Wavefront Sensing and Control on JWST During CommissioningM. Meléndez (melendez[at]stsci.edu)
"It is difficult to say what is impossible, for the dream of yesterday is the hope of today and the reality of tomorrow." ― Robert H. Goddard
NASA's Webb will be the first space observatory with an actively controlled segmented primary mirror. The Wavefront Sensing and Control (WFS&C) working group, a collaboration between Ball Aerospace and Technologies Corp, STScI, and NASA Goddard scientists and engineers, will have the responsibility to perform the precise optical alignment of the telescope. In the WFS&C office at the institute, we have been actively rehearsing, practicing, writing, and developing the procedures, algorithms, and tools necessary to carry out the processes required to go from a misaligned telescope after deployment to a fully phased single 6.5-meter telescope. This article summarizes the wavefront activities and how we are preparing for OTE commissioning including the milestones successfully reached thus far and the upcoming activities before the launch of Webb in early 2021.
Wavefront Sensing and Control on JWST
Imagine having access to an instrument so advanced and sensitive that you could detect the heat signature of a bumblebee at the distance to the Moon (Dr. John Mather, https://jwst.nasa.gov/faq_tweetchat2.html). A golden time machine steadily floating in space 1.5 million kilometers away from Earth with a shiny sunshield, undisturbed in the cold temperatures of space. Certainly, this would have been an impossible dream for Galileo, but in a couple of years it will be our reality. NASA's James Webb Space Telescope will be our unobscured window into the history of our Universe—from detecting the signature of the first galaxies that illuminated the nascent Universe to the formation of solar systems capable of supporting life on planets like Earth. In this unimaginable cosmic puzzle, the Space Telescope Science Institute (STScI) in Baltimore, MD, will be home to the flight operations and science center, crucial for the success of one of NASA's most ambitious space missions.
Webb is an infrared telescope providing both imaging and spectroscopy in the wavelength range from 0.6 to 28.5 μm with a 6.5-meter primary mirror (PM) composed of 18 hexagonal-shaped segments (about 1.32 meter each). A convex secondary mirror (SM) steers light from the PM into a Cassegrain focus before the light is collected and focused by a fixed concave tertiary mirror and directed to the instruments by a flat Fine Steering Mirror. These mirrors and the supporting structure and subsystems comprise the Optical Telescope Element (OTE), gathering light from space and feeding it into the science instruments. At the heart of Webb is the Integrated Science Instrument Module (ISIM) which holds the main science payload containing four science instruments as well as the fine guidance sensor. The Near-Infrared Camera (NIRCam), Near-Infrared Spectrograph (NIRSpec), Mid-Infrared Instrument (MIRI) and the Fine Guidance Sensor/Near InfraRed Imager and Slitless Spectrograph (NIRISS). The specifications and science goals for Webb and each of the instruments are explained in great detail elsewhere, for example, check JWST Observatory and Instrumentation Documentation and JWST Observer website.
NASA's Webb will be the first space observatory with an actively controlled segmented primary mirror. To achieve this, the development, testing, and implementation of a complex set of technologies and algorithms were required. All this complexity, wrapped up with a great deal of ingenuity, will be crucial to ensure that we will see the image produced by a large 6.5-meter mirror and not 18 bumblebees randomly floating in space. The Wavefront Sensing and Control (WFS&C) working group, a collaboration between Ball Aerospace and Technologies Corp, STScI, and NASA Goddard scientists and engineers, will have the responsibility to perform the precise optical alignment of the telescope, so the wavefront from each of the 18 individual segments coherently add up in order to form a diffraction-limited image of a full 6.5-meter mirror telescope. In the WFS&C office at the institute, we have been actively rehearsing, practicing, writing, and developing the procedures, algorithms, and tools necessary to carry out the processes required to go from a misaligned telescope after deployment to a fully phased single 6.5-meter telescope ready to unveil some of the mysteries of the Universe.
Wavefront Activities During Commissioning
Roughly 36 days after launch, deployment, and with temperatures cool enough for the instruments to start taking the first images, the WFS&C commissioning activities begin. Detailed descriptions of the commissioning activities are described in various articles, e.g., Acton et al. 2004; Perrin et al. 2016. One way to think about commissioning is to split it into three different processes (Acton et al. 2012): PM segment location and positioning, segment-level wavefront control, and global phasing of all the controllable mirror segments.
The first process involves a tense game of hide and seek, trying to find each of the primary mirror segments as they are misaligned after initial deployment. In addition, this first process will yield a better determination of where the telescope is pointing (i.e., its boresight). For this, we need to be clever about systematically pointing the telescope until we find all the 18 segments in a series of mosaic images covering a large area surrounding a bright and isolated star. Once we find all the 18 bumblebees, we need to identify them, initiating a well-coordinated ballet as each of the mirrors are commanded, in turn, to perform a pirouette. At the end of this choreography, we arrange the segments into an array on a single NIRCam detector. This will allow us to better quantify the alignment of the telescope and the segment-level wavefront errors in the ensuing measurements.
During the segment-level wavefront control process, we try to improve the wavefront error for each segment by adjusting a total of 131 degrees of freedom among the primary mirror segments and the secondary mirror (113 unique degrees of adjustment). During this stage, we start with the first set of Global Alignment (GA) corrections. During GA, the SM is moved in piston to create defocused images. Using phase retrieval algorithms these images can be analyzed in order to reconstruct the phase for each individual segment. GA is an iterative process in which at each step the leading terms of wavefront error, e.g., power and astigmatism, are corrected locally by moving the PM segments and globally, by moving the SM. After GA, we stack the segments in preparation for the initial iteration of Coarse Phasing. To achieve sub-pixel stacking, caution must be taken when moving the mirrors in order to avoid large uncertainties associated with the coarse mechanism of the segment actuators. In Coarse Phasing, large piston errors are corrected by using the Dispersed Hartmann Sensors (DHS) in NIRCam, which enable measurement of the piston difference between 20 pairs of segments. In practical terms, piston errors translate into a fringe pattern in the DHS spectra; by analyzing these fringes, relative piston offsets can be determined accurately (for piston differences smaller than about 350 microns).
After global alignment and coarse phasing, the main cause of misalignment in the telescope would be an incorrect position of the secondary mirror. To address this issue, we need to determine the WF performance across the field of view (FOV). We begin this process using only the NIRCam FOV, with a program dubbed "Coarse MIMF." During Coarse MIMF we can determine the SM position error by measuring the relative centroid positions of an unstacked PM configuration (small array) moving around to different field locations on a single NIRCam detector. Following Coarse MIMF, another step of Global Alignment is executed to correct any residual wavefront errors without any further adjustment of the secondary mirror.
Finally, towards the end of commissioning, we start the global phasing steps in which we execute an iterative process, a loop of coarse and fine phasing corrections resulting in a wavefront error of less than 50 nanometers. During fine phasing, phase retrieval algorithms are applied to weak lens images from NIRCam (internal defocus images) permitting reconstruction of the phase of the PM and correction for small wavefront errors between PM segments. At the end of the WFS&C commissioning process there is the multi-instrument and multi-field component (MIMF) where corrections to ensure low wavefront error over the entire field of view are calculated and implemented. In order to determine these errors, phase retrieval algorithms are applied to a set of defocus images (about +/- 100 microns of SM piston) collected at five different field points in each of the science instruments. After MIMF, we expect to have reduced the wavefront error of a large 6.5-meter telescope into the range of tens of nanometers.
How are we preparing for OTE Commissioning?
For the past couple of years, we have been participating in various successful rehearsals and team practice exercises involving large collaborations between the different subsystems that comprise the SOC. To first order, the objectives of these rehearsals are to evaluate the ability of the SOC personnel to sustain operation over a period of time and to exercise the various channels of communication, team interactions, and coordination between the different subsystems and operators.
In March 2018, we participated in Wavefront Rehearsal #1 where we used accurate optical simulations to practice our procedures, tools and analyses for some of the OTE commissioning activities, namely, identifying the primary mirror segments (Segment ID), applying the first set of segment-level corrections to reduce the wavefront errors of the telescope (Global Alignment #1), and correcting for large piston errors after deployment (Coarse Phasing #1). In October 2018, we participated in Wavefront Rehearsal #2, where we incorporated lessons learned from the previous exercises and practiced the process with which we will determine the WF performance across the NIRCam field of view (multi-field corrections, Coarse MIMF).
In 2019, we'll follow a WF rehearsal schedule that includes a third WF rehearsal in April followed by a fourth (and possibly last) rehearsal towards the end of the year, around October 2019. In our last rehearsal, we are planning to practice the final step on the commissioning, i.e., applying MIMF corrections. In addition to these rehearsals, the WFS&C group will also participate in the Launch Readiness Exercises sometime in 2020.
Beside these full-scale rehearsals, involving various SOC subsystems, the WF team has been busy organizing team practice activities, scheduled approximately every six months in between rehearsals. The last two team activities took place in June 2018 and January 2019 where, using simulated images, we exercised our procedures, analyses, and internal communications between the different team members operating inside the WFS&C office. In a couple of years, we won't need to imagine such an amazing and ambitious telescope, it will be our reality. Until then, we need to practice and prepare for the unexpected until we are ready to discover the Universe with the reality of tomorrow.
Acton, D. S., Atcheson, P. D., Cermak, M., et al. 2004, "James Webb Space Telescope Wavefront Sensing and Control Algorithms." In Mather, J. #C., editor. Optical, Infrared, and Millimeter Space Telescopes (Proc SPIE; Vol. 5487); Oct.; 2004. p. #887–896. Acton et al. 2004, SPIE 5487, 887.
Acton, D. S., Knight, J. S., Contos, A., et al. 2012, "Wavefront Sensing and Controls for the James Webb Space Telescope." In: Space Telescopes and Instrumentation 2012: Optical, Infrared, and Millimeter Wave (Proc SPIE; Vol. 8442); Sep.; 2012. p. #84422H. Acton et al. 2012, SPIE 8442, 84422H.
Perrin, M. D., Acton, D. S., Lajoie, C. P., et al. 2016, "Preparing for JWST Wavefront Sensing and Control Operations." In: Space Telescopes and Instrumentation 2016: Optical, Infrared, and Millimeter Wave (Proc SPIE; Vol. 9904); Jul.; 2016. p. #99040F. Perrin et al. 2016, SPIE 9904, 99040F.