In this chapter, each of the “large,”“medium,” and “small” recommendations of New Worlds, New Horizons in Astronomy and Astrophysics1 (NWNH) are considered in turn for the ground-based program. The progress that has been made toward NWNH goals is evaluated, including the programs adopted by the agencies and their plans for the remainder of the decade. Following the committee’s evaluation of the individual NWNH recommendations, an overview of the ground-based program implementation is provided, and its balance is considered.
For reference, Table ES.3 of NWNH is reproduced below, listing the priorities for large-scale ground-based activities (Table 3.1). These included, in rank order, the Large Synoptic Survey Telescope (LSST), a Mid-Scale Innovations Program (MSIP) augmentation at the National Science Foundation (NSF), a Giant Segmented Mirror Telescope (GSMT), and the Atmospheric Cerenkov Telescope Array (ACTA).
LSST is an integrated survey system consisting of a new 8-meter class wide-field telescope, a 3.2-gigapixel camera with a 3.5-degree field of view, and an automated
1 National Research Council (NRC), 2010, New Worlds, New Horizons in Astronomy and Astrophysics, The National Academies Press, Washington, D.C.
TABLE 3.1 Priorities for Large-Scale Ground-Based Activities from the 2010 Astronomy and Astrophysics Decadal Survey
|Recommendationb||Science||Technical Riskc||Appraisal of Costs Through Constructiona (U.S. Federal Share, 2012-2021)||Appraisal of Annual Operations Costsd (U.S. Federal Share)||Cross-Reference in Chapter 7|
|1. LSST —Science late 2010s —NSF/DOE||Dark energy, dark matter, time-variable phenomena, supernovae, Kuiper belt and near-Earth objects||Medium low||$465M ($421M)||$42M ($28M)||Page 223|
|2. Mid-Scale Innovations Program —Science mid-to-late 2010s||Broad science; peer-reviewed program for projects that fall between the NSF MRI and MREFC limits||N/A||$93M to $200M||Page 225|
|3. GSMT —Science mid-2020s —Immediate partner choice for ~25% federal share||Studies of the earliest galaxies and galactic evolution; detection and characterization of planetary systems||Medium to medium high||$1.1B to $1.4B ($257M to $350M)||$36M to $55M ($9M to $14M)||Page 228|
|4. ACTA —Science early 2020s —NSF/DOE; U.S. join European Čerenkov Telescope Array||Indirect detection of dark matter; particle acceleration and active galactic nucleus science||Medium low||$400M ($100M)||Unknown||Page 232|
a The survey’s construction-cost appraisals for the Large Synoptic Survey Telescope (LSST), Giant Segmented Mirror Telescope (GSMT), and Atmospheric Čerenkov Telescope Array (ACTA) are based on the survey’s cost, risk, and technical readiness evaluation (i.e., the cost appraisal and technical evaluation, or CATE, analysis) and project input, in FY2010 dollars; cost appraisals for the Mid-Scale Innovations Program augmentation are committee-generated and based on available community input. For GSMT the cost appraisals are $1.1 billion for the Giant Magellan Telescope (GMT) and $1.4 billion for the Thirty Meter Telescope (TMT). Construction costs for GSMT could continue into the next decade, at levels of up to $95 million for the federal share. The share for the U.S. government is shown in parentheses when it is different from the total.
b The survey’s appraisals of the schedule to first science are based on CATE analysis and project input.
c The risk scale used was low, medium low, medium, medium high, and high.
d The contractor had no independent basis for evaluating the operations cost estimates provided for any ground-based project. The survey’s appraisals for operations costs, in FY2010 dollars, were constructed by the survey committee on the basis of project input and the experience and expertise of its members. For GSMT the range in operations costs is based on estimates from GMT ($36 million) and TMT ($55 million). The share for the U.S. government is shown in parentheses when it is different from the total.
SOURCE: National Research Council, 2010, New Worlds, New Horizons in Astronomy and Astrophysics, The National Academies Press, Washington, D.C., Table ES.3.
data processing system. LSST was NWNH’s highest-ranked large-class ground-based initiative:
The Large Synoptic Survey Telescope (LSST) would employ the most ambitious optical sky survey approach yet and would revolutionize investigations of transient phenomena. It would address the pressing and fundamental question of why the expansion rate of the universe is accelerating, and would tackle a broad range of priority science questions ranging from understanding the structure of our galaxy to elucidating the physics of stars. LSST . . . opens a new window on the time-variable universe and therefore promises discoveries yet to be imagined. LSST’s observations repeatedly cover large areas of sky following a preordained and optimized sequence to create a data set that addresses a majority of SFP-identified questions.2
The top ranking of LSST in NWNH was “a result of its capacity to address so many of the identified science goals and its advanced state of technical readiness.”3 The expected scientific performance of LSST is essentially unchanged since NWNH, and there has been major progress programmatically. LSST is an excellent example of successful interagency cooperation, and it also leverages the relative strengths of public and private funding. NSF, which is the lead agency, supports the telescope and site facility, data management system, and education and public outreach, primarily via a Major Research Equipment and Facilities Construction (MREFC) award in August 2014. The camera is provided by the Department of Energy (DOE), along with additional contributions from international partners. Authority to begin camera fabrication was granted by DOE in August 2015. Private support enabled the LSST project to retire several major risks through early fabrication of the primary/tertiary mirror and the secondary mirror blank, preliminary site preparation, and early sensor studies. Engineering First Light is planned in 2020, with the 10-year science survey commencing in late 2022.
FINDING 3-1: LSST planning and construction have progressed well and are on schedule and within budget, successfully bringing together NSF funding, DOE funding, and private funding.
Data products of LSST will include the following:
- A nightly stream of ~10 million time-domain events detected and transmitted to event distribution networks within 60 seconds of observation;
- Catalogs of orbits for ~6 million bodies in the solar system; and
- Catalogs of ~37 billion objects (20 billion galaxies, 17 billion stars) and trillions of single-epoch detections.
2 NRC, 2010, New Worlds, New Horizons, p. 223.
3 NRC, 2010, New Worlds, New Horizons, p. 225.
Services and computing resources will be provided by data access centers to enable user-specified custom processing and analysis and interface software for higher-level science analyses. To support community-based brokering of the tremendous transient stream (advanced filtering services that add contextual information—for example, correlation with external catalogs or other alert steams, or additional analyses, for follow-up observation decisions and coordination), LSST will also provide an alert-filtering service.
LSST data will be transformational in scale and impact, especially for time-domain astronomy. However, the cost of operations and data analysis will be substantial. The estimated operations costs are about $50 million (in fiscal year [FY] 2023 dollars) per year over the 10-year planned lifetime of the project, of which NWNH proposed that 70 percent be borne by the U.S. funding agencies—two-thirds by NSF and one-third by DOE.4 The budgets presented to the committee by NSF’s Division of Astronomical Sciences (NSF-AST) cannot accommodate LSST operating costs without a reduction in support for other activities, and the current individual investigator programs do not have the resources to carry out the scientific program enabled by LSST.
FINDING 3-2: Current projections for LSST performance and data products promise transformational scientific impact, as envisioned by NWNH. To realize the full scientific potential of this great new facility, funding that enables individual investigators and groups of investigators to deliver the scientific results will be critical.
NWNH recommended the creation of a new Mid-Scale Innovations Program “that would enable moderate-scale projects to be frequently selected through peer review.”5 The survey committee likened this mid-scale instrumentation and facility program at NSF to NASA’s Explorer program. This “mid-scale” level is that which falls between the limits of the NSF Major Research Instrumentation and MREFC programs, $4 million to $135 million. NWNH recognized the many highly promising projects in this funding category that might achieve innovative and timely scientific goals. NWNH recommended that the program be funded at an annual funding level of $40 million per year, or approximately double the amount being spent on projects in this size category through a less formal programmatic structure, and issue “roughly annual” calls for proposals.
4 Jim Ulvestad, National Science Foundation, presentation to the committee on October 8, 2015.
5 NRC, 2010, New Worlds, New Horizons, p. 226.
As reported to the committee, NSF-AST funded midscale activities at a level of $31 million in FY2010, similar to the level at the end of the previous decade. Among the competed programs were the University Radio Observatories (URO), with a $6 million to $10 million per year budget, the Telescope Systems Instrumentation Program (TSIP), with a $3 million per year budget administered by the National Optical Astronomy Observatory (NOAO), and Renewing Small Telescopes (ReSTAR), a short-term program with a 2009 start and a total budget of $5.4 million. The unsolicited programs ranged from $10 million to $15 million per year for 10 different projects and included the Very Energetic Radiation Imaging Telescope Array System, the Sloan Digital Sky Survey, the Atacama Cosmology Telescope, POLARBEAR, the Hobby-Eberly Telescope Dark Energy Experiment, the Virtual Astronomical Observatory, the Dark Energy Survey, the Murchison Widefield Array, Precision Array to Probe Epoch of Reionization, and LSST design and development.
The NSF-AST Portfolio Review of 2012 recommended a vigorous divestment from many facilities in order to accommodate new priorities, including the Atacama Large Millimeter/submillimeter Array (ALMA) operations, MSIP, and future LSST operations. The Portfolio Review recommended that MSIP include telescope open access time, laboratory astrophysics, MREFC design and development, and long-term mid-scale facilities operations, in addition to the NWNH new telescopes and instruments recommendations. The Portfolio Review also recommended merging the solicited and unsolicited programs into MSIP. While NSF-AST has implemented MSIP, with a first call in 2013 and a second in 2015, it has (consistent with the Portfolio Review recommendation) discontinued the URO, TSIP, and ReSTAR and unsolicited proposal programs. Overall, NSF-AST mid-scale funding dropped from $31 million in FY2010 to a nadir of $15.5 million in FY2015, recovering to $21 million in FY2016.
FINDING 3-3: Implementation of the NWNH recommendation of MSIP has been possible only by subsuming previous programs into MSIP and by aggressive divestment from older facilities. The total NSF-AST funding for mid-scale initiatives has dropped by nearly a factor of two since the start of the decade, in stark contrast to the NWNH recommendation of MSIP as a new initiative that would expand opportunities for mid-scale projects.
Following the Portfolio Review, NSF-AST issued a first call for proposals for MSIP in June 2013 and did not separately compete the URO, TSIP, and ReSTAR programs. The identified funding was for 2 fiscal years. Four categories of proposals were accepted: mid-scale science projects, mid-scale facilities, development investments, and open access to telescopes. The maximum request per project was $40 million. NSF-AST received 38 pre-proposals requesting $400 million. Individual project budgets ranged from $3 million to $40 million and were mostly 5-year
projects. Twelve full proposals were invited with a total request of $177 million. Six awards were made with a total from NSF-AST FY2014/FY2015 of $27.1 million and other NSF support of $20 million. The awards were as follows: Zwicky Transient Facility ($9 million), Advanced ACTPol ($10 million), the Hydrogen Epoch of Reionization Array (HERA) ($2.1 million), EHT ($6.5 million), POLARBEAR ($5 million), and the North American Nanohertz Observatory for Gravitational Waves (NANOGrav) ($14.5 million).
Unranked examples of MSIP projects considered promising were identified by NWNH. In the largest funding category, between $40 million and $120 million, were the Big Baryon Oscillation Spectroscopic Survey (BigBOSS), the Frequency Agile Solar Radiotelescope (FASR), HERA, and NanoGrav. In the $12 million to $40 million range were cosmic microwave background (CMB) measurements, Exoplanet Initiatives, Next-Generation Adaptive Optics Systems, and Next-Generation Instruments for Solar Telescopes. Finally, in the $4 million to $12 million group, NWNH mentioned the High-Altitude Water Cherenkov Observatory (HAWC).
HAWC has been built with funds from DOE and the NSF Physics Division (NSF-PHY). BigBOSS is now called the Dark Energy Spectroscopic Instrument (DESI) and is being supported by DOE and NSF-AST. CMB programs at the South Pole are being supported by NSF-Polar, NSF-Physics, DOE, and private funds (South Pole Telescope, Keck, and BICEP). MSIP grants were awarded to HERA, AdvACT (CMB), and POLARBEAR (CMB); and NanoGrav is being funded by NSF-AST-MSIP and NSF-PHY Physics Frontiers Center. Of the mid-scale projects identified by NWNH, FASR, Exoplanet Initiatives, Next-Generation Adaptive Optics Systems, and Next-Generation Instruments for Solar Telescopes have not yet been funded through MSIP.
FINDING 3-4: Despite limited resources for MSIP, NSF-AST has funded an exciting set of highly ranked proposals in a heavily oversubscribed competition. Some mid-scale programs recommended by NWNH have also moved forward with funding from DOE and from the NSF Physics and Polar Programs. The scientific promise of these projects confirms the NWNH expectation that a mid-scale program would enable major advances that respond nimbly to opportunities on a diverse range of science topics.
Participation in one of the U.S. GSMT projects was the third-ranked, ground-based, large-scale priority in NWNH. As stated by NWNH, “[t]hese Giant Segmented Mirror Telescopes (GSMTs) will be essential to understanding the distant galaxies discovered by JWST and to obtaining spectra of the faint transients found
by LSST, and they will be transformative for a broad range of science aimed at understanding targets ranging from stars and exoplanets to black holes.6 The report notes that the lower ranking was not due to scientific capabilities but primarily due to concerns about technical readiness and cost and schedule risk. The recommendation called for an “immediate selection by NSF of one of the two U.S.-led GSMT projects” with a goal of a 25 percent share. NWNH noted that “[t]his share could be secured through whatever combination of construction (that is, MREFC), operating funds, and instrumentation support is most favorable.”7
The two GSMT projects in the United States, the Giant Magellan Telescope (GMT) and the Thirty Meter Telescope (TMT), have made considerable progress in the first half of this decade. Both have passed Preliminary Design Reviews, retiring many of the technical risks identified earlier in the decade. Both have advanced the technology for manufacturing mirrors and mirror segments. GMT is a private consortium consisting primarily of university and research-institute partners. With funding pledged for 70 percent of the budget, the GMT Board recently approved proceeding immediately with construction of its first stage. Site preparation work in Chile is well under way, and ground breaking occurred in November 2015. The first stage consists of four of the seven primary segments, a secondary mirror without adaptive optics capability, and two first-light instruments; when the remaining funding is identified, the project would complete all seven segments and the full first-light suite, including adaptive optics. TMT is a collaboration between the University of California, the California Institute of Technology, and international partners India, China, Japan, and Canada. Similar to GMT, approximately 80 percent of the budget has been pledged by these partners. TMT began construction at its site on Mauna Kea in 2014, but construction has been interrupted until the project can address concerns raised by the Hawaii Supreme Court regarding the permitting process.
Both GSMT projects are actively engaging the optical and infrared (OIR) community through a series of workshops and other outreach activities. The lack of fully identified construction funding, which in turn can partially be traced to the lack of NSF engagement, represents a risk for both programs. If one or both projects are not fully completed, the U.S. public will be left with relatively little GSMT capability, while European astronomers will have the European Extremely Large Telescope.
The highly constrained budget environment has, so far, prevented any significant involvement by NSF in either of the two GSMT projects. NSF did issue a solicitation for planning a partnership model for a GSMT and providing community engagement, to which TMT responded. A preliminary summary of the plan
6 NRC, 2010, New Worlds, New Horizons, p. 228.
7 NRC 2010 New Worlds New Horizons p. 232.
being developed under this award was presented to the committee. It included a scenario for MREFC support of ~20 percent of TMT capital costs beginning at the end of this decade, with operations funding ranging from a similar share to a scenario where the operating costs are fully assumed by other partners. Although the latter would result in a lower allocation of telescope time to the U.S. community, it would still provide a governance role in the project and is an interesting possibility in the current environment where operations funds are highly constrained. The GMT project has also considered public-private partnerships, and the GMT operating agreement is structured to allow flexibility in combinations of capital, operating, and instrumentation funds. NSF participation could come through MSIP instrumentation funding, with allocation of nights to the community, through capital construction support with provision for community access, or though operations support.
FINDING 3-5: The GMT and TMT projects have both made major progress since 2010, and both offer technically feasible routes to achieving the GSMT science goals set forth by NWNH. However, programmatic hurdles remain, and neither project has secured the funding needed to complete construction at its full intended scope. NSF budget constraints have prevented NSF’s implementation of the NWNH recommendation that NSF-AST select one partner and participate in GSMT construction.
FINDING 3-6: A selection process leading to MREFC commitment to construction of a U.S.-led GSMT project, without commitment of NSF funds to GSMT facilities operations, would partially address the NWNH recommendation of U.S. federal participation in a GSMT, while retaining flexibility in the NSF-AST budget for implementation of other priorities in the next decade.
With NSF so far unable to meaningfully participate in a GSMT effort, concerns have been raised about broad community access to optical facilities and about the health and balance of the OIR system. The National Research Council report Optimizing the U.S. Ground-Based Optical and Infrared Astronomy System8 addresses these issues and makes several recommendations for adjusting the system within the limited resources now available. These include an expanded role for NOAO, coordinating exchanges of time between telescopes, organizing workshops to define new instrumentation priorities, and operating the U.S. facilities, particularly in the southern hemisphere, in a coordinated fashion. The report highlights the need for a coordinated system to respond to transient events from new facilities, particularly
8 NRC, 2015, Optimizing the U.S. Ground-Based Optical and Infrared Astronomy System, The National Academies Press, Washington, D.C.
LSST. Several other recommendations, including development of new instruments and a large share in a GSMT, are challenging to implement in the current environment. In a “Dear Colleague” letter response, NSF highlighted work to respond to these recommendations where practical.9
ACTA was ranked fourth among ground-based, large-scale activities in NWNH. At that time, U.S. groups had proposed a U.S.-led ACTA effort, named the Advanced Gamma-ray Imaging System (AGIS). NWNH recommended that “the U.S. AGIS team collaborate as a partner with the European CTA team and that a U.S. budget for construction and operations of approximately $100 million over the decade be shared between DOE, NSF-Physics, and NSF-Astronomy”10 The 2014 Particle Physics Project Prioritization Panel (P5) recommended that NSF-PHY and DOE “invest in CTA as part of the small projects portfolio if the critical NSF Astronomy funding can be obtained.”11
The Čerenkov Telescope Array (CTA) is planned as an open observatory with one array in the northern hemisphere and one in the southern hemisphere. The three leading present-day ground-based gamma-ray efforts (VERITAS (located in Arizona, United States), MAGIC (Canary Islands, Spain), and HESS (Namibia) have largely joined together to form CTA, which is designed to increase the sensitivity to gamma-ray sources by an order of magnitude over a wide range of energies (tens of GeV to above 100 TeV). The main science drivers are the study of high-energy astrophysical particle acceleration in a variety of galactic and extragalactic sources and searches for signals of dark matter interactions. In 2015, the CTA consortium selected its sites: the southern array at European Southern Observatory (ESO), Paranal, Chile, and the northern array at Observatorio del Roque de los Muchachos (ORM), La Palma, Spain. Construction could start as early as 2016, and the U.S. groups would benefit from a timely initiation of U.S. participation. The U.S. groups have developed a plan for participation in CTA, but at a lower level than that originally proposed at the time of NWNH. This plan seeks funding at a level appropriate for the mid-scale programs of the Astronomy and Physics divisions of NSF.
9 National Science Foundation, Dear Colleague Letter: NSF/AST Response to the NRC Report “Optimizing the U.S. Ground-Based Optical and Infrared Astronomy System,” September 1, 2015.
10 NRC, 2010, New Worlds, New Horizons, p. 24.
11 Particle Physics Project Prioritization Panel, 2014, Building for Discovery: Strategic Plan for U.S. Particle Physics in the Global Context, Department of Energy, Washington, D.C., p. 15, http://science.energy.gov/~/media/hep/hepap/pdf/May-2014/FINAL_P5_Report_Interactive_060214.pdf.
FINDING 3-7: U.S. participation in CTA at budget levels below those recommended by NWNH would still have a significant positive impact on the scientific productivity of the observatory and would give U.S. scientists leadership roles in the CTA program. If the U.S. CTA proposal competes successfully in the MSIP and NSF-Physics mid-scale programs, the NWNH recommendation can be implemented, albeit at a level lower than anticipated in 2010.
For reference, Table ES.2 of NWNH is reproduced below (Table 3.2). The single priority in this category was the Cerro Chajnantor Atacama Telescope (CCAT, formerly the Cornell-Caltech Atacama Telescope), a submillimeter-wave survey telescope led by a consortium of U.S., Canadian, and German universities.
Despite the high initial priority given to CCAT in NWNH, the project has not secured significant federal funding. CCAT was initially envisioned as a 25-meter telescope equipped with large-format cameras to enable surveys at millimeter wavelengths. The planned site for CCAT is at 5,600-meter elevation, some 500 meter above the nearby ALMA site. NWNH recommended an NSF-AST investment of $37 million in the $140 million cost of construction of the project, with the remaining funding coming from the university and international partners. NWNH further recommended an annual contribution to operations of $7.5 million from NSF-AST. The NSF contribution so far this decade to CCAT has been support for the design at $4.75 million from 2011 to 2015.
The Portfolio Review reiterated support for an NSF investment in CCAT, but only in the event that MSIP was funded at an annual level of $30 million. Since that goal has not been met, NSF-AST did not support CCAT as a separate line-item and instead encouraged the CCAT consortium to submit a proposal to the MSIP competition, which the consortium did. This proposal, however, was not funded. In the current budget climate, NSF will only contribute to CCAT through future MSIP competitions. The project is now being rebaselined, and a number of international potential partners have expressed interest, including the host country of Chile.
FINDING 3-8: In the current budget climate, NSF-AST has not been able to fund CCAT beyond an initial contribution to the design. This is because the NSF-AST budget increases anticipated by NWNH did not materialize, and NSF-AST, consistent with the Portfolio Review’s guidance, gave higher priority to funding the MSIP program within the constraints imposed by the budget.
TABLE 3.2 Medium-Scale, Ground-Based Recommended Activities from the 2010 Astronomy and Astrophysics Decadal Survey
|Recommendationb||Science||Technical Riskc||Appraisal of Costs Through Constructiona (U.S. Federal Share, 2012-2021)||Appraisal of Annual Operations Costsd (U.S. Federal Share)||Cross-Reference in Chapter 7|
|CCAT —Science early 2020s —University-led, 33% federal share||Submillimeter surveys enabling broad extragalactic, galactic, and outer-solar-system science||Medium||$140M ($37M)||$11M ($7.5M)||Page 234|
a The survey’s construction-cost appraisal for CCAT is based on the survey’s cost, risk, and technical readiness evaluation (i.e., the cost appraisal and technical evaluation, or CATE, analysis) and project input, in FY2010 dollars.
b The survey’s appraisal of the schedule to first science is based on CATE analysis and project input.
c The risk scale used was low, medium low, medium, medium high, and high.
d The survey’s appraisal of operations costs, in FY2010 dollars, is based on project input.
SOURCE: National Research Council, 2010, New Worlds, New Horizons in Astronomy and Astrophysics, The National Academies Press, Washington, D.C., Table ES.2.
For small-scale programs at NSF, NWNH recommended augmentations of $8 million per year (17 percent) to the Astronomy and Astrophysics Research Grants (AAG) program and $5 million per year (50 percent) to the Advanced Technologies and Instrumentation (ATI) program, which are the two primary individual investigator programs in NSF-AST. NWNH additionally recommended DOE and NSF funding of Theoretical and Computational Astrophysics Networks (TCAN) at a level of $1.0 million and $2.5 million per year, respectively. From FY2011 to FY2015, the combined budget of AAG and ATI has instead declined 3 percent, from $58.5 million to $56.6 million in real-year dollars, with a greater drop in purchasing power.12 The AAG budget in the years between 2012 and 2014 was approximately 10 percent less than it was in 2011, and, although there was an increase in AAG in 2015, it is anticipated that the funding will return to the 2012 to 2014 level in 2016. The ATI budget was further decreased to reallocate funds to MSIP and now stands at a level that is 20 percent below what it was in 2011 and half of what was envisioned
12 Unless otherwise noted, budget numbers in this section are taken from James Ulvestad, Director, Division of Astronomical Sciences, National Science Foundation (NSF-AST), “Responses to MidTerm Assessment Questions,” provided to the committee on December 4, 2015, referred to hereafter as “Ulvestad Responses,” p. 8.
in NWNH. NSF funded the TCAN program at $1.5 million per year in FY2014 and FY2015, although there is no TCAN support in the FY2016 budget. DOE did not provide funding for TCAN.
In its list of recommended small-scale activities, NWNH included an augmentation of TSIP to a level of $5 million per year. NWNH recognized that TSIP as a vital component of the OIR system that enabled new instrumentation on privately funded telescope in exchange for community access. As noted previously, TSIP has instead been discontinued as a separate program and subsumed into MSIP, which is, itself, underfunded relative to FY2011 mid-scale programs.
FINDING 3-9: Because the NSF-AST budget did not grow at the rate assumed by NWNH, NSF-AST has not implemented the majority of the NWNH recommendations for small-scale projects or for expanded support for individual investigator programs. Support for the individual investigator programs has decreased during the first half of the decade.
The international (U.S.-managed) Gemini Observatory was an area of concern for the NWNH committee, highlighting community dissatisfaction with the alignment of its capabilities and management to U.S. priorities. NWNH called for a $2 million per year increase in NSF’s contribution to the Gemini international partnership to enhance the community’s efforts in exoplanets, dark energy, and early-galaxy studies. Language in the FY2011 congressional appropriation led to an increase in the U.S. contribution. With the deployment of the Gemini Planet Imager and plans for a high-resolution optical spectrograph, the instrumentation suite is becoming more capable, and the management structure has been changed to be more responsive to community concerns. Four proposed next-generation Gemini instruments are high-throughput, rapid-response OIR spectrographs similar to those envisioned in the NRC report Optimizing the U.S. Ground-Based Optical and Infrared Astronomy System13 and well-matched to LSST followup. The greatest change in U.S. access to Gemini came in 2012 with the withdrawal of the United Kingdom from the partnership. This has resulted in the U.S. share of Gemini going from 50 to 65 percent with no increase in the U.S. contribution.
NWNH placed strong emphasis on expanding support for individual investigator grants: “Individual investigator programs are paramount in realizing the science potential of existing facilities, in pathfinding for future space missions and ground-based projects, and in training the current and future workforce. A healthy enterprise in astronomy and astrophysics requires a vigorous research grants program. . . . One of the most important secondary products is people who are trained in the broad discipline of science and who have skill in quantitative
13 NRC, 2015, Optimizing the U.S. Ground-Based Optical and Infrared Astronomy System.
thinking and analysis, numerical computation, instrumentation and engineering, teaching, and project management.”14
The success rate of AAG proposals fell from 22.4 percent in FY2010 to a low of 14.8 percent in FY2012, and recovered to 18.0 percent in FY2015.15 Relative to a decade ago, AAG proposal success rates have fallen by about a factor of two (e.g., the average success rate from FY2000 to FY2005 was 33 percent). While a number of factors contribute to declining proposal success rates,16 the mismatch between growing scientific opportunities and shrinking available support is ultimately responsible. Feedback can amplify the effect of such a mismatch: as even highly rated proposals are declined for lack of funding, investigators resubmit similar proposals in successive years or to multiple programs, driving proposal pressure up and success rates down. The committee concurs with NWNH that “the data analysis and dissemination and theoretical work performed by both individual scientists and science teams are ultimately responsible for the amazing results witnessed in astronomy in the past few decades.”17 Without adequate support at the level of individual investigators and research groups, the United States cannot reap the rewards of its investment in advanced observational and computational facilities.
FINDING 3-10: The core grants programs AAG and ATI have declined in real-year dollars and dropped still further in purchasing power over the first half of the decade. This reduction in funding has contributed to a substantial decline in grant funding rates, threatening the scientific productivity of the U.S. ground-based astronomy program.
As previously noted, NWNH emphasized the importance of a balanced ground-based astronomy program, with a mix of large and medium-scale initiatives and strengthening of core research infrastructure through individual grants and support for instrumentation, technology development, and theory. The combination of an approximately flat NSF-AST budget with rising costs for facility operations has made maintaining such a balance extremely difficult. Averaged over FY2011 to FY2015, the annual NSF-AST budget has been $237 million, equal to its value in
14 NRC, 2010, New Worlds, New Horizons, p. 132.
15 James Ulvestad, Director, NSF-AST, “NSF Division of Astronomical Sciences (AST) Activities and Plans Pertaining to NWNH,” presentation to committee on October 8, 2015, slide 29.
16 Astronomy and Astrophysics Advisory Committee, 2016, Competed Grant Success Rates in US Astronomy and Astrophysics, National Science Foundation, Arlington, Virginia.
17 NRC, 2010, New Worlds, New Horizons, p. 132.
FY2011.18 The annual cost of facilities operations within that budget has risen from $130 million to $146 million over this period because of the ramp-up of ALMA operations (from $23 million in FY2011 to $40 million in FY2015) and the beginning of Daniel K. Inouye Solar Telescope (DKIST) operations ($5 million in FY2015), partly offset by $4 million in reduced operations funding for Arecibo and NOAO. With the NSF-AST individual investigator grants program declining 3 percent in real-year dollars over this period, the increases in facilities operations have come at the expense of mid-scale initiatives. As noted previously, when MSIP began in FY2014, it subsumed previous mid-scale support, and the total annual funding for mid-scale projects has declined by a factor of two over the decade. Given the necessity of operating NSF-AST’s powerful new facilities and the importance of maintaining support at the individual investigator level, squeezing the mid-scale program was probably the best choice available within a flat budget scenario, but it is opposite to the augmentation of mid-scale research funding envisioned by NWNH. Furthermore, the diminished funding of mid-scale programs has the collateral effect of reducing the number of future instrument builders needed for the next generation of cutting-edge initiatives.
FINDING 3-11: The combination of a flat NSF-AST budget (in real-year dollars) with new operations costs for ALMA and DKIST, and the need to sustain the individual investigator program, have led to sharp reductions in funding for mid-scale initiatives during the first half of the decade.
NWNH emphasized the importance of a “vigorous periodic senior review” of facilities to maintain “the appropriate balance in NSF’s astronomy and astrophysics research portfolio.”19 It recommended that NSF-AST complete a senior review before mid-decade to “determine which, if any, facilities NSF-AST should cease to support in order to release funds for (1) the construction and ongoing operation of new telescopes and instruments and (2) the science analysis needed to capitalize on the results from existing and future facilities.”20 Responding quickly to this recommendation, NSF carried out its Portfolio Review. The review recommended divestment of several facilities as described above. NSF-AST has moved to implement the Portfolio Review’s recommendations, and in several cases, it has been able to implement them by transferring operations to new partners rather than closing facilities. DOE is building DESI for the Mayall telescope and will assume most of the costs of Mayall operations after the DESI survey begins in 2019. NASA
18 At the next decimal place, the FY2011 to FY2015 average has been $237.2 million in real-year dollars, just above the FY2011 value of $236.8 million, and significantly below the FY2010 value of $246.5 million (James Ulvestad, Director, NSF-AST, “Responses to Mid-Term Assessment Questions,” provided to the committee on December 4, 2015.)
19 NRC, 2010, New Worlds, New Horizons, p. 32.
is partnering with NOAO to support development of an extreme radial velocity precision spectrograph for the 3.5-meter Wisconsin/Indiana/Yale/NOAO (WIYN) Telescope and operate exoplanet searches with the NOAO share of WIYN. The NANOGrav consortium (with funding from MSIP and an NSF Physics Frontier Center) and the Breakthrough Listen initiative are supporting Byrd Telescope operations in Green Bank by buying observing time. The committee notes that both arrangements are for finite periods of time and do not preclude the closure of those facilities in the 2020s.
Painful though they are, the divestments recommended by the Portfolio Review are essential to maintaining other key aspects of the NSF-AST program. However, the division’s new facilities—ALMA, DKIST, and (at the end of the decade) LSST—are more sophisticated and more complex than the facilities being divested, and they are correspondingly more expensive to operate. Divestment alone will not resolve the budget stresses imposed by rising facilities costs.
Looking ahead, the vigorous, periodic senior review advocated by NWNH may become necessary again in the future. While there are a number of long-lived space missions (e.g., Chandra and Swift), major ground-based facilities have historically remained operational longer than space missions, in part because they can be made continually more powerful with instrumentation that takes advantage of developing technologies and can be repaired more readily. An example of this is the Hubble Space Telescope—its unique long life as the world’s leading space observatory is a consequence of the servicing missions that renewed its instrumentation over the span of two decades. Nonetheless, NSF could construct major new astronomical facilities with an identified “prime operations lifetime,” and continued operations funding beyond this lifetime could be a positive decision informed by the senior review, not a default. The burden of operation costs from MREFC-funded facilities has been recognized as a problem facing other NSF divisions besides AST,21 and it is likely to grow worse with time as world-class facilities become more expensive.
One important obstacle to closing facilities is the cost of environmental mitigation and returning sites to their pre-construction state. Under the present NSF structure, the costs of facility closeout and environmental mitigation fall to the operating division. It is not difficult to imagine circumstances in which closing a facility would impose crippling costs on a division for a large fraction of a decade, and where continued operation could be preferred for financial reasons alone. This problem is also likely to affect NSF divisions other than AST, and finding ways to address it is an NSF-level policy challenge that cannot be resolved by NSF-AST alone.
21 Barry C. Barish, NSB Consultant, “The NSF MREFC Program: Perspectives on DUSEL as a proposed MREFC project,” presentation to the Committee to Assess the Deep Underground Science and Engineering Laboratory (DUSEL) on December 15, 2010.
Unfortunately, the squeeze between rising facility operations costs and a flat NSF-AST budget is on track to become far worse over the next 5-10 years. While projections of out-year facility operations costs are only notional, the estimates presented to the committee by the NSF-AST director suggest increases of about $10 million in ALMA operations (U.S. share) and $9 million in DKIST operations between FY2016 and FY2022.22 Projected NSF operations costs for LSST are estimated at $7.5 million in FY202123 and approximately $25 million by the time LSST is fully operational in the next decade (Figure 3.1). While full implementation of the Portfolio Review divestment recommendations is an essential step, it will not come close to offsetting this increase in facility operations costs. Unless the NSF-AST budget rises accordingly, there will not be adequate resources to operate the new facilities, even with severe cuts to the individual investigator program and continued restriction of mid-scale funding. Without adequate support for data analysis and theoretical interpretation, many of the discoveries enabled by LSST, DKIST, ALMA, and other powerful facilities, such as Gemini and the Jansky Very Large Array (formerly the VLA), will be made outside of the U.S. community or with U.S. scientists in supporting rather than leading roles. Conversely, the extraordinary science reach of these facilities, which led to their high-priority rankings in decadal surveys, means that even moderate augmentations in the NSF-AST budget will have highly leveraged science impact.
FINDING 3-12: Even following the divestment recommended by the Portfolio Review, the operations costs of ALMA, DKIST, and LSST will compromise the ability of the U.S. community to reap the scientific return from its premier ground-based facilities. Moderate increases in the NSF-AST budget would have highly leveraged science impact as a consequence of these powerful new facilities.
RECOMMENDATION 3-1: NSF should proceed with divestment from ground-based facilities that have a lower scientific impact, implementing the recommendations of the NSF Portfolio Review, which is essential to sustaining the scientific vitality of the U.S. ground-based astronomy program as new facilities come into operation.
22 FY2016 estimates are based on the FY2016 Budget Request; FY2017 estimates are based on the FY2017 Budget Request; FY2018-FY2022 estimates are based on the FY2017 Budget Request and are notional.
23 James Ulvestad, Director, NSF-AST, “Responses to Mid-Term Assessment Questions,” provided to the committee on December 4, 2015, referred to hereafter as “Ulvestad Responses,” p. 4.
Even following divestment at the maximum feasible level, anticipated facilities operations costs will constrain the mid-scale and individual investigator programs by an amount that could rise to $10 million to $20 million per year by the end of the decade. The LSST operations cost of $8 million at first, growing to $25 million, will be an additional burden on the AST budget in the first half of the next decade. The committee strongly supports the goal of a balanced program that includes facilities, medium-scale initiatives, and small-scale initiatives. Maintaining this balance is a challenge at the current level of funding.
RECOMMENDATION 3-2: NSF and the National Science Board should consider actions that would preserve the ability of the astronomical community to fully exploit NSF’s capital investments in ALMA, DKIST, LSST, and other facilities. Without such action, the community will be unable to do so because at current budget levels the anticipated facilities operations costs are not consistent with the program balance that ensures scientific productivity.
This committee is of the opinion that the optimum allocation of funds between facilities and mid- and small-scale activities is probably different for different levels of funding. Therefore, this balance is best left to the discretion of the NSF-AST Director once budget levels are known, and it would be appropriate for the Director to seek advice from the community by convening a panel for this purpose.