The Gemini Near-infrared Integral-Field Spectrograph (NIFS) is a moderate resolution, near-infrared, integral-field spectrograph intended for use with the ALTAIR facility adaptive optics system on Gemini North. NIFS science is adaptive optics science, so much of the Gemini core science that has been described in the science case for ALTAIR will be realized using NIFS in combination with ALTAIR. NIFS is a fast-tracked, limited capability spectrograph and so will not reproduce all of the GNIRS capabilities. The key features of NIFS are its high spatial resolution integral-field unit (IFU) and its moderate spectral resolution in the 1-2.5 mm wavelength range; imaging spectroscopy at the maximum spatial resolution attainable with Gemini will be the primary role of NIFS. The application of this technique to the study of the demographics of massive black holes in nearby galactic nuclei, and the related study of the excitation and dynamics of the inner narrow-line regions of nearby Seyfert galaxies are defined to be NIFS core science. NIFS will excel in all observations requiring moderate spectral resolution data of spatially complex regions having high surface brightness in either a spatial or spectral sense; spectroscopy of compact, high surface brightness continuum sources and imaging of extended, narrow emission-line regions are examples of these two extremes. These themes feature strongly in the broader science programs described below.

The science requirements of NIFS cannot be treated in isolation from the requirements imposed by ALTAIR. We therefore attempt to estimate AO natural guide star and performance requirements where possible.

4 NIFS Core Science

4.1 Massive Black Holes in Nearby Galactic Nuclei

One of the most profound results from the Hubble Space Telescope (HST) is the evidence for the existence of massive (10⁷ to 10⁹ M_ʘ) black holes in the nuclei of many nearby early-type galaxies (e.g., Kormendy & Richstone 1995; Lauer et al. 1995; Faber et al. 1997). An apparent relationship exists between the black hole mass and bulge mass which suggests that either central black holes grew by accreting inner bulge stars, or else that the central black hole and the bulge formed coevally in major merger events. However, this correlation suffers from strong observational selection effects (Ford 1997), and it is yet to be determined whether elliptical galaxies and spiral galaxies follow the same or different correlations. Given the potential close link between black hole formation and galaxy evolution, it is important to define the mass distribution and frequency of occurrence of central black holes in both classes of galaxies. These are still poorly known; most particularly in late-type spiral galaxies in which the nuclear regions are obscured by dust and the bulge masses are smaller so the black hole masses may also be smaller. Observations with NIFS will help determine the demographics of massive black holes in galactic nuclei. Spatially resolved, high resolution dynamical studies of the innermost nuclear stellar populations (Figure 1) and LINER-like gaseous accretion disks (Figure 2) at near-infrared wavelengths are necessary to do this. Observations of both surface brightness distributions, mean rotation, and radial velocity dispersion profiles with spatial resolutions of a few parsecs and spectral resolutions of 3000-5000 are required to model the stellar dynamics or gaseous accretion disk dynamics and infer properties of the central black hole. The high spatial resolution required dictates the use of adaptive optics correction. This, and the presence of obscuring dust in the central regions of many galaxies which complicates the interpretation of optical data from HST, dictates the use of near-infrared observations. The CO (2-0) absorption bandhead at 2.3 mm is ideal for measuring velocity dispersions of cool stellar populations in the nuclei of low redshift galaxies (e.g., Gaffney, Lester, & Doppmann 1995; Shier, Rieke, & Rieke 1996; Böker, van der Marel, & Vacca 1999). The presence of a mass concentration is indicated by a rising stellar velocity dispersion profile near the nucleus. The emission-lines of H I Pb 1.282 mm, H I Brg 2.166 mm, [Fe II] 1.257 mm, and [Fe II] 1.644 mm are expected from shock-excited gas in circumnuclear LINER-like accretion disk. The enclosed mass is inferred from the rotational velocity, assuming that the gas follows Keplerian orbits about the mass concentration.

Figure 1: HST/WFPC2 F606W images of the central 18²´18² regions of spiral galaxies (Carollo, Stiavelli, & Mack 1998). Many have cuspy cores, and many are obscured optically.

Figure 2: Dusty circumnuclear LINER-like disk in the black hole galaxy NGC 7052 (van der Marel, & van den Bosch 1998).

The velocity dispersion profile of M32 rises from s_V ~ 60 km s^-1 at 1.0² radius to s_V ~ 95 km s^-1 at 0.1² radius (Bender, Kormendy, & Dehnen 1996; Figure 3). Stellar velocity dispersions are therefore expected to be ~ 50 km s^-1 in the outer parts and ~ 100 km s^-1 at the centers of galaxies containing low mass black holes. These velocity dispersions correspond to Gaussian FWHMs of ~ 118 km s^‑1 and ~ 235 km s^-1, respectively. A FWHM velocity resolution of ~ 100 km s^-1 will therefore suffice to measure stellar velocity dispersions from the CO (2-0) absorption bandheads in the K band. The LINER gas disk in the elliptical galaxy M84 has peak rotational velocities of ±400 km s^‑1 (Bower et al. 1998), and the gas disk in NGC 4261 has peak rotational velocities of ±200 km s^‑1 (Ferrarese, Ford, & Jaffe 1996). A velocity resolution of < 100 km s^-1 will be required to measure disk rotational velocities in a range of lower mass objects. Similarly high spectral resolving powers of R ~ 4000-5000 are required to significantly separate individual OH airglow lines in the J and H bands in order to perform sensitive measurements of the emission-lines from circumnuclear LINER-like disks.

Figure 3: Velocity dispersion and rotational velocity profiles for M32 which contains a 3´10⁶ M_ʘ black hole (Bender, Kormendy, & Dehnen 1996).

High Strehl ratios are required for these observations. Galaxy nuclei are in general too faint and diffuse to be efficiently used as guide objects for ALTAIR. Approximately 10% of nearby galaxies in the Shapley-Ames catalog have suitably bright AO guide stars within ~ 30² of the nucleus. However, the laser guide star facility will allow this sample to be greatly extended and will deliver near optimal Strehl ratios on objects with suitable OIWFS guide stars. It will be necessary to accurately determine the PSF by frequent measurement of a PSF star, by reconstructing the PSF from OIWFS frames, or by modeling based on the AO control loop output (Véran et al. 1997).

The performance of NIFS for measurements of stellar velocity dispersions in galactic nuclei has been modeled in NIFSSIM (NIFS Performance Model, SDN0004.01) using empirical K band surface brightness and velocity dispersion distributions for our Galactic center. Saha, Bicknell, & McGregor (1996) fit the Galactic center cool star and hot star surface brightness distributions with Reynolds-Hubble law profiles and the stellar rotational velocity and velocity dispersion profiles with quartic polynomials. These profiles have been adopted for our model galactic nuclei after shifting them to distances ranging up to 20 Mpc and convolving with the instrumental point spread function, P(q). Each spatial position in the nuclear profile is assumed to have an M1III type spectrum in the K band, characterized by the Kleinmann & Hall (1986) spectrum of 75 Cyg, before convolution with a Gaussian velocity dispersion profile.

Figure 4 shows a simulated 3600 s K grating exposure on the nucleus of the model spiral galaxy shifted to 10 Mpc with no interstellar extinction. The simulated spatial image of the galaxy (compressed in the spectral direction) is shown in Figure 5. The peak K band surface brightness in the central 0.1²´0.1² region is 10.53 mag arcsec^‑2. Spectra with continuum signal-to-noise-ratios of ~ 30 are generally required to measure stellar velocity dispersions for dynamical studies (Kormendy, priv. comm.). The simulated observation has a signal-to-noise ratio of ~ 100 in the central 0.1²´0.1² region (excluding systematic effects). The K band surface brightness at a fiducial radius of 1.0² (~50 pc at 10 Mpc) is 12.8 mag arcsec^‑2. Figure 6 shows the K grating spectrum extracted from the central 0.1²´0.1² region of the simulated exposure in Figure 4 after subtraction of a 3600 s sky exposure, transformation for 2D wavelength calibration, and division by a smooth spectrum star. The stellar velocity dispersion has been estimated from this frame using a least-square fit to a Gaussian-smoothed template spectrum which is identical to the simulated galaxy spectrum. Various subtleties complicate the analysis of real data including interstellar extinction, smearing due to the PSF, incomplete sky subtraction and correction for terrestrial atmospheric absorption, and systematic differences between the galaxy and template spectra. None of these is considered here. However, this simple analysis is sufficient to demonstrate the potential of this type of observation. The stellar velocity dispersions measured for each 0.1²´0.05² spatial pixel are shown in Figure 7. The green dots are azimuthal averages of the individual velocity dispersion measurements, and the red curve is the velocity dispersion profile from which the data were simulated. The formal velocity dispersion errors are more than adequate for the purpose of detecting a central massive black hole, but these are likely to be under-estimates of the uncertainties in real data. The K band surface brightness profile in the Galactic center declines as r^-0.8 so the signal-to-noise ratios for the annular azimuthal averages should be proportional to ~ r^0.2, thus accounting for the approximate constancy of the signal-to-noise ratios with radius in the simulation. This highlights the benefits of using an integral field unit for observations of this type.

Figure 4: Simulated 3600 s K grating exposure of the nucleus of a spiral galaxy at 10 Mpc. The CO Dv=2 absorption bands in the galaxy spectrum are apparent at right.

Figure 5: Spatial image of the galaxy nucleus shown in Figure 4. The K band surface brightness in the central 0.1²´0.1² region is 10.53 mag arcsec^-2. Each vertical slitlet is 0.1² wide. The field-of-view is 3.1²´3.3².

Figure 6: Extracted spectrum of the central 0.1²´0.1² region of the simulated galaxy at 10 Mpc shown in Figure 4.

Figure 7: Radial stellar velocity dispersion profile derived from the spectra in Figure 4. The black points indicate the velocity dispersion measurements for each 0.1²´0.05² spatial pixel. The lighter dots are azimuthal averages of these data. The solid curve shows the velocity dispersion profile from which the simulated data were derived.

The radial velocity dispersion profile inferred from a simulated 3600 s exposure with the K grating of the same simulated galaxy now shifted to 20 Mpc is shown in Figure 8. Two effects are apparent here; the linear spatial resolution is degraded, and the signal-to-noise ratios of the measurements are worse. The poorer spatial resolution will be the limiting factor in detecting massive black holes at distances beyond ~ 20 Mpc. The velocity dispersion uncertainties are expected to limit attempts to infer total stellar masses, and hence mass-to-light ratios, in objects such as ultra-luminous IRAS galaxies which are typically at distances between 20 and 100 Mpc. The K band surface brightness (with no extinction) in the 20 Mpc simulation at a fiducial radius of 1² (~ 100 pc at 20 Mpc) is 13.4 mag arcsec^‑2. A signal-to-noise ratio of 30 in 0.1²´0.1² will be reached in ~ 4.5 hr of on-source integration with NIFS (NIFS Performance Model, SDN0004.01) at this surface brightness. Clearly, integration times longer than 3600 s will be required to reduce the velocity dispersion uncertainties.

Figure 8: Radial stellar velocity dispersion profile for a simulated 3600 s observation of a spiral galaxy at 20 Mpc using the K grating. The black points indicate the velocity dispersion measurements for each 0.1²´0.05² spatial pixel. The lighter dots are azimuthal averages of these data. The solid curve shows the velocity dispersion profile from which the simulated data were derived.

The finite spatial resolution achieved with NIFS is the main factor limiting our ability to detect massive nuclear black holes. The enclosed stellar mass at any radius is derived from the velocity dispersion profile and the light distribution using the collision-less Boltzmann equation. Under the reasonable assumptions that the velocity dispersion is approximately constant in the inner region of the galaxy and that the volume mass density is proportional to ~ r^-2, the mass enclosed within radius r is given very approximately by

where D is the distance to the galaxy, s is the velocity dispersion at radius r, and M_enc is in units of M_ʘ. Figure 7 indicates that a galaxy like the Milky Way at a distance of 10 Mpc will have a velocity dispersion of ~ 70 km s^-1 at r ~ 0.1² (~ 5 pc at 10 Mpc), the innermost measurement. This corresponds to an enclosed mass of » 1.2 ´ 10⁷ M_ʘ according to the above equation. The K band luminosity of the Milky Way within a radius of 5 pc is ~ 3.5 ´ 10⁷ L_ʘ, so the mass-to-light ratio of our simulated galaxy within the region probed by NIFS is M/L_K » 0.34. Normal stellar populations can have M/L_K in the range 0.0-2.5 (Figure 9), so we will require mass-to-light ratios in excess of at least ~ 5.0 to unambiguously indicate the presence of a massive black hole. This will be the case if the mass enclosed within r ~ 0.1² is > 1.8 ´ 10⁸ M_ʘ due to the presence of a > 1.7 ´ 10⁸ M_ʘ black hole. If this was the case, the velocity dispersion at r = 0.1² would be ~ 270 km s^-1. We therefore expect that NIFS will unambiguously detect black holes with masses > 2 ´ 10⁸ M_ʘ in spiral galaxies out to distances of ~ 10 Mpc. The 2´10⁶ M_ʘ black hole in the Milky Way nucleus is uncharacteristically small; the best estimate of the mean relation between black hole mass and bulge mass (Magorrian et al. 1998) predicts that the 2´10¹⁰ M_ʘ Milky Way bulge would typically be associated with a 1.2´10⁸ M_ʘ nuclear black hole mass. If this mean relation proves to be appropriate for late-type spiral galaxies, our simulations suggest that it will be possible to detect only the more massive of the black holes that may be present in late-type spiral galaxies to distances of at least 10 Mpc with NIFS.

Figure 9: M/L_K versus cluster age for a exponentially declining star formation rate with 10⁹ yr time constant (Thatte et al. 1997).

4.2 Nearby Active Galactic Nuclei

Many nearby galaxies possess active nuclei which are characterized by broad (FWHM ~ 500 km s^-1) emission-lines originating in their central regions over size scales of 100 pc up to ~ 2 kpc. This is the so-called narrow-line region (NLR). The ultimate energy source is believed to be accretion onto a massive black hole in most objects, although intense starbursts in dense regions may be responsible for some LINER-like activity (Terlevich & Melnick 1985). Emission from the immediate vicinity of the accretion disk produces the broad-line region (BLR) which remains unresolved with existing telescopes. Understanding the nature of the central energy source, its interaction with the host galaxy, and the global implications for the evolution of galaxies are continuing themes in the study of Active Galactic Nuclei (AGN). High spatial resolution optical studies of AGN with HST (e.g., Capetti et al. 1996; Winge et al. 1997; Axon et al. 1998; Figure 10; Figure 11) have revealed a wealth of information about the structure and excitation of the inner NLR. While it has traditionally been believed that the NLR clouds are photoionized by the central source (Ferland & Netzer 1983; Wilson & Tsvetanov 1994), these recent high spatial resolution imaging and dynamical studies have demonstrated that NLR clouds may instead be predominantly shock-excited by energetic thermal and non-thermal mass outflows from the central object. Strong dynamical interactions between the emission-line gas and radio-emitting ejecta can be explained if the NLR is formed from shells of ambient interstellar medium swept up and compressed by the supersonic expansion of hot-gas heated by interactions with the advancing radio jet (Pedlar, Dyson, & Unger 1985; Taylor, Dyson, & Axon 1992; Steffen et al. 1997). The nuclear regions of Seyfert galaxies are invariably obscured by dust clouds making near-infrared observations of the inner NLR desirable. The near-infrared region also offers the best ground-based spatial resolution using adaptive optics correction. [Fe II] 1.257 mm, [Fe II] 1.644 mm, H I Pb 1.282 mm, and H I Brg at 2.166 mm emission-lines are well-suited to excitation and dynamical studies of the high-excitation precursor zones associated with fully radiative shocked regions (Figure 12). Strong coronal emission-lines are the primary initial coolants of hot gas in partially radiative shocks. With NIFS, the [Si VI] 1.961 mm coronal line will become accessible at modest redshift. The mechanical energy flux from the jet can be estimated from the [Fe II] and H I Pb lines in less obscured regions, and from H I Brg in more obscured regions. H₂ 1-0 S(1) 2.122 mm emission in Seyfert galaxies is also collisionally-excited, but generally has a smaller velocity width of ~300 km s^‑1 suggesting that it may arise in a different emission region (Veilleux, Goodrich, & Hill 1996). X-ray-heating from the AGN core, shock-heating by the interaction of the radio jets with the interstellar medium, and shock-excitation in outflows from star formation regions may all contribute to the H₂ emission from Seyfert galaxies. High spatial resolution dynamical studies may provide a means of distinguishing between these alternatives.

Figure 10: HST/WFPC2 image in [O III] l5007 of the NLR clouds near the nucleus of the Seyfert 1 galaxy NGC 4151 (Hutchings et al. 1998).

Figure 11: HST/WFPC2 image of NGC 1068 in [O III] l5007 light (grayscale) and 6 cm radio emission (contours) from Capetti, Macchetto, & Lattanzi (1997). Note how the radio jet penetrates between [O III] clouds. The 3²´3² field-of-view is similar to that of NIFS.

Figure 12: Near-infrared K band spectra of the Circinus galaxy obtained with the 3D integral-field spectrograph in 0.5-0.6² seeing and with spectral resolving powers of 1000 (top) and 2000 (bottom) (Maiolino et al. 1998).

Seyfert activity is frequently associated with circumnuclear starbursts (Figure 13), often in rings, but the role these play in fuelling or refuelling the active nucleus is still unclear. The presence of circumnuclear starburst rings demonstrates that large quantities of gas have been channelled into the region close to the nucleus. This gas may accrete directly onto the black hole, but it must lose its remaining angular momentum to do this. It may be the stars, or their remnants, formed in the starburst ring or the central star cluster that feed the central black hole. High angular resolution spectral imaging of Seyfert galaxy cores will reveal structure interior to the starburst ring. The morphology and dynamics of the emission-line regions will permit new insight into how gas is funnelled into the core and the role possibly played by stellar bars in driving such gas flows. Measurement of stellar mass-to-light ratios, M/L_K, will probe the star formation histories of the regions. Detailed comparison of spatially-resolved spectra with starburst models (e.g., Leitherer et al. 1999) will provide estimates of the starburst ages, masses, and star formation histories. These can then be compared to particular models for AGN fuelling (e.g., Norman & Scoville 1988). The potential of AO-corrected imaging of Seyfert galaxy cores in addressing these issues is beginning to be explored (Marco, Alloin, & Beuzit 1997; Chapman, Walker, & Morris 1998; Rouan et al. 1998; Marco & Alloin 1998, 1999; Figure 14).

Figure 13: K band images of NGC 7469 obtained with the AOB on CFHT (left) and without AO correction (right). The FWHM with AO correction is ~ 0.13². The field is 10²´10². Note that the bright central Seyfert core has been subtracted from the AO corrected image (ALTAIR PDR).

Figure 14: Raw AO-corrected K band image of the core of NGC 1068 obtained with CFHT (Rouan et al. 1998). The field-of-view of 2.2²´2.2² and pixel size of 0.0344 are similar to those of NIFS. The image is plotted on a logarithmic scale.

Measurement of the central black hole masses in Seyfert galaxies is also highly desirable. Although, these can be estimated using reverbation-mapping techniques for a few objects (e.g., Wandel, Peterson, & Malkan 1999), stellar velocity dispersions provide a more direct determination. Stellar velocity dispersions can be measured using the CO (2-0) absorption bandheads and interpreted in the same way as for normal galaxies (§4.1) to place limits on the enclosed mass and hence detect or constrain black hole masses. The analysis is complicated in the case of common Seyfert galaxies by the intense Seyfert core emission (Figure 14). Minimizing this contamination will depend on achieving high Strehl ratios in the telescope, ALTAIR, and NIFS. However, approximately 43% of nearby galaxies show a detectable level of nuclear activity (Ho, Filippenko, & Sargent 1997). These galaxies may contain either lower mass black holes or massive black holes that are currently accreting at well below their Eddington limit. These galaxies will be prime candidates for observation with NIFS. The core of the nearest Seyfert galaxy, Circinus, has a radius of < 1.5 kpc at K (Maiolino et al. 1998), corresponding to < 0.08². The stellar velocity dispersion within ~ 40 pc (~ 2²) of the nucleus is ~ 75 km s^-1, corresponding to a Gaussian FWHM of ~ 180 km s^-1. We require a velocity resolution of ~ 100 km s^-1 to confidently measure such velocity dispersions.

The AO requirements on Seyfert galaxy programs are less severe than for normal galaxies. Many Seyfert nuclei are bright enough and sufficiently compact to use as AO guide objects for ALTAIR. A large number of nearby Seyfert galaxies in the Shapley-Ames catalog were checked for nearby stars. Approximately 20% have suitably bright guide stars within ~ 30² of the nucleus. Many of the 486 low luminosity Seyfert galaxies of Ho, Filippenko, & Sargent (1997) also have suitable AO guide stars and are at distances < 20 Mpc. High Strehl ratios are needed to measure black hole masses, and are desirable when studying emission from NLR clouds.

The feasibility of measuring emission-line profiles for Seyfert galaxy NLR clouds at high spatial resolution has been assessed using NIFSSIM (NIFS Performance Model, SDN0004.01). Near-infrared emission-lines from the NLR are seen against stellar emission from the galaxy and thermal dust emission from the Seyfert core. We use the signal-to-noise ratio predictions for point sources observed with a Strehl ratio of 1.0 (NIFS Performance Model, SDN0004.01) to infer the signal-to-noise ratios per 0.1²´0.1² aperture obtained in 3600 s for different continuum surface brightnesses, and from these estimate 3s detection limits for emission-lines seen against these continua. We adopt an emission-line FWHM of 500 km s^-1 typical of [Fe II] emission in Seyfert galaxies (Fig. 14 of Veilleux, Goodrich, & Hill 1997). H₂ 1-0 S(1) emission in Seyfert galaxies typically has a FWHM of 300 km s^-1, so our surface brightness detection limits should be decreased by 3/5 for this line. The 3s emission-line detection limit predictions for a 3600 s integration time are listed in Table 1 for each of the NIFS gratings at their full resolving power and for a smoothed two-pixel resolving power of R = 1000 matched to the adopted emission-line width. The latter is better suited to detecting NLR emission-lines, but would provide little profile information.

Table 1: 3s emission-line detection limits in 3600 s integrations for 500 km s^-1 line width.

J1 grating			J2 grating
m_J	R = 5090	R = 1000	m_J	R = 6100	R = 1000
(mag arcsec^-2)	(W cm^‑2 arcsec^‑2)	(W cm^‑2 arcsec^‑2)	(mag arcsec^-2)	(W cm^‑2 arcsec^‑2)	(W cm^‑2 arcsec^‑2)
9.0	1.7´10^-21	7.4´10^-22	9.0	1.4´10^-21	5.7´10^-22
10.0	1.2´10^-21	5.2´10^-22	10.0	8.9´10^-22	3.6´10^-22
11.0	7.4´10^-22	3.3´10^-22	11.0	5.8´10^-22	2.4´10^-22
12.0	4.8´10^-22	2.1´10^-22	12.0	3.7´10^-22	1.5´10^-22
13.0	3.1´10^-22	1.4´10^-22	13.0	2.4´10^-22	9.9´10^-23
14.0	2.2´10^-22	9.7´10^-23	14.0	1.7´10^-22	7.0´10^-23
15.0	1.5´10^-22	6.8´10^-23	15.0	1.3´10^-22	5.5´10^-23

H grating			K grating
m_H	R = 5340	R = 1000	m_K	R = 5340	R = 1000
(mag arcsec^-2)	(W cm^‑2 arcsec^‑2)	(W cm^‑2 arcsec^‑2)	(mag arcsec^-2)	(W cm^‑2 arcsec^‑2)	(W cm^‑2 arcsec^‑2)
9.0	8.8´10^-22	3.8´10^-22	9.0	5.2´10^-22	2.3´10^-22
10.0	5.4´10^-22	2.4´10^-22	10.0	3.6´10^-22	1.5´10^-22
11.0	3.4´10^-22	1.5´10^-22	11.0	2.3´10^-22	1.0´10^-22
12.0	2.2´10^-22	9.6´10^-23	12.0	1.9´10^-22	8.1´10^-23
13.0	1.4´10^-22	6.1´10^-23	13.0	1.5´10^-22	6.3´10^-23
14.0	1.0´10^-22	4.3´10^-23	14.0	1.4´10^-22	6.0´10^-23
15.0	7.9´10^-23	3.4´10^-23	15.0	1.3´10^-22	5.7´10^-23

J grating			HK grating
m_J	R = 2790	R = 1000	m_K	R = 2530	R = 1000
(mag arcsec^-2)	(W cm^‑2 arcsec^‑2)	(W cm^‑2 arcsec^‑2)	(mag arcsec^-2)	(W cm^‑2 arcsec^‑2)	(W cm^‑2 arcsec^‑2)
9.0	1.2´10^-21	7.2´10^-22	9.0	3.6´10^-22	2.3´10^-22
10.0	7.4´10^-22	4.4´10^-22	10.0	2.4´10^-22	1.5´10^-22
11.0	4.7´10^-22	2.8´10^-22	11.0	1.7´10^-22	1.0´10^-22
12.0	3.2´10^-22	1.9´10^-22	12.0	1.3´10^-22	8.3´10^-23
13.0	2.1´10^-22	1.3´10^-22	13.0	1.2´10^-22	7.4´10^-23
14.0	1.6´10^-22	9.3´10^-23	14.0	1.1´10^-22	6.9´10^-23
15.0	1.3´10^-22	7.7´10^-23	15.0	9.9´10^-23	6.2´10^-23

The underlying continuum surface brightness in AO observations of Seyfert galaxies can be estimated in several ways. Veilleux et al. have measured near-infrared spectra of 33 Seyfert galaxies through 3²´3² or 1.5²´1.5² apertures. The average emission-line and continuum surface brightnesses from Veilleux et al. are listed in Table 2. The J and K average surface brightnesses provide lower limits to the values that will be detected from the Seyfert core at AO resolution. NICMOS images suggest central H band surface brightnesses of < 10 mag arcsec^-2 in typical Seyfert galaxies (Quillen et al. 1999), and AO images of NGC 3227 obtained with the Adaptive Optics Bonnette on CFHT have surface brightnesses at 0.1² radius of J ~ 7.5 mag arcsec^-2, H ~ 6.5 mag arcsec^-2, and K ~ 6.2 mag arcsec^-2 (Chapman, Morris, & Walker 2000). Peletier et al. (1999) have compiled a database of ground-based subarcsecond resolution J, H, and K surface photometry of a sample of CfA Seyfert galaxies[1]. The K band surface brightnesses of this representative sample of Seyfert galaxies at a fiducial radius of 1² (which largely avoids the non-stellar core emission) are shown in Figure 15 as a function of redshift. Most of this sample have K surface brightnesses at 1² radius in the range 12-16 mag arcsec^-2. Clearly, the underlying core surface brightness depends on the galaxy distance and on the Strehl ratio achieved in the observations, and may vary from < 10 mag arcsec^-2 to > 15 mag arcsec^-2 over the NIFS field-of-view.

Table 2: Average emission-line surface brightnesses (from Veilleux, Goodrich, & Hill 1997).

Object	Aperture (²)	J Continuum (mag arcsec^-2)	K Continuum (mag arcsec^-2)	Pb 1.282 mm (W cm^‑2 arcsec^‑2)	[Fe II] 1.257 mm (W cm^‑2 arcsec^‑2)	H₂ 1-0 S(1) 2.122 mm (W cm^‑2 arcsec^‑2)	H I Brg 2.166 mm (W cm^‑2 arcsec^‑2)

MCG-05-23-16	3´3	13.2	12.0	2.2´10^-22	1.4´10^-22	…	…
Mrk 1	1.5´1.5	…	14.3	…	…	2.0´10^-22	1.8´10^-22
Mrk 3	3´3	14.6	…	1.3´10^-21	1.8´10^-21	…	…
Mrk 78	3´3	15.6	…	1.5´10^-22	2.2´10^-22	…	…
Mrk 176	3´3	14.3	14.6	4.2´10^-23	1.8´10^-22	2.4´10^-23	6.2´10^-24
Mrk 266SW	3´3	14.6	15.2	6.1´10^-22	2.9´10^-22	3.1´10^-23	4.1´10^-23
Mrk 266NE	3´3	14.8	…	4.5´10^-22	5.7´10^-22	…	…
Mrk 348	3´3(J) 1.5´1.5(K)	14.9	13.9	1.5´10^-22	2.6´10^-22	5.2´10^-23	3.9´10^-23
Mrk 403	3´3	15.3	…	6.7´10^-23	…	…	…
Mrk 463E	3´3	14.9	12.6	3.3´10^-22	2.3´10^-22	5.8´10^-23	3.0´10^-23
Mrk 477	3´3	15.5	14.5	1.2´10^-21	8.6´10^-22	1.0´10^-22	1.1´10^-22
Mrk 533	3´3(J) 1.5´1.5(K)	14.5	12.5	7.3´10^-22	4.1´10^-22	1.5´10^-22	4.6´10^-22
Mrk 573	3´3(J) 1.5´1.5(K)	14.3	13.0	2.9´10^-22	2.7´10^-22	5.6´10^-23	1.2´10^-22
Mrk 622	3´3	14.4	…