3.1

Moderate resolution spectroscopy of directly imaged exoplanets

With over 1000 planetary systems now identified, the ubiquity of planets in our Galaxy has been spectacularly confirmed.^12,13 Much of the focus of the next decades will now shift to characterization of planetary systems in order to understand the way in which they form and evolve. Spectroscopy of exoplanets offers a powerful tool to not only constrain their atmospheric properties as a function of mass, age, and separation, but also to reveal the fingerprints of planet formation. With high enough spectral resolution, individual atomic and molecular features can be resolved to perform detailed abundance analysis that can offer clues about the composition of exoplanets.^14,15 IRIS astrometric accuracy in the imager mode will also play a crucial role in astrometric monitoring of directly imaged planets and 3D orbital analysis, that will be coupled with the radial velocity measurements from spectroscopy.

Direct imaging is the latest technique for detecting and studying extrasolar planets (e.g., HR 8799¹⁶, Beta Pic¹⁷, HD95096b¹⁸). This technique has the distinct advantage, compared to the Doppler radial velocity and transit detection, of being able to directly measure the photometric and spectroscopic properties of the planets themselves. With the recently commissioned Gemini Planet Imager (GPI¹⁹) and SPHERE²⁰ on VLT a new population of widely separated (~5-100 AU) extrasolar planets are expected to be discovered in the coming years. However, characterization will initially be limited to low spectral resolution (R~50-100) by the instrumental design (GPI and SPHERE), scattered light artifacts (current-generation AO), and sheer faintness. At the resolutions of 4000 or more offered by IRIS, multiple absorption lines from neutral atoms (e.g. Na and K) and molecules (e.g. FeH and VO) can be resolved and used to assess surface gravity (a proxy for age and mass) and metallicity (a key diagnostic for the formation of gas-giant planets). In addition, detailed modeling of the dominant opacity sources of H₂O, NH₃, CO, and CH₄ will provide more accurate temperature measurements and also probe non-equilibrium chemistry though the relative abundance of CO to CH₄, providing important clues on planetary formation.^21,22,23 Figure 2 provides an example of the quality of spectroscopy that will be routinely achievable with IRIS on directly imaged exoplanets. The higher SNR and photometric accuracy achieved with IRIS spectroscopy on directly imaged planets, compared to GPI and SPHERE, will be important for studying cloud variations and atmospheric weather, as well as spin rotation.²⁴ IRIS will therefore offer an unequaled probe of the physics of planetary atmospheres compared to current and upcoming facilities.

Figure 2:

The observed R=3800 spectrum of the HR 8799c planet (black) from OSIRIS at Keck Observatory (Konopacky et al. [23]). The green curve represents the best-fit atmospheric model. A binned version of this spectrum (R~10, purple dots) illustrates what the spectrum would look like if it had the typical resolution of space-based transit spectroscopy. This is one of the highest spectral resolution observations of any extrasolar planet atmosphere. The R=3800 resolution allowed for detailed measurements of the strengths of the molecular features H₂O and CO, and the lack of CH₄. This is one of the few targets observable with current instrumentation due to its brightness and relatively wide separation from the host star¹⁶. IRIS will be able to routinely achieve this level of spectroscopy on the newly discovered extrasolar planets at lower masses, cooler temperatures, and closer separations.

3.2

Crowded field spectroscopy: resolved stellar populations at the Galactic Center

One of the leading science cases for IRIS is to study the Galactic Center with unprecedented sensitivity and spatial resolution. The Galactic Center is the closest laboratory for studying the environments and fundamental physics of supermassive black holes and offers several unique science cases for TMT, e.g., Yelda et al. [25]. Both the IRIS imager and IFS have been carefully designed to characterize the surroundings of the supermassive black hole (SMBH), SgrA*, at the Milky Way’s center.^26,27 The high relative astrometric accuracy of 30 μas will offer a test of General Relativity and probe the distribution of dark matter through orbital monitoring of the stars surrounding SgrA*.^28,29 Characterizing the stellar population at the Galactic Center is of great interest to ascertain how stars form and evolve around a SMBH.^30,31 The young stellar population near SgrA* has puzzled astronomers, as young massive stars should have difficulty forming in close proximity to a SMBH. Researchers have thus far been limited to studying only the most luminous stars in the area, including OB main sequences stars, red giants, and Wolf-Rayet stars. Currently, OSIRIS³² on Keck is able to achieve spectroscopy with sufficient SNR to measure spectral types and radial velocities for Kp < 15.5 mag stars. In contrast, IRIS is predicted to have the sensitivity to allow for high SNR spectroscopy on Kp = 20 - 21 mag stars (Figure 3). These sensitivities and high angular resolutions will allow researchers to study the low mass-end of the main sequence, which will be crucial for investigating the stellar population. They will also provide essential radial velocity measurements to couple with the proper motion monitoring via imaging for deriving 3D orbital solutions. Currently, the shortest period star identified around the black hole has a period of 11.5 years.³³ IRIS is expected to identify multiple sources with shorter orbital periods (1-2 years), which will be instrumental for fundamental physics studies like testing General Relativity.

Figure 3:

We present the theoretical luminosity function of the stellar population within 1’’ of Sgr A* including both the young and old stellar population in the Galactic Center that has been explored by Lu et al. [31] and Do et al. [34], respectively. The fraction of all stars as a function of Kp magnitude (left) illustrates the number of stars observable with TMT (right) compared to the current capabilities of Keck AO imaging (middle). The Keck LGS-AO has a magnitude limit of 15.5 magnitude for spectroscopy (dotted line in left figure), while TMT will easily observe low mass stars on the main sequence and probe the initial mass function, which is essential for understanding the mysterious origins of the stellar population surrounding the SMBH. The IFS will be able to do spectroscopy with a signal-to-noise ratio of 20 in 2 hours of observations for Kp=20 mag stars. This is important for investigating the star formation history, as well as coupling radial velocity measurements for 3D orbital analysis to investigate General Relativity and the dark matter distribution.

3.3

Black hole mass measurements: probing M_BH-σ and M_BH-L

Understanding the interplay between supermassive black hole (SMBH) growth and formation of its host galaxy is currently one of the most outstanding astrophysical questions. Some of the largest puzzles are the origins of the M_BH-σ relationship,^35,36 the role of active galactic nuclei (AGN) feedback, mergers of black holes and their effect on the centers of galaxies, and how to effectively fuel black hole growth. The angular resolution of IRIS represents the next major leap in observational capabilities for SMBH detection. These measurements will greatly impact our understanding of galaxy and SMBH formation. TMT will make it possible to expand the mass range over which black holes can be detected, into the largely unexplored low to high-mass range of the M_BH-σ relation (see Figure 4). Quiescent galaxy black hole masses have been primarily measured from stellar or gas dynamics, with only ~80 galaxies measured to-date³⁶. Most recently, the use of adaptive optics on 8-10m class telescopes has allowed some of the most precise measurements of SMBH masses,^38,39,40 since the angular resolution is sufficient for sampling the sphere of influence of the SMBH and near-infrared wavelengths is less susceptible to dust extinction. Do et al. [3] have recently explored the prospects for measuring SMBH masses in the range of 10⁴ to 10¹⁰ M_⊙ using IRIS. They demonstrated that IRIS will provide the SNR to easily measure intermediate SMBH masses (>10⁶ M_⊙) out to z~0.2 and will have the capability of investigating the elusive intermediate black holes in globular clusters and dwarf galaxies. In contrast, the spatial scales offered by JWST and HST are insufficient to probe the sphere of influence of the SMBH, which is necessary to make mass measurements and explore the full M_BH-σ phase space. It is also worth noting that the most massive black holes (10^9-10 M_⊙) have the largest sphere of influence, and IRIS has the angular resolution necessary to resolve the stellar or gas dynamics around such behemoth black holes even early in the universe (z > 1). These would be incredibly challenging observations, but TMT’s angular resolution makes this project potentially viable.

Figure 4:

The current black hole mass and stellar velocity dispersion (M_BH vs. σ) phase space (values obtained from McConnell & Ma [37]). The highlighted regions show the capabilities for measuring black hole masses with TMT and IRIS. For black hole masses in the range 10⁶ - 10⁷ M_⊙, very few detections exist outside of the Local Group. TMT will make it possible to resolve the black hole sphere of influence in nearby late-type spiral galaxies and dwarf ellipticals (red and green regions), allowing measurements that were previously only possible for Local Group galaxies to reach greater distances (see Do et al. [3]). IRIS will even be able to explore the elusive intermediate black hole mass (10³ to 10⁴ M_⊙) phase space (shaded blue).

3.4

Resolved spectroscopy of high-redshift galaxies (1 < z < 5)

IRIS will expand our knowledge of galaxy formation in the young universe. While we are beginning to compile measures of the global parameters (e.g., luminosity, color, star formation rate, gas and dust content, and stellar mass) of very high redshift galaxies, there is still a gap in our knowledge of the processes that regulate galaxy growth and evolution even at modest redshifts of z ~1 – 3. This is an extremely important epoch in the lives of normal galaxies, mirrored in the precipitous drop in the cosmic star formation rate density below z ~ 2.⁴¹ Recent groundbreaking observations have made use of IFSs and AO (e.g., OSIRIS and SINFONI) on 8-10m class telescopes, and have been able to probe the dynamical processes of individual high-redshift galaxies (z > 1.5),^42,43,44,45 but they are limited to the most luminous galaxies (10 – 100 M_⊙ yr^-1) and coarse sampling (≥ 0.5 kpc) compared to the sampling capabilities of IRIS and TMT (e.g., 34 pc at z = 2).

IRIS will be able to spatially resolve nebular emission lines such as Hß, [O II], [O III], [N II], and [S II] in addition to Ha. With these measurements, it will be possible to construct spatially resolved maps of diagnostic nebular line ratios on a range of galaxy masses and redshifts. In addition to spatially isolating regions dominated by different excitation mechanisms (photoionization, shock heating, etc.), the spatial resolution of IRIS at high-z will also probe the star formation history and differential enrichment of large H II regions and individual super star clusters. It will therefore be possible to compare the abundances of different components of high redshift galaxies, differentiating (for instance) between a central AGN,^46,47 bulge, and the nascent stellar disk. This will offer strong constraints on the origin of the mass-metallicity relation^48,49 and the primary sources of feedback responsible for regulating star formation during the epoch when the majority of stars in the visible universe were formed.

IRIS will also be able to probe lower integrated star formation rates, well below 1 M_⊙ yr^-1. IRIS will achieve sensitivities that are 20-30 times better than current 8-10m telescope AO-fed IFSs. This is illustrated in Figure 5, where the 0.05’’ scale will allow observations of z = 1.5 star-forming galaxies with integrated star formation rates of 0.1 – 10 M_⊙ yr^-1. The 0.025’’ and 0.05’’ scales are advantageous for detecting fainter emission lines (e.g., [OII], Hß, [OIII], [NII]) to study the 2D metallicity and dynamics of high-redshift galaxies (>10 M_⊙ yr^-1), while the finer spatial scales 0.004’’ and 0.009’’ will be used to study luminous star forming galaxies at high spatial resolution. Our team has already explored sensitivities and this science case for high-redshift star-forming galaxies in greater detail (see Wright et al. [2]).

Figure 5:

A simulated z = 2.3 galaxy, with Ha redshifted to K-band and an integrated star formation rate of 10 M_⊙ yr^-1 (left). This is compared to Keck observations of the same simulated galaxy (right) using the same spatial sampling of 0.05’’ per spaxel. Both simulations were for a R=4000 IFS with a 2 hour (8x900s) integration in K-band. Complex structure with multiple knots can be seen with IRIS and TMT, while only the brightest portion of the galaxy is barely detectable (SNR~3) with Keck. The higher SNRs for IRIS are also essential for detecting other emission lines, like [NII], [OIII], [SII], for diagnostics of their ionization parameters and chemical abundances. The finest spatial scale of 0.004’’ and 0.009’’ will be able to probe down to individual giant molecular cloud regions in this galaxy.

Characterizing Lya emitters (LAEs) and Lyman continuum at moderate redshifts (z = 2 - 4) will also be essential for understanding the epoch of reionization (z > 6). The Lyman continuum flux escaping from high-z galaxies is responsible for the bulk of the ionizing photons necessary to initiate and maintain reionization.^50,51,52 Theoretical interpretations that predict the fraction of ionizing photons are based on the observed relationship between the escape fraction of Lyman continuum and optical nebular emission lines, such as [OII], [OIII], and Balmer lines.⁵³ Galaxies with the highest observed fraction of escaping Lyman continuum flux are faint, blue galaxies and the LAEs that have typical star formation rates of < 1 M_⊙ yr^-1.^54,55 Therefore, the sensitivity of TMT+IRIS is necessary to detect faint high-z nebular emission lines to directly calibrate against Lyman continuum measurements. Empirically establishing these relationships at z = 2 - 4 will be essential for deducing the ionizing escape fraction in the peak era of reionization (see Section 3.5). The angular resolution of IRIS will enable an understanding of the physical mechanisms and conditions that allow the Lyman continuum to escape in high-z galaxies, and therefore its subsequent effect on galaxy formation.

3.5

Ly-α spectroscopy of first light galaxies (6 < z < 12)

Understanding the first galaxies in the Universe is a crucial problem in cosmology and astrophysics. The epoch of “first light” - between 300 to 900 million years after the Big Bang - is a key period in cosmic evolution for understanding the first seeds of stellar, supermassive black hole, and galaxy formation. It is the epoch where we can probe the building blocks of the first galaxies and understand their stellar population and intergalactic medium. The science case for “first light” galaxies is one of the prime drivers for upcoming facilities like JWST and the GSMTs.

JWST will have unprecedented sensitivity that will allow for the discovery of first light galaxies.⁵⁶ Through the use of accurate photometric redshifts, JWST will likely discover the most distant and luminous galaxies yet observed. IRIS will be uniquely positioned to conduct spatially resolved near-infrared spectroscopy of these distant young galaxies. Resolved spectroscopy will be essential for spatially resolving rest-frame ultraviolet emission lines, like Lya, from the star-forming regions of the galaxies and/or any potential AGN. Observations of the Lya line profile will reveal a great deal about the dynamics of the gas in these extremely young galaxies. In many redshift intervals, the He II (164.0 nm) line is also accessible. If a strong HeII feature is discovered, the observations will provide compelling evidence for a hard radiation field from massive stars formed from low metallicity gas, thus helping to map out the chemical enrichment histories of these elusive galaxies.

IFS observations of the Lyα emission line will be crucial for tracing the kinematics and morphologies of first light galaxies. The spatial distribution will potentially allow us to rule out AGN, and the line profile of the Lya feature will yield strong constraints on the ionization state of the nearby intergalactic medium. Although NIRSPEC on JWST has an IFS, the spatial sampling (0.1’’ per spaxel, R<2700) is not as well suited for the sizes of these first light galaxies compared to IRIS with TMT. The surface brightness profiles of the most distant galaxies observed to-date (z = 7 - 8)

have typical half light radii of 0.4 - 0.8 kpc (0.09 - 0.18’’ at z=9).⁵⁸ Therefore to resolve the internal structure of first light galaxies it is necessary to have high angular resolution and sensitivity to observe the velocity structure and widths of Lya. Our team has explored a range of expected Sersic indices (n = 1 - 4) and integrated light fluxes of Lya at varying redshifts.⁵⁸ Figure 6 illustrates how the SNR is impacted by the surface brightness profiles (total flux of 0.1-2×10^-18 erg s^-1 cm^-2) of Lya across a z=8.9 first light galaxy. The “clumpier” the surface brightness profiles the higher SNR achieved on these sources, as has been observed in present IFS observations of high-z galaxies.

Figure 6:

We investigated IFS sensitivities of Lyα emission from young distant galaxies with varying surface brightness profiles and redshifts (6 < z < 12). Using the measured half light radius of first light galaxies, between 0.4 - 0.8 kpc (Oesch et al. [56]), we explored a range of Sersic profiles and the SNR per FWHM versus the total Lya flux of the galaxy⁵¹. (LEFT) We present a simulated image of a z=8.9 galaxy with Lyα of n=3 Sersic index with normalized flux with a half-light radius of 0.6 kpc. (RIGHT) A range of Sersic profiles with their corresponding SNR per FWHM is presented for the same galaxy. This is using the 0.05’’ IFS plate scale at R=4000 with a Lya velocity dispersion of 25 km s^-1 observed for a total of 5 hours (20 × 900s) in H-band.

The detectability of Lyα emission will largely be dominated by the escape fraction and optical depth to Lya. Recent spectroscopic studies of z = 8 Y-dropout candidates imply that the optical depth at these redshifts may already be high enough to limit their detection.^59,60 Treu et al. [59] used MOSFIRE⁶¹ to reach a 5σ limit of 0.4-0.6×10^-17 erg s^-1 cm^-2, which is still 2 to 20 times brighter than sources detectable with IRIS. However, even if the optical depth is considerably higher at z > 8, observing redder lines (e.g., CIII 190.9 nm) may prove to be a better probe of first light galaxies. For instance, CIII emission emanating from z = 8 - 10 galaxies is redshifted into the J, H, and K bands, where the NFIRAOS performance is best. Although, CIII is intrinsically fainter than Lya, it may still have an increased SNR based on lower extinction and better instrumental performance of IRIS and NFIRAOS at these longer wavelengths. IRIS and TMT will provide the necessary sensitivity and spatial resolution to sample the dynamics of first light galaxies. This is a key science driver for all the GSMTs, and IRIS with NFIRAOS is well suited for this science objective.

3.6

Strong gravitational lensed systems: investigating first light galaxies and measuring dark matter substructure

Gravitational lensing will play an essential role in studying the most distant galaxies and will be crucial for mass estimates of galaxies at intermediate redshifts. The combination of an IFS and AO on GSMTs is ideally suited for resolved spectroscopy of lensed z = 8 - 10 galaxies. First light gravitational lensed sources have yet to be discovered, but hundreds of high-redshift (z > 1) strong lenses will be found by future surveys with JWST and LSST. A recent example is the discovery of a lens at z = 1.5 using HST.⁶² Resolved spectroscopy using IRIS and TMT will provide the necessary angular resolution to decouple the emission of the lens and background galaxy, even in the most complicated systems.

These observations will provide measurements for both the lens (z = 1 - 2 massive galaxy) and lensed source (z = 8-10 galaxies of Lyα emission) with a direct estimate of the total mass of the lens and high angular resolution velocity profiles of Lyα of the lensed galaxy. For instance, for a z=9 un-lensed galaxy, the angular size scale of 0.05’’ corresponds to 228 pc. Assuming a typical lensing magnification factor of 10, the spatial sampling in the source plane will be 22.8 pc per spaxel. For the 0.009” IFS plate scale, the sampling for lensed z = 9 galaxy would be 4.1 pc, which remarkably, is equivalent to ground-based observations of Virgo cluster galaxies.

Strong gravitational lensed systems also offer powerful probes of the substructure in dark matter halos at a range of redshifts. With the unprecedented resolution of IRIS, we will be able to probe dark matter substructures in lensing galaxies with mass a factor ~100 lower than current measurements, which will provide a stringent test of a generic prediction of the Λ Cold Dark Matter model. Local studies of nearby galaxies and their sub-halos rely heavily on baryonic physics (e.g. tidal streams⁶³), whereas strong gravitational lensing is able to directly probe the dark matter distribution and measure lower mass sub-halos. Photometric and astrometric studies have been used to investigate the gravitational potential of the lens system based on the source image time delay, position, and anomalous flux ratios.^64,65,66,67 Recently, Keck OSIRIS AO observations have been able to use observed flux differentials in a quadruple lensed background QSO (z=3.6) to measure the mass of sub-halos in a z=0.4 lens elliptical galaxy (see Figure 7).⁶⁸ As previously stated, thousands of quadruple lensed systems will be discovered with JWST and LSST at even greater redshifts (z = 0.5 - 2). With the use of TMT and IRIS we will therefore be able to extend current studies to a larger number of systems at higher redshifts and with lower mass sub-halos (~10⁶ M_⊙).

Figure 7:

Results of narrow-line emission analysis using Keck AO with OSIRIS on a gravitational lensed z=3.6 QSO (Courtesy Nierenberg et al. [68]). (LEFT) HST NICMOS F160W image of B1422+231. (MIDDLE) OSIRIS field of view of three lensed images (A, B, C) in [OIII] emission, which were used to measure differential flux ratios between the lensed images to derive the mass of subhalos in the lens system (z=0.4 elliptical galaxy). (RIGHT) OSIRIS extracted spectra for each lensed image. This was the first demonstration of using an AO + IFS to measure sub-halo masses with flux anomaly measurements in quadruply lensed system.