Optimizing the U.S. Ground-Based Optical and Infrared Astronomy System (2015)

5
Realizing the Full Science Potential of LSST

The path from Large Synoptic Survey Telescope (LSST) data to science results is not simple. The scope of the LSST survey, the volume and rate of the data, will require additional capabilities for the community to maximize the science from LSST. While a great amount of research can be pursued with LSST data alone, the availability of other instrumentation, specific precursor studies,¹ and development of new software systems will result in a huge increase in the quantity and quality of science results from LSST. This section describes instrumentation, software, and coordination needs that will enhance the returns from LSST data.

5.1 FOLLOW-UP TELESCOPE AND INSTRUMENTATION NEEDS

LSST will carry out extremely powerful deep, fast, and wide multi-band photometric surveys in the Southern Hemisphere. The requirements for follow-up observations, both photometric and spectroscopic, will span a range of timescales and telescope sizes. While significant capability could be contributed by existing small- and medium-aperture telescopes, the depth of a single LSST observation suggests that the majority of follow-up observations of all types will require large telescopes. An optimized spectroscopic system would exploit and enhance all three

_______________

¹ See white paper discussion by Willman et al. on the need for precursor Level 3 data products and tool development (B. Willman, K. Olsen, J. Bochanski, N. Brandt, A. Burgasser, W. Clarkson, M. Cooper, K. Covey, H. Ferguson, E. Gawiser, M. Geha, et al., 2014, “Enabling a Diverse User Community to Produce Cutting-Edge Science with LSST,” white paper submitted to the committee).

of LSST’s capabilities. Individual very faint sources will be followed up by 6- to 30-meter-class telescopes, pushing the frontiers at high redshift and in the local universe down the luminosity function. A high-throughput, moderate-resolution spectrograph on Gemini South or on another large southern telescope accessible to the community would help satisfy the spectroscopic needs. Optical transients, depending on the timescale and flux requirements, can be followed by small-aperture robotic arrays and by medium-aperture telescopes through target-of-opportunity (ToO) observations or dedicated programs, addressing decadal survey priorities in time-domain astronomy. LSST will produce only two-dimensional maps of the cosmos (providing precise measurements of fluxes, colors, and shapes); this gives rise to the need for wide survey follow-up. The Sloan Digital Sky Survey (SDSS) is by some measures the most highly cited astronomical facility of the past decade, with some 6,250 papers cited about 275,000 times, for an h-index of 205.² SDSS enabled a rich array of astronomical discoveries through its combination of photometric (two-dimensional, 2D) and spectroscopic (three-dimensional, 3D) surveys. LSST promises much more.

The science reach of LSST could be substantially enhanced by developing for the U.S. astronomy community a very-wide-field, massively multiplexed, spectroscopic capability. This facility should be capable of overlapping the majority of the sky area covered by the LSST surveys.³ Such a wide-field instrument or instruments should be sufficiently multiplexed to enable spectroscopic surveys of tens of millions of objects over several years. The science case for such a capability is rich for objects of a wide range of brightnesses, as already indicated in Section 4.1.

A primary systematic error for extragalactic and cosmological studies with LSST will be uncertainty in photometric redshift estimates (based on colors and fluxes).⁴ Deep, wide redshift surveys that substantially overlap LSST are needed to both train and calibrate photometric redshift estimates. The sample sizes here are much smaller than those needed for the measurements of spatial clustering (hundreds of thousands versus tens of millions), but they have much more stringent requirements for sample completeness in order to mitigate systematic errors and therefore will likely need to include larger-aperture facilities.

Worldwide, the currently planned, highly multiplexed (N roughly 500 or larger) spectroscopy facilities include DESI on the KPNO 4-meter, HETDEX on the HET, WEAVE on the William Herschel Telescope (WHT), EMIR on the GTC,

_______________

² Recovered from ADS in February 2015 (Smithsonian Astrophysical Observatory/NASA Astrophysics Data System (ADS), Query Results from the Astronomy Database, http://tinyurl.com/42jxy).

³ The time spent within different regions of the LSST footprint on the sky is still under discussion and should be chosen with a view to increasing its scientific value. Conversely, future space missions like WFIRST may find it valuable to choose survey fields that overlap with LSST.

⁴ A. Abate, J.A. Newman, and S.J. Schmidt, 2014, “Spectroscopic Needs for Training of LSST Photometric Redshifts,” white paper submitted to the committee.

Prime Focus Spectrograph⁵ (PFS) on Subaru, the Maunakea Spectroscopic Explorer (MSE) to replace CFHT,⁶ and LAMOST (China) in the Northern Hemisphere, with 4MOST on VISTA, MOONS at the VLT, and, if it proves possible to correct its wide field, GMACS on the GMT in the Southern Hemisphere.⁷ Except for DESI, these are all on private telescopes or at non-U.S. facilities to which U.S. astronomers do not have access. There is a lack of wide-field, multiplexed spectroscopic capability in the Southern Hemisphere to which the U.S. community has ready (public) access. This deficiency will negatively impact the ability to achieve maximum science from LSST.

For cosmological studies, massive, sufficiently deep and broad redshift surveys of galaxies and quasars enable measurement of the baryon acoustic oscillation feature in the spatial two-point correlation function and of the redshift-space anisotropy of clustering due to redshift-space distortions. The former provides a geometrical probe of the cosmic expansion history, and the latter constrains the growth rate of large-scale structure over cosmic time. Together these techniques provide constraints on dark energy, neutrino masses, and modified gravity. In recent years, it has been recognized that the constraints on dark energy can be tremendously enhanced by the synergy between these two 3D clustering measure-

_______________

⁵ H. Murayama, R. Ellis, T. Heckman, M. Seiffert, and D. Spergel, 2014, “Prime Focus Spectrograph on Subaru to Follow Up LSST Targets,” white paper submitted to the committee.

⁶ A.W. McConnachie, J. Bullock, P. Garnavich, P. Guhathakurta, G. Hasinger, M. Mateo, M. Strauss, and B. Tully, 2014, “The Maunakea Spectroscopic Explorer (MSE) Status Update,” white paper submitted to the committee.

⁷ Acronyms, especially those denoting individual instruments and missions, are defined in Appendix C.

ments, and the complementary 2D weak-lensing measurements that will come from LSST. That synergy is enhanced by maximizing the sky overlap of the 3D and 2D samples and by having the galaxies responsible for the 2D lensing well sampled in the spectroscopic survey. DESI on the KPNO 4-meter, HETDEX on HET, and the PFS on Subaru will carry out such surveys in the north; they overlap about half of the LSST extragalactic survey area.

A complementary capability in the Southern Hemisphere would significantly extend this synergy.⁸ Such a large, deep spectroscopic survey would also reduce systematic errors associated with weak lensing intrinsic alignments in LSST. This capability would enable other science as well, such as getting spectroscopic redshifts for supernovae and constraining maps of faint galaxies, dark matter halos, and the dark matter environment of clusters. Having an instrument on the Blanco 4-meter such as the DESI spectrograph could also be beneficial in targeting brighter objects. For example, DESI could in principle be moved to the Blanco 4-meter after its Northern Hemisphere survey if that proved cost-effective, technically feasible, and timely.

As discussed in Section 4.1, 4-meter-class telescopes fill a niche for a wide range of science, such as narrow band imaging, polarimetry and spectropolarimetry, spectroscopy, and special cadences. Although the Portfolio Review Committee calls for divestment of the Mayall 4-meter,⁹ that report notes that it is well suited for wide-field, Northern Hemisphere imaging but that its instrumentation is inferior to DECam. If the Mayall telescope were to continue operating beyond DESI, there would be a benefit to moving DECam there after its survey work on the Blanco

_______________

⁸ S.J. Schmidt, J.A. Newman, and A. Abate, 2014, “Spectroscopic Needs for Calibration of LSST Photometric Redshifts,” white paper submitted to the committee.

⁹ National Science Foundation (NSF), 2012, Advancing Astronomy in the Coming Decade: Opportunities and Challenges. Report of the National Science Foundation Division of Astronomical Sciences Portfolio Review Committee, http://www.nsf.gov/mps/ast/portfolioreview/reports/ast_portfolio_review_report.pdf.

4-meter is complete. DECam on the Mayall would provide data that are otherwise unavailable for the northern sky. DECam would go deeper than the SDSS or ZTF, cover more area in a given time than the CFHT Lensing Survey or Subaru Hyper Suprime-Cam, and could in principle be made available to the community (Mayall, along with the now-divested KPNO 2.1-meter, provided most of the Northern Hemisphere public access to U.S. astronomers). DECam could provide photometry for the 20 million galaxies studied by DESI and could be used for a very-wide-area northern weak lensing/cluster survey that would in part complement LSST.

There are other important U.S.-based resources that have the potential for significant scientific interactions with LSST. The Magellan Consortium¹⁰ operates two 6.5-meter telescopes at Las Campanas. These telescopes have superb imaging, wide fields of view, and a full complement of state-of-the-art imaging and spectroscopic instruments in the optical and near-infrared (IR), including a very successful adaptive optics secondary. Many of the partner institutions have strong connections to LSST. In the past, public access to the Magellan Telescopes was provided by NSF through the Telescope System Instrumentation Program. For very faint or distant objects, there will be strong scientific motivations for supplementary observations with the largest facilities available. These will include GMT in the south, fruitfully located just 1 degree of latitude north of LSST, and TMT along with Gemini North, the Keck 10-meter telescopes, LBT, HET, and MMT in the north, which will overlap with a large percentage of LSST sky coverage.¹¹ These facilities could be accessed by the community through a telescope time exchange, which is described and recommended in Section 6.2.

LSST Follow-up Observations Beyond OIR

Observatories operating outside the OIR bands are already planning for LSST follow-up observing. The queue-scheduled VLA now spends 10-15 percent of its observing time on time domain projects. The VLA is dynamically scheduled, and

_______________

¹⁰ Led by the Carnegie Observatories, with participation from Harvard University, University of Arizona, University of Michigan, and MIT, the consortium includes more than 200 senior astronomers, 100 postdoctoral astronomers, and nearly 100 Ph.D. students.

¹¹ The white paper submitted to the committee by McConnachie notes that at Maunakea about 1,500 square degrees of the nightly sky covered by LSST is instantaneously available; about 50% of LSST’s total 20,000 square degrees sky coverage is accessible at Maunakea (McConnachie et al., 2014, “The Maunakea Spectroscopic Explorer (MSE) Status Update”).

its observing schedule can be rapidly altered in response to time-sensitive events. New modes of operation are being designed for the VLA to complement LSST. An interrupt mode is under development that will allow the VLA to be commanded by external triggers from event brokers, giving an on-sky response that is limited only by the slew time of the antennas (typically several minutes). A preliminary version of this mode will be tested in the optical when ZTF comes online.

5.2 COORDINATING TRANSIENT OBSERVATIONS IN THE LSST ERA

The time domain, that is, the detection and study of variable sources, is a strong theme in New Worlds, New Horizons in Astronomy and Astrophysics¹² (NWNH) and an extremely important driver for LSST. A major aspect of maximizing the science in the LSST era will be spectroscopic follow-up of transient events. The Lick Observatory Supernova Search (LOSS), Palomar Transient Factory (PTF), the Panoramic Survey Telescope and Rapid Response System (Pan-STARRS), and the All-Sky Automated Survey for Supernovae (ASAS-SN) have already produced an abundance of new transient discoveries. Soon to come online, the Asteroid Terrestrial-impact Last Alert System (ATLAS), a NASA-funded high-cadence, all-sky monitor, will discover every supernova brighter than 19th magnitude in V-band and all Type Ia supernovae to a redshift z = 0.01. The ZTF will feature a camera that can scan 3,750 square degrees of the sky each hour.

LSST will bring a new chapter to time domain astronomy with an unprecedented set of challenges because of its huge rate of discovery of transient objects. It will be both desirable and necessary to achieve a follow-up response time as small as possible to capture the physics of the quickest transient events. As an example, long gamma-ray bursts have 30-second outbursts. Currently, there is virtually no spectroscopy or spectropolarimetry of these objects during their outbursts. There are important and controversial issues about the nature of Type Ia supernova progenitors, the production and distribution of radioactive elements, and the supernova explosions, which need to be addressed within the first hours and days. Perhaps most importantly, but by nature unpredictably, there will be events of unknown origin. LSST will drive changes in the sociology of astronomy in ways that are beginning to happen now, since there will be too much data for any one group to handle. There is a need to develop the software to recognize different types of transients and to coordinate the facilities that can be used for ancillary observations.

_______________

¹² NRC, 2010, New Worlds, New Horizons in Astronomy and Astrophysics, The National Academies Press, Washington, D.C.

Event Brokers

The required response to transient objects can be divided into three regimes according to the scientific goals and technical requirements. Several different categories of transients, with their brightnesses and decay times, are shown in Figure 5.1.

1. Observations within days to weeks. These might be traditional targets of opportunity for which an event has occurred that requires an unforeseen observation, such as asteroids, comets, and stars and galactic nuclei in outburst. These events require a policy that allows someone to identify such a need and to make a decision to direct that the observation be obtained.
2. Observations within hours to a day. Examples might be supernovae, explosive events that fall between novae and supernovae in peak luminosity and timescale, microlensing detections of exoplanets, and tidal disruption events. For such events, active human intervention to make decisions is appropriate. Current telescopes that claim quick response to targets of opportunity operate in this way. This requires policies that allow intervention and fairly rapid communication.
3. Observations as rapidly as possible (minutes or less). Examples are long and short gamma-ray bursts, fast radio bursts, supernova shock breakout, and gravity wave detections. These events require a facility to be connected in such a way that requests for immediate observations flow without human intervention. This is new ground that requires the development of new policy, procedures, and infrastructure for observatory operation essentially as a time domain system.¹³

The PTF, Pan-STARRS, and LCOGT groups have found that complex, dedicated software must be developed to maximize the scientific output of their programs. The task of selecting what is interesting is now the bottleneck, requiring human intervention. With a discovery rate of many events per night, the task of weeding out false-positive events, identifying new variables, deciding which are important for scientific follow-up, and deciding what sort of follow-up is needed cannot be handled efficiently by humans. These groups are developing software known as “event brokers” (or “marshals”) that can automate these processes. This is a complex task that has required many person-years to construct in relatively

_______________

¹³ T. Tyson, Large Synoptic Survey Telescope, “LSST Time Domain Data Products,” Transient Phenomena in Astronomy and Astrophysics Workshop, October 2014, http://www.gmtconference.org/Tyson_LSST_GMT.pdf.

FIGURE 5.1 Different categories of transient objects, with their R-band peak magnitudes as a function of their characteristic decay timescales. SOURCE: A. Rau, S.R. Kulkarni, N.M. Law, J.S. Bloom, D. Ciardi, G.S. Djorgovski, et al., 2009, Exploring the optical transient sky with the Palomar Transient Factory, Publications of the Astronomical Society of the Pacific 121:1334.

restricted environments in which each group operates its own discovery and followup facilities.¹⁴

LSST will detect changes in position or flux for around 2 million objects per night and produce an alert for each of them. While the great majority of these will be associated with known objects (variable stars, quasars, main belt asteroids), LSST will also detect a large number of new transient sources per night. There will also be artifacts due to poor subtractions, glints, diffraction spikes, Poisson fluctuations in the background, and other instrumental effects. The LSST project expects to use

_________________

¹⁴ See the following white papers submitted to the committee: T. Matheson, S. Ridgway, K. Olsen, and A. Saha, 2014, “Optical/Infra-red Spectroscopy of Transients and Variables in the LSST Era”; R.A. Street, C. McCully, T.A. Lister, D.A. Howell, and J. Parrent, 2014, “Time Domain Astronomy in the Era of LSST”; W.T. Vestrand and P.R. Wozniak, 2014, “The Follow-Up Crisis: Optimizing Science in an Opportunity-Rich Environment.”

a combination of improved pixel-level algorithms and, where necessary, machine-learning techniques to reduce the rate of false positives to an acceptable level. Event brokers must aggregate diverse information for each transient detection, allowing the filtering of the huge LSST alert stream into many manageable streams, one for each science project, and generate requests for follow-up observations.

Writing and operating such an event broker is a huge task that must be done to make maximum utility of LSST. There are several current research efforts,¹⁵ such as ANTARES, but the problem is not yet solved. What will be needed for LSST is a single system or a small number of systems that provide a framework that could accommodate a broad range of science applications simultaneously. It is desirable that this work be coordinated so that redundancy is minimized. Ideally, event brokers will identify the necessary follow-up observations and the proper facilities to make those observations, allowing for local weather conditions and other contingencies, and then assign those observing tasks to a range of facilities around the globe within minutes. Coordinating heterogeneous telescopes and instruments will be harder than current smaller, homogeneous efforts.

Follow-up of transients will be much more effective if automated, even roboticized. The need to respond more quickly to gamma-ray bursts and supernovae drove the development of robotic telescopes that could respond with no human intervention. Remote observing would also serve to provide fast response.

The most efficient ancillary facilities to respond to the call from an event broker will enable interruption of observing programs already in place. These calls may be in response to targets of opportunity, which LSST will have in great abundance, or to synoptic observations for which precise timing is key. Other observations may require coordination of several facilities over a variety of wavelengths, on

_______________

¹⁵ J.S. Bloom, J.W. Richards, P.E. Nugent, R.M. Quimby, M.M. Kasliwal, D.L. Starr, D. Poznanski, E.O. Ofek, S.B. Cenko, N.R. Butler, S.R. Kulkarni, et al., 2012, Automating discovery and classification of transients and variable stars in the synoptic survey era, Publications of the Astronomical Society of the Pacific 124(921):1175-1196; E. Terziev, N.M. Law, I. Arcavi, C. Baranec, J.S. Bloom, K. Bui, M.P. Burse, H.K. Das, R.G. Dekany, A.L. Kraus, et al., 2013, Millions of multiples: Detecting and characterizing close-separation binary systems in synoptic sky surveys, The Astrophysical Journal Supplement 206(2):11; H. Brink, J.W. Richards, D. Poznaski, J.S. Bloom, J. Rice, S. Negahban, and M. Wainwright, 2013, Using machine learning for discovery in synoptic survey imaging data, Monthly Notices of the Royal Astronomical Society 435(2):1047-1060; T. Matheson, X. Fan, R. Green, A. McConnachie, J. Newman, K. Olsen, P. Szkody, and W.M. Wood-Vasey, 2013, Spectroscopy in the era of LSST, arXiv:1311.2496 [astro-ph.CO]; A. Sahu, T. Matheson, R. Snodgrass, J. Kececioglu, G. Narayan, R. Seaman, T. Jenness, and T. Axelrod, 2014, ANTARES: A prototype transient broker system, Proceedings of the SPIE 9149, Observatory Operations: Strategies, Processes and Systems V, 914908; S.T. Ridgway, T. Matheson, K.J. Mighell, K.A. Olsen, and S.B. Howell, 2014, The variable sky of deep synoptic surveys, The Astrophysical Journal 796(1):53; G. Elan Alvarez, K. Stassun, D. Burger, R. Wiverd, and D. Cox, 2015, A prototype external event broker for LSST, American Astronomical Society, AAS Meeting 225, 336.37.

the ground and in space. Some targets may require new scheduling of multiple observations of the same target for a significant period of time. Queue scheduling may be a special advantage for the required response. On the other hand, the astronomical community is burdened with ToOs; even now there are too many, and they are too disruptive of telescope schedules. There may be a need to develop other strategies. One possibility is to concentrate on one area of the sky. A followup telescope could be trained on an LSST field and dedicated for a set period of time so that rapid follow-up of a new discovery could occur with only moderate repositioning of the telescope. Current, smaller-scale surveys (e.g., Gaia, PTF) are already forcing new approaches to many of these problems, and new software tools and policy changes are allowing research groups to enhance coordination between discoveries and follow-up.

Coordination is required to maximize the science in the LSST era.¹⁶ It will not happen without early and intense support. At the national level, the need is to use science requirements to set standards and protocols for event brokers, not their details. This could be a task for the future OIR System coordinator (e.g., National Optical Astronomical Observatory, as described in Chapter 6). Broker efforts need to be coupled with statisticians and computer scientists to learn how to handle sparse data in parameter space and other research-level technical issues. Care must also be taken not to put so much structure in place that the broker system is inflexible, stifling individual initiative. There is a need to correlate with archives, with SDSS being an example of a successful archive in the context of contemporary transient searches. There is also a need for coordination with multi-wavelength observations.

A modular approach to the development of event broker systems, interacting with LSST, would allow different groups to collaborate and build on one another’s efforts. Development of such systems could be helped through a solicitation or an existing program, or through fostering public-private collaborations. The committee notes the important self-organization that has occurred in the form of a series of workshops¹⁷ over the past few years.

_______________

¹⁶ See, for example, the discussion on coordination in the white papers submitted to the committee by Walkowicz et al. (L. Walkowicz, A. Mahabal, M. Agüeros, A. Becker, H. Bond, B. Frye, J. Grindlay, V. Kalogera, S. Kanbur, K. Long, M. Moniez, et al., 2014, “OIR Time Domain Astronomy in the Era of LSST”; M. Blanton, K.G. Stassun, and R. Walterbos).

¹⁷ “Hotwiring the Transient Universe,” the first of which was held in 2007 and the fourth of which is scheduled for May 2015, organized by LCOGT.

Needed Capabilities for Transient Follow-up

It will be important to have an appropriate suite of facilities and instruments to respond to significant LSST transients and to coordinate follow-up observations; NSF can employ facilities under its control to maximize the science of transient studies. Especially important are those facilities co-located with LSST, namely, SOAR, Blanco, and Gemini South. Although the focus here is on the southern facilities due to their proximity to LSST, facilities in the Northern Hemisphere, including Gemini North, will cover 30-50 percent of the sky that will be surveyed by LSST, and hence will also have a potential role in LSST follow-up.

Events that pass the LSST event broker system may be relatively rare and thus single objects not requiring a highly multiplexed instrument. A response time of minutes shepherded by an appropriate event broker may be especially valuable for some transient science. Low-dispersion spectrographs are the instrument of choice for first, quick-look spectroscopy and much of the required follow-up.

Spectropolarimetry

Spectropolarimetry is employed in many areas, including studies of active galactic nuclei, galactic and extragalactic magnetic fields, dust and extinction, circumstellar and proto-stellar disks, exoplanets, active stars, and supernovae.¹⁸ It is important that spectropolarimetry remain part of the suite of responses in the LSST era. Current studies show, for instance, that not only are supernovae not round in aspect, but that different chemical species are ejected with different geometries. Total flux spectra cannot provide this sort of information. Coupling the depth information from total flux spectra with the 2D information projected

_______________

¹⁸ B.-G. Andersson, A. Adamson, K.S. Bjorkman, J.E. Chiar, D.P. Clemens, D.C. Hines, J.L. Hoffman, T.J. Jones, A. Lazarian, C. Packham, J.E. Vaillancourt, et al., 2014, “The Need for General-Use Polarimeters in the Era of LSST,” white paper submitted to the committee.

on the sky with spectropolarimetry yields full 3D, time-dependent, “tomographic” information on the object of study.

Near-IR Capability

Near-IR capability is also very important for transient studies, even when the transient is detected in the optical. As an example, the spectra of supernovae are much less blended in the near-IR than in the optical. This makes identification of certain elements—for instance, helium—much easier to detect in the near-IR. The near-IR is also important to detect weak signatures of hydrogen in otherwise hydrogen-deficient events and to study molecular features such as the CO band head.

SOAR

With its 4.1-meter aperture and potential for rapid response, SOAR could become a prime instrument for early characterization of LSST transients.¹⁹ The critical ingredient provided by SOAR would be early (but then repeated with appropriate, science-driven cadence) low-resolution spectroscopy and spectropolarimetry. This capability would not simply characterize any new event; it would also provide important data on the physical properties and geometrical shape of the outburst at the very times that have been especially elusive with today’s technology. The early and ongoing data from SOAR would be archived for later analysis, but the information would also be fed automatically back into the event broker to aid in the subsequent study of the most interesting events with other facilities.

_______________

¹⁹ See the following white papers submitted to the committee: S. Heathcote, 2014, “Cerro Tololo Inter-American Observatory in the LSST Era”; J.M. Strader, E.F. Brown, L. Chomiuk, E.D. Loh, and S.E. Zepf, 2014, “A View of Astronomy at Universities in the LSST Era.”

Gemini South

The operation, especially the queue scheduling, and the near-IR capability of the Gemini telescopes makes them especially powerful facilities to respond to transient targets of opportunity. They need to be coupled to the event broker system to maximize the science output in the LSST era. Gemini South could be an especially important ingredient in the subsequent chain of observations of key transients. With its large aperture, Gemini South could provide timely, higher-resolution spectroscopy, particularly with an instrument with high throughput and moderate resolution. For objects that occur sufficiently far north, Gemini North can also play an important role in this follow-up. For objects in the equatorial swath, the complementary instrument suites in the two Gemini telescopes may be a positive aspect in the follow-up campaigns.

GSMTs

Follow-up of LSST discoveries will undoubtedly be a rich scientific area for Giant Segmented Mirror Telescopes (GSMTs) as well, and ideally these would factor in to the transient response system. The Giant Magellan Telescope Organization sponsored a community science meeting in Washington, D.C., in October 2014 focusing on transient phenomena²⁰ that would benefit from GSMT follow-up. As just one example, neutron star-neutron star collisions are associated with short gamma-ray bursts, but they should also produce optical transients that would be very faint and brief. Even when there is a successful strategy for localizing such events, only the largest telescopes will be able to make effective observations in the limited time they are detectable. For this task, the GSMTs will need to be equipped with the capability to do rapid, single-object spectroscopic and spectropolarimetric observations.

_______________

²⁰ Second Annual GMT Community Science Meeting, 2014, http://www.gmtconference.org/, accessed February 1, 2015.