UC Observatories is the lead institution for TMT’s Wide Field Optical Spectrograph (WFOS). Complementing the near-IR AO capabilities of IRIS, WFOS will be TMT’s workhorse instrument in dark time and natural seeing. The WFOS team has been making rapid progress on some exciting new design options for WFOS. We are studying two different WFOS instrument concepts and aiming to make a down-select in March 2018. The “Slicer-WFOS” concept is a monolithic slit-mask spectrograph (think DEIMOS or LRIS) that provides similar capabilities as the previous MOBIE design but in a different way that mitigates many of the challenges we faced before. The “Fiber-WFOS” concept introduces a radically different design that uses optical fibers to increase the number of simultaneous targets that could be observed to 700 (compared to 25-100) over a larger field of view. Fiber-WFOS would also provide resolved spectroscopy by deploying ~70 IFUs over a central area that could be corrected to 0.3’’ FWHM by ground-layer adaptive optics (GLAO).

What is WFOS for?

WFOS is primarily a survey instrument but will also provide rapid followup of transient sources. It offers R~5000 spectral resolution from 310-1000nm (an R~1500 option with greater multiplex is an option for the monolithic concepts) and is optimized for the faintest sources accessible by TMT in natural seeing. The science case includes IGM tomography and high-z galaxy studies at z=2-5, stellar observations as probes of M31 and other local group galaxy assembly histories, resolved (GLAO-enabled) spectroscopy of galaxies at z~1, and transient followup.

What happened to MOBIE?

For many years, UCO was actively pursuing a WFOS concept known as MOBIE (PI: Rebecca Bernstein). Starting in 2013, a number of detailed studies and improvements were made under leadership from Sandy Faber, Nick Konidaris, Chuck Steidel, Matt Radovan, and Luc Simard. These efforts, combined with further analysis in 2016, enabled a comprehensive review of the MOBIE-inspired architecture in May 2017. A number of risks, challenges, and limitations became increasingly clear. Most of these center around the way in which these designs dispersed light and from the desire to achieve greater spectral resolution through cross dispersion. The Fiber-WFOS and Slicer-WFOS concepts emerged as exciting alternatives that could achieve these goals in different ways.

Did you say optical fibers?

Fibers have often been avoided as an option for instruments on large telescopes out of fear of poor throughput and systematics, but an improved understanding based on existing fiber spectrographs and technological progress have dramatically changed this picture in recent years. Subaru’s fiber-based Prime Focus Spectrograph will be commissioned next year and various fiber designs for 10m-class and larger telescopes are being pursued by ESO, CFHT, and GMT. With MaNGA on the Sloan Telescope, we recently achieved a surface brightness sensitivity of 27.6 AB / arcsec^2 using ~14 hour fiber integrations (


A fiber-based WFOS concept is new to TMT and is therefore a major part of our design work in this phase, especially here at UCO. I attach a diagram showing the schematic layout. Roughly 700 robotic positioners would be mounted to a rotating focal plate at the TMT Nasmyth focus, spanning a 10 arcmin diameter field. Each actuator would position a fiber bundle (19-fibers each) on desired targets. Having multiple fibers in each bundle is a necessity given TMT’s fine plate scale (2.1mm per arcsec) but also offers advantages in optimal extraction of source flux. The roughly 10,000 fibers emerging from the backend would be injected into ~10 mounted spectrographs, each with 3 or 4 channels providing full wavelength coverage at R~5000.

In such a design, the cost of the fiber and associated bundles is a small fraction of the spectrograph and robotic positioner costs. This makes it possible to deploy multiple sets of fibers with different configurations at the focal plane at little extra cost.

In Fiber-WFOS, we would deploy a second set of fibers arranged in ~70 IFUs, each with 127 fibers and spanning 1-2’’ on the sky, but configured to spatially sample scales of 0.15’’, appropriate for resolved spectroscopy across GLAO’s 0.3’’ FWHM-corrected field.


The Slicer-WFOS concept provides capabilities more similar to the MOBIE-based designs, but instead of achieving R~5000 with a cross-dispersing prism, Slicer-WFOS would do so by utilizing narrower effective slits. This monolithic spectrograph would have a single VPH grating that would deliver R~1600 for 0.75’’ wide slits and R~5000 for 0.25’’ slits.

To gain back the slit aperture losses with 0.25’’ wide slits, one could deploy ~25 "slit slicers” across the field. Each would slice a 0.75’’ wide slit into three slitlets. Using a series of small mirrors, the upper and lower slitlets get re-imaged to either side of the central slit. From the point of view of the spectrograph, the flux incident on one 0.75’’ wide slit appears as three adjacent 0.25’’ slits, all of which could be combined later.

As with MOBIE, the observer could trade greater spectral resolution for lower multiplex. At R~1600, it would be possible to observe ~100 objects using a standard slit mask. At R~5000, ~25 objects could be observed in slicer modules positioned across the field.

The WFOS Team

We have a growing international team working on WFOS. Our partner institutions include Caltech, the Indian Institute for Astrophysics (IIA), the Nanjing institute of Astronomical Optics & Technology (NIAOT), and the National Astronomical Observatory of Japan (NAOJ).

At UCO, Kevin Bundy is serving as the Principal Investigator with Maureen Savage as Project Manager, Matt Radovan as Lead Engineer, Reni Kupke as Senior Optical Designer, and Nick MacDonald as Senior Engineer. A number of other researchers, engineers, and UCSC graduate students are contributing to the effort, including Drew Phillips, Kyle Westfall, Brian DiGirogio, Namrata Roy, Dave Cowley, and Jerry Cabak."

At Caltech, Chuck Steidel serves as the WFOS Project Scientist and Jason Fucik as Optical Designer. Work at IIA is led by T. Sivarani, with contributions from S. Sriram, Devika Divakar, and Arun Surya. Hangxin Ji and Zhongwen Hu are participating from NIAOT, and Satoshi Miyazaki and Shinobu Ozaki are leading efforts at NAOJ.