Spatially resolved study on star and planet formation in Local Group galaxies

We are investigating the environmental effects of star and planet formation. Particularly, we have been focusing on metallicity dependence. Metallicity, or the abundance of metal, is known to increase with the cosmic evolution owing to the element synthesis of stars and supernovae. Even presently, only 2 % (in mass) of baryon exists in our solar system. Nevertheless, metallicity is believe to be one of the most critical factors for star and planet formation because dust is necessary to form planet cores despite the very small mass fraction in disks ($\sim$1 %) and also because metal is sensitive to heating and cooling in star-forming processes and is directly related to radiative transfer. If any dependences are observed, these can have a strong influence on theories of star and planet formation (e.g., [Elmegreen et al.(2008), Elmegreen et al. 2008], [Ercolano & Clarke(2010), Ercolano & Clarke 2010]).

In the investigation of metallicity dependence of star and planet formation, we are focusing on the IMF and the lifetime of protoplanetary disks. Considering that stars are fundamental components in galaxies and that stellar mass determines their evolutionary path, the IMF is one of the most fundamental parameters that determines the physical and chemical evolution of stellar systems. Because planets are formed in protoplanetary disks and disk duration can regulate planet formation, disk lifetime is one of the most fundamental parameters directly connected to the planet formation probability. Whether the IMFs and disk lifetime derived in the solar neighborhood are universal or sensitive to environmental conditions is under debate ([Lada & Lada(2003), Lada & Lada 2003]). In both these science cases, resolved study is crucial for avoiding blending/stochastic effects.

Our observational targets are star-forming clusters that satisfy the following two criteria: i) young star-forming regions ($<$5Myr old) because massive stars still exist, age spreads are small, and most of the protoplanetary disks are likely present and ii) moderate-scale clusters, with a cluster mass ($M_{\rm cl}$) of $\sim$$10^{2}$–$10^{3}$ $M_{\odot}$. In the solar neighborhood, $M_{\rm cl}$ of $\sim$$10^{2}$ $M_{\odot}$ class is most common ([Adams et al.(2006), Adams et al. 2006]). Moreover, even the largest star-forming region in the solar neighborhood, that is, Orion Nebula Cluster (ONC; Fig. 2, left), has $M_{\rm cl}$ of $\sim$1000 $M_{\odot}$ ([Lada & Lada(2003), Lada & Lada 2003]). In the derivation of the absolute metallicity dependence, we focus on the young clusters with $M_{\rm cl}$ of $<$$10^{4}$ $M_{\odot}$ to combine the results with that of nearby clusters. Indeed, non-standard IMF is indicated for starburst clusters ($M_{\rm cl}$ of $>$$10^{4}$ $M_{\odot}$; e.g., Arches, Westerlund 1, NGC 3603; [Bastian, Covey, and Meyer(2010), Bastian, Covey, & Meyer 2010]) although it may be owing to blending effects. Other possible environmental effects, aside from than metallicity, are suggested on disk lifetime (e.g., cluster mass and cluster density; references in [Yasui et al.(2016a), Yasui et al. 2016a]). Note that the solar system is also most likely to be formed in $M_{\rm cl}\sim 10^{3}$ $M_{\odot}$ class cluster (e.g., [Adams(2010), Adams 2010]).

8–10 m class telescopes. The distances to the target clusters and their stellar separations are summarized in Table 1. In this study, we assumed that the typical stellar separation of nearby cluster members is $\sim$0.1 pc, which is estimated from young embedded clusters in the solar neighborhood ([Adams et al.(2006), Adams et al. 2006]). The distance of $\sim$50–100 kpc (LMC and SMC) is the current limit for spatially resolved studies that used 8–10 m class telescopes with an adaptive optics (AO) that assures decent Strehl ratio ($\sim$0.1) even for extragalactic objects.

As a first step, we are focused on the outer Galaxy, which we defined as a region with $R_{g}$ of $\gtrsim$15 kpc because it is known to have a metallicity of as low as $\sim$$-$1 dex despite its proximity, in comparison with the galaxies in the local universe. The environments are similar to that in nearby dwarf galaxies, damped Lyman-alpha systems, and that in the early stage of the the Galactic disk formation (e.g., [Ferguson et al.(1998), Ferguson et al. 1998], [Kobayashi et al.(2008), Kobayashi et al. 2008]). We have obtained NIR images of more than 10 star-forming clusters in the outer Galaxy through the 8.2-m Subaru Telescope (e.g., Yasui et al. 2008, 2010, 2016b, 2016c). High spatial resolution (i.e. $\lesssim$0.5^′′) and relatively small distances to the clusters enabled us to resolve individual cluster members in each cluster even without AO. Hence, the limiting magnitudes of $K_{S}\sim 18$–21 mag (10$\sigma$) are achieved. This magnitude corresponds to the mass of $\sim$0.1 $M_{\odot}$, which is much smaller than that achieved in nearby dwarf galaxies, such as LMC and SMC ($\sim$1 $M_{\odot}$). Based on the fitting of K-band luminosity functions, we determined that the IMF in the outer Galaxy is consistent with that in the solar neighborhood with regard to the high-mass slope and IMF peak ($\sim$0.3 $M_{\odot}$; Yasui et al. 2017, 2008), suggesting that the IMF down to substellar mass regime ($\sim$0.1 $M_{\odot}$) have no dependence on the metallicity down to $\sim$$-$1 dex. However, we identified that the fraction of stars with a K-band excess (originated from the inner circumstellar dust disk at a radii of $r\leq 0.1$ AU) is significantly lower than that in the solar neighborhood ([Yasui et al.(2009), Yasui et al. 2009]), suggesting a metallicity dependence of the disk lifetime ([Yasui et al.(2010), Yasui et al. 2010]).

JWST. We are planning on the NIR and mid-infrared (MIR) imaging with MIRI/NIRCam as a part of the Guaranteed Time Observation (GTO 1237; PI: Michael Ressler (JPL)). The expected limiting magnitudes with 30 min integration (Vega, 5$\sigma$) are $\sim$23 mag at 4.4 $\mu$m and 17.3 mag at 12.8 $\mu$m, whereas the spatial resolution will be $\sim$0.4^′′ at 10 $\mu$m. The sensitivities are significantly higher compared with existing telescopes (Fig. 2; Thirty Meter Telescope (TMT) Planning instructions (2012): https://tmt.nao.ac.jp/document/brochure/tmtbb-ols.pdf) both in NIR and MIR. However, the spatial resolution is comparable. Therefore, we will focus on the outer Galaxy for the JWST study. The mass detection limit will be $\sim$8 $M_{J}$ in NIR (Lyons/DUSTY evolutionary models of [Chabrier et al.(2000), Chabrier et al. 2000]; cf. $\sim$0.1 $M_{\odot}$ for LMC and SMC). Thus, we can determine whether stars down to 0.5 $M_{\odot}$ still have disks in MIR (calculated based on the case of IC 348 at $D=320$ pc by [Lada2006, Lada et al. 2006]). By using the JWST, IMF in this low-metallicity environment down to substellar mass regime ($<$0.1 $M_{\odot}$) can be derived. The IMF in the very low mass end can be a sensitive function of the formation environment ([Lada & Lada(2003), Lada & Lada 2003]). This study will be the first to research on brown dwarfs and planetary mass objects in low-metallicity environments, enabling us to investigate whether they are common in such environments. We will also gain new insights related to the outer ($\sim$0.1–5 AU) circumstellar disk evolution thanks to the high sensitivities of JWST in MIR. Hence, a comprehensive comparison of disk lifetimes in low-metallicity regions with that in the solar neighborhood that has been extensively characterized by Spitzer (e.g., [Hernández et al.(2007), Hernández et al. 2007]) will be possible.

ELTs offer significantly high sensitivities in $\lambda\lesssim 2$ $\mu$m compared with existing telescopes, but they are comparable to JWST in long NIR and lower MIR sensitivities (Fig. 2). In the case of NIR imaging with TMT/IRIS, limiting magnitudes with 5-h integration (Vega, 10$\sigma$) will be 27.6 mag at the $K$-band (2.2 $\mu$m) ([Moore et al.(2014), Moore et al. 2014]). Meanwhile, the spatial resolution will be as small as $\sim$0.02^′′ with AO at the K-band, which is significantly higher than before, enabling us to extend the spatially resolved studies to the Local Group for the first time (Table 1). For 1 Myr-old targets at $D=770$ kpc, the mass detection limit of $\sim$0.1 $M_{\odot}$ (assuming that $A_{V}=0$ mag) / $\sim$0.5 $M_{\odot}$ (assuming that $A_{V}=10$ mag), based on the isochrone model of [Siess, Dufour, & Forestini(2000), Siess, Dufour, & Forestini (2000)], and stellar separation of 0.08 pc, which is less than the typical stellar separations in clusters (0.1pc; [Adams et al.(2006), Adams et al. 2006]), will be achieved.

The target star-forming regions are as follows:

• •

  Spiral Galaxies (M31, M33): Many star-forming clusters exist. We can derive the absolute metallicity dependence by observing the same type of galaxies as our Galaxy.

• •

  Dwarf galaxies, e.g., IC 10 (Fig. 2, right), NGC 6822, and Leo I. We can explore wider metallicity range down to $\sim$$-$2 dex.

The spatially resolved study on the Local Group galaxies should be one of the most cutting-edge science themes with ELTs. However, even ONC located at $D=400$ pc (covering $\sim$$30^{\prime}\times 30^{\prime}$, corresponding to $\sim$$3.6\times 3.6$ pc²), covers only $\sim$1^′′ square (Fig. 2) if it is located at $D=660$ kpc (i.e. the distance of IC 10). Therefore, the achievement of the spatially resolved study on the Local galaxy is heavily dependent on how effective the AO systems on the ELTs will work.