DOMAINS OF OBSERVABILTY
IN THE NEAR-INFRARED
(ADAPTIVE OPTICS AUGMENTED) LARGE GROUND-BASED TELESCOPES
A summary study solicited by STScI in preparation for HST Cycle 12
Steward Observatory, University of Arizona
With contributions by and consultations with:
Eric Becklin, UCLA
Laird Close, University of Arizona
Don Figer, STScI
James Lloyd, UC, Berkeley
Bruce Macintosh, LLNL
Dean Hines, Steward Observatory
Claire Max, LLNL
Daniel Potter, University of Hawaii
Marcia Rieke, University of Arizona
Nicholas Scoville, Caltech
Rodger Thompson, University of Arizona
Alycia Weinberger, Carnegie Inst. Of Washington
Rogier Windhorst, Arizona State University
This report has been prepared at the request of STScI. It is intended to serve as a guide for HST proposers and TAC panel reviewers in evaluating the need for HST/NICMOS to accomplish the scientific goals of programs under consideration in specific observational domains.
In the years since the first incarnation of NICMOS, ground-based technical capabilities in support of challenging astronomical observations have made great strides, notably in the arena of large telescopes augmented with adaptive optics systems. Some programs, which could once only have been attempted from space, might now be possible with suitable instrumentation on such ground-based facilities. Yet, many of these systems are still in there early evolutionary stages and their theoretical potentials as applied to any particular observation is far from assured, and performance is far from repeatable given the temporally variable line-of-sight conditions encountered at ground-based observatories. The delineation between programs best suited for space and ground, and the complementary nature of others has, in some cases, become blurred. In this report, we review the capabilities of HST/NICMOS and the current state-of-the-art of operational AO systems, as applied to astronomical investigations of contemporary import, and suggest areas of suitability and uniqueness for each where applicable.
CONTENTS / EXECUTIVE SUMMARY
CONTENTS / EXECUTIVE SUMMARY
I SPACE & GROUND
I.I HST/NICMOS & ADAPTIVE OPTICS
I.2 NICMOS OVERVIEW
High-level review of NICMOS instrument concept and science goals.
I.2.1 Optical Channels
Description of 3 NICMOS cameras: (1) f/80 11"x11 43mas/pix, (2) f/45 19"x19" 76mas/pix, (3) f/17 52"x52" 200mas/pix.
Description of 256x256 pixel HgCdTe arrays, readout electronics, and performance.
I.2.3 Observing Modes
Overview of observing modes: Deep imaging, Wide-field imaging, Line imaging, High dynamic range imaging, Coronagraphy, Polarimetry, Grism spectrophotometry.
I.3 ADAPTIVE OPTICS OVERVIEW
General overview of adaptive optics.
I.3.1 Control Authority and Atmospherics
Low and high order AO systems & requirements, r0, t0 drivers for near-IR.
I.4 STREHL RATIO - A Fundamental (But Incomplete) Metric
Definition and wavelength dependence of Strehl, invariance on HST, variability (instability) on ground-based AO systems.
I.4.1 NICMOS Strehl Ratios
98% diffraction-limited Strehl in cameras 1 & 2, 84—87% in camera 3.
I.4.2 AO Strehl Ratios
Systemic dependencies. Uncontrolled (and uncontrollable) light in the PSF halo.
I.5 ADAPTIVE OPTICS LIMITATIONS
The theoretical potential of an AO system is realized for only a small fraction of targets, regions of the sky, and for only very limited periods of time.
AO images degrade (Strehl, PSF FWHM and 50%EE) off-axis from wavefront reference guide star limiting the correctable field-of-view.
I.5.2 Guide Star Brightness
AO images degrade with guide star brightness. Limits for closed-loop control: V ~ 14 for Shack-Hartmann sensors, V ~ 18 for curvature sensors.
I.5.3 Natural Guide Star (NGS) Sky Coverage
Statistically ~ 99% of the sky lacks guide stars of sufficient brightness for isoplanatic AO operability with Shack-Hartmann sensors, and > 90% with curvature sensors.
I.5.4 Laser Guide Stars (LGS)
Artificial (laser) guide stars systems are not yet mature (or robust) but can improve sky coverage to 10— 50% (when necessary tip/tilt reference stars are available).
AO images degrade with seeing conditions. Performance scaling with seeing is discussed.
I.5.6 Zenith Distance (Airmass)
AO images degrade with increasing airmass; hour angle, so temporal dependence and declination, so spatial dependence. Performance scaling with airmass is discussed.
I.6 "TYPICAL" STREHLS - The Fallacy of Simplification
AO performance is often characterized by Strehl, but characterizing Strehl is complex and "typical" performance metrics are often not realized. An illustrative study on the variability of Strehl with NGS brightness and target declination is discussed.
I.6.1 AO System Descriptions and Performance Characteristics
Current, advertised as typical, AO performance metrics for CFHT, Keck II, Gemini North, La Sillia, Palomar, Lick, Subaru and VLT UT4 AO systems with facility instruments are summarized.
II SCIENCE CASES AND CONSIDERATIONS
We discuss NICMOS and AO applicability for representative, but diverse, astronomical investigations, and identify the observational domains of uniqueness overlap, and complementary science.
II.1 Sky-Limited Observations: Background vs. Aperture
NICMOS is significantly more sensitive than ground-based telescopes at < 1.8 m due to very low sky backgrounds (several hundred times lower in H-band). At K-band large AO telescopes have point-source sensitivities competitive with NICMOS.
II.1.1 Faint Galaxies
NICMOS is advantageous because of very low J—H band backgrounds, sharp, repeatable and near perfect Strehl with 110—160 mas spatial resolution, spatial and temporal invariance of its PSF over large (52"x52" FOV), and very large dynamic sampling range.
II.1.1.1 A “Historical Note” of Significance
The importance of very high Strehls and dynamic range is reviewed in the context of WF/PC-1 faint galaxy science with the aberrated HST.
II.1.2 Deep Imaging
NICMOS imaging in the Hubble Deep Field reached 1 per-pixel noise levels of AB mag ~ 31.8 in F110W and F160W, in 36.5 h of integration time, with reliable detections of galaxies as faint as AB mag = 30.0 and were not yet approaching the systematic sensitivity limits. Such performance is unmatched from the ground.
II.2 Crowded Field Photometry/Astrometry
We discuss the relative merits of AO and NICMOS observations in optically obscured, but near-IR crowded fields in subsections II.2.1 and II.2.2.
II.2.1 The Arches Cluster
Studies of the mass-function of the Arches Cluster with Gemini/Hokupa'a and NICMOS from H and K band photometry, with comparable spatial resolutions, are reviewed. The Gemini data suffer from incompleteness and probe the mass function to a lower limit of ~ 13 solar masses, compared to 2 solar masses for NICMOS.
II.2.2 The Galactic Center
The highly stable NICMOS PSF enables very accurate multi-color photometry and searches for stellar variability in the crowded region such as the galactic center, and (by example) set an upper limit to the dereddened flux from SgrA*. Paschen imaging (not possible from the ground) is presented. VLT/NAOS(CONICA) imaging with a currently unique near-IR WFS produced Ks-band images possibly astrometrically superior to NICMOS and significantly improved upon earlier ground-based position measures of the galactic center stars.
II.3 CORONAGRAPHY OF CIRCUMSTELLAR ENVIRONMENTS
NICMOS PSF-subtracted coronagraphy achieves per-pixel H-band background rejection from the PSF wings of an occulted target of 107 at r = 1". We discuss the utility of this observing mode for imaging circumstellar debris disks and substellar companions in sections II.3.1 — II.3.3.
II.3.1 Debris Disk Imaging
Spatially resolved imaging of debris disks from the ground is extremely difficult, and in general, AO augmentation provides little or only modest benefits. We illustrate this with comparative ground- and NICMOS imaging of debris disks (sections II.3.1.1, HD 141569A; II.3.1.2, HR 4796A; II.3.1.3, Tw Hya) which were first imaged by, and not particularly challenging for, NICMOS. In section II.3.1.4 we discuss the problem of "false positive" disk detections which have arisen from ground-based imaging.
II.3.2 Polarimetry & High Contrast Imaging: Nulling Polarimetry
Nulling polarimetry on ground-based AO systems provides an avenue for imaging the polarized light component of circumstellar disks. We discuss this method comparatively with NICMOS coronagraphy and Gemini/Hokupa'a nulling polarimetric imaging of the GM Aurigae disk. We also comment on the uniqueness of NICMOS high spatial resolution polarimetry over a wide (19" x 19") fields with embedded targets.
II.3.3 Imaging Brown-Dwarf and Young Jovian Planet Companions
We present results from substellar companion imaging programs with NICMOS coronagraphy and the Keck and Gemini AO systems. On average NICMOS is at least 2 to 5 times more sensitive than Keck in the regions from 0.3" to 1.5" from the star, and delivers highly repeatable performance over the entire sky. In addition, direct comparison of TWA6"B" (H = 13.2 at r = 2.5") observations (section II.3.3.1) shows much higher detection sensitivity of NICMOS relative to Keck AO. We also discuss NICMOS non-coronagraphic companion imaging for separations < 0.3" and complementary ground-based AO imaging in this spatial domain.
II.4 High Resolution Broad Band and Line Imaging
For programs where spatial resolution is the primary scientific requirement, large telescopes with AO systems can outperform NICMOS at comparable wavelengths. We illustrate broadband and neutral hydrogen line imaging of the ultraluminous IR galaxy NGC 6240 ,with NICMOS and the Keck II AO system.
Notes To Readers:
1) This report may be downloaded in PDF, PostScript, or MS Word formats from:
2) Many of the figures (images and illustrations) in this document are intended for color display and printing. Hard-copies of these figures, if reproduced in black and white (gray scale), may lose their intended conveyance of information.
The Hubble Space Telescope provides a unique enabling platform for near-infrared astronomy, which has been exploited by NICMOS, opening avenues for scientific investigations with high efficiency that remain unmatched by ground-based facilities in many observational domains. Today’s large (10 meter class) ground-based telescopes deliver raw light-gathering power exceeding HST's by a factor of twenty, and with augmented technologies such as adaptive optics even smaller telescopes (3-5 meter class) can, under the right conditions, yield superior spatial resolutions. Such facilities are beginning to challenge the observational uniqueness space, which once was wholly in the domain of a space-based instrumental platform (i.e. HST).
Despite the great advances in recent technological achievements implemented at many ground-based facilities, the stability of the HST platform and its atmosphere-free environment giving it unfettered diffraction-limited access over the entire sky and across all accessible wavelengths, continue to make it the facility of choice, and indeed necessity, for many scientific investigations. Understanding when proposed ground-based programs may succeed using new observing technologies, and when would likely fall-short of (if not fail to) achieve their intended goals has become an increasingly complex task. Here we attempt to lend some clarity to this issue by reviewing the capabilities repeatably demonstrated by HST + NICMOS in comparison to those claimed, but achieved with varying degrees of success for particular applications, with the current generation of ground-based facilities.
This report is organized into two principle sections. In section I we provide a brief overview of the NICMOS instrument, its observing modes, and its technical capabilities as integrated with HST and as it is currently being operated by the NICMOS Cooling System. We then provide an overview of Adaptive Optics (AO) technologies as applied to large ground-based telescopes. AO on large-aperture telescopes has been seen as the pathway to enabling observations that at one time could only be considered from space. In doing so we discuss the principle metrics by which AO systems are often evaluated, and the merits of those metrics (which often are inherently unstable) for comparing anticipated performance for actual applications and in light of what has been achieved by HST+NICMOS. In section II we examine results from programs with similar scientific goals and with similar (sometimes identical) targets, as undertaken by HST NICMOS and with large ground-based telescope systems. Here we have drawn demonstrable results from the published literature, which are particularly germane in assessing comparative performance in a diversity of observational domains.
I - SPACE & GROUND
I.1 – HST/NICMOS & ADAPTIVE OPTICS
In this section we provide overviews of NICMOS, adaptive optics, fundamental metrics for intercomparison, environmentally and instrumentally imposed limitations on AO systems, and a summary of performance characteristics for currently (or recently) operating AO augmented telescopes and instruments.
I.2 - NICMOS OVERVIEW
After a more than three-year hiatus the Near-Infrared Camera and Multi-Object Spectrometer (NICMOS) was returned to full operational status in HST Cycle 11. The three cameras in the NICMOS instrument provide near-infrared (NIR) imagery with spatial resolutions from ~ 0.1" to 0.25" in spatially adjacent (but non-contiguous) narrow and wide fields-of-view. Working in space in the NIR (from 0.8 – 2.4 m), NICMOS peers through dusty regions obscured at optical wavelengths that are otherwise compromised by the intrinsic brightness, wavelength-dependent opacity, and turbidity of the Earth's atmosphere. The diversity of instrumental observing modes, including deep-imaging, wide-field imaging, coronagraphy, polarimetry and slitless (grism) spectrophotometry, have allowed NICMOS to address a wide range of astrophysical investigations from within our own solar system to the most distant reaches of the currently observable universe. Emissions from young objects embedded in optically-opaque star and planet forming regions may be studied by NICMOS individually in our own galaxy and collectively in others, and light from the oldest observable galaxies is red-shifted from the rest-frame optical into the NICMOS pass bands.
Originally installed in February 1997, NICMOS was used heavily by astronomers in HST Cycle 7/7N. These capabilities ceased on January 4, 1999 when its supply of solid Nitrogen cryogen, used to cool the instrument detectors and filters, was exhausted. During HST servicing mission 3B (March, 2002) a pathfinder “experiment” in space/cryogenic technology culminated in the subsequent re-cooling and resurrection of NICMOS. With the installation of a reverse Brayton-cycle micro-turbine cooler, heat-exchanger, and external radiator on HST, cryogenic Neon-gas circulating through the NICMOS cooling coils is now maintaining the NICMOS detectors and cold-optics at nearly optimum and highly stable temperatures. NICMOS became the first space cryogenic instrument to ever have been re-activated after cryogen depletion. The success of SM3B has literally moved space infrared (IR) astronomy out of the ice age.
NICMOS extends the Hubble Space Telescope's UV/Optical panchromatic vision into the NIR fulfilling a Level-1 requirement of the HST mission. With NIR light we can see into regions, which are obscured at visible and ultra-violet (UV) wavelengths. Dust, which is pervasive and prevalent in many forms throughout the universe, obscures many of the most interesting objects in optical/UV light and limits our ability to understand the dynamical, evolutionary, and energetic processes in many others. NIR light can penetrate this dust to reveal the birthplaces of planets, stars, and galaxies, as well as the centers of the most powerful galaxies in the universe.
We can also see further back into the ancient history of the universe with NIR light than in the visible and UV. With the global expansion of the universe, electromagnetic radiation is redshifted to longer wavelengths. The intrinsically more distant an object is, the greater the redshift, such that the most distant (currently) observable objects - those with the greatest redshifts - have their optical/UV light redshifted into the infrared. Distant quasars and galaxies provide, in the NIR, a look-back in time to the early epochs of the assemblies of large structures in the universe.
I.2.1 - Optical Channels
NICMOS contains three independently re-imaged optical channels presenting different image scales and field sizes to its NIR detectors over the 0.8–2.4 m wavelength regime. All three cameras carry nineteen spectral elements, each with sixteen wide, medium-band, and atomic/molecular line filters (~25%, 10%, and 1% pass bands), with a core set of “standard” photometric filters in each camera. Camera 1 provides high-resolution imagery, diffraction limited at and longward of 1m, over an 11" x 11" field-of-view. This f/80 camera is sampled by its focal plane detector with 0.043" pixels. Camera 1 is equipped for polarimetric imaging at 0.8–1.1 m with < 1% instrumental polarization. Camera 2 provides a wider field (19.3" x 19.4") at f/45, and is diffraction limited to a wavelength of 1.75 m as sampled by 0.076" wide pixels in its image plane. Camera 2 contains coronagraphic optics for high contrast imaging in the regions around point sources, optimized for H-band (1.6 m) and shorter wavelengths. Instrumentally diffracted and scattered light in the wings of the PSF of coronagraphically occulted targets is reduced significantly when a bright object is placed in the 0.3" radius coronagraphic occulting spot. Imaging polarimetry at wavelengths longer than in camera 1 (1.9 – 2.1 m) is available in camera 2 with similar low instrumental polarization as in camera 1. Camera 3 enables wide field imaging at “short” wavelengths (< 1.9 m), with a 52" x 52" field-of-view in its f/17 optical channel. Although spatially undersampled with the filters provided, camera 3 is well suited for deep imaging. Camera 3 also provides three grisms for slitless spectrophotometry, with spectral resolutions of ~ 100, covering the full wavelength range of sensitivity of the NICMOS detectors. The fields-or-view of the three cameras are nearly adjacent on the sky, but not spatially contiguous. Simultaneous imaging with all three cameras is possible, but camera 3 (the “wide field” imager) in not simultaneously confocal with cameras 1 and 2.
I.2.2 - Detectors
NICMOS, the instrument, employs three NICMOS-3/PACE technology detectors, which were manufactured for space-qualified use by the Rockwell Electro-Optics Center under the direction of the University of Arizona's Steward Observatory. The three 256x256 pixel HgCdTe photodiode arrays (one at the re-imaged focal plane of each camera) are indium-bump bonded to a sapphire substrate and addressed through a silicon multiplexor. Each 128 x 128 pixel quadrant is read out through independent on-chip amplifiers (located at the four corners of the focal plane array), and sampled with 16-bit analog-to-digital converters. The NICMOS detectors can be read-out non-destructively while integrating from 0.2 to several thousand seconds, providing a sampling (data quantization) dynamic range of 22 stellar magnitudes in a single exposure. The per-pixel detector read-noise is typically ~ 22 electrons when read-out in this non-destructive “multiaccum” mode. A “bright object mode” supports per-pixel integrations as short as a millisecond. Dark currents are ~ 0.1 e-1 s-1 pixel-1 with the arrays at a temperature of 61K (where they were operated on-orbit from 1997 – 1998 with a solid nitrogen (SN2) coolant) and ~ 0.2 – 0.3 e-1 s-1 pixel-1 at 77K (where they are currently being operated, cooled by recirculating cold neon gas). The arrays exhibit wavelength-dependent detective quantum efficiencies (dQE) of ~ 30 – 80%, with dQE increasing at longer wavelengths where the inter-pixel dispersion in dQE is greatly reduced. The devices reach hard saturation with well-depths of ~ 200,000 electrons, and respond with a very high degree of linearity to ~ 75% of full well depth.
I.2.3 - Observing Modes
NICMOS opens up many other avenues of scientific investigation through its diversity of observing modes, capabilities and, in cameras 1 and 2, diffraction-limited imaging, such as:
• Deep imaging with very low instrumental plus dark sky backgrounds unmatched from the ground (e.g. 0.024 e-1 s-1 pixel-1 in camera 1's 43 mas x 43 mas pixels).
• Wide-field imaging with camera 3's 52"x52" field-of-view and field-independent point spread function, which is a pathway to uniformly complete and deep imaging surveys.
• Molecular and/or atomic line imaging in bands significantly affected by, or unavailable because of, atmospheric telluric absorption (e.g. Paschen- at 1.87 m, a critical tracer of star formation in active galaxies, and typically ~ 30x more efficient than H- in the optical).
• High dynamic range imaging (e.g. for young exoplanet and circumstellar disk imaging) which, when coupled with Point-Spread-Function (PSF) subtracted coronagraphy, yields per-pixel H-band background rejections in the unocculted wings of the PSF of an occulted point-target of
10-7 that of the flux density of the target at a radial distance of 1".
• Polarimetric imaging at high spatial resolution (0.1", critically sampled in camera 1, and 0.2" over a 19.3" x 19.4" field in camera 2), and very low (< 1%) instrumental polarization, which helps elucidate the nature of circumstellar and interstellar grains and their interactions with their environments.
• Full-field low-spectral resolution (R ~ 100) grism (slitless) spectrophotometry at all wavelengths accessible to NICMOS (0.8 – 2.4 m in three grisms).
NICMOS has no slit or high spectral resolution spectrophotometric capabilities of any kind. Observing programs requiring near-IR high-resolution spectroscopy should look to ground-based facilities and instrumentation (e.g. Keck II/NIRSPEC, VLT/NAOS+CONICA).