_{a}^{-1}\right]}^{-1},\end{eqnarray} \tag{ 6 }$$

where the sum is over all relevant species interacting with ${{\rm{H}}}_{3}^{+}$ as in Equation (4). The timescale changes throughout the atmosphere as the density of each species changes. We show the comparison of these timescales in Figure 12, using the model example atmospheres of the late M dwarf (left) and T dwarf (right) and the known reaction rates taken from the STAND2019 chemical network (Rimmer & Helling 2016; Rimmer & Rugheimer 2019). In Figure 12, for each model atmosphere and at each pressure level, we show the corresponding destruction timescale for each species (dashed lines, etc.) and the total effective timescale (solid blue line). The radiative timescale (inverse of the Einstein A-coefficient) for the strongest ${{\rm{H}}}_{3}^{+}$ IR emission lines is ∼10⁻² s (Neale et al. 1996; Tao et al. 2011). Deeper in the atmosphere at higher pressures, the destruction timescales are smaller than the radiative timescale, indicating that at those heights ${{\rm{H}}}_{3}^{+}$ does not survive long enough to radiatively de-excite and produce the IR emission-line features. For the brown dwarfs these pressures correspond to energy deposition and ionization heights for electron beams with energies ≳2–10 keV. Only beams with low-energy electrons (similar to Jovian and weaker) are likely to generate any ${{\rm{H}}}_{3}^{+}$ from the lowest-pressure regions of the atmosphere. At these low-pressure regions (<10⁻⁶ to 10⁻⁷ bars) the mixing of the molecular species has likely decoupled (Woitke et al. 2020), above the homopause, so destruction timescales are actually longer than modeled in this study, favoring ${{\rm{H}}}_{3}^{+}$ emission at those heights.

Zoom In Zoom Out Reset image size

If the beams are very energetic, as anticipated by the modeling of the brown dwarf magnetospheric–ionospheric coupling currents (Nichols et al. 2012; Turnpenney et al. 2017), then brown dwarf atmospheres are unlikely to exhibit any ${{\rm{H}}}_{3}^{+}$ emissions. This atmospheric chemistry would explain this work and literature observations showing no evidence for any IR ${{\rm{H}}}_{3}^{+}$ emission lines. Our analysis further suggests a rough lower limit on typical late M dwarf and T dwarf electron beam energies of 10 and 2 keV, respectively, with beams needing less than this energy typically to generate long-lasting levels of ${{\rm{H}}}_{3}^{+}$ in the brown dwarf upper atmospheres. Well defined luminosity limits for the ${{\rm{H}}}_{3}^{+}$ lines would map directly to upper limits on the electron beam flux, F_ε , at low energies; however, a full model of the ionization and emissions in varied substellar atmospheres will be required to take this next step (e.g., Tao et al. 2011, 2012).

Discovered by Reid et al. (2003), LSR J1835+3259 (hereafter LSR 1835), an M8.5 dwarf with a K-band magnitude of 9.171 at a distance of 5.67 pc, has been studied extensively for its magnetic activity (Skrutskie et al. 2006). LSR 1835 is known to emit sinusoidally varying Hα emission and displays steady long-term photometric variability with a period of 2.84 hr (Harding et al. 2013a; Hallinan et al. 2015). LSR 1835 also shows both quiescent radio emission indicative of a strong radiation belt (Kao et al. 2023) and periodically pulsed highly polarized radio emission, spanning 4–8 GHz (Hallinan et al. 2008, 2015). Despite the presence of strong magnetic fields, as indicated by the radio emission, no X-ray detections have been observed from LSR 1835+32 (Berger et al. 2008, 2010). Berger et al. (2008) also report near-ultraviolet (λ_eff = 2600 Å) measurements from Swift consistent with photospheric emission. Filippazzo et al. (2015) report an effective temperature, T_eff, of 2316 ± 51 K and gravity, log g, of 5.22 ± 0.11 (cgs) base on a semiempirical approach using evolutionary models and measured bolometric luminosities from multiwavelength observations spanning the bulk of the spectral energy distribution of the target. Gibbs & Fitzgerald (2022) report emission limits of <(6.3–6.7) × 10¹⁶ W.

TVLM 513–46546 (hereafter TVLM 513) is an M8.5 dwarf with a K-band magnitude of 10.706 at a distance of 10.59 pc (Kirkpatrick et al. 1995; Dahn et al. 2002; Skrutskie et al. 2006). TVLM 513 has been studied extensively for its magnetic activity, including both Hα and radio emission, and has become a benchmark object for activity at the end of the main sequence. TVLM 513 displays highly polarized periodic pulses at a rotation period of 1.958 hr and quiescent radio emission (Hallinan et al. 2007, 2008). Extensive follow-up at radio wavelengths has shown this emission and period to be very stable and long-lasting (Wolszczan & Route 2014). TVLM 513 also shows stable photometric variations over long time periods at optical wavelengths, which could be connected to the process emitting the radio emission (Harding et al. 2013a; Wolszczan & Route 2014). TVLM 513 has an effective temperature and log gravity of 2242 ± 55 K and 5.22 ± 0.11 (cgs), respectively (Filippazzo et al. 2015).

2MASS J07464256+2000321 (hereafter J0746+20AB) is a binary system comprising L0 and L1.5 dwarfs at a distance of 12.2 pc with a K-band magnitude of 10.468 (Dahn et al. 2002; Bouy et al. 2004; Skrutskie et al. 2006). Extensive multiwavelength monitoring of this target has revealed periodic radio pulses and Hα emission that are offset in phase by 0.25 within a period of 2.07 hr, as well as nondetections in the X-ray and UV (Berger et al. 2009). Photometric monitoring of the binary by Harding et al. (2013b) indicates that the primary component is photometrically variable at optical wavelengths and rotates with a period of 3.3 hr, whereas the faster-rotating secondary is the source of the radio emission.

Discovered by Delfosse et al. (1997), DENIS J1058.7–1548 (hereafter DENIS 1058–15) is an L3 dwarf at a distance of 17.3 pc with a K-band magnitude of 12.532 (Kirkpatrick et al. 1999; Dahn et al. 2002; Skrutskie et al. 2006). Spectroscopic studies detected weak Hα emission from DENIS 1058–15 (e.g., Kirkpatrick et al. 1999. This target also displays J-band variability in photometric monitoring with similar periodic variation in the Spitzer bands at 3.6 and 4 μm (Heinze et al. 2013; Metchev et al. 2015). DENIS 1058–15 has an effective temperature and log gravity of 1809 ± 68 K and 5.20 ± 0.19 (cgs), respectively (Filippazzo et al. 2015).

2MASS J00361617+1821104 (hereafter J0036+18) is an L3.5 dwarf at a distance of 8.76 pc with a K-band magnitude of 11.058 (Kirkpatrick et al. 2000; Reid et al. 2000; Dahn et al. 2002; Skrutskie et al. 2006). Initial studies demonstrated that J0036+18 displayed strong radio emission and periodic highly circularly polarized pulses (Berger 2002; Berger et al. 2005; Hallinan et al. 2008). However, simultaneous monitoring campaigns showed no X-ray emission and an apparent lack of Hα (Berger et al. 2005, 2010). Pineda et al. 2016, however, did detect Hα emission, indicating variability timescales much longer than the rotation period of 3.08 hr (Hallinan et al. 2008). Harding et al. (2013a) detected periodic variability, consistent with the rotational period in optical photometric monitoring. Additionally, Metchev et al. (2015) detected irregularly periodic photometric variability in Spitzer monitoring of J0036+18 at both 3.6 and 4 μm. J0036+18 has T_eff = 1869 ± 64 K and log g = 5.21 ± 0.17 (cgs; Filippazzo et al. 2015).

Discovered by Leggett et al. (2000), 2MASS J12545393–0122474 (hereafter J1254–01) is a brown dwarf with an IR spectral type of T2 at a distance of 13.21 pc with a K-band magnitude of 13.84 (Vrba et al. 2004; Burgasser et al. 2006b; Skrutskie et al. 2006). Burgasser et al. (2003) report the detection of weak Hα emission from J1254–01. Further monitoring at radio wavelengths yielded no detections in two 2 hr observing blocks (Kao et al. 2016). Spitzer observations looking for photometric variability at 3.6 and 4 μm also did not yield positive results (Metchev et al. 2015). J1254–01 has T_eff = 1219 ± 94 K and log g = 5.02 ± 0.47 (cgs; Filippazzo et al. 2015).

SIMP J013656.5+093347.3 (hereafter SIMP 0136+09), discovered by Artigau et al. (2006), is a T dwarf at the L/T transition, with a K-band magnitude of 12.562, displaying strong photometric variability in the J band interpreted as heterogeneous cloud cover in its atmosphere (Skrutskie et al. 2006; Artigau et al. 2009). At a distance of 6.1 pc, SIMP 136+09 is one of the closest cool brown dwarfs and a benchmark object for atmospheric modeling (Weinberger et al. 2016). Efforts to find Hα emission have only produced upper limits despite multiple observations at different rotational phases (Pineda et al. 2016). However, radio monitoring at 4 GHz by Kao et al. (2016) has detected quiescent emission and multiple highly circularly polarized radio pulses. Kao et al. (2016) use an updated version of the spectral index methods of Burgasser et al. (2006a) to measure T_eff = $1089{\pm }_{54}^{62}$ K and log g = $4.79{\pm }_{0.33}^{0.26}$ (cgs) for SIMP 0136+09. Kinematic analysis of this target has also shown that it is a likely member af the Carina-Near moving group, has an age less than 950 Myr, and has an estimated mass of 12M_Jup (Gagné et al. 2017).

Since its discovery, 2MASS J12373919+6526148 (hereafter J1237+65), a brown dwarf at 10.4 pc with a K-band magnitude of 16.4, has been an unusual T6.5 (IR spectral type) because of its exceedingly strong Hα emission (Burgasser et al. 1999, 2003, 2006b; Vrba et al. 2004; Skrutskie et al. 2006). Many follow-up efforts have attempted to understand the cause of the emission; however, the initial studies were inconclusive (Burgasser et al. 2002; Liebert & Burgasser 2007). More recently, the search for ECM radio emission in cool brown dwarfs has revealed J1237+65 to be a radio source with moderate levels of circular polarization, likely connected to the strong Hα emission (Kao et al. 2016). J1237+65 has T_eff = 851 ± 74 K and log g = 4.95 ± 0.5 (cgs; Filippazzo et al. 2015).

First reported in Burgasser et al. (1999), 2MASS J10475385+2124234 (J1047+21 hereafter) is a brown dwarf with a spectral type of T6.5 in the IR, at a distance of 10.56 pc, with a K-band magnitude of 16.2 (Vrba et al. 2004; Burgasser et al. 2006b; Skrutskie et al. 2006). Burgasser et al. (2003) detected weak Hα emission from this target, as one of a few T dwarfs with this emission feature. J1047+21 was also the first T dwarf to be detected in radio emission (Route & Wolszczan 2012). Follow-up efforts have since confirmed the radio detection, measured quiescent emission levels, and characterized the periodic pulses of the object with a period of 1.77 hr (Williams & Berger 2015; Kao et al. 2016; Route & Wolszczan 2016). J1047+21 has an effective temperature and log gravity of 880 ± 76 K and 4.96 ± 0.49 (cgs), respectively (Filippazzo et al. 2015).