Imaging Molecular Gas in a Nearby Starburst Galaxy NGC 3256, a nearby luminous infrared galaxy, as imaged by the SMA. (Left) Integrated CO(2-1) intensity at 2”x1” resolution, (Middle) peak CO(2-1) intensity along each line of sight, (Right) mean velocity along each line of sight derived from the CO line. During Early Science, ALMA should be able make comparable observations of nearby starburst galaxies in only a few hours on source. Science Goal: Measure the distribution and kinematics of molecular gas in a nearby starburst galaxy. The starbursts in luminous and ultraluminous infrared galaxies tend to be small (~1 kpc) and even the nearest of these galaxies are still quite distant in absolute terms (10s of Mpc). Exploring the distribution, kinematics, and physical conditions of the gas and dust in these systems is a natural application for sub-mm interferometers. ALMA's Early Science capabilities are well suited to study the nearest such galaxies. In this example, we replicate an SMA study of CO emission from NGC 3256 by Sakamoto et al. (2006). Receivers: Band 6 (230 GHz). Molecular gas in starbursts tends to be fairly excited, making either CO(3-2) in Band 7 or CO(2-1) in Band 6 fine tools to probe the gas. The whole source is ~5 kpc in size or ~30” at 35 Mpc, which is perfectly matched to the field of view in Band 6. Angular Resolution: ~1”. We want to resolve the central disk of molecular gas hosting the starburst. These disks are often ~1 kpc in size, and this is the case NGC 3256. At 35 Mpc, this is only ~5”. To clearly resolve the starburst, we target a resolution of ~1”. Spectral Resolution: ~5 km s -1 . We want to be able to measure the dispersion along individual lines of sight. Sakamoto et al. measure velocity dispersions (1σ) of 10-60 km s -1 and we target 5 km s -1 to ensure at least 5 resolution elements across the line at each position. Channel (Line) Sensitivity: 0.02 K at 20 km s -1 .Sakamoto et al. (2006) report peak 12 CO(2-1) intensities of 2-10 K in 10 km s -1 channels at ~1.5” resolution. If we cared only about this line we might target a 1σ sensitivity of 0.2 K per 5 km s -1 channel. However, we would also like to observe less optically thick 13 CO line, which is often 5-10 times fainter than 12 CO. We design our observations assuming a line ratio of 10 and so target 0.02 K. If 13 CO is much weaker (as Sakamoto et al. actually found), we can improve the SNR by degrading the resolution. The 12 CO data meets our kinematic science goals, so we target a lower spectral resolution of 20 km s -1 (i.e., 1 channel across the line) for 13 CO. This 0.02 K at 20 km s -1 drives our time request. Continuum Sensitivity: N/A. Continuum could easily be a goal of this project, especially if we had chosen Band 7, but we will not focus on it in this example except to say that any unused correlator resources can be put to good use observing the 1mm (Band 6) dust continuum.