Low EUV Luminosities Impinging on Protoplanetary Disks
I. Pascucci, L. Ricci, U. Gorti, D. Hollenbach, N. P. Hendler, K. J. Brooks, Y. Contreras
aa r X i v : . [ a s t r o - ph . S R ] A ug Low EUV Luminosities Impinging on Protoplanetary Disks
I. PascucciLunar and Planetary Laboratory, The University of Arizona, Tucson, AZ 85721, USA [email protected]
L. RicciDepartment of Astronomy, California Institute of Technology, MC 249-17, Pasadena, CA91125, USAU. Gorti and D. HollenbachSETI Institute, 189 Bernardo Ave., Mountain View, CA 94043, USAN. P. HendlerLunar and Planetary Laboratory, The University of Arizona, Tucson, AZ 85721, USAandK. J. Brooks and Y. ContrerasAustralia Telescope National Facility, PO Box 76, Epping, NSW 1710, AustraliaReceived ; accepted NASA Ames Research Center, Moffett Field, CA 94035, USA 2 –
ABSTRACT
The amount of high-energy stellar radiation reaching the surface of protoplan-etary disks is essential to determine their chemistry and physical evolution. Here,we use millimetric and centimetric radio data to constrain the EUV luminosityimpinging on 14 disks around young ( ∼ photons/s forall sources without jets and lower than 5 × photons/s for the three oldersources in our sample. These latter values are low for EUV-driven photoevap-oration alone to clear out protoplanetary material in the timescale inferred byobservations. In addition, our EUV upper limits are too low to reproduce the[Ne ii ] 12.81 µ m luminosities from three disks with slow [Ne ii ]-detected winds.This indicates that the [Ne ii ] line in these sources primarily traces a mostlyneutral wind where Ne is ionized by 1 keV X-ray photons, implying higher pho-toevaporative mass loss rates than those predicted by EUV-driven models alone.In summary, our results suggest that high-energy stellar photons other than EUVmay dominate the dispersal of protoplanetary disks around sun-like stars. Subject headings: protoplanetary disks – radio continuum: planetary systems – stars:pre-main sequence
1. Introduction
Gas-rich dust disks around young stars (hereafter, protoplanetary disks) provide theraw material to build up planets. Hence, it is critical to understand when and how theydisperse. Observations of protoplanetary disks suggest that they follow a two-timescaleevolution. In the first few Myr their dispersal is thought to be driven by viscous evolution,accretion of disk gas onto the central star. This stage is followed by a rapid ( ∼ yr)clearing attributed mainly to photoevaporation driven by the central star, which heatsgas in the disk surface to thermal escape velocity (e.g. Alexander et al. 2013 for a recentreview).While observations of blueshifted lines tracing the disk surface demonstrate thatphotoevaporation is occurring in some systems (e.g. Pascucci & Sterzik 2009), its role inclearing protoplanetary material is still debated. This is mainly because photoevaporativemass loss rates are poorly constrained observationally. At the same time predicted valuesspan two orders of magnitude for sun-like stars, from ∼ − to ∼ − M ⊙ /yr, dependingon the high-energy photons dominating the photoevaporation (Alexander et al. 2006;Gorti & Hollenbach 2009; Owen et al. 2010). The smallest mass loss rate implies thatphotoevaporation contributes only to the latest stages of disk clearing, while the largestis close to the median mass accretion rate of ∼ ∼ . Thus, the stellar chromosphereand/or corona are thought to dominate their X-ray emission (e.g. G¨udel & Naze 2009 fora review). Hard X-rays (1-10 keV) are also not easily absorbed by circumstellar matter(e.g. Ercolano et al. 2009), such as accretion columns or magnetically driven winds, hencemeasured stellar X-ray luminosities should provide a good estimate for the hard X-rayradiation reaching the disk surface. At the opposite side of the high-energy stellar spectrum,the H -dissociating far-ultraviolet (FUV; 6-13.6eV) luminosity of young sun-like starshas been found to be proportional to their accretion luminosity (e.g. Ingleby et al. 2011;Yang et al. 2012). This suggests that the FUV emission mostly traces disk gas accretingonto the star and shocked at the stellar surface. Finally, extreme-UV radiation (EUV;13.6-100eV) is poorly constrained both for young (e.g. Alexander et al. 2005) as wellas for old stars (e.g. Ribas et al. 2005; Linsky et al. 2014). This is because interstellargas easily absorbs EUV photons thus hampering their direct detection. Similarly, stellarEUV photons may be absorbed in accretion columns or magnetically driven jets launchedfrom the disk surface near the star (e.g. Alexander et al. 2004; Hollenbach & Gorti2009). Hence, the EUV radiation reaching the disk may be substantially lower than thatemitted by the star. It is also debated if most of the EUV emission originates in theaccretion shock or in the chromosphere (e.g. Alexander et al. 2004; Herczeg 2007). Onlywith a rather large chromospheric component ( > photons/s), which would not fadeaway as accretion declines, could EUV radiation shut off disk accretion and clear theprotoplanetary material in a timescale consistent with that observed (e.g. Alexander et al.2006; Alexander & Armitage 2009).Ground-based observations have demonstrated that the surface of protoplanetary However, a soft excess, identified as anomalously high fluxes in lines forming at temper-ature of only a few MK, is seen in all accreting young stars (e.g. G¨udel & Naze 2009). 5 –disks can be ionized by the central star high-energy photons, e.g. via the detectionand characteristic profile of the [Ne ii ] emission line at 12.8 µ m (Herczeg et al. 2007;Pascucci & Sterzik 2009; Najita et al. 2009; Baldovin-Saavedra et al. 2012; Sacco et al.2012). In a previous contribution we showed that a fully or a partially ionizedprotoplanetary disk surface emits free-free cm radiation that should also be detectablewith current astronomical facilities (Pascucci et al. 2012). We also derived analytic scalingrelations between the ionizing radiation impinging on the disk and the free-free diskemission. Recent hydrodynamical model calculations of Owen et al. (2013) agree with ourderived relations. Hence, free-free cm emission can be used to constrain the high-energyradiation actually reaching the disk and photoionizing H.Building on these findings, here we provide stringent upper limits on the EUV photonluminosity reaching the disk for 14 young ( ∼ −
10 Myr) stars. The paper is organized asfollows. In Sect. 2 we summarize new and nearly simultaneous cm observations of six youngstars with disks obtained with the Australia Telescope Compact Array (ATCA). These sixtargets were chosen because they are relatively nearby, far away from massive stars, andhave ancillary evidence of an ionized disk surface from the detection of the [Ne ii ] line at12.81 µ m (see Table 1 for the main properties). We also preferred sources with no knownjets because shocked ISM gas is known to produce [Ne ii ] and free-free cm emission (e.g.Anglada et al. 1998; van Boekel et al. 2009). In Sect. 3 we discuss the immediate resultsand show that all ATCA sources have cm emission in excess to the dust thermal emission.We then measure the excess cm emission for the ATCA sources as well as for eight otherdisks in the literature which have a good coverage at millimeter and centimeter wavelengths.Finally, in Sect. 4 we assume that all excess cm emission is due to free-free disk emissionand derive upper limits on the EUV photon luminosity impinging on the disk. We discussthe main implications of our findings in Sect. 5. 6 –
2. Observations and Data Reduction
Centimeter continuum observations were carried out with the 6 ×
22 m antennas ATCAinterferometer between 2012 October 18 and October 21. We used the hybrid H214 arrayconfiguration, where five antennas are arranged with baselines between 82 and 247 m, andthe sixth antenna is on a 4.5 km baseline. However, the atmospheric phase stability was toopoor to calibrate the longest baselines therefore we discarded all the data from the sixthantenna. Our observations were conducted with the Compact Array Broadband Backend(CABB), 2049 channels with a total bandwidth of 2 GHz, dual sideband with frequencypairs centered at 33+35 GHz (8.8 mm) and 17+19 GHz (17 mm), and simultaneousobservations at 9.0+5.5 GHz (3.3 and 5.5 cm, respectively). In addition to our six sciencetargets we also observed the source 1934-638 for flux calibration and 1921-293 for bandpasscalibration. For each target we also identified a nearby and bright gain/phase calibratorwhose exposures were interleaved with the science target exposures (see Table 2).The data reduction followed the standard CABB procedure described in the ATCAuser guide and was carried out with the software package MIRIAD version 1.5 (Sault et al.1995). In brief, we checked individual exposures and flagged bad baselines, antennas, and/ortimes. For the 8.8 and 17 mm data we used the option opcor in atlod and then tsyscal=any in atfix to correct the fluxes for atmospheric opacity. We then used the task mfcal onthe bandpass calibrator to determine bandpass corrections taking antenna 2 as referenceantenna. The bandpass solution was transferred to the flux and gain/phase calibratorsusing gpcopy . The fluxes of the bandpass and gain/phase calibrators were then scaled to theabsolute flux units using gpboot and mfboot . Finally, we copied the phase calibrations to thescience targets with gpcopy . For the 17 mm data we preferred to use 1934-638 as bandpasscalibrator because 1921-293 had large phase amplitudes (between +10 ◦ and -10 ◦ ) even after invert , clean , and restore in MIRIAD).Cleaned ATCA maps using uniform weighting are shown in Figs. 1 and 2. The resulting fitsimages were loaded in the Common Astronomy Software Applications (CASA) package tomeasure flux densities or 3 σ upper limits. For this last step, we first computed the rms ina polygonal area around the expected location of the target, which is given by the 2MASScoordinates (Table 1). If the source is not detected we report an upper limit equal to 3times the computed rms. If a source is detected we provide the flux density within the3 rms closed contour except in a few instances where our procedure is clearly missing somesignificant flux , hence we report the 2 rms closed contour flux density. All flux densitiesand upper limits are listed in Table 3. The flux density within the 2 rms contour is larger than that within the 3 rms plus the10% absolute flux calibration uncertainty 8 –
3. Immediate Results
We detect all ATCA sources at 0.9 cm, all but CS Cha at 1.7 cm, all but SZ Cha at3.3 cm, and only V892 Tau and CS Cha at 5.5 cm (Figs. 1 and 2). Several other radiosources are detected in the larger areas covered at 3.3 and 5.5 cm. The most relevant onefor the interpretation of our data is Hubble 4, a young star in Taurus with no circumstellardisk but strong cm emission (see discussion in Appendix A).To determine if sources are extended we fit their emission with a 2D gaussian (task gaussfit in CASA) and compare the size of the gaussian with the restored beam FWHMreported in Table 3. With this approach we find that SZ Cha and CS Cha are marginallyextended at 0.9 cm (gaussian widths of 12 ′′ × ′′ and 10 ′′ × ′′ respectively), in agreementwith Ubach et al. (2012) who report extended emission for these two sources at a slightlyshorter (0.7 cm) wavelength. SZ Cha is also spatially resolved at 1.7 cm (gaussian width of29 ′′ × ′′ ) while CS Cha is clearly extended at 3.3 cm (gaussian width of 35 ′′ ×
26, positionangle ∼ ◦ ). V892 Tau is only marginally resolved at 3.3 cm but its emission is clearlyextended at 5.5 cm (see Fig. 1).Disk emission should be confined within the disk size, rarely larger than 1,000 AU (e.g.Vicente & Alves 2005). Hence, the extended emission at 3.3 and 5.5 cm for V892 Tau andCS Cha, which corresponds to sizes & α emission line as amicro-jet of ∼
10 mas in size and NE-SW direction, almost perpendicular to the elongation 9 –we see at 3.3 cm. However, another possibility is that the H α displacement indicates thedirection of the recently discovered stellar companion to CS Cha (Guenther et al. 2007 andTable 1). Typical spectroastrometric displacement in the H α line for known binaries haveangular scales of ∼ + , and S + (e.g. Hartigan et al. 1995), jets/outflows remain the most plausible explanationfor the extended cm emission seen in the ATCA images of these two sources. Based onthese images the orientation of the jet is almost N-S for V892 Tau and NW-SE for CS Cha.When the jet emission is not spatially resolved we report in Table 3 the flux density withinthe 3 rms contour near the source and note that the emission is likely associated with ajet/outflow.The presence of a jet is more difficult to ascertain for SZ Cha. The marginal extensionsat 7 mm from Ubach et al. (2012) and at 0.9 cm from our ATCA image indicate linearscales of ∼ iii ] emission line at 15.55 µ m resultingin an unusually high [Ne iii ]/[Ne ii ] line flux ratio of ∼ EUV ∝ ν − , Hollenbach & Gorti 2009)impinging on the disk, which Espaillat et al. attribute to the central star. However, shockscan locally heat the ISM gas to very high temperatures ( > iii ] emission toward SZ Cha is not produced in the disk surface but in a jet(see also the recent report of [Ne iii ] emission from the Sz 102 microjet, Liu et al. 2014). 10 –Additional observations are clearly required to establish if SZ Cha is powering a jet, hencewe will continue classifying this source as jet-free in our study. In this paper we call excess centimeter emission any long wavelength emission ontop of the thermal dust disk emission. The first step in identifying excess cm emission isto assemble the source spectral energy distribution (SED) and subtract off the thermalcontribution from dust grains.In assembling the SEDs of our ATCA sources we gather additional millimetricand centimetric fluxes from the literature. For V892 Tau we find millimeter fluxesin Andrews & Williams (2005) and Ricci et al. (2012). CS Cha and SZ Cha havebeen observed at 870 µ m with APEX/LABOCA (Belloche et al. 2011) and additionalmillimeter/centimeter fluxes are reported in Ubach et al. (2012). MP Mus has a 1.2 mmdetection from Carpenter et al. (2005) as well as 3 mm and centimeter data in Cortes et al.(2009). SR 21 has millimeter fluxes from various compilations (Andre & Montmerle1994, Ricci et al. 2010a, Andrews et al. 2011, Ricci et al. 2012, Ubach et al. 2012).Finally, V4046 Sgr has been observed by Jensen et al. (1996), Rodriguez et al. (2010),and Oberg et al. (2011) at millimeter wavelengths. A summary of literature fluxes andreferences is provided in Table 5.In addition to our six ATCA targets, we include in our study other young stellarobjects from the literature that have: i) a good SED coverage at millimeter and centimeterwavelengths; and ii) stellar X-ray luminosities, since X-rays contribute to ionize the disksurface. These criteria result in eight additional sources: DG Tau, DK Cha, GM Aur,HL Tau, RY Tau, TCha, TW Hya, UZ TauE. Source properties that are relevant to our 11 –study are summarized in Table 4 while literature fluxes and references to assemble theirSEDs are provided in Table 5. All sources except DK Cha have also infrared spectroscopycovering the [Ne ii ] line at 12.81 µ m (see Sect. 4.2).Figs. 3 and 4 show the long wavelength portion of the SED of our ATCA targets andliterature sources. Because thermal dust emission is mostly optically thin at millimeterwavelengths, the flux density can be written as F ν ∝ ν α mm . In fitting this relation we seta minimum uncertainty of 10% for each flux density. The fit to the millimeter fluxes ,ignoring the cm excess emphasized here, results in slopes ranging from 2.4 to a maximumvalue of 3.5 for SR 21, with uncertainties between ∼ α mm issimilar to that found in nearby star-forming regions (e.g. Ricci et al. 2010b; Ubach et al.2012). The important result is that for none of the sources studied here can thermal dustemission account for the measured centimeter fluxes.Several physical mechanisms are known to produce radio emission. In Sect. 3 wediscussed two ATCA sources with clearly extended cm emission and concluded that mostof the emission is likely coming from shocked gas in jets. This gas can be ionized andproduce cm free-free emission (e.g. Anglada et al. 1998). Similarly, a fully or partiallyionized disk surface emits free-free cm radiation (e.g. Pascucci et al. 2012). Non-thermal(gyrosynchrotron) cm emission originating in magnetic fields has been also detected in latetype dwarfs and several classes of active stars (e.g. G¨udel 2002). Finally, a populationof very large (cm-size) grains can produce extra cm thermal emission (e.g. Wilner et al.2005) while very small (nm-size) spinning grains produce electric dipole emission in themicrowave range (Rafikov et al. 2006). These different mechanisms can, and likely do,operate concurrently in young accreting stars surrounded by disks. Multi-epoch andmulti-wavelength radio observations can be used to assess the dominant physical mechanism The longest wavelength we include in our fit is 10 mm. 12 –(see Sect. 3.2) but the contribution of each process cannot be quantified with certainty.Therefore, in Sect. 4 we will assume that all excess cm emission is due to the ionized disksurface and thus compute upper limits on the stellar EUV luminosity reaching the disk. Wewill show that even upper limits place interesting constraints on disk dispersal theories.
Thermal bremsstrahlung (free-free) radiation and non-thermal gyrosynchroton emissionare characterized by power-law spectra of the form F ν ∝ ν α cm . Cm- and nm-dust emissionhave a more bell-like spectral shape, which however cannot be easily recognized withtwo-three data points typically available at cm wavelengths. Therefore, we follow commonpractice and investigate the origin of the cm emission assuming that it has a power-lawspectrum and thus compute the radio spectral index α cm .It is worth pointing out that a large range of spectral indices is expected both forfree-free and gyrosynchroton emission. In the case of free-free emission α cm can be as low as-0.1 for optically thin emission (e.g. for an ionized disk surface, Pascucci et al. 2012) and upto +2 for optically thick emission. Values of ∼ α cm in each sub-class but average values that decrease with evolutionarystage, from ∼ . α cm > +0 . − . ≥ α cm ≤ +0 . α cm < − . α cm include the uncertaintyon the slope of the dust thermal emission.Seven of our sources have flat α cm , consistent with free-free emission from opticallythin to moderately thick plasma, thus including jet emission. V892 Tau and CS Cha(IDs 1 and 3), whose ATCA cm emission is extended and most likely dominated by jetemission (Sect. 3), fall in this category. Only three sources (SR 21-ID 5, RY Tau-ID 9,and T Cha-ID 10) have flat α cm and no evidence of jets. Of them, only SR 21 and T Chahave spectral slopes consistent with optically thin free-free disk emission while emissionfrom RY Tau may be moderately thick as expected in the X-ray irradiated disk model ofOwen et al. (2013). Two sources, MP Mus and DG Tau have negative spectral indices.Dzib et al. (2013) also find Class II sources in Ophiucus with negative spectral indices,indicating that substantial contribution from gyrosynchroton emission can be present insome disk sources. Multi-epoch cm observations would be helpful to further assess thenon-thermal origin of the cm emission in these sources.Finally, we wish to discuss the case of TW Hya, which sports one of the largest positive 14 – α cm in our sample and has no evidence of a jet (ID 7, filled circle in Fig. 5). This α cm is computed from the excess emission at 3 wavelengths: 3.5, 4.1, and 6.3 cm. The lattertwo points are new VLA band-integrated flux densities from Menu et al. (2014). Theseauthors note that the spectral slope in the 1-GHz-band centered at 4.1 cm is different fromthat at 6.3 cm, it becomes flatter at the longer wavelength. They interpret this result asan indication of dust still contributing to the 4.1 cm emission. We have thus re-computed α cm neglecting the 3.5 cm flux density, where dust would contribute even more to theemission. Indeed, we find that the new α cm (empty circle in Fig. 5) is substantially reduced,confirming the interpretation of Menu et al. (2014). Thus, even this new α cm should beconsidered an upper limit and it is likely that TW Hya has also a flat cm spectral index.Sensitive observations at wavelengths longer than 6 cm are needed to test this hypothesis.Although in some cases α cm is not consistent with free-free disk emission, we willnevertheless assume that all the excess cm emission is from free-free so that we can deriveupper limits to the EUV luminosity impinging on the disk (Sect. 4.1).
4. Ionizing Radiation Reaching the Disk
A necessary condition to estimate the stellar ionizing luminosity reaching the diskatmosphere and ionizing the disk surface is that the associated free-free emission is opticallythin. In Pascucci et al. (2012) we showed that even a partially ionized wind at 5,000 Kbecomes optically thick at wavelengths longer than 20 cm for plausible wind values. We canalso compute the EUV photon luminosity above which the free-free emission would becomeoptically thick using the expression for the continuum optical depth in Bell & Seaquist(1978) and relating it to the emission measure as in eq. 5 from Hollenbach & Gorti (2009). these flux densities are not published 15 –In doing this calculation we assume a fully ionized region at 10,000 K, as appropriate forthe EUV case, with an emitting radius equal to the gravitational radius , since most of theemission measure comes from regions close to this radius. As done in previous papers (e.g.Hollenbach & Gorti 2009) we also assume that the fraction of the stellar photons interceptedby the disk is 0.7 and restrict ourselves to solar-mass stars . With this approach we findthat the 3.3 and 5.5 cm free-free emission becomes optically thick for EUV luminosities ≥ × s − and ≥ × s − respectively. We will see shortly that the upper limitswe derive from the excess cm emission are lower than these values, meaning that free-freedisk emission is optically thin at these wavelengths and we can use it to constrain the EUVluminosity impinging on the disk. We also note that neglecting the absorption of EUVphotons by dust grains is justified. In fact, ISM dust provides an optical depth to Lymancontinuum photons greater than one in the ionized surface zone only for ≥ s − (eq. 6.3in Hollenbach et al. 1994), well above the luminosities that we will derive here. In addition,because of dust growth and settling in disks (e.g. Testi et al. 2014), the opacity of the dustat the disk surface will be reduced with respect to the ISM value, further increasing theluminosity above which dust significantly absorbs EUV photons. Since several physical processes can produce radio emission (see Sects. 3 and 3.2), byassuming that all excess cm emission is due to free-free disk emission we will obtain upper The gravitational radius is where the hydrogen thermal speed is equal to the escapespeed from the star gravitational field, e.g. Hollenbach et al. (1994) The value 0.7 corresponds to a disk vertical extent z max = r where r is the midplaneradial distance from the star. This is the vertical extent appropriate for EUV ionization. 16 –limits to the ionizing radiation reaching the disk. Both EUV and X-rays can photoionizeH atoms. In Pascucci et al. (2012) we used a gas temperature of 5,000 K to estimate thefree-free contribution from soft X-rays for the nearby disk of TW Hya, a source with anexceptionally soft X-ray spectrum (e.g. Kastner et al. 2002). While hard X-rays heat the gasat lower temperatures thereby producing less free-free cm emission (Pascucci et al. 2012),the relative contribution of soft- vs hard-X-rays depends on the source X-ray spectrum,which is not always well characterized. Hence, we prefer to provide here conservativeupper limits on the EUV radiation by not subtracting off the X-ray free-free contribution.Hereafter, we will use the notation Φ
EUV , cm to refer to these upper limits derived from theexcess cm emission.To estimate Φ EUV , cm we proceed as follows. First, we fit the millimeter fluxes tocalculate the contribution from dust thermal emission at cm wavelengths as discussedin Sect. 3. Next, for each cm wavelength where we detect excess emission we estimateΦ EUV , cm using a generalization of eq. 2 in Pascucci et al. (2012) from the measured cmfluxes minus the dust thermal emission: the excess cm emission is directly proportionalto the EUV luminosity reaching the disk (Pascucci et al. 2012; Owen et al. 2013). Themain assumptions here are that the characteristic temperature of the gas is 10,000 K andthe fraction of photons absorbed by the disk is 0.7. This factor accounts only for the diskgeometry, a further reduction in the ionizing radiation can occur because of absorption inthe circumstellar matter, e.g. accretion columns and/or magnetically driven winds.The most stringent upper limits on the EUV luminosity reaching the disk aresummarized in Table 6 together with the wavelength providing such limits. The mostsensitive wavelengths to place such upper limits are typically around 3 and 6 cm, see lastcolumn of Table 6. Note that the new 6.3 cm flux density for TW Hya reduces the upperlimit on Φ EUV , cm by a factor of ∼ EUV , cm for GM Aur is consistent with thatreported by Owen et al. (2013). The main uncertainty in these upper limits comes from theuncertainty associated with the slope of the dust thermal emission ( α mm ). Thus, we havealso computed Φ EUV , cm assuming a steeper dust SED with α mm minus the 1 σ uncertaintyon the dust spectral slope. We find that upper limits derived at the longest wavelength( ∼ EUV , cm can be up to a factor of 2. Thus, in discussing ourresults we will assume that the Φ EUV , cm derived with our approach can be at most off bya factor of 2. In other words, the upper limits could be at most a factor of 2 higher thanthose provided in Table 6.We find a broad range of Φ EUV , cm from ∼ × to 10 s − with no obvious correlationwith the stellar X-ray luminosity (Fig. 6). Sources with known jets/outflows (red symbolsin Fig. 6) have higher Φ EUV , cm than those without, on average by an order of magnitude.In these sources most of the cm emission is likely arising from shocked ISM material anddoes not trace the ionized disk surface or the ionizing luminosity from the star. This isconfirmed in a few instances, such as in CS Cha, where the cm emission is found to bespatially extended and does not peak at the stellar location (see discussion in Sect. 3). Ifwe exclude the sources with jets and average the other upper limits we find that Φ EUV , cm isat most 2 × s − .How do these upper limits compare with the ionizing radiation emitted by the star?Ribas et al. (2005) used a small sample of solar analogs with ages between ∼ . ∼ × s − for solar analogs that are 100 Myr old. This 18 –luminosity is close to the upper limits we estimate for the three ∼ L FUV /L star of ∼ − in non-accreting stars. The behavior of the EUV emission in the 1-10 Myr age rangewill depend on whether it mostly traces accretion (as FUV, in which case it should decreasewith time) or the chromosphere (as X-ray, in which case it could be flat).In the pre-main sequence regime, Alexander et al. (2005) estimated order-of-magnitudeEUV luminosities (between 700-912 ˚A) by modeling literature emission measures fromfive sun-like stars that are a few Myr old. They find a broad range of luminosities(10 − s − ) which we show as a dashed region in our Fig. 6. For the only source wehave in common, RY Tau, our EUV upper limit reaching the disk is only a factor of ∼ EUV , cm lie on the lowest side of the stellar ionizingluminosities they infer. In addition, the Φ EUV , cm for the ∼ s − . Herczeg (2007) estimatedan ionizing photon luminosity of ∼ × s − from the accretion shock on TW Hya butnoted that only ∼ phot s − could reach the disk if the emission is buried under the ∼ × cm − column of neutral hydrogen gas inferred by X-ray and FUV data. Ourupper limit on the Φ EUV , cm of 1.5 × s − suggests that part of the stellar EUV luminosityis indeed absorbed by circumstellar matter before reaching the disk even in this relativelyold system. More estimates of the ionizing radiation emitted by pre-main-sequence starsare necessary to evaluate the typical extent of circumstellar extinction. 19 – All but one (DK Cha) of the 14 sources studied here have fluxes or upper limits in the[Ne ii ] line at 12.8 µ m. This transition is relevant to our study because ionized Neon isknown to also probe the disk atmosphere of some young stars (e.g. Sacco et al. 2012). Inaddition, this line is found to be slightly ( ∼
10 km/s) blueshifted in several disks pointing tounbound gas in a photoevaporative wind (e.g. Pascucci & Sterzik 2009). Ne + could eithertrace the uppermost layer of the disk surface fully ionized by EUV photons or rather alower mostly neutral layer where Ne is ionized by 1 keV X-rays (Glassgold et al. 2007). Thesecond configuration would imply larger mass loss rates ( > − M ⊙ /yr) than the first, e.g.the example of TW Hya in Gorti et al. (2011) and Pascucci et al. (2011). We are now inthe position to answer the question: is the Φ EUV , cm estimated in Sect. 4.1 large enough, i.e.are there enough ionizing photons reaching the disk atmosphere, to reproduce the observed[Ne ii ] luminosities?We first assemble [Ne ii ] fluxes from the literature, giving higher priority tofluxes obtained from spectrally resolved line profiles using ground-based facilities(Pascucci & Sterzik 2009; Pascucci et al. 2011; Baldovin-Saavedra et al. 2012; Sacco et al.2012). When ground-based observations are not available we take the Spitzer/IRS fluxes(G¨udel et al. 2010; Baldovin-Saavedra et al. 2011; Espaillat et al. 2013, see also Tables 1and 4). If Neon atoms are ionized by EUV photons, the [Ne ii ] luminosity is directlyproportional to the EUV luminosity reaching the disk, hence we can convert the measured[Ne ii ] fluxes into a Φ EUV , NeII . This quantity represents the EUV luminosity necessary toreproduce the observed [Ne ii ] luminosities and can be compared to the upper limits on theEUV luminosity estimated from the excess cm emission (Φ EUV , cm ). To compute Φ EUV , NeII we use eq. 19 in Hollenbach & Gorti (2009) and take the fraction of neon in the singlyionized state equal to unity. 20 –Fig. 7 shows Φ
EUV , cm as a function of Φ EUV , NeII . The uncertainty in Φ
EUV , NeII is drivenby the flux calibration uncertainty at mid-infrared wavelengths, ∼
20% for ground-basedobservations. For three sources (CS Cha-ID 3, V4046 Sgr-ID 6 and TW Hya-ID 7) Φ
EUV , cm is clearly not sufficient to reproduce the observed [Ne ii ] luminosities, even when accountingfor a factor of 2 uncertainty on the estimated EUV upper limits. Hence, at least in thesesources stellar X-rays must contribute to the ionization of Ne atoms. Note that for thesethree sources there are high-resolution spectra demonstrating that the [Ne ii ] line traces aslow ( ∼ −
10 km/s) photoevaporative wind (Pascucci & Sterzik 2009; Pascucci et al. 2011;Sacco et al. 2012). This, in combination with the low Φ
EUV , cm upper limits, implies that the[Ne ii ] is not tracing a fully ionized disk layer but rather a lower region only partially ionizedby X-ray photons. The presence of a partially ionized and unbound disk layer implies largermass loss rates than those that can be achieved via EUV-driven photoevaporation alone.Rigliaco et al. (2013) recently re-analyzed the low-velocity component of the [O i ]optical forbidden lines from young sun-like stars. Based on the line fluxes, profiles, and peakcentroids they argued for the presence of a slow partially molecular photoevaporative flow(driven by X-ray and/or FUV photons) where oxygen is produced by FUV dissociation ofOH molecules. Their result also implies higher photoevaporation rates than those producedby EUV-ionization only. The fact that these two independent approaches lead to the sameconclusion shows that X-ray and FUV irradiation of the disk surface must be taken intoaccount to estimate realistic photoevaporative mass loss rates.
5. Conclusions and Implications
This contribution explores the use of cm data to constrain the high-energy radiationreaching the surface of protoplanetary disks and photoionizing H. Because free-free emissionfrom a partially or fully ionized disk layer is optically thin, the free-free flux density is 21 –directly proportional to the photon EUV luminosity, Φ
EUV . By identifying the cm emissionin excess to the thermal dust emission and attributing that to free-free disk emission weobtain upper limits to the Φ
EUV impinging on the disk of 14 young ( ∼ −
10 Myr) stars.Our approach results in two main findings:1. The average Φ
EUV upper limit reaching the disk is 2 × s − in sources without jetsand several times lower than 10 s − for the older systems TW Hya, MP Mus, andV4046 Sgr.2. The inferred Φ EUV upper limits are not sufficient to reproduce the [Ne ii ] luminositiesfrom three disks, hence stellar X-rays must contribute to the ionization of Ne atomsin these systems.These two results have interesting implications for our understanding of disk evolutionand dispersal. The first result shows that the EUV photon luminosity received by the disk ison the low side of the range of stellar EUV luminosities inferred for pre-main sequence stars.Such low Φ EUV luminosities do not appear to be sufficient alone to disperse protoplanetarydisks in the timescale that is required by observations. We also note that accounting forgyro-synchrotron and other sources of cm emission would further reduce our estimates (seeSect. 3.1 and Appendix A.1), making it more difficult for EUV photoevaporation alone toclear out protoplanetary material.The second result demonstrates that, at least in three systems, the [Ne ii ] emissionat 12.81 µ m primarily traces a mostly neutral disk region where Ne atoms are ionizedby 1 keV X-ray photons. This, in combination with blueshifts in the peak emissionpointing to a wind, indicates that disk gas is photoevaporated deeper in the disk atrates larger than those predicted by EUV irradiation alone. How much higher dependson the relative contribution and evolution of stellar FUV and X-rays in driving and 22 –maintaining photoevaporative winds (Gorti et al. 2009; Owen et al. 2011). In line with the[O i ] 6300 ˚A observations (Rigliaco et al. 2013), our findings demonstrate that star-drivenphotoevaporation contributes to disk dispersal at higher disk masses than previouslythought.In the next years the synergy between ALMA and sensitive cm interferometers suchas the EVLA will enable extending these studies to larger samples of disks in nearbystar-forming regions. These data will show which are the typical upper limits on the EUVluminosity reaching the disk. Such values, in combination with measurements of the massaccretion rate (as suggested in Owen et al. 2013) or direct tracers of the disk ionized surface(as the [Ne ii ] line discussed here) will enable to firmly establish if EUV photons play aminor role in the dispersal of protoplanetary material.The authors thank the anonymous referee for a prompt and useful report. I.P. thanksCathie Clarke for stimulating discussion. I.P., U.G., and D.H. acknowledge support froman NSF Astronomy & Astrophysics Research Grant (ID: 1312962). Facilities:
ATCA
A. Selected Sources Identified in the ATCA Fields
The 3.3 and 5.5 cm ATCA images cover a field of over 30 arcminutes, hence severalother radio sources are detected in these fields. It is beyond the scope of this paperto characterize all these radio sources but we mention those that are of interest for theinterpretation of the excess cm emission and demonstrate how our data reduction recoversknown radio sources. The most relevant of these sources is Hubble 4, a young star with nodisk but strong cm emission. We dedicate a subsection to this source. 23 –The 3.3 and 5.5 cm fields of SZ Cha and CS Cha cover a well known BL Lac-typeobject (PMN J1057-7724, e.g. Veron-Cetty & Veron 2010). This source dominates the cmemission in its surrounding and even enhances the background at 5.5 cm at the locationof SZ Cha (see Table 3). In the 3.3 cm and 5.5 cm image of SR 21 the source located at-100 ′′ RA is a known X-ray source in the ρ Ophiuchus cloud core (GDS J162702.1-241928,Gagne et al. 2004) while the two 3.3 cm sources at ∼ -50 ′′ DEC are close to the young stellarobject candidate BKLT J162707-242009 (Barsony et al. 2001). Finally, the 3.3 and 5.5 cmemission at about (+100 ′′ ,+25 ′′ ) RA, DEC of V4046 Sgr is not associated with a knownsource. The closest object to this radio emission is a nearby M1-type emission-line star(2MASS J18142207-3246100) with a large proper motion (7 ×
40 mas/yr, Zacharias et al.2003).
A.1. Hubble 4
Our ATCA 1.7, 3.3 and 5.5 cm exposures of V892 Tau also cover the Taurus memberHubble 4 (V1023 Tau, 2MASS J04184703+2820073). Hubble 4 is a single K7 star classifiedas a weak-line T Tauri star based on its low H α EW (-3 ˚A which gives an upper limit onthe mass accretion rate of < × − M ⊙ /yr, White & Ghez 2001). Furlan et al. (2006)compiled its SED and found no evidence of excess emission out to ∼ µ m. The system isclassified as Class III (no disk) even when extending the SED at 24 µ m with Spitzer /MIPS(Luhman et al. 2010; Rebull et al. 2010) and millimeter observations place an upper limiton the dust disk mass of only 4 × − M ⊙ (Andrews & Williams 2005). Hubble 4 is a strongX-ray emitter (4-6.5 × erg/s) with modest absorption (3.1 × cm − , G¨udel et al.2007). The X-ray luminosity is about 8 times larger than the average X-ray luminosityof Taurus sources and, although the X-ray spectrum peaks at ∼ gaussfit task in CASA and find that the 1.7, 24 –3.3 and 5.5 cm emission is not clearly spatially extended beyond the beam (FWHMs of25 ′′ × ′′ , 36 ′′ × ′′ and 63 ′′ × ′′ respectively). We measure flux densities of 0.19 ± ± ± α as F ν ∝ ν α , we find α =-0.4 using the 3.3 and 5.5 cmdata and α =-1.7 when including the 1.7 cm datapoint. Such negative spectral indexesexclude thermal free-free emission from a wind or jet and rather point to non-thermalgyro-synchrotron radiation (see discussion in Sect. 3.2). Recently, Dzib et al. (2013) haveshown that the Gudel-Benz relation between radio and X-ray emission of old magneticallyactive stars holds even for young stellar objects but with a slightly less steep power lawof the form L X /L radio ∼ ± . The example of Hubble 4 demonstrates that strong cmemission of the order of several hundred µ Jy can be produced by magnetic activity in youngstars that lack disks and jets. We note that this emission could account for most of theexcess cm emission we measure toward our targets if we scale the flux density of Hubble 4at 3 cm by source distance and X-ray luminosity. However, in sources with multiple cmdetections the spectral index is mostly positive (see Sect. 3.2) suggesting that in stars withdisks the level of magnetic activity is perhaps lower than in stars without. 25 –
REFERENCES
Alexander, R. D., Clarke, C. J., Pringle, J. E. 2004, MNRAS, 348, 879Alexander, R. D., Clarke, C. J., Pringle, J. E. 2005, MNRAS, 358, 283Alexander, R. D., Clarke, C. J., Pringle, J. E. 2006, MNRAS, 369, 229Alexander, R. D. & Armitage, P. J. 2009, ApJ, 704, 989Alexander, R., Pascucci, I., Andrews, S., Armitage, P., Cieza, L. 2013, as-troph(arXiv1311.1819A)Andre, P. & Montmerle, T. 1994, ApJ, 420, 837Andrews, S. M. & Williams, J. P. 2005, ApJ, 631, 1134Andrews, S. M., Wilner, D. J., Espaillat, C., Hughes, A. M., Dullemond, C. P., McClure,M. K., Qi, C., Brown, J. M. 2011, ApJ, 732, 42Andrews, S. M., Rosenfeld, K. A., Kraus, A. L., Wilner, D. J. 2013, ApJ, 771, 129Anglada, G., Villuendas, E., Estalella, R., Beltr´an, M. T., Rodr´ıguez, L. F., Torrelles, J.M., Curiel, S. 1998, AJ, 116, 2953Baldovin-Saavedra, C., Audard, M., G¨udel, M., Rebull, L. M., Padgett, D. L., Skinner, S.L, Carmona, A., Glauser, A. M., Fajardo-Acosta, S. B. 2011, A&A, 528A, 22Baldovin-Saavedra, C., Audard, M., Carmona, A., Guedel, M., Briggs, K., Rebull, L. M.,Skinner, S. L., Ercolano, B. 2012, A&A, 543A, 30Barsony, M., Kenyon, S. J., Lada, E. A., Teuben, P. J. 2001, A&A, 372, 173Bell, M. B.& Seaquist, E. R. 1978, ApJ, 223, 378 26 –Belloche, A., Schuller, F., Parise, B., Andr´e, Ph., Hatchell, J., Jorgensen, J. K., Bontemps,S., Weiss, A., Menten, K. M., Muders, D. 2011, A&A, 527A, 145Brickhouse, N. S., Cranmer, S. R., Dupree, A. K., Luna, G. J. M., Wolk, S. 2010, ApJ, 710,1835Briceno, C., Luhman, K. L., Hartmann, L., Stauffer, J. R., Kirkpatrick, J. D. 2002 ApJ,580, 317Calvet, N., Hartmann, L., Strom, S. E. 2000, in Protostars and Planets IV (Book - Tucson:University of Arizona Press; eds Mannings, V., Boss, A.P., Russell, S. S.), p. 377Carpenter, J. M., Wolf, S., Schreyer, K., Launhardt, R., Henning, Th. 2005, AJ, 129, 1049Cortes, S. R., Meyer, M. R., Carpenter, J. M., Pascucci, I., Schneider, G., Wong, T., Hines,D. C. 2009, ApJ, 697, 1305de Geus, E. J., de Zeeuw, P. T., Lub, J. 1989, A&A, 216, 44Dzib, S. A., Loinard, L., Mioduszewski, A. J. et al. 2013, ApJ, 775, 63Ercolano, B., Clarke, C. J., Drake, Jeremy J. 2009, ApJ, 699, 1639Espaillat, C., Ingleby, L., Furlan, E., McClure, M., Spatzier, A., Nieusma, J., Calvet, N.,Bergin, E., Hartmann, L., Miller, J. M., Muzerolle, J. 2013, ApJ, 762, 62Frank, A., Ray, T. P., Cabrit, S. et al. 2014, in Protostars and Planets VI (arXiv:1402.3553),in press.Furlan, E., Hartmann, L., Calvet, N. et al. 2006, ApJS, 165, 568Furlan, E., Luhman, K. L., Espaillat, C. et al. 2011, ApJS, 195, 3Gagne, M., Skinner, S. L., Daniel, K. J. 2004, ApJ, 613, 393 27 –Getman, K. V., Feigelson, E. D., Grosso, N., McCaughrean, M. J., Micela, G., Broos, P.,Garmire, G., Townsley, L. 2005, ApJS, 160, 353Ghez, A. M., McCarthy, D. W., Patience, J. L., Beck, T. L. 1997, ApJ, 481, 378Glassgold, A. E., Najita, J. R., Igea, J. 2007, ApJ, 656, 515Gorti, U. & Hollenbach, D. 2009, ApJ, 690, 1539Gorti, U., Dullemond, C. P., Hollenbach, D. 2009, ApJ, 705, 1237Gorti, U., Hollenbach, D., Najita, J., Pascucci, I. 2011, ApJ, 735, 90Grosso, N., Montmerle, T., Bontemps, S., Andre, P., Feigelson, E. D. 2000, A&A, 359, 113G¨udel, M. 2002, ARA&A, 40, 217, 61G¨udel, M., Briggs, K. R., Arzner, K. et al. 2007, A&A, 468, 353G¨udel, M. & Naze, Y. 2009, The Astronomy and Astrophysics Review, Volume 17, Issue 3,pp.309-408G¨udel, M., Lahuis, F., Briggs, K. R., Carr, J., Glassgold, A. E., Henning, Th., Najita, J.R., van Boekel, R., van Dishoeck, E. F. 2010, A&A, 519A, 113Guenther, E. W., Esposito, M., Mundt, R., Covino, E., Alcala, J. M., Cusano, F., Stecklum,B. 2007, A&A, 467, 1147Hamaguchi, K., Yamauchi, S., Koyama, K. 2005, ApJ, 618, 360Hartigan, P., Edwards, S., Ghandour, L. 1995, ApJ, 452, 736Herczeg, G. 2007, Proceedings of the International Astronomical Union, Volume 243,147-154 28 –Herczeg, G. J., Najita, J. R., Hillenbrand, L. A., Pascucci, I. 2007, ApJ, 670, 509Hollenbach, D., Johnstone, D., Lizano, S., Shu, F. 1994, ApJ, 428, 654Hollenbach, D. & Gorti, U. 2009, ApJ, 703, 1203Ingleby, L., Calvet, N., Hernandez, J., Briceno, C., Espaillat, C., Miller, J., Bergin, E.,Hartmann, L. 2011, AJ, 141, 127Jensen, E. L. N., Mathieu, R. D., Fuller, G. A. 1996, ApJ, 458, 312Kastner, J. H., Huenemoerder, D. P., Schulz, N. S., Canizares, C. R., Weintraub, D. A.2002, ApJ, 567, 434Kenyon, S. J., G´omez, M., Whitney, B. A. 2008, Handbook of Star Forming Regions,Volume I: The Northern Sky ASP Monograph Publications, Vol. 4. Edited by BoReipurth, p. 405Kim, K. H., Watson, Dan M., Manoj, P. et al. 2009, ApJ, 700, 1017Linsky, J. L., Fontenla, J., France, K. 2014, ApJ, 780, 61Liu, C.-F., Shang, H., Walter, F. M., Herczeg, G. J. 2014, ApJ, 786, 99Luhman, K. 2007, ApJS, 173, 104Luhman, K. 2008, Handbook of Star Forming Regions, Volume II: The Southern Sky ASPMonograph Publications, Vol. 4. Edited by Bo Reipurth, p. 169Luhman, K. L., Allen, P. R., Espaillat, C., Hartmann, L., Calvet, N. 2010, ApJS, 186, 111Mamajek E. E., Meyer, M. R., Liebert, J. 2002, AJ, 124.1670MMenu, J., van Boekel, R., Henning, Th. et al. 2014, A&A, 564A, 93 29 –Najita, J. R., Doppmann, G. W., Bitner, M. A. et al. 2009, ApJ, 697, 957Oberg K. I., Qi, C., Fogel, J. K. J., Bergin, E. A., Andrews, S. M., Espaillat, C., Wilner, D.J., Pascucci, I., Kastner, J. H. 2011, ApJ, 734, 98Owen, J. E., Ercolano, B., Clarke, C. J., Alexander, R. D. 2010, MNRAS, 401, 1415Owen, James E., Ercolano, Barbara, Clarke, Cathie J. 2011, MNRAS, 412, 13Owen, J. E., Scaife, A. M. M., Ercolano, B. 2013, MNRAS, 434, 3378Pascucci, I. & Sterzik, M. 2009, ApJ, 702, 724Pascucci, I., Sterzik, M., Alexander, R. D., Alencar, S. H. P., Gorti, U., Hollenbach, D.,Owen, J., Ercolano, B., Edwards, S. 2011, ApJ, 736, 13Pascucci, I., Gorti, U., Hollenbach, D. 2012, ApJ, 751L, 42Prato, L., Simon, M., Mazeh, T., Zucker, S., McLean, I. S. 2002, ApJ, 579L, 99Prato, L., Greene, T. P., Simon, M. 2003, ApJ, 584, 853Preibisch, T., Kim, Y.-C., Favata, F. et al. 2005, ApJS, 160, 401Rafikov, R. R. 2006, ApJ, 646, 288Rebull, L. M., Padgett, D. L., McCabe, C.-E. et al. 2010, ApJS, 186, 259Ribas, I., Guinan, E. F., G¨udel, M., Audard, M. 2005, ApJ, 622, 680Ricci, L., Testi, L., Natta, A., Brooks, K. J. 2010a, A&A, 521A, 66Ricci, L., Testi, L., Natta, A., Neri, R., Cabrit, S., Herczeg, G. J. 2010b, A&A, 512A, 15Ricci, L., Trotta, F., Testi, L., Natta, A., Isella, A., Wilner, D. J. 2012, A&A, 540A, 6 30 –Rigliaco, E., Pascucci, I., Gorti, U., Edwards, S., Hollenbach, D. 2013, ApJ, 772, 60Rodmann, J., Henning, Th., Chandler, C. J., Mundy, L. G., Wilner, D. J. 2006, A&A, 446,211Rodriguez, D. R., Kastner, J. H., Wilner, D., Qi, C. 2010, ApJ, 720, 1684Sacco, G. G., Flaccomio, E., Pascucci, I., Lahuis, F., Ercolano, B., Kastner, J. H., Micela,G., Stelzer, B., Sterzik, M. 2012, ApJ, 747, 142Sacco, G. G., Kastner, J. H., Forveille, T., Principe, D., Montez, R., Zuckerman, B.,Hily-Blant, P. 2014, A&A, 561A, 42Sault, R. J., Teuben, P. J., Wright, M. C. H. 1995, ASPC, 77, 433Skrutskie R.M., Cutri, R., Stiening, M.D. et al. 2006, AJ, 131, 1163Stempels, H. C. & Gahm, G. F. 2004, A&A, 421, 1159Takami, M., Bailey, J., Chrysostomou, A. 2003, A&A, 397, 675Testi, L., Birnstiel, T., Ricci, L. et al. 2014, PPVI, in press. (arXiv:1402.1354)Torres, C. A. O., Quast, G. R., Melo, C. H. F., Sterzik, M. F. 2008, Handbook of StarForming Regions, Volume II: The Southern Sky ASP Monograph Publications, Vol.4. Edited by Bo Reipurth, p. 757Ubach, C., Maddison, S. T., Wright, C. M., Wilner, D. J., Lommen, D. J. P., Koribalski,B. 2012, MNRAS, 425, 3137van Boekel, R., G¨udel, M., Henning, Th., Lahuis, F., Pantin, E. 2009, A&A, 497, 137Veron-Cetty, M.-P. & Veron, P. 2010, A&A, 518A, 10Vicente, S. M. & Alves, J. 2005, A&A, 441, 195 31 –White, R. J. & Ghez, A. M. 2001, ApJ, 556, 265Wilner, D. J., D’Alessio, P., Calvet, N., Claussen, M. J., Hartmann, L. 2005, ApJ, L626,109Yang, H., Herczeg, G. J., Linsky, J. L., Brown, A., Johns-Krull, C. M., Ingleby, L., Calvet,N., Bergin, E., Valenti, J. A. 2012, ApJ, 744, 121Zacharias, N., Urban, S. E., Zacharias, M. I., Wycoff, G. L., Hall, D. M., Germain, M. E.,Holdenried, E. R., Winter, L. 2003, CDS/ADC Collection of Electronic Catalogues,1289, 0This manuscript was prepared with the AAS L A TEX macros v5.2. 32 –
V892 Tau 0.9cm SZ Cha 0.9cm CS Cha 0.9cmV892 Tau 1.7cm SZ Cha 1.7cm CS Cha 1.7cmV892 Tau 3.3cm SZ Cha 3.3cm CS Cha 3.3cmV892 Tau 5.5cm SZ Cha 5.5cm CS Cha 5.5cm
Fig. 1.— Cleaned ATCA maps using uniform weighting. In all panels the contours are 3, 6,12, and 24 times the image rms. 33 –
MP Mus 0.9cm SR 21 0.9cm V4046 Sgr 0.9cmMP Mus 1.7cm SR 21 1.7cm V4046 Sgr 1.7cmMP Mus 3.3cm SR 21 3.3cm V4046 Sgr 3.3cmMP Mus 5.5cm SR 21 5.5cm V4046 Sgr 5.5cm
Fig. 2.— Cleaned ATCA maps using uniform weighting. In all panels the contours are 3, 6,12, and 24 times the image rms. 34 –Fig. 3.— SEDs of our targets with fluxes (filled circles) and upper limits (downward triangles)from this work and from the literature. In each panel the red dashed line is a linear fit tothe millimeter fluxes between 0.8 and 10 mm. These fits represent the contribution from thedust thermal emission. Note that for all sources there is an excess emission longward of 1 cm. 35 –Fig. 4.— Literature sources with good SED coverage at mm and cm wavelengths and knownX-ray luminosities (see Sect. 3.1). Symbols are as in Fig. 3. 36 –Fig. 5.— Radio spectral slopes for sources with more than one excess cm flux measurement.Note that Sz Cha (ID 2), V4046 Sgr (ID 6), and GM Aur (ID 8) have detected cm excessemission only at one wavelength, hence they are not included in the figure. Red stars aresources with known jets. For TW Hya (ID 7) we also compute the spectral slope using onlythe two longest wavelengths at 4.1 and 6.3 cm (empty circle), see discussion in Sect. 3.2.Flat sources are consistent with free-free emission from optically thin, as from an ionizeddisk surface, to slightly thick plasma. Negative slopes are suggestive of gyro-synchrotronemission. The very positive slope of source 13 points to optically thick free-free jet emission. 37 –Fig. 6.— Upper limits to the EUV photon luminosity reaching the disk (Φ
EUV , cm ) as afunction of stellar X-ray luminosity (L X ). Φ EUV , cm is estimated from the excess cm emission.Sources with known jets are marked in red. The grey area shows the range of stellar EUVluminosities derived by Alexander et al. (2005). Assuming an EUV photon energy of 13.6 eVthe conversion factor from photons/s to erg/s is ∼ . × − , meaning that an EUVluminosity of 10 s − corresponds to 2 . × erg/s. 38 –Fig. 7.— Upper limits to the EUV photon luminosity reaching the disk (Φ EUV , cm ) as afunction of the EUV luminosity needed to reproduce the [Ne ii ] line luminosities or upperlimits (Φ EUV , NeII ). Symbols are as in Fig. 6. The dot-dashed line gives the one-to-one relation.Several sources lie below this relation suggesting that EUV ionization alone is not sufficientto reproduce the observed [Ne ii ] luminosities. Table 1. ATCA sources: properties relevant to this study
ID Source R.A. Decl. SpTy Distance Companion(s) L X F [NeII] Ref(J2000.0) (J2000.0) (pc) ( ′′ ) (10 erg/s) (10 − erg/s/cm )1 V892 Tau 04 18 40.62 +28 19 15.5 B9 140 0.05, 4 9.21-7.94 17.9 w w w w w w Sources have spectrally resolved [Ne ii ] line profiles with blueshifted peak emission pointing to a slow photoevaporative wind Table 2. Log of the observations
ID Source Wavelength Obs. Date Integration a Gain/Phase(cm) (min) Calibrator1 V892 Tau 0.9 18Oct2012 64 0510+1801.7 18Oct2012 64 0333+3213.3, 5.5 19Oct2012 197 0400+2582 SZ Cha 0.9 18-19Oct2012 148 1057-7971.7 18Oct2012 126 1057-7973.3, 5.5 20-21Oct2012 460 1057-7973 CS Cha 0.9 18Oct2012 95 1057-7971.7 19Oct2012 63 1057-7973.3, 5.5 19-20Oct2012 644 1057-7974 MP Mus 0.9 18Oct2012 63 J1147-67531.7 19Oct2012 32 J1147-67533.3, 5.5 20-21Oct2012 247 1251-7135 SR 21 0.9 19Oct2012 63 1622-2531.7 19Oct2012 63 1622-2533.3, 5.5 19-20Oct2012 170 1622-2536 V4046 Sgr 0.9 19Oct2012 32 1759-391.7 19Oct2012 32 1759-393.3, 5.5 21Oct2012 126 1817-254 a Total time spent on target before flagging and calibration of data
Table 3. ATCA continuum fluxes and rms for our science targets. We also report the restored beam FWHMs of thecleaned images.
ID Source F . (RMS) Beam F . (RMS) Beam F . (RMS) Beam F . (RMS) BeammJy (mJy beam − ) (arcsec) mJy (mJy beam − ) (arcsec) mJy (mJy beam − ) (arcsec) mJy (mJy beam − ) (arcsec)1 V892 Tau 1.8 (0.04) 10 × ×
11 0.78 a (0.02) 35 ×
24 0.57 b (0.03) 60 ×
402 SZ Cha 0.17 c (0.02) 8 × c (0.02) 18 × < × < d (0.3) 49 ×
423 CS Cha 0.35 c ,e (0.03) 8 × < . e (0.04) 15 ×
11 0.38 b (0.01) 30 ×
26 0.30 b (0.04) 51 ×
434 MP Mus 0.80 (0.02) 8 × ×
11 0.136 (0.013) 30 × < ×
405 SR 21 0.13 c (0.028) 6 × g (0.022) 12 ×
10 0.057 (0.011) 29 × < ×
346 V4046 Sgr 2.4 (0.02) 6 × ×
10 0.21 (0.011) 28 × < × σ RMS closed contourn. Note that the absolute flux calibration uncertainty (notincluded in the quoted RMS) is ∼
10% of the measured flux densities. a The target and jet (located ∼
80 N) are separated, the reported flux density does not include the jet emission. The 3.3 cm flux density of the jet is 0.094 mJy. b The target and jet are not separated, the reported flux density includes the jet emission. Note that in the case of CS Cha the peak emission is offset fromthe 2MASS source coordinates suggesting that is dominated by the jet. c Flux densities measured within the 2 σ RMS contour. The 3 σ RMS contour flux is less than the one reported by more than the 10% absolute flux calibrationuncertainty suggesting that significant emission is lost with the 3 σ cutoff (see also text). d High background due to emission from the BL Lac-type object PMN J1057-7724 (see Sect. A). e The jet is separated from the star/disk and detected at a ∼ σ level. The jet flux density within the 2 σ RMS contour is 0.061 mJy at 0.9 cm and 0.096 mJyat 1.7 cm. f The jet is not separated from the star/disk system hence the reported flux densities include the jet emission. g We discarded the 19 GHz dataset because it has a rms that is 50% higher than the 17 GHz dataset.
43 –Table 4. Literature sources: properties relevant to this study
ID Source SpTy Distance L X F [NeII] Jet? Ref(pc) (10 erg/s) (10 − erg/s/cm )7 TW Hya K6 51 1.5 3.8 ∗ ,w n 1,2,3,48 GM Aur K7 140 1.6 1.2 n 5,6,79 RY Tau G1 140 5.5 < w n 10,11,1211 DG Tau K6 140 0.55 26 yes 5,7,612 HL Tau K7 140 3.8 < < ∗ Average of several values w Sources have spectrally resolved [Ne ii ] line profiles with blueshifted peak emission pointing to aslow photoevaporative wind Table 5. Millimeter and centimeter measurements compiled from the literature.
ID Source λ Flux Ref(cm) (mJy)1 V892 Tau 0.085, 0.13, 0.29 638, 234, 55, 1, 22 SZ Cha 0.087, 0.12, 0.32, 0.68 314, 77.5, 5.8, 0.7 3, 43 CS Cha 0.087, 0.12, 0.32, 0.67, 0.7, 1.56, 1.56, 1.76, 1.76 197, 128, 9.4, 1.4 ∗ , 1.3 ∗ , 0.4, < < < < < < ∗ Mean of several measurements
45 –Table 6. Upper limits on the EUV photon luminosity (Φ
EUV , cm ) reaching the disk. Theseupper limits could be at most a factor of 2 higher. ID Source Φ
EUV , cm λ EUV , cm (10 s −1