WFIRST was the No. 1 rated large-scale mission in the 2010 decadal survey for astrophysics. Credit: NASA illustration

Jon Morse is the chief executive of the nonprofit BoldlyGo Institute. He was director of NASA’s Astrophysics Division in the Science Mission Directorate from 2007 to 2011, and was project scientist for the Cosmic Origins Spectrograph at the University of Colorado from 1997 to 2003.

I write in response to the recent op-ed piece by Donald F. Robertson regarding a possible servicing mission to extend the life of the Hubble Space Telescope.

As a card-carrying “Hubble hugger” who has done a large fraction of my scientific work for over 20 years using Hubble data, who has helped design and build the Cosmic Origins Spectrograph instrument that was installed onboard Hubble during the 2009 servicing mission, who has managed NASA’s astrophysics portfolio during the most recent science community-based decadal planning effort in 2009-2010, and who has spent significant time and energy considering what to do when Hubble reaches the end of its lifetime, I will discuss reasons why the servicing mission idea may not be in the best interests of the astronomical science community or the taxpaying public, and describe an alternative to servicing Hubble — namely, build Hubble 2.0.

Hubble Space Telescope
Hubble Space Telescope. Credit: NASA

To be clear, I am pleased about Mr. Robertson’s interest in the topic. I and many of my colleagues recognize that the loss of Hubble will create a large gap in our space ultraviolet and visible light capabilities for astronomy. But we part ways when considering how to react to this eventuality.

While Mr. Robertson reports mostly on the engineering challenges, I bring the additional perspectives of a scientist and portfolio manager — this is not merely a matter of how, but also what, why and how to pay for it. The technical approaches are interesting to speculate about, but the scenarios that rely on future spacecraft with orders of magnitude less carrying capacity than the space shuttle appear to me to understate the scale, complexities, costs and effectiveness of another Hubble servicing mission.

Conducting another servicing mission to Hubble flies in the face of the National Research Council’s recommendations in the so-called decadal surveys in astronomy and astrophysics. This in itself is the reason why NASA’s astrophysics program is not pursuing any sort of Hubble servicing options within its highly constrained budget.

The 2001 decadal survey (“Astronomy and Astrophysics in the New Millennium”) stated, “The committee endorses the current plans that call for [Hubble] to continue operation until the end of the decade, with reduced operating costs after completion of the final servicing mission.” The timetable for these events was delayed after the final servicing mission — Servicing Mission (SM) 4 — was canceled, reinstated, then finally performed in May 2009. (As Mr. Robertson noted, NASA counts four maintenance missions rather than five because mission three was split into two flights.)

Subsequently, the 2010 decadal survey (“New Worlds, New Horizons in Astronomy and Astrophysics”) stated, “Following the fourth servicing mission, the Hubble Space Telescope (HST) is now more capable than ever before and is enabling spectacular science, including observation at ultraviolet wavelengths. No more servicing missions are planned, and NASA intends to deorbit HST robotically at the end of the decade. The committee endorses this decision.”

Conducting another servicing mission to Hubble flies in the face of the National Research Council’s recommendations in the so-called decadal surveys in astronomy and astrophysics. This in itself is the reason why NASA’s astrophysics program is not pursuing any sort of Hubble servicing options within its highly constrained budget.

In other words, in a tight fiscal environment that will clearly persist at least for the rest of this decade, the community is prepared to press on to new scientific priorities and new observatories that will expand our capabilities beyond Hubble’s.

The 2010 decadal survey was carefully crafted by a committee and scientific panels replete with avid Hubble users. Their recommendations consciously omitted mounting a whole new infrastructure to service Hubble again — the science community understands that Hubble’s lifetime is limited and that it’s time to move on.

While the science community indeed hopes that Hubble will continue to operate long enough to overlap with the James Webb Space Telescope toward the end of the decade, either way JWST is poised to assume the mantle as NASA’s flagship observatory.

Saying that JWST is Hubble’s “successor” is both accurate and deliberate — it indicates the passing of the torch to a next-generation facility designed to exceed Hubble’s scientific reach, especially into the early universe. JWST’s optimization at infrared wavelengths is science-driven, and it is also why it is not called a “replacement” for Hubble.

JWST artist's concept
JWST artist’s concept. Credit: NASA
JWST artist’s concept. Credit: NASA

Meanwhile, the top recommended large mission in the 2010 decadal survey to follow JWST is the Wide-Field Infrared Survey Telescope (WFIRST), which Mr. Robertson does not mention. Any diversion of resources toward a currently unplanned Hubble servicing mission would necessarily impact this top-rated priority.

I’ll return to the idea of replacing Hubble’s unique capabilities later, but first we can add more information to Mr. Robertson’s narrative for clarity and completeness. For example, Hubble’s overlap with JWST is called “slight” even though the description of the overlap from orange wavelengths through the near-infrared contradicts such a designation. Quantitatively, JWST will overlap the current Hubble capabilities over wavelengths from about 0.6 to 1.6 microns, whereas Hubble’s unique coverage is from about 0.1 to 0.6 microns. So the overlap is two times larger than the unique region — hardly slight. The current WFIRST implementation calls for using a donated 2.4-meter (Hubble-sized) telescope package. With its silver-coated optics, WFIRST will be optimized for the near-infrared but is expected to be sensitive down to 0.43 microns.

Regarding Hubble’s limited on-orbit lifetime, we know that orbital decay and failed gyroscopes are two considerations that would eventually contribute, but the situation for these may not be dire. In speaking with knowledgeable officials at NASA, including on the Hubble project team, recent analysis indicates that Hubble is not expected to re-enter the atmosphere in the early 2020s, as stated by Mr. Robertson. Hubble’s orbit should remain stable through at least another solar cycle into the 2030s. Of the six new gyroscopes that were installed during SM4, one has failed, one is a bit flaky but could be used, and the other four are essentially working nominally. Three gyroscopes are needed for normal science operations, but the Hubble operations team has also devised and tested operating the observatory in two-gyro and one-gyro modes.

Perhaps the biggest intellectual miscue in Mr. Robertson’s treatise is the statement that, “Replacing instruments with more up-to-date cameras would be nice, but is not essential to keeping Hubble in business and productive.” The notion that Hubble can be serviced and remain productive “indefinitely” without new science instruments simply belies the reality of Hubble’s entire 25-year history. There would have been no servicing missions at all if new scientific instruments were not integral to them. This is the reason that Hubble’s productivity has increased over time; a graph of scientific papers published per year shows that every time new instrumentation is installed, there is a significant increase in papers published.

Astronauts last serviced Hubble in 2009. Credit: NASA
Astronauts last serviced Hubble in 2009. Credit: NASA

Moreover, Hubble’s past operational history demonstrates that instruments tend to degrade and/or fail after several years. The SM4 upgrades and repairs have exceeded their goal of enabling Hubble science operations for five additional years, but it is unreasonable to expect that the observatory would only need new gyros and an orbit boost, and can avoid instrumentation and other electronic systems failures beyond 2020 and certainly not “indefinitely.”

Finally, despite any analyses of servicing feasibility using SpaceX Crew Dragon or other future spacecraft, where is the money going to come from to configure a new spacecraft, design and install suitable robotic arms, practice the servicing operations, build the maintenance hardware, and all of the other aspects of an actual servicing mission?

The Hubble servicing budget just for the astrophysics portfolio portion of the missions (i.e., not including space shuttle launch and operations) ran over $100 million per year. Despite the “low-cost” description reported by Mr. Robertson, such a robotic or crewed mission, meaningfully thought through with due consideration of all the complexities inherent to Hubble, is more likely billion-dollar class than Explorer class.

The astrophysics community has already made its choice: NASA’s priorities for the future of astrophysics are JWST, WFIRST, Explorer missions, strategic contributions to key foreign-led missions such as Euclid, missions of opportunity on the space station and other platforms such as Japan’s Astro-H mission, suborbital balloon and sounding rocket experiments, and healthy budgets for data analysis, technology development, data archiving and theory.

Mr. Robertson mentions the idea that SpaceX could mount a self-funded Hubble servicing mission, take possession of the observatory, and then “rent” the facility back to scientists. Aside from SpaceX showing no inclinations toward any such activity or investment, this notion has two characteristics we share with our approach to conducting frontier space science missions at the BoldlyGo Institute: that restoring the lost Hubble capabilities could be funded privately, and that scientists could put “skin in the game” toward helping to build and operate a space telescope.

We believe, based on a simple analysis, that investing the equivalent capital in a new observatory optimized for the ultraviolet-visible wavelengths would be far more scientifically productive than maintaining Hubble with 14-to-22-year-old instruments that are well beyond their design lifetimes.

We have considered the “Hubble for rent” model and conclude it is more expedient and productive to build a new observatory.

First, it would be very difficult to conduct science operations for Hubble anywhere except the Space Telescope Science Institute (STScI) in Baltimore, and it’s not clear that it could be done without the involvement of NASA’s Goddard Space Flight Center in Greenbelt, Maryland. So even if operations were maintained at STScI, what would the cost be? Hubble’s current annual budget is about $97 million, with $25 million to $30 million of that supporting the research of the scientists who use Hubble. That leaves about $70 million for operating the observatory, which includes (but is not limited to) monitoring Hubble’s on-orbit health and status; soliciting, processing and selecting proposed observing programs; planning, scheduling and commanding sequences for each science and housekeeping operation on the satellite; conducting the ongoing calibration program; downlinking and processing the data through the data reduction pipelines; and maintaining and upgrading proposal software tools and instrument handbooks for the observers. And this is in the era of “cheap ops” that was planned for the post-SM4 time frame since the late 1990s and began executing during my tenure as NASA’s astrophysics director.

Even if some further efficiencies could be found, charging for Hubble usage is an expensive proposition. If 1,000 scientists used Hubble (about today’s number who are involved in the selected observing programs), you’d have to charge every one of them $60,000 each year just to keep operating, with no research budget and no recovery of the capital investment for the servicing mission. (There is a similar calculus if one were to charge by the hour or orbit.)


A new observatory with a 10-times-larger field of view ultraviolet-visible camera can be operated in a Lagrange point or heliocentric orbit for about a third of this cost, similar to NASA’s Kepler and Spitzer missions during this decade, and would have almost a factor of two higher observing efficiency (time observing targets) compared with Hubble in its low Earth orbit; the science per dollar is thus 60 times greater for a new facility.

In conclusion, NASA will not “wastefully abandon” Hubble. Barring a major financial crisis in the government, we can expect that Hubble will be operated as long as it is scientifically productive, as determined through rigorous periodic review.

Let’s look forward to Hubble’s 30th anniversary in space in April 2020, and hope that by that time we are well on our way toward restoring and enhancing its ultraviolet and visible light imaging and spectroscopic capabilities with a true replacement mission. Our aim is to privately fund this replacement and use a commercial development model that leverages extensive industry experience. This path minimizes NASA’s obligations, allowing it to focus resources on WFIRST and a future large-aperture mission after WFIRST is launched. Then let’s utilize the servicing capabilities mentioned by Mr. Robertson to upgrade and/or repair this new observatory in a much higher orbit, as is also being contemplated for WFIRST. And we can even call the new observatory “Hubble” in order to maintain continuity, if desired — the way a car manufacturer comes out with a new model under the same name.