SUPPORTED We have recently discovered a strong candidate for a planetary-mass (~10 Mjup) companion to a young Sun-like star, based on near-IR imaging and spectroscopy with the Keck and Subaru adaptive optics (AO) systems. While the ground-based data strongly suggest that the candidate has a very low effective temperature, and hence a very low mass, they are not definitive. We propose to obtain NICMOS coronagraphy to measure the companion's 1.9um water-band absorption. This feature is a distinct signature of very cool objects and is unobservable from the ground. The combined ground-based and space-based data set will dermine whether the companion has a very low temperature, and hence if it is the lowest mass companion found to date by direct imaging.

<<CORONGRAPHIC IMAGING>> We will acquire NIC2+F160W coronagraphic imaging of the candidate companion to a bright star using the "two-rolls-in-one-orbit" strategy. The observation strategy will require two visits scheduled as one orbit, with a spacecraft roll of 25-25.9 degrees in between the two visits. During each visit, we will acquire F110W, F171M, F190N, and F215N images of the companion to easure its colors and the depth of the 1.9um water absorption band. We have chosen ORIENTs such that the companion is at least 30D (+/-3D) away from the PSF diffraction spikes in any pair of visits. For each visit of a two-roll orbit, we assume 22 min are available. The observing breakdown is as follows: guide star acquisition and NICMOS setup (6.3 min); Mode-2 Acquisition to place the object under the coronagraphic hole (3 min); three MULTIACCUM integrations in each of the F110W, F171M, F190N, and F215N filters, with total integrations of 9s (STEP1), 48s (STEP8), 312s (STEP8), and 240s (STEP16), respectively (10.2 min total); MULTIACCUM overheads (0.8 min); and filter change overheads (1 min). We have chosen a conservative approach in the MULTIACCUM sampling to recover the largest dynamic range in the images (i.e., to acquire unsaturated data of the inner halo of the primary star, which is needed for good photometry of the companion). Photometry of the companion will be dominated by systematic errors from the PSF subtraction. Hence, we will use 3 orbits to observe the companion (visits 10+11, 20+21, and 30+31). Having multiple independent datasets will provide an estimate of the systematic errors associated with PSF subtraction, photometry and astrometry at such large flux ratios. We have verified that the primary star is bright enough (H=7.7 mag) to be acquired with short (1.0 sec) ACQ observations using the F171M filter. These images should have a peak pixel S/N=250 and fill the pixel to ~60% of the saturation level, according to the NICMOS ETC. Also, by using one of the "science" filters for ACQ, we can avoid the overhead associated with one filter change. <<DIRECT IMAGING>> In order to connect the 1.9um water absorption measurements to an effective temperature scale, we will obtain photometry in the same filters for a sample of field M and L dwarfs. These stars are bright (K~8-12 mag), so integration times to achieve sufficient S/N (>100) will be short. Instrument setup, integration, dithering and filter changes will lead to a total time for each star of typically 8~min. However, the time to slew to a new target and acquire new guide stars will limit us to 2 stars per orbit. We have 3 orbits (visits 80, 81, and 82) for the dwarf calibrators. STAR T TAURI STAR K V-IV J2000 K = 7.1 STAR M V-IV High proper motion. J2000 J = 11.2, H = 10.6, K = 10.3 STAR M V-IV Coordinates come from 2MASS catalog. Object is maybe barely seen on GSC plates --- position on the plates is consistent with the 2MASS position. Comparison of the 2MASS coords and those in Dahn et al (2002), which are epoch 1984, show proper motion is negligible. J2000 I = 15.7, J = 12.5, H = 11.8, K = 11.3 STAR M V-IV High proper motion. J2000 J = 9.9, H = 9.3, K = 8.9 STAR M V-IV Coordinates are from SIMBAD. Unable to identify object on GSC plates. Should check 2MASS catalog but their server is down at the moment. J2000 J = 11.8, H = 11.2, K = 10.7 STAR BROWN DWARF This object is an L1 dwarf, not necessarily a brown dwarf, but the Description fields don't include L dwarfs. Coordinates are from the literature. Object is too faint to be seen on GSC plates, and the 2MASS server is currently down. J2000 J = 12.7, H = 12.0, K = 11.5 STAR BROWN DWARF Coordinates are from discovery literature, since object is too faint to be seen on GSC and the 2MASS server is down. J2000 J = 14.2, H = 13.4, K = 12.6 1 NIC-SPIRAL-DITH SPIRAL DITHER 3 2.0 POS-TARG 0.0 false false Observations will be obtained with "two-roll" technique in one orbit, i.e. imaging in the 1st visit, rolling the telescope by 30D, and then imaging in the 2nd visit. The NICMOS orientations are chosen such that the companion does not land on the diffraction spikes. With a 30D roll between observations, this means placing the companion at the +/-15D positions relative to the NICMOS detector x/y axes - such a choice would put the companion 30D away from the two relevant diffraction spikes. Given the companion sky PA of 280D (east of north), the allowable orientations for the first position of the two-roll observation are 40D, 130D, 220D, and 310D. I include a +/-3D allowable range in APT to ease scheduling. For the second visit of the two-roll observation, I believe the correct *sign* to rotate the telescope from the 1st to the 2nd roll position (i.e. I don't want to rotate onto the diffraction spike) is positive. The primary star is H=7.7. We choose a 1.0 sec integration in F171M, which should give a peak pixel S/N=250 which fills the pixel to ~60% of the saturation level given by the NICMOS ETC. Using one of the "science" filters also allows us to avoid the overhead associated with a filter change. Second orientation of "two-roll" observation. Must be done in the same orbit as Visit 10. Identical to Visit 10 --- see comments there. The primary star is H=7.7. We choose a 1.0 sec integration in F171M, which should give a peak pixel S/N=250 which fills the pixel to ~60% of the saturation level given by the NICMOS ETC. Using one of the "science" filters also allows us to avoid the overhead associated with a filter change. Second orientation of "two-roll" observation. Must be done in the same orbit as Visit 20. Identical to Visit 10 --- see comments there. The primary star is H=7.7. We choose a 1.0 sec integration in F171M, which should give a peak pixel S/N=250 which fills the pixel to ~60% of the saturation level given by the NICMOS ETC. Using one of the "science" filters also allows us to avoid the overhead associated with a filter change. Second orientation of "two-roll" observation. Must be done in the same orbit as Visit 31. Observe two field dwarfs in one orbit. LHS 2471 is "bright" target and BRI 1222-122 is the "faint" one. Integration times are adjusted to account for the difference in brightness. Observe two field dwarfs in one orbit. LHS 2471 is "bright" target and BRI 1222-122 is the "faint" one. Integration times are adjusted to account for the difference in brightness. Observe two field dwarfs in one orbit: T513-46 (bright) and 2M1439+19 (faint). The overheads to slew between targets is MUCH lower for this visit, so we can double the number of exposures. Observe two field dwarfs in one orbit: T513-46 (bright) and 2M1439+19 (faint). Observe two field dwarfs in one orbit: LHS 3003 (bright) and Den1058-18 (faint) Observe two field dwarfs in one orbit: LHS 3003 (bright) and Den1058-18 (faint)