JWST: The Early Years


Read Garth Illingworth's reports around the December 25, 2021 JWST Launch. These posts where distributed as updates to the astronomy community.


JWST Presentation

Garth Illingworth's talk "The Next Generation Space Telescope – NGST – Conceptualizing what comes beyond Hubble, before Hubble!".

NGST: The Early Days of JWST

A STScI Newsletter article by Garth Illingworth on the early development of JWST in the 1985-1991 period, plus a short set of slides that summarize the early developments.
The Next Generation Space Telescope

JWST Presentation

Garth Illingworth's talk at the Harvard's CfA ITC "Thoughts on Flagship Missions: JWST & the Implications for Getting What's Next(?)".

JWST Project History

A timeline of the concept and development of JWST from 1989 to 2010. The timeline from 1989 to 2001 is reproduced below.

JWST Budget Troubles

In 2011 the US House of Representatives' appropriations committee on Commerce, Justice, and Science moved to cancel the JWST project.

JWST: Early Years

30th Anniversary of the First Big Science and Technology Workshop on the Next Generation Space Telescope (NGST)

A note from Garth Illingworth re that first NGST (now JWST) workshop.

September 2019 marked the 30th anniversary of the first major science and technology conference on the Next Generation Space Telescope NGST. The NGST workshop was held during the week of 13-15th September 1989. The concept for NGST was driven by the recognition that a large passively-cooled telescope in high-earth orbit with IR capability would hugely complement and extend the capabilities of Hubble. Riccardo Giacconi, then director of the Space Telescope Science Institute (STScI), pushed several of us (Peter Stockman, Pierre Bely and me) in the 1987 timeframe to develop a new concept since “it takes a long time to do these large missions”. Very true! It has been a long haul, but JWST will truly be worth it when it flies in 2021. Note that this 1989 workshop meeting preceded the launch of Hubble. The workshop was also sponsored by NASA (by Ed Weiler) as well as being supported by AURA/STScI.

I was fascinated to find that the Proceedings from the NGST workshop meeting have been archived:


And amazingly this is also available on Amazon:


It is remarkable how long it can take to bring something like this to fruition. Following the 1989 workshop, NGST was the focus of some studies organized by JPL with NASA headquarters funding (through Mike Kaplan and AstroTech21), as well as being discussed in the 1990 Astronomy Decadal. The panel I chaired in that Decadal recommended a 6 meter passively-cooled UV/Optical/Infrared telescope in high earth orbit for launch in 2009 at a cost in 1990 dollars of $2 billion. A bit low but not a bad estimate at such an early stage (corresponding to about $4B in current dollars). NGST did not make it through the Decadal that time since there was a focus on getting Spitzer launched. But it was a good start to have that discussed in the 1990 Decadal, and then to have NGST recommended in the 2000 Decadal.

It is certainly interesting to contrast the 1989 concept with the reality of 2019. The core features of the concept remain but the implementation is very different. I saw JWST with sunshield layer 5 tensioned a few weeks ago and it was incredibly impressive. The telescope size is in the ballpark we wanted in the late 80’s early 90’s (6-10 meters), and it is passively-cooled like we wanted, but implementing that concept has been an enormous technical challenge. But that is how it should be. We carry out many important scientific missions of smaller scale, but our Flagship missions should push us into new technological regimes. If a Flagship turns out to be easy, we just have not been ambitious enough. And for the most technologically-advanced nation on Earth, such ambitious missions should always be on our plate. No other nation on Earth can do such a mission. Exercising leadership and enabling science beyond what can be done elsewhere is what we should be doing. And when we include international partners, as JWST has with the Europeans and Canadians, we make such missions a human endeavor that is unifying. JWST is just such an advanced, cutting-edge and incredibly impressive international scientific capability!

You can read more about the early days of JWST here.

— Garth Illingworth

James Webb Space Telescope Fully Assembled

After successfully assembling the entire observatory, technicians and engineers moved on to fully deploy and tension all five layers of its tennis court sized sunshield, which is designed to keep its optics and sensors in the shade and away from interference.

NGST: The Early Years

Prior to September 10, 2002, the JWST was known as the Next Generation Space Telescope (NGST). In the section below we will reference the name of the observatory that was in use at the time of a given milestone.

The links below provide the history of the conception and development of JWST thus far. These pages are partly based on a presentation given by Peter Stockman (JWST/STScI Project Scientist) at the 2001 Hubble Fellows Symposium.

The Early Years


Marshall Space Flight Center (MSFC, the NASA center responsible for the Hubble satellite) studied the possibility of launching an 8-m telescope in the external tank of the space shuttle (SPIE, volume 228).


Perkin-Elmer, the company that manufactured the Hubble mirror, described a variety of advanced concepts of large space telescopes in an Astronomical Interferometric Systems Technology requirements study (ER 991, 1986).


The Space Science Board recommended that a large area, high-throughput UV-IR telescope (8-16m) be developed as Hubble’s successor.


In 1989 Riccardo Giacconi (STScI Director) was concerned that Hubble was to be deployed in 1990 and that its successor was not on NASA’s radar screen (large space science projects often take about 20 years from first development to launch). He suggested that a workshop should be held about a followup to HST. The Next Generation Space Telescope Workshop, organized by Garth Illingworth and Pierre Bely was hosted by STScI with support from NASA/Goddard. It focused on the science drivers and technical capabilities of an HST follow-up telescope at the end of the HST lifetime, which was then estimated to be 2005. An obvious science driver was studying galaxies at high redshift, which was at that time a redshift of z ~ 1. At the end of the workshop, it was proposed that NASA should investigate the feasibility of an 8m passively cooled near-IR telescope, in a high earth orbit, or a similar telescope based on the moon.


The discovery of the spherical aberration problem of the Hubble Space Telescope's primary mirror slowed the development of the NGST. While all effort was put into fixing HST and working around the spherical aberration problem, only a few dared think about the long term future. STScI and MSFC had been working on a large (10m) telescope in space and a larger telescope (16m) on the Moon. These reports were completed in 1991 (MSFC LLT 001) and 1992 (MSFC LLT003) respectively.

The Astronomy and Astrophysics Survey Committee (NAS, 1991) recommended that NASA fund technology development for large space telescopes (> 6-m), either free-flyers or lunar based.


JPL led a series of meetings with academia and industry to identify the critical technologies for large telescopes in space. (Large Aperture Telescopes in Space, Astrotech 21, JPL 1991)


The Association of Universities for Research in Astronomy Inc. (AURA) and NASA chartered the "HST and Beyond" committee, chaired by Alan Dressler, to consider the needs of the astronomical community after the nominal mission lifetime of HST, ending in 2005. The committee came up with three recommendations:

  1. Extend the lifetime of HST to 2010 to allow a overlap and smooth transition to the new telescope,
  2. Study the feasibility of a 4m space telescope in a low background orbit,
  3. Support science based on the Origins theme, i.e. investigations that seek to understand the formation processes of galaxies, stars, planets and life.

Motivation for the report's emphasis on observing early galaxies (z ~2) came from deep HST images that still clearly showed structure in cluster galaxies at a redshift of 0.4, even with spherical aberration. Furthermore, there were strong indications that galaxies were visible in these images out to higher redshifts. Astronomers had not expected this performance from the 2.4m Hubble Telescope, after their ground based observations of galaxies extended over 1 arcsec due to atmospheric seeing. Galaxies, particularly those with active star formation, turned out to be more compact and have more structure. As a result, Hubble could be used to resolve them at much higher redshifts (z ~ 5) as the Hubble Deep Fields later showed.


Based on these developments, the STScI proposed studying a passively cooled space observatory called "High-Z", which had a monolithic 4m mirror, was still fully baffled and was envisioned to be in a 1x3 AU elliptical orbit.

Going for 8 Meters


Partly in response to the Dressler Report, Joe Rothenberg (then GSFC Director) and Ed Weiler (NASA HQ) encouraged a GSFC/STScI collaboration to study a passively cooled successor to the Hubble. John Mather and Peter Stockman were the two Project Scientists and John Campbell (HST Project Manager) led the initial engineering effort. As the Project grew, Bernie Seery replaced Campbell and led the project through the early formative years.

Dan Goldin, NASA Administrator, gave a presentation pushing his "Faster, better, cheaper" motto at the San Antonio winter AAS meeting. He was disappointed by the lack of development in space telescope technology, both for astronomy as well as for the military. He urged the astronomical community to think big, challenging them to come up with an 8m telescope design at a lower cost than previous telescopes.


Combining Goldin's challenge with the realization that great science could be done at high redshifts (z=1-5 and higher), the GSFC/STScI team studied more radical telescope designs. The higher redshifts goal pushed explorations of very low infrared backgrounds, meaning observatory orbits much further away from Earth.

Portions of the very early designs from Pierre Bely and Peter Stockman are present in the final design of NGST/JWST: a large deployable mirror (then still at 8m), L2 point orbit, and an "open telescope design" (no external baffle) with passive cooling behind a large multi-layer sunshield. This concept was still envisioned with one large near-IR camera in the focal plain.

Four corporations joined the feasibility study, and came up with designs that were generally very similar:

  • NASA/Goddard with a telescope at L2, with a large sunshield and a deployable 8m mirror,
  • TRW with a telescope at L2, with a large sunshield and an articulated 8m mirror,
  • BALL Aerospace with a telescope at L2, with a large sunshield and a deployable 8m mirror, and
  • Lockheed with a monolithic 4m mirror at 3AU (which provides low background, but also low solar power and more data transmission problems).

The trade studies suggested that these designs could be developed for about $500 million, under the assumption that the whole telescope would be built by one contractor, as would the science instruments. The idea of one contractor, like many other assumptions going into the studies, turned out not to be feasible, especially for the instruments.

Based on these studies, a comprehensive report was written, titled Next Generation Space Telescope, Visiting a Time When Galaxies Were Young. This report also presented a technological roadmap for the succesful development of NGST in the next decade.

In the meantime, scientists created more detailed simulations to provide a better scientific basis for NGST and to drive its instrument capabilities. For example, Myungshim Im simulated deep images for an 8m NGST, suggesting feasibility to detect galaxies out to redshifts of 15, if they exist. These simulations highlighted the need for diffraction limited operations at 1-2 microns to a high wavelength in order to be feasible, and a strong push for mid-IR capabilities. The simulations from, for instance Massimo Stiavelli, indicated the need for Multi-Object Spectroscopic capability for NGST. Where the spectroscopic followup of deep HST imaging like the Hubble Deep Fields was possible from the ground with 8m class ground-based telescopes, this would no longer be true for the very high redshift galaxies that will be discovered by NGST in the near and mid IR. Atmospheric emission would swamp the faint signals from these galaxies; astronomers assumed that 30m ground telescopes would not be completed until after the NGST mission. If any science was to be done on very high redshift galaxies beyond morphology (shapes) and photometry (colors), NGST would have to do the spectroscopy itself.

Buoyed by all these studies, NASA agreed to fund additional studies that further refined the technical and financial requirements for building the telescope.

Reality Hits

During this period, the astronomical community at large began to join up to define more carefully and realistically the science drivers for NGST. GSFC and STScI chartered the Ad-Hoc Science Working Group (ASWG). The ASWG proposed science goals centered around five themes:

  1. Cosmology and the Structure of the Universe
  2. The Origin and Evolution of Galaxies
  3. The History of the Milky Way and Its Neighbors
  4. The Birth and Formation of Stars
  5. The Origins and Evolution of Planetary Systems

Within these themes, the ASWG developed the Design Reference Mission (DRM), a suite of 21 hypothetical key science observing programs for NGST, to which each design of NGST could be tested. The ASWG furthermore came to an agreement as to what the core instrument package of NGST should be in order to maximally cover the science proposed in the DRM:

  • a large field of view near-IR camera,
  • a multi object near-IR spectrograph with at least 100 elements,
  • a general purpose mid-IR camera/spectrograph.

The Ad Hoc Science Working Groups was a key player in cementing the collaboration between NASA, the European Space Agency and the Canadian Space Agency.

In the same 1997-2000 period reality started to hit. Initial technology development was started, with a focus on:

  • lightweight mirrors,
  • wavefront sensing and control,
  • detectors, and
  • cryogenic actuators.

GSFC and aerospace companies performed trade studies, looking at a configuration with:

  • an 8m diameter, deployable mirror,
  • an Atlas V Launcher,
  • a satellite in L2 orbit, and
  • a large deployable sunshield.

August 2000

NASA designated the STScI to be the Science and Operations Center for NGST. The GSFC/STScI collaboration, the success of the Hubble operations team, and the potential for reuse of Hubble software formed the basis for NASA’s selection.

At the end of 2000 the NGST project performed a careful Cost & Schedule analysis based on these trade studies and concluded that the envisioned design was over budget by several $100 million and that the schedule of mirror development would not allow a 2008 launch.


Realizing the budget and technology problems, the project had to accept the painful but necessary re-scope of NGST. To address both problems, the project management decided to reduce mirror size to 6m+. Two contractors put in a bid to be the prime contractor to build the telescope: TRW/Ball Aerospace and Lockheed-Martin.

September 2001

The Jet Propulsion Laboratory was selected by NASA as the implementing center for the NGST Mid-Infra-red Instrument (MIRI). (Press Release)

James Webb Space Telescope Cryogenic Mirror Testing

Ball Aerospace's Jake Lewis is reflected in one of the mirrors on a James Webb Space Telescope Array that was in the X-ray and Cryogenic Facility for Testing.

Early NGST Publications

A collection of early NGST publications.

The Next Generation Space Telescope

The Next Generation Space Telescope

Proceedings of a workshop jointly sponsored by the National Aeronautics and Space Administration and the Space Telescope Science Institute, Baltimore, Maryland, 13-15 September 1989

Technologies for Large Filled-Aperature Telescopes in Space

The Next Generation Space Telescope

Proceedings from the Astrotech 21 workshop series held at JPL in Pasadena, California, 4-5 March 1991

A digital version of this very relevant NGST study has finally been obtained. Only a few hard copies exist. PDF was not invented in 1991 and digital storage was not easily available. Jim Cutts at JPL was the Advanced Instruments Program Manager in those days and responsible to NASA for the Astrotech 21 study. Jim recently found that the JPL library had a microfilm(!) copy and the JPL library folks recovered a pdf version for us. Their help is much appreciated. Given that it was microfilm the quality is not good, but it is readable and valuable to have available. Over time we will scan a paper version for some of the hardest-to-read material. But this is a key piece of the NGST history with many thoughtful, perceptive and forward-looking discussions by an experienced group of engineers and scientists. As such we wanted to make this available to the engineering and scientific communities.

JWST Vibrational Testing

Vibrational testing on the James Webb Space Telescope at NASA's Goddard Space Flight Center, in Greenbelt, Maryland.