Weather on Other Worlds. IV. H$\alpha$ emission and photometric variability are not correlated in L0$-$T8 dwarfs

TL;DR

This study investigates the relationship between magnetic activity and photometric variability in ultra-cool dwarfs, finding no correlation and suggesting dust clouds drive variability rather than magnetic phenomena.

Contribution

It provides the first extensive analysis showing that Hα emission and photometric variability are uncorrelated in L and T dwarfs, emphasizing dust clouds as the primary cause of variability.

Findings

Hα emission is common in early L dwarfs but rare in later types.

Photometric variability occurs across all L and T spectral types with no clear correlation to Hα emission.

Dust clouds are likely the main cause of variability, not magnetic activity.

Abstract

Recent photometric studies have revealed that surface spots that produce flux variations are present on virtually all L and T dwarfs. Their likely magnetic or dusty nature has been a much-debated problem, the resolution to which has been hindered by paucity of diagnostic multi-wavelength observations. To test for a correlation between magnetic activity and photometric variability, we searched for H$\alpha$ emission among eight L3$-$T2 ultra-cool dwarfs with extensive previous photometric monitoring, some of which are known to be variable at 3.6 $\mu$m or 4.5 $\mu$m. We detected H$\alpha$ only in the non-variable T2 dwarf 2MASS J12545393$-$0122474. The remaining seven objects do not show H$\alpha$ emission, even though six of them are known to vary photometrically. Combining our results with those for 86 other L and T dwarfs from the literature show that the detection rate of H$\alpha$ emission is very high (94$\%$) for spectral types between L0 and L3.5 and much smaller (20$\%$) for spectral types $\ge$L4, while the detection rate of photometric variability is approximately constant (30$\%-$55$\%$) from L0 to T8 dwarfs. We conclude that chromospheric activity, as evidenced by H$\alpha$ emission, and large-amplitude photometric variability are not correlated. Consequently, dust clouds are the dominant driver of the observed variability of ultra-cool dwarfs at spectral types at least as early as L0.

Figures

Tables

Table 1. Table 1: Targets information and observing log

Object	SpTaafootnotemark:	Var. periodaafootnotemark: (h)	$J$ (mag)	UT Date	Inst.	Config.	texpbbfootnotemark: (s)	Air Mass
2MASS J01033203$+$1935361	L6	$2.7\pm 0.1$	$16.29\pm 0.08$	2013 Sep 02	GMOS-N	Blue	6$\times$900	1.06$-$1.00
				2013 Sep 29	GMOS-N	Blue	10$\times$900	1.00$-$1.11
2MASS J11263991$-$5003550	L4.5	$3.2\pm 0.3$	$14.00\pm 0.03$	2014 Mar 05	GMOS-S	Red	8$\times$900	1.12$-$1.08
				2014 Mar 07	GMOS-S	Red	8$\times$900	1.07$-$1.25
2MASS J12545393$-$0122474	T2	$\cdots$	$14.89\pm 0.04$	2014 Mar 09	GMOS-S	Red	6$\times$925	1.14$-$1.19
				2014 May 04	GMOS-S	Red	8$\times$925	1.30$-$1.14
				2014 May 05	GMOS-S	Red	8$\times$925	1.24$-$1.14
2MASS J14162408$+$1348263	L6	$\cdots$	$13.15\pm 0.03$	2014 Apr 23	GMOS-N	Blue	8$\times$780	1.07$-$1.40
2MASS J17210390$+$3344160	L3	$2.6\pm 0.1$	$13.63\pm 0.02$	2014 May 01	GMOS-N	Blue	6$\times$900	1.06$-$1.03
				2014 May 02	GMOS-N	Blue	1$\times$900	1.04$-$1.05
				2014 May 06	GMOS-N	Blue	1$\times$900	1.07$-$1.08
2MASS J17534518$-$6559559	L4	$>50$	$14.10\pm 0.03$	2013 Aug 30	GMOS-S	Blue	6$\times$900	1.23$-$1.25
				2013 Aug 31	GMOS-S	Blue	12$\times$900	1.23$-$1.45
2MASS J18212815$+$1414010	L4.5	$4.2\pm 0.1$	$13.43\pm 0.02$	2013 Aug 26	GMOS-N	Blue	8$\times$900	1.01$-$1.09
				2013 Aug 29	GMOS-N	Blue	6$\times$900	1.04$-$1.23
				2013 Aug 30	GMOS-N	Blue	2$\times$900	1.09$-$1.13
2MASS J21481633$+$4003594	L6	$19\pm 4$	$14.15\pm 0.03$	2013 Aug 08	GMOS-N	Blue	8$\times$900	1.21$-$1.07
				2013 Sep 01	GMOS-N	Blue	5$\times$900	1.21$-$1.10
				2013 Sep 02	GMOS-N	Blue	3$\times$900	1.06$-$1.08
				2013 Sep 29	GMOS-N	Blue	5$\times$900	1.07$-$1.09

Table 2. Table 2: Pseudo-equivalent widths and 3 σ 𝜎 \sigma upper limits for H α 𝛼 \alpha and Li I

Object	H$\alpha$	Li I
	(Å)	(Å)
J0103$+$1935	$>$-0.6	12.9$\pm$0.2
J1126$-$5003	$>$-0.3	$\le$0.3
J1254$+$0122	-26.6$\pm$0.2	$\le$0.5
J1416$+$1348	$>$-0.3	$\le$0.3
J1721$+$3344	$>$-0.3	$\le$0.3
J1753$-$6559	$>$-0.3	$\le$0.3
J1821$+$1414	$>$-0.3	13.5$\pm$0.1
J2148$+$4003	$>$-0.3	12.6$\pm$0.1

Table 3. Table 4: H α 𝛼 \alpha and photometric variability detections and non-detections in the literature.

Object	SpT	H$\alpha$?	Phot. var.?	Rot. Per. (h)	Filter	Duration (h)	${\sigma }_{\mathrm{phot}}$ ($\%$)	Ref.
2MASS J00584253$-$0651239	L0	Y	N	$\cdots$	$I$ Cousins	2.6	1.2	3, 29
2MASS J03140344$+$1603056	L0	Y	N	$\cdots$	$I$ Cousins	2.0	0.7	21, 29
2MASS J11593850$+$0057268	L0	Y	N	$\cdots$	$I$ Cousins	3.0	0.6	21, 29
2MASS J22000201$-$3038327	L0	Y	N	$\cdots$	$I$ Cousins	3.2	4.4	21, 29
2MASS J07464256$+$2000321A	L0.5	Y	Y	3.32$\pm$0.15	$I$	61.6aafootnotemark:	0.3a	21, 30
2MASS J14122449$+$1633115	L0.5	Y	N	$\cdots$	$I$ Cousins	3.3	1.6	21, 7
2MASS J14413716$-$0945590	L0.5	Y	N	$\cdots$	$I$ Cousins	8.2	1.2	28, 29
2MASS J23515044$-$2537367B	L0.5	Y	Y	$\cdots$	$I$ Cousins	7.0	0.9	21, 29
2MASS J10224821$+$5825453	L1	Y	Y	$\cdots$	Sloan $i$	3.5	1.4	21, 38
2MASS J1045240$-$014957	L1	Y	N	$\cdots$	$I$ Cousins	6.7	0.6	28, 29
2MASS J10484281$+$0111580	L1	Y	Y	$\cdots$	$I$ Cousins	2.6	1.0	21, 29
2MASSW J1108307$+$683017	L1	Y	N	$\cdots$	Sloan $i$	4.5	1.0	4, 38
2MASS J14392836$+$1929149	L1	Y	N	$\cdots$	$R$, $I$ Cousins	6.8, 1.2	1.1, 0.9	5, 9, 15, 21, 29
2MASS J15551573$-$0956055	L1	Y	N	$\cdots$	$I$ Cousins	11.1b	0.5	16, 21
WISEP J190648.47$+$401106.8	L1	Y	Y	8.9 h	Kepler	2.3 yr	0.5	40
2MASS J1300425 $+$ 191235	L1	Y	Y	$\cdots$	R, $I$ Cousinsc	5.6, 22.9	1.1, 1.4	7, 15, 17, 28, 33
2MASSI J0829066$+$145622	L1	Y	N	$\cdots$	$I$ Cousins	3.2	0.8	29, 36
DENIS J090957.1$-$065806	L1	Y	N	$\cdots$	$I$ Cousins	6.3	0.9	8, 29, 41
2MASS J11455714$+$2317297	L1.5	Y	Y	$\cdots$	$I$ Cousins	9.3	1.8	5, 21
2MASS J1334062$+$194035	L1.5	Y	Y	$\cdots$	$I$ Cousins	6.8	1.8	5, 28
2MASS J16452211$-$1319516	L1.5	Y	N	$\cdots$	$I$ Cousins	3.1	1.0	11, 21
2MASS J20575409$-$0252302	L1.5	Y	N	$\cdots$	$I$ Cousins	4.4	1.0	11, 17, 21
2MASSW J0832045$-$012835	L1.5	Y	N	$\cdots$	$I$ Cousins	2.0	1.1	3, 29
2MASSW J0135358$+$120522	L1.5	Y	N	$\cdots$	Sloan $i$	2.2	1.8	3, 38
DENIS J174534.6$-$164053	L1.5	Y	N	$\cdots$	$I$ Cousins	2.7	1.4	29, 41
2MASS J08283419$-$1309198	L2	Y	Y	2.9	$I$ Cousins	11.4d	1.3	14, 21
2MASS J0921141$-$210444	L2	Y	Y	$\cdots$	$R$, $I$ Cousins	1.4, 1.4	1.8, 1.0	28, 29
2MASS J11553952$-$3727350	L2	Y	Y	$\sim$8	$I$ Cousins	23.8e	0.5	11, 21
Kelu$-$1	L2	Y	Y	1.8$\pm$0.05	$I$	13.0f	0.1	8, 12, 21
2MASS J00154476$+$3516026	L2	Y	N	$\cdots$	Sloan $i$	2.8	0.5	3, 38
2MASSW J0030438$+$313932	L2	Y	N	$\cdots$	$I$	4.9	2.3	1, 5
2MASSI J0847287$-$153237	L2	Y	N	$\cdots$	$I$ Cousins	7.6	1.0	29, 41
2MASSI J1726000$+$153819	L2	N	N	$\cdots$	[3.6], [4.5]	14, 7	0.29, 0.49	3, 37
2MASSW J2208136$+$292121	L2	N	Y	3.5$\pm$0.2	[3.6], [4.5]	14, 7	0.07, 0.11	3, 37
2MASS J05233822$-$1403022	L2.5	Y	N	$\cdots$	$I$ Cousins	7.0	0.8	21, 29, 33
2MASS J10292165$+$1626526	L2.5	Y	N	$\cdots$	$R$, $I$	2.5, 2.2	1.8, 1.1	19, 21
2MASS J1047310$-$181557	L2.5	Y	N	$\cdots$	$R$, $I$ Cousins	3.5, 13.6g	2.3, 1.0	28, 29
DENIS J081231.6$-$244442	L2.5	Y	N	$\cdots$	$R$, $I$ Cousins	3.4, 5.7	2.5, 0.9	41, 29
2MASS J10584787$-$1548172	L3	Y	Y	4.1$\pm$0.2	[3.6], [4.5]	8, 6	0.2, 0.2	1, 37
2MASS J0913032$+$184150	L3	Y	N	$\cdots$	$I$	5.4	3.5	5, 28
2MASS J1203581$+$001550	L3	Y	N	$\cdots$	$I$	6.8	0.6	5, 28
2MASS J15065441$+$1321060	L3	Y	N	$\cdots$	$R$, $I$	4.5, 5.2	1.8, 0.9	4, 19, 21
2MASS J1615441$+$355900	L3	Y	N	$\cdots$	$I$ Cousins	3.6	6.0	7, 28
2MASS J2104149$-$103736	L3	Y	N	$\cdots$	$I$ Cousins	16.4h	1.1	28, 29
2MASS J08355829$+$0548308	L3	Y	N	$\cdots$	$I$ Cousins	6.5	1.5	29, 36
2MASS J11463449$+$2230527	L3	Y	N	$\cdots$	$I$	3.9	1.0	5, 36
2MASS J03261367$+$2950152	L3.5	Y	N	$\cdots$	$I$	2.5	1.6	1, 5
2MASS J00361617$+$1821104	L3.5	Y	Y	2.7$\pm$0.3	$I$, [3.6], [4.5]	10.5, 8, 6	0.9, 0.1, 0.1	30, 37, 43
2MASS J1705483$-$051646	L4	Y	N	$\cdots$	$J$	3.04	0.02	28, 45
2MASS J161542552$+$49532117	L4	N	Y	24	$I$	3.6	5.9	7, 28
DENIS 1228$-$1547	L4.5	Y	N	$\cdots$	$I$ Cousins, $J$	4.3, 3.4	1.3, 0.7	12, 29, 33, 44
2MASSW J2224438$-$015852	L4.5	Y	N	$\cdots$	[3.6], [4.5]	12, 8	0.14, 0.14	3, 37
2MASS J13153094−2649513	L5	Y	N	$\cdots$	$I$ Cousins	2.9	1.3	27, 29
2MASS J01443536$-$0716142	L5	Y	Y	aperiodic	$I$ Cousins	14.5k	1.5	13, 29
2MASSI J0421072$-$630602	L5	N	N	$\cdots$	[3.6], [4.5]	14, 7	0.2, 0.2	24, 37
2MASSW J0820299$+$450031	L5	N	N	$\cdots$	[3.6], [4.5]	14, 7	0.4, 0.4	3, 37
DENIS$-$P J142527.97$-$365023.4	L5	N	Y	3.7$\pm$0.8	$J$	3.24	0.02	18, 45
2MASSW J1507476$-$162738	L5	N	Y	2.5$\pm$0.1	[3.6], [4.5]	12, 8	0.14, 0.14	3, 22, 28, 37
2MASS J11501322$+$0520124	L6	N	N	$\cdots$	[3.6], [4.5]	14, 7	0.5, 0.6	26, 37
2MASSI J0825196$+$211552	L7.5	N	Y	7.6	[3.6], [4.5]	12, 9	0.14, 0.14	21, 37
2MASS J15450901$+$3555271	L7.5	N	N	$\cdots$	[3.6], [4.5]	14, 7	0.8, 0.9	26, 37
Luhman 16A	L7.5	N	Y	$\sim$4.5–5.5	$RIzYJH{K}_s$	several epochs	0.1–1.5	31, 34, 46, 47, 48
SDSS J042348.57–041403.5	L7.5	Y	N	$\cdots$	$J$	3.57	0.02	43, 45
2MASS J01075242$+$0041563	L8	N	Y	Irregular	[3.6], [4.5]	14, 7	0.3, 0.2	18, 37
2MASS J16322911$+$1904407	L8	N	Y	3.9$\pm$0.2	[3.6], [4.5]	8, 6	0.4, 0.3	1, 37
2MASSI J0328426$+$230205	L9.5	N	N	$\cdots$	[3.6], [4.5]	14, 7	0.5, 0.6	3, 37
2MASS J15203974$+$3546210	T0	N	N	$\cdots$	[3.6], [4.5]	14, 7	0.3, 0.4	26, 37
2MASS J15164306$+$3053443	T0.5	N	Y	6.7	[3.6], [4.5]	14, 7	0.4, 0.5	26, 37
Luhman 16B	T0.5	N	Y	4.87$\pm$0.01	$RIzYJH{K}_s$	several epochs	0.1–1.5	31, 34, 46, 47, 48
2MASS J08583467$+$3256275	T1	N	N	$\cdots$	[3.6], [4.5]	14, 7	0.3, 0.4	26, 37
2MASS J21392676$+$0220226	T2	N	Y	7.72	$J$	2.5	0.05	43, 45
SDSS J075840.33$+$324723.4	T2	N	Y	4.9$\pm$0.2	$J$	3.54	0.02	43, 45
2MASS J120956131$-$10040081	T2$+$T7.5	N	N	$\cdots$	[3.6], [4.5]	14, 7	0.7, 0.7	37, 43
SIMP J013656.5$+$093347.3	T2.5	N	Y	2.3895$\pm$0.0005	$J$	4 nights	0.5	25, 43
SIMP J162918.41$+$033537.0	T3	N	Y	6.9$\pm$2.4	$J$	4.03	0.01	43, 45
SDSS J102109.69$-$030420.1	T4	N	N	$\cdots$	$J$	3.17	0.7	23, 33, 43
2MASS J05591914–1404488	T4.5$/$T5	N	N	$\cdots$	$J$	3.52	0.07	43, 45
2MASSI J2254188$+$312349	T5	N	N	$\cdots$	[3.6], [4.5]	14, 7	0.7, 0.7	37, 43
2MASS J12255432$-$2739466	T6	N	N	$\cdots$	$J$	2.85	0.7	33, 43
2MASSI J1534498$-$295227	T6	N	N	$\cdots$	$J$	3.88	0.7	33, 43
SDSSp J162414.37$+$002915.6	T6$/$T6	N	N	$\cdots$	$I$	4 nights	4.4	35, 43
2MASSW J1047539$+$212423	T7	Y	N	1.77ll	$J$	9.9l	1.1	39, 43
2MASS J12171110$-$0311131	T7	N	Y	$\cdots$	$J$	3.2	0.07	43, 45
2MASS J12373919$+$6526148	T7	Y	Y	$\cdots$	$J$	2.5	2.3	6, 43
2MASSI J0415195$-$093506	T8	N	Y	$\cdots$	$J$	3.79	0.01	43, 45
2MASS J07271824$+$1710012	T7$/$T8	N	N	$\cdots$	$J$	5.0	1.1	32, 43

Full text

brown dwarfs — stars: activity — stars: low-mass — stars: rotation — stars: variables

1 Introduction

Observational studies have shown that ultra-cool dwarfs (spectral types $\geq$M7) can display spectro-photometric variability with typical timescales of a few hours (Bailer-Jones & Mundt, 2001; Gelino et al., 2002; Littlefair et al., 2008; Artigau et al., 2009; Radigan et al., 2012; Apai et al., 2013; Koen, 2013; Radigan et al., 2014; Crossfield et al., 2014; Yang et al., 2016). Accurate spectroscopic time series using the Hubble Space Telescope and precise 3–5 $\mu$m monitoring with the Spitzer Space Telescope, as part of the Storms and Weather on Other Worlds campaigns, indicate that atmospheric spots responsible for this variability are ubiquitous on L3–T8 dwarfs (Buenzli et al., 2014; Metchev et al., 2015). Some of these observations, combined with state-of-the-art radiative transfer models, have shown that the detected spots can generally be attributed to atmospheric dust clouds that modulate the object’s brightness as it rotates (Artigau et al., 2009; Marley et al., 2010; Radigan et al., 2012; Morley et al., 2012; Apai et al., 2013; Yang et al., 2015). However, the consistent detection of variability in the warmest spectral types of ultra-cool dwarfs (late-M and early L) also points to another possibility: magnetically induced chromospheric activity and hot or cold star spots, similarly to F-M stars, for which magnetic activity can produce spots (Strassmeier, 1994) that lead to photometric variability (e.g., Hooten & Hall 1990; Henry et al. 1995, or Strassmeier et al. 1997).

Magnetic activity can be revealed in different ways, for example, as emission lines from the chromosphere (e.g., Ca II H and K lines) or from the transition region (e.g., C IV), coronal X-rays, radio emission, and/or spots and flares (Reid, 2013, and references therein). In most cases the observation of these activity indicators is extremely challenging for ultra-cool dwarfs: observations at X-rays and radio wavelengths are limited to the closest ultra-cool dwarfs; while Ca II H and K emissions ($\lambda$3968, $\lambda$3933 Å) are almost undetectable given the low fluxes of ultra-cool dwarfs at $\leq$6000 Å. Several works on F-M stars have shown that H$\alpha$ emission is also an appropriate tracer of magnetic activity as it correlates with the Ca II H and K lines (Zarro & Rodgers, 1983; Pasquini & Pallavicini, 1991; Montes et al., 1995), the C IV line (Rutten et al., 1991), and/or X-rays emission (Doyle, 1989; Young et al., 1989). H$\alpha$ emission is one of the most suitable activity indicators in ultra-cool dwarfs, which emit most of their flux at red optical and near-infrared wavelengths. Observations of H$\alpha$ emission have revealed that virtually all SDSS late-M dwarfs are chromospherically active (Schmidt et al., 2007; West et al., 2011). While for L dwarfs, Schmidt et al. (2007, 2015) observed a decreasing fraction of H$\alpha$ emitters into the L s; and Pineda et al. (2016) found that 9.2$\pm^{3.5}_{2.1}\,\%$ of L4–T8 dwarfs show H$\alpha$ emission. Star spot-like cool regions with frozen-in magnetic field lines are not expected in the highly neutral atmospheres of ultra-cool dwarfs (Mohanty & Basri, 2003). However, energetic magnetic field discharges ($\sim$10 kG) are a potential source of chromospheric heating (“hot spots”) as seen in some late-M and early-L dwarfs (e.g., Berger, 2002, 2006; Berger et al., 2008; Yu et al., 2011). More recently, optical and radio aurorae have also been proposed as likely drivers of some of the observed photometric variability at optical and near-infrared wavelengths (Hallinan et al., 2015; Kao et al., 2016).

The likely magnetic or dusty nature of the atmospheric inhomogeneities of ultra-cool dwarfs has been a much-debated problem almost since the discovery of the first brown dwarfs (Tinney & Tolley, 1999; Martín et al., 2001; Bailer-Jones, 2002; Lane et al., 2007). Recent multi-wavelength photometric studies in early L dwarfs have shown sinusoidal periodicities that are well-explained by a thick long-lived cloud. Heinze et al. (2013) presented data for the L3 dwarf DENIS-P J1058.7$-$1548 in the $J$, 3.6 $\mu$m, and 4.5 $\mu$m bands, and showed that it is difficult to account for the observed photometric variability by magnetic phenomena unless they are combined with cloud inhomogeneities. Also, Gizis et al. (2015) showed that the optical light curve of the L1 dwarf WISEP J190648.47$+$401106.8 has remained stable for $\sim$2 years and in phase with other light curves at 3.6 $\mu$m and 4.5 $\mu$m, while its variable H$\alpha$ emission is not synchronized with the light curves. Such observational evidence points to the existence of a long-lived cloud.

Nonetheless, with chromospheric activity expected to produce star spot-like inhomogeneities, it is important to address the question whether star spots, rather than dust clouds, may dominate the photometric variability of active ultra-cool dwarfs. We address this problem through a combination of deep optical spectroscopy of a sample of eight ultra-cool dwarfs, whose photometric light curves have been studied to a high precision with Spitzer, and a literature sample of 86 photometrically and spectroscopically observed L0-T8 dwarfs. We seek to answer whether the combined sample shows a correlation between magnetic activity—probed by the detection of H$\alpha$ emission—and photometric variability. We briefly describe our sample in Section 2 and the observations and data reduction in Section 3. The main results of our survey and its combination with literature data are presented in Sections 4–5. We discuss their implications for the correlation between observed chromospheric activity and photometric variability in Section 6, and summarize our findings in Section 7.

2 Sample selection

We selected seven L3–L6 dwarfs and one T2 brown dwarf from the sample of Metchev et al. (2015). Our targets are among the brightest (J$$\sim13.2–16.3 mag) dwarfs studied in that work, in which they were continuously monitored in the 3.6 $\mu$m and 4.5 $\mu$m bands for a total of 16–21 h by using the IRAC instrument on Spitzer.

Five objects exhibited photometric periodicities in the 2.7–19 h range, attributed to rotation. Another of our targets seems to be a long periodicity variable with a time scale greater than 50 h, and the remaining two did not show any photometric variability within $\pm$0.91% in either of the 3.6 $\mu$m and 4.5 $\mu$m bands.

The full names of our targets, their spectral types, variability period (if measured), and $J$-band magnitudes are listed in columns 1–4 of Table 1. Henceforth we will use abridged names for the targets. For more details about the objects we refer to Metchev et al. (2015).

3 Observations

We used the two copies of the Gemini Multi-Object Spectrograph (GMOS; Hook et al., 2004) mounted on the 8 m Gemini North and Gemini South telescopes to collect optical spectra of our targets. Observations were carried out in queue mode between August 2013 and May 2014.

Five of our targets (J0103$+$19, J1416$+$13, J1721$+$33, J1821$+$14, and J2148$+$40) were observed from the North using the R831*-G5302 grating and the remaining three (J1126$-$50, J1254$-$01, and J1753$-$65) from the South with the R831-*G5322 grating. At both telescopes we used a slit of 0.75$\arcsec$ (pixel scale of 0.08″) with a binning of 2 pixels, which yielded a resolution of $\sim$6 Å in our spectra. In each campaign we also used the second order-blocking filter GG455 to avoid contamination of our data with stray light from wavelengths $\leq$5000 Å. For six of our targets we used a central wavelength of 5712 Å (“blue configuration”), and for the other two a central wavelength of 7300 Å  (“red configuration”). Data collected with the blue and red configurations cover 4670–6820 Å and 6270–8460 Å, respectively. For each target we collected between 8 and 21 individual spectra, by dithering along the spectroscopic slit, with typical exposure times of 780–900 s. On each night we also recorded spectra of Cu$+$Ar lamps with the same instrumental setup as used for the science targets. No standard spectrophotometric stars were observed in any observing epoch, thus, our final spectra were not calibrated in flux. In columns 5–9 of Table 1 we provide for each target: dates of observations, instrument used, instrumental configuration, number of individual spectra collected multiplied by individual exposure time, and the range of airmass covered.

We used the Gemini package within the Image Reduction and Analysis Facility software (IRAF) for bias subtraction and flat fielding of the data. We extracted the spectra using the IRAF apextract standard routines. We employed typical aperture widths of 20 pixel centered on the spectroscopic traces, which we fit with second- to fourth-order Legendre polynomials. We estimated sky backgrounds from two 20-pixel-wide bands, centered 20–30 pixels away from the trace of the object. Figure 1 shows the final wavelength-calibrated (with a typical uncertainty of $\pm$0.1–0.2 Å) and median-combined spectra for all eight targets.

4 Results from our survey

4.1 H$\alpha$ emission

The main motivation of our survey is the search for H$\alpha$ emission in a set of targets for which we have exquisitely precise determinations of the presence or absence of photometric variability. As seen in Figure 1, we detect H$\alpha$ in only one of the eight targets: the field T2 dwarf J1254$-$01, which is not known to vary photometrically. H$\alpha$ emission from this T2 dwarf was already reported by Burgasser et al. (2003). The remaining seven targets in our sample—all L dwarfs—do not show any sign of H$\alpha$ emission, even though five of them exhibit photometric variability compatible with rotation.

Figure 2 shows the locations of the H$\alpha$ $\lambda 6563$ Å  and Li I $\lambda 6708$ Å  lines. H$\alpha$ emission is not detectable even at low levels among the seven L dwarfs. We collected spectra of most of our targets, with the exception of the L6 dwarf J1416$+$13, on at least two different nights. Weak signs of H$\alpha$ are not seen in the individual spectra, either. Therefore, it is unlikely that we have missed a phase of emission. Note that the photospheric H$\alpha$ absorption—seen from F to mid M stars—is not expected in the spectra of ultra-cool dwarfs ($\geq$M7) since their photospheres are too cool to populate the $n=2$ level of hydrogen significantly. Thus, H$\alpha$ can only be present as an emission line if there is a chromospheric heating (Cram & Mullan, 1985; Giampapa, 1985; Giampapa & Liebert, 1986; Basri, 2000).

We measured the pseudo-equivalent width (pEW) of the H$\alpha$ emission line detected in J1254–01 by first subtracting the neighboring continuum determined from a 45 Å-wide region (excluding the 10 Å  centered in the H$\alpha$ line), and then summing the flux contained in the central 10 Å  as done in Pineda et al. (2016). We also integrated the flux in different regions of the continuum and adopted the average value as the uncertainty in our pEW. For the remaining 7 targets without H$\alpha$ emission, we adopted the same procedure for estimating 3$\sigma$ upper limits. These values are given in Table 4.1.