A Strategic Vision for NSF Investments in Antarctic and Southern Ocean Research (2015)

CHAPTER THREE

Suggested Research Priorities for the Coming Decade

This chapter discusses the first two major components in the Committee’s strategic vision for NSF investments in Antarctic and Southern Ocean science. The first section describes our rationale for recommending continuation of the current “bottom-up” mode of support for research across a broad base of topics. The second section offers our recommendations for a very limited set of research priorities to support as large-scale community-driven initiatives in the coming decade.

MAINTAIN A CORE PROGRAM OF INVESTIGATOR-DRIVEN RESEARCH ACROSS A BROAD BASE OF TOPICS

A number of previous NRC reports (including Rising Above the Gathering Storm [NRC, 2007], chaired by Norman Augustine) have outlined why it is in the nation’s best interest to support basic science research across a broad spectrum of disciplinary fields. Although there are strategic reasons for the United States to take a leadership role in targeted areas of science and engineering, it is impossible to predict where the next major breakthroughs or advances will happen. Thus to ensure that the nation is well positioned to take advantage of such breakthroughs, it is important to be engaged in all core areas of scientific research.

This recommendation also holds true for research supported in the U.S. Antarctic Program (USAP). It is in the national interest for U.S. researchers to be active in all basic areas of Antarctic and Southern Ocean science, since it is unknown when or where critical advances will occur in the future, and likewise unknown when or where future major environmental changes may emerge. The Committee thus fully endorses NSF’s Division of Polar Programs (NSF/PLR’s) well-tested model of supporting research across a broad spectrum of topics, in response to the ideas that emerge in proposals from the research community. Indeed, numerous participants in our community forums voiced concerns about the dangers of NSF moving away from this bottom-up model of research support. For instance, one respondent stated: “NSF must remain nimble and responsive. More mission-driven, large-team based science . . . will be at the expense of being responsive to new ideas and exciting discoveries that we don’t even know about yet. These discoveries are often made by single or small groups of

PIs thinking outside the box, or with a crazy new idea, or even just making the first observations from a new place. There has to be continued support for that, financially and logistically.”

The Committee agrees with this sentiment and notes that many of the important scientific breakthroughs that have occurred in Antarctic and Southern Ocean science have come from relatively small-scale, hypothesis-driven, curiosity-inspired research (see Box 3.1). Nonetheless, the balance of “small” versus “large” science takes on a unique cast in the context of USAP field research. There is no truly “small science” in the Antarctic—given the significant logistical needs and costs associated with getting any researchers into the field in that part of the world.

In this context, we suggest that NSF’s traditional broad-based research support strategies be balanced with more directed, larger-scale efforts aimed at ensuring that a critical mass of human and financial resources is concentrated on meeting key research goals over a limited period of time. With the large costs and logistical challenges

involved in accessing certain regions of interest (e.g., remote coastal or mountainous terrain, subglacial environments), large organized campaigns can provide major economy-of-scale benefits. NSF/PLR has always maintained some degree of balance between the smaller bottom-up research activities and the larger programmatic-level campaigns; and thus our Committee’s overall vision does not differ significantly from PLR’s existing strategies in this regard.

The suggested priority topics discussed below focus on relatively large research initiatives—efforts that seem likely to require a coordinated “push” in order to move forward at a sufficient scale and time frame. In other words, these are not the types of research activities that are likely to proceed of their own accord through the traditional process of proposals submitted by individuals or small groups of investigators. PI-driven proposals should indeed continue to receive support, and many of the research ideas raised in our community input process are the appropriate scale (and are indeed compelling candidates) for this category of bottom-up research. Those ideas are not all individually discussed here in this report, but the full collection of that input may serve as a rich database of ideas and inspiration for PLR program managers.

A theme that arose often in our community discussions and is relevant to a wide array of PI-driven research is the need for improved integration and coordination in the collection, management, and analysis of in situ observations. Two good examples of this relate to coordination of science in McMurdo Sound and the Ross Sea region—areas that have been the focus of intense study for decades by U.S. researchers (see example in Box 3.2). For instance, the integration of existing research activities could be greatly aided by creating a Ross Sea “Research Coordination Network.” This could provide a mechanism to support workshops that foster new collaborations and address interdisciplinary topics, to identify novel networking strategies and collaborative technologies, and to develop community standards for data and metadata. This would also link well with efforts of the international Southern Ocean Observing System, which has identified the Ross Sea region as a key area for enhancing regional-scale cross-disciplinary research.

A related example raised in our community discussions illustrates how coordinated data collection and management can eliminate duplication and improve research efficiency. Many ecological studies around McMurdo Sound (and likewise around Palmer Station) require collecting basic physical oceanographic observations such as temperature, salinity, and pH profiles, as context for interpreting biological observations. Typically, individual research groups each collect such observations themselves; but a possible alternative is to maintain a few autonomous observing assets on moorings and floats that collect standard physical oceanographic data in key locations over the

course of a field season. If properly managed, these data could then be used by numerous studies, and without the burden of collecting such duplicative information themselves, those studies would have lower overall costs.

Recommendation: NSF should continue to support a core program of broad-based, investigator-driven research and actively look for opportunities to improve coordination and data sharing among independent studies.

SUPPORT A FEW STRATEGICALLY CHOSEN LARGE RESEARCH INITIATIVES

The Committee was charged to identify priorities for NSF/PLR research support, drawing upon the ideas collected from the research community at large and upon the judgment of the Committee itself. This type of prioritization exercise of course presents a daunting challenge—especially given the huge diversity of compelling research topics that are addressed in Antarctic and Southern Ocean research. But in a time of relatively flat budgets and rising infrastructure and operational costs, hard

choices must be made by NSF. This study aims to ensure that those choices are driven by a strategic perspective and by a broad base of ideas and opinions.

This sort of prioritization exercise is inevitably a subjective process to some degree. To provide some structure and objective rigor to this analysis, the Committee agreed upon a set of criteria to evaluate the many different options that arose in our community outreach. This list was developed by examining all the many types of evaluation criteria that have been used in similar exercises carried out by other NRC study committees over the past several years, and tailoring that list to best fit the context of Antarctic and Southern Ocean science. The selected criteria are listed below. “Compelling science” was seen as the primary filter, as all candidate topics must meet this standard. The subsequent criteria were viewed as important but not always necessary in every case.

• Compelling science: research that has the potential for important, transformative steps forward in understanding and discovery;
• Potential for societal impact: research that yields information of near-term and/or long-term benefits for society;
• Time-sensitive in nature: involves systems/processes undergoing rapid change that need to be observed sooner rather than later; research that could help inform current public policy concerns;
• Readiness/feasibility: research that is poised to move forward quickly within the coming decade, in terms of needed technologies and community readiness;
• Key area for U.S. and NSF leadership: research for which the United States (and NSF in particular) is advantageously positioned to lead.

Additional factors that were considered:

• Partnership potential: research that will allow NSF/PLR to leverage investment by other federal agencies, other parts of NSF, other international partners;
• Impacts on program balance: research that will not cause significant adverse impacts on other projects by requiring disproportionate funding or logistical support;
• Potential to help bridge existing disciplinary divides: research that will provide opportunities to bring together disciplinary communities that seldom work together.

Weighing the ideas that arose in the community input against these evaluation criteria, and building on their own expertise and judgment, the Committee identified

a short list of topics to recommend as priorities for NSF/PLR investment, listed below. The first of these topics is the largest of the three, and in the Committee’s judgment is the most urgent to pursue.

Recommendation: NSF should pursue the following as strategic priorities in Antarctic and Southern Ocean research for the coming decade:

1. How fast and by how much will sea level rise?
   • A multidisciplinary initiative to understand why the Antarctic ice sheets are changing now and how they will change in the future.
   • Using multiple records of past ice sheet change to understand rates and processes.
2. How do Antarctic biota evolve and adapt to the changing environment? Decoding the genomic and transcriptomic bases of biological adaptation and response across Antarctic organisms and ecosystems.
3. How did the universe begin and what are the underlying physical laws that govern its evolution and ultimate fate? A next-generation cosmic microwave background program.

For each of these topics, we offer below a brief description of the fundamental scientific questions to be addressed, and the ways in which the topic meets our key evaluation criteria. We also offer some details on how these different initiatives might conceivably move ahead in the coming years—although in all cases, working out the detailed implementation plans will require subsequent rounds of engagement and deliberation between NSF leadership and staff and the relevant research communities.

Strategic Priority I How Fast and by How Much Will Sea Level Rise? The Changing Antarctic Ice Sheets Initiative

Background Context and Motivation

Twenty thousand years ago, massive ice sheets covered Chicago, Albany, Toronto, and Oslo. Rapid episodes of ice retreat and subsequent rise in global sea level are recorded by proxy indicators in terrestrial deposits and sediments throughout the North Atlantic, and in ice and sediment cores from Greenland and Antarctica. At times the rates of sea level rise were at least 1 m per century, and probably faster—but exactly how fast is not well known.

Today three large ice sheets remain on our planet, one in Greenland and two in Antarctica—the West Antarctic Ice Sheet (WAIS) and the East Antarctic Ice Sheet (EAIS). Mapping of the ice has long shown that large areas of the WAIS and some regions of the EAIS lie on bedrock well below sea level. These regions, known as marine-based ice sheets, are the most responsive to climate change. The coastal areas of the northern WAIS and major portions of the EAIS (as well as the deep glaciated fjords of Greenland) are all in contact with the warming ocean today, and a rapid, large-scale response has begun. Increased ice flow, thinning ice, and a major fraction of the ongoing sea level rise are all consequences of this interaction. If the marine-based ice from both East and West Antarctica were fully melted, it would contribute over 20 m to global sea level rise. Understanding how and how fast these ice sheets could collapse is thus critical for understanding how future sea level rise might proceed.

Theory predicts certain marine-based ice sheets are subject to a runaway collapse process known as marine ice sheet instability (MISI) (Hughes, 1981; Mercer, 1978). For MISI to occur, the bedrock must slope toward the interior of the ice sheet, so that the water depth increases as the ice front retreats away from the sea. If warming water or other changing climatic conditions initiate a retreat of the ice front, the water depth beneath the front and the thickness of the ice both increase, and this in turn increases the speed of the ice flow. Faster flow leads to thinning of the ice sheet, and the ice becomes ungrounded (floating), causing further retreat and setting up a positive feedback mechanism that leads to even more rapid ice loss.

In West Antarctica, particular interest thus centers on areas where the bedrock slopes most strongly toward the interior of the ice sheet, making it susceptible to MISI. No modern analog to WAIS collapse has actually been observed, and so the detailed physical processes by which collapse occurs are not well understood. This lack of understanding translates directly into model uncertainty on the predicted speed and extent of WAIS collapse. But satellite observations have confirmed that in certain areas of the WAIS, glacier loss is accelerating, the ice sheet surface is lowering rapidly, and overall the ice sheet is losing mass at an accelerating pace. (See Figure 3.1.) Evidence is also building that these key areas are becoming unstable and beginning to collapse (Joughin et al., 2014; Rignot et al., 2014). A region of particular concern is Thwaites Glacier in the Amundsen Sea sector, because available modeling shows that loss of Thwaites would in turn lead to the loss of all marine ice in West Antarctica (Pollard et al., 2015). Some models suggest that such a scenario is now inevitable and could occur rapidly, but there is still active debate on this question (Alley et al., 2015; Favier et al., 2014; Joughin et al., 2014; Parizek et al., 2013; Pollard et al., 2015). Most importantly, it is unclear whether this collapse might occur within 200 years or 2,000 years.

FIGURE 3.1 Top: Change in thickness and volume of Antarctic ice shelves from 2002 to 2012. Rates of thickness change (meters/decade) are color-coded from –25 (thinning) to +10 (thickening). Circles represent percentage of thickness lost (red) or gained (blue). SOURCE: Paolo et al., 2015. Bottom: Change in the mass of the Antarctic ice sheet from 2003 to 2014, based on data from the GRACE satellite mission which measures gravity changes. In the past 11 years, the Antarctic ice sheet lost 92 billion tons of ice per year. The vast majority of ice loss was from West Antarctica’s Amundsen Sea region (box a) and the Antarctic Peninsula (box b). The ice sheet on East Antarctica (box c) primarily thickened during that time. Color scale indicates mass (equivalent to centimeters of water) of the land ice, with red denoting the largest loss and blue denoting the largest gain. SOURCE: Harig and Simons (2015).

Understanding how sea level may rise as the WAIS responds to a changing climate is critical for evaluating the level and urgency of societal response that will ultimately be required. Estimates of overall sea level rise from the most recent Intergovernmental Panel on Climate Change report range from 26 to 98 cm in this century (IPCC, 2013), but recent modeling studies suggest that far more rapid rates are possible and perhaps even likely (Alley et al., 2015; Bamber and Aspinall, 2013). Slower collapse scenarios (extending over millennia) may mean that society will have significant time to adapt, whereas rapid collapse scenarios imply a potential need for massive investment in infrastructure to protect coastal infrastructure and ecosystems in the United States and globally (Box 3.3)—as well as attention to issues such as displaced populations from flooded areas and intrusion of saltwater into freshwater aquifers.

Moreover, the rise of sea level around the globe will not be uniform. Relative to the global average, the United States is particularly sensitive to a change in sea level from West Antarctic ice loss. When ice mass is lost from an ice sheet, the ocean is pulled less strongly toward that coastline, which lowers sea level near the ice sheet and raises sea level elsewhere by a larger amount. Studies have found that due to these gravitational effects, ice loss from the WAIS will be amplified along the East and West coasts of North America by about 25 percent (Mitrovica et al., 2009; NRC, 2012).

West Antarctica was first identified as a critical and potentially unstable part of Antarctica through USAP-supported research on the Ross Ice Shelf, the Siple Coast, and the ice sheet interior. Subsequent research investments in the West Antarctic region—including campaigns on Pine Island Glacier, the Whillans Ice Stream, WAIS Divide, and the Amundsen Sea Polynya International Research Expedition—have provided a strong foundation for the United States to lead the critical research challenges that must still be addressed in this region. For example, the United States:

• Has a research community that has published numerous high-profile scientific studies on the region’s evolving climatic, cryospheric, and oceanic conditions (e.g., Dutrieux et al., 2014; Holland et al., 2010; Jacobs et al., 2011; Joughin et al., 2014; Steig et al., 2009); and a new generation of scientists trained with the critical skills needed to study the ice sheets as part of a complex coupled system, and to integrate remote sensing with field measurements;
• Has a strong base of infrastructure support at McMurdo Station (which will be further strengthened with the currently planned modernization program), and has a strong track record of support for deep-field research in West Antarctica—thanks in part to unique airlift access to the ice sheet interior via LC-130, Basler/DC-3, and DHC-6 Twin Otter aircraft;
• Has developed the capacity to efficiently drill through the ice shelves and to install complex sensors and instrumentation on, in, and beneath the ice, and in the surrounding oceans;
• Has significant research support infrastructure in the region, including proven runways, summer research camp installations, and a partial array of ocean moorings and automatic weather stations, and established overland traverse capability (aided by advances in crevasse detection techniques and discovery of crevasse-free routes);
• Has conducted numerous research cruises around West Antarctica that have yielded much of the detailed bathymetry, ocean measurements, and marine sediment cores acquired in the region (Jacobs et al., 2012).

To build on these past successes, the Committee suggests that the NSF launch an ambitious new integrated, interdisciplinary, international research program referred to here as the Changing Antarctic Ice Sheets Initiative (Changing Ice Initiative for short). This initiative will improve understanding of how the marine ice sheets of Antarctica have changed in the past, how they are changing now, and how they will change in the future; and it will advance knowledge of how ice sheets work as part of the global ice–ocean–atmosphere system. The ultimate goal of this work is to improve quantitative projections of how much and how fast sea level will rise in the future.

The initiative will involve two distinct but linked components that address different aspects of the same underlying challenge. One component aims to better understand the drivers, mechanisms, and projections for the changes currently taking place in Antarctica’s ice sheets. The other component aims to better understand the rates and processes of future ice sheet change by examining the past behavior of the ice sheets in response to natural forcings. Together this work will comprise a large, decadal-scale initiative, involving a broad cross section of the Antarctic and Southern Ocean research community and will be more extensive in scope and duration than West Antarctic research campaigns to date. As discussed later, expanded cooperative efforts—within NSF, among the USAP agencies, and at the international level—will be needed to advance this “grand challenge” of Antarctic and Southern Ocean science.

The proposed effort, component i: A multidisciplinary initiative to understand why the Antarctic ice sheets are changing now and how they will change in the future

Ice sheets are part of a complex atmosphere–ocean–climate system, and changes in both atmospheric and oceanic forcing can trigger faster flow and melt of Antarctic ice. Increases in circumpolar winds observed over the Southern Ocean (thought to be due to both polar and tropical influences) are driving warm deep ocean water closer to the areas where ice contacts the ocean. Particular concern centers on the warm water reaching into the deepest troughs where the largest glaciers enter the ocean, and into the “grounding line” where the ice sheet begins to float on the ocean as an ice shelf. As noted earlier, these are the sites where the key processes of MISI can occur, if there is landward-sloping bedrock (i.e., sloping toward the interior of the ice sheet). This warm ocean water drives increased melt at the flotation point and on the underside of the ice shelf. (See Figure 3.2.) Questions remain about the factors controlling rates of change, but the underlying causes of retreat of the grounding line, ice shelf thinning, and increased ice flow are nonetheless clear (e.g., Dutrieux et al., 2014; Rignot, 2008; Shepherd et al., 2004).

Major gaps remain in understanding these processes, because of a lack of observations in the critical areas in the ocean and beneath the ice shelves and limited understanding of the causes of the changing Antarctic atmospheric and ocean circulation. As most of the fundamental processes driving Antarctic ice sheet change are hidden beneath the surface of the ocean and the ice sheet, they cannot be imaged from space. Understanding them requires a coordinated research effort on-site in Antarctica and the Southern Ocean, with measurements taking place over an extended time in the critical regions affecting change. In addition, numerical modeling efforts and studies to elucidate past ice sheet dynamics are needed to place these observations in a broader context.

FIGURE 3.2 A simplified schematic of the forces that affect growth and loss of an ice shelf, including ice that flows from the ice sheet, warm ocean water that melts the underside of the ice shelf, and wind patterns (e.g., coastal easterlies, sub-antarctic westerlies) that affect ocean and sea ice dynamics. The marine ice sheet instability (MISI) process occurs where the bedrock behind the grounding line slopes inward toward the interior of the ice sheet.

Historically, ice sheets have been studied in isolation from atmospheric and oceanic drivers. Proposed here is a more integrated effort involving both ocean- and ice sheet-based investigations. This effort will require investment in new instrumentation for detailed mapping of the seabed and subglacial bedrock, for long-term on-site observations, and for exploration of the regions where key processes occur—including explorations of the ocean cavities beneath the ice shelves and the edges of the continental shelf.

The main elements of this initiative include interdisciplinary studies of key processes, systematic measurements of the atmospheric and oceanic drivers of change, mapping of unknown submarine and subglacial terrains that modulate rates of change, and use of these new data streams in coupled models. Each of these elements is discussed further below:

1. (i)   Advancing our understanding of the complex ice, oceanic, and atmospheric processes driving the observed Antarctic ice sheet changes, through interdis-

1. ciplinary studies of key processes such as grounding-line dynamics, ice shelf collapse, and atmospheric drivers of ocean waters.

2. (ii) Providing the first systematic measurements and understanding of atmospheric and ocean drivers of change in West Antarctica. This requires a continuous data collection presence in the region and closer integration of observations and process models. In situ observations related to atmospheric and ocean circulation, formation and transformation of ocean water masses, sea ice changes and influences, ice sheet flow and accumulation, the sub-ice-shelf and grounding-line environment, and near-coastal ocean hydrography and biogeochemistry are needed. Retrospective investigations based on targeted sediment cores, shallow ice cores, and ice borehole logging are needed to understand the century-scale context for today’s climate changes. The relative contributions of tropical and high-latitude forcings to atmospheric and oceanic behavior that affects West Antarctica need to be better understood.
3. (iii) Mapping the unknown terrains beneath the major ice shelves and the critical regions beneath the ice sheet. Technologies such as airborne

1. imaging, active seismic surveys, gliders and autonomous vehicles (AUVs), as well as traditional coastal and on-ice surveys, will be essential to determine the boundary conditions that control the fluxes and processes determining ice sheet change.

2. (iv) Advancing coupled atmosphere–ocean–sea ice–ice sheet models. These models should be optimized for the Antarctic environment and designed to build closer connections between observations, processes, and ice and ocean physics. They should be run at spatial resolutions that can resolve critical physical processes. This will help inform and improve lower-resolution global-scale coupled models being developed elsewhere in NSF, NOAA, DOE, and NASA.

These interdisciplinary studies of critical processes that control the rates of ice sheet change will require improved access to logistically challenging coastal regions of Antarctica. Some key locations for this work are Thwaites Glacier, Pine Island, and ice shelves along the northern WAIS coast. There are currently no coastal stations in the critical zone between McMurdo and the Antarctic Peninsula, a distance of roughly 4,000 km. As a result of local meteorological dynamics, the automatic weather stations

in this area of the continent are not representative of conditions over the critical ocean and sea ice zones—thus there is a need for drifting oceanic and sea ice buoys to obtain the needed measurements. Studies based around the Ross Ice Shelf and Siple Coast would also be valuable for assessing current structure and establishing a baseline for evaluation of future changes. The Ronne Filchner Ice Shelf that bounds the WAIS is an excellent target for international programs advanced by nations with strong logistics hubs in the region, including the United Kingdom, Chile, and Germany.

Working in any of these areas requires research vessel, airlift, aerogeophysical, and over-snow traverse capabilities; these infrastructure and logistical needs are discussed further in Chapter 4. New sampling approaches and technologies will also be needed, to extend the temporal and spatial reach of observations into unique environments. The needed remote sampling strategies are within reach due to recent advances in ocean and ice sheet instrumentation, including use of autonomous instrumented submarines, and advanced moorings, buoys, and gliders developed by the NSF Ocean Observatories Initiative. Attention is needed for widespread deployment of existing technologies that enable sub-ice AUV location and navigation, for instance, using acoustic transponders because, currently, AUVs cannot communicate their position to satellites from underneath the ice, and thus they have limited scientific use. A recent German program in the Weddell Sea demonstrates the success of this approach.

To help address these needs, the initiative could include supporting calls for proposals for targeted engineering development of specialized moorings, AUVs and gliders, unmanned airborne vehicles (drones or UAVs), and autonomous surface sensor stations. Emphasis for this engineering should be placed on reliability and low maintenance, small environmental footprint, and ease of removal.

The proposed effort, component ii: Using multiple records of past ice sheet change to understand rates and processes

As the second major component of the Changing Ice Initiative, the Committee proposes an integrated program of sediment and ice core observations that can directly inform and test the models used to predict future WAIS melting. This includes studies aimed at the question “How fast did WAIS collapse in the past?” and studies aimed at the question “How much sea level rise was caused by past WAIS collapse?” Both of these are discussed in turn below.

How Fast Did WAIS Collapse in the Past?

Today’s rate of climate forcing is thought to be higher than the forcing rates that triggered WAIS collapse in the past; yet the timescale over which past collapse events

occurred may be similar to what one could expect in the future. That is, the MISI process described above may have a standard characteristic timescale—and this timescale may be recorded in a variety of paleoclimatic and paleoceanographic records.

The most recent warm interval in which WAIS collapse is thought to have happened is the last interglacial period about 125,000 years ago, a time when sea level was 5-9 m higher than present and global mean surface temperature was 2-3°C higher (Kopp et al., 2009). In Northern Hemisphere records, this time interval is often referred to as the Eemian. Although it is not certain that the WAIS collapsed at this time, circumstantial evidence suggests that it did (Alley et al., 2015; Pollard et al., 2015). Availability of high-resolution ice cores that can provide the needed chronological accuracy diminishes rapidly if one looks further back in time, and so, ice core studies targeting the last interglacial are the most likely to be successful in answering the question of how fast sea level might rise. For this reason, the last interglacial has been targeted as a priority for future research by the International Partnerships in Ice Core Sciences.¹

In recent years, a wide array of paleorecords have been recovered from the last interglacial period; but in many cases, the low temporal resolution and poor dating of these samples has frustrated attempts to further constrain questions about the rate of ice sheet collapse. The WAIS Divide ice core showed it is possible to count annual layers of fine dust concentration in the ice back more than 60,000 years (McConnell et al., 2007). During interglacials when high snow accumulation rates compensate for age-driven thinning, it should be possible (in places where the ice survived) to count annual layers back to 130,000 years. High snow accumulation rates would also be expected at coastal sites that are proximal to the collapse region, which increases the chances of successfully identifying annual layers during a collapse interval. Thus pursuing these types of annually resolved ice core samples may yield particularly valuable evidence for establishing how fast Antarctic ice has melted.

Annually or near-annually resolved intervals from past ice sheet collapse phases can also plausibly be recovered in carefully selected marine sediment records. Rapidly accumulating marine sediments have been recovered from shelf basins from nearly all sectors of the Antarctic margin (Boldt et al., 2013; Escutia et al., 2011; Michalchuk et al., 2009; Milliken et al., 2009), and many of these samples are sufficiently finely layered to provide annual-scale resolution (Leventer et al., 2006; Maddison et al., 2012; Swann et al., 2013). Such records can document the timing and location of rapid ice margin retreat during intervals of warming (Mackintosh et al., 2011, 2014). New targeted collections of marine sediments from key time periods can provide new insights on how

_________

¹ See http://www.pages-igbp.org/workinggroups/endorsed-and-affiliated/ipics/.

fast and how much Antarctic ice melted during previous episodes of warmer temperatures, high sea level, and rapid deglaciation, each with different solar and oceanic forcing characteristics.

Targeted coring and drilling efforts can provide samples with chemical indicators that also allow us to infer whether collapse has occurred or not over time. These indicators can also provide estimates of environmental conditions associated with periods of rapid ice loss—such as sea ice/open-water feedbacks or the penetration of relatively warm seawater onto the Antarctic continental shelf. For instance, to detect marine ice sheet presence or absence and amount of collapse recorded in ice cores, the presence of nearby open water can be inferred from high concentrations of sulfur-containing compounds (diagnostic of nearby marine biological productivity) and high-precision seawater isotope indicators (deuterium excess, ¹⁷O excess) (Steig et al., 2015). Marine sedimentary strata typically record depositional events during periods of ice margin retreat, and thus provide insights into the timing and rates of Antarctic ice sheet advance and retreat (e.g., Anderson, 1999; Anderson et al., 2002; Gohl et al., 2013). Marine sediment cores also contain a variety of indicators of open water and geochemical recorders of seawater temperature, which provide insights on the timing and rate of ice margin retreat in specific regions and the role of ocean thermal forcing.

This initiative should include drilling one or more new ice cores from sites on the margin of the suspected WAIS collapse region. These sites should be downwind of where open marine water would have replaced portions of the WAIS, so that the ice core would contain airborne chemical and isotopic indicators of WAIS status. During our community outreach sessions, an often-mentioned sampling site is Hercules Dome, located at the boundary between East and West Antarctica (86°S/105°W; Jacobel et al., 2005). At this site, ice from the last interglacial (120-130 ka) appears to be present in the bottom 200 m of the ice sheet, based on interpretation of radar layering combined with ice flow modeling. The relatively high snow accumulation rate at Hercules Dome, combined with low mean annual surface temperatures, implies that downward advection of cold ice may have precluded loss of the target ice by basal melting. A recent modeling study suggests that the diagnostic chemical fingerprint of WAIS collapse should be detectable in a Hercules Dome ice core (Steig et al., 2015), allowing a precise estimate of the rate of collapse. Other possible candidate sites should not be excluded, however (e.g., the Whitmore Mountains), and thus an appropriate survey of suitable sites is suggested.

This initiative should also include high-resolution sediment cores from marine basins within and adjacent to the suspected WAIS collapse region, in carefully chosen locations where erosion of soft young sediment by ice flow is unlikely. Sites should be

chosen based on seismic surveys and the identification of seabed features that provide accommodation space for rapidly accumulating deposits and also limit subsequent glacial erosion.

Determining the speed of past ice sheet collapse at a level of accuracy that is directly useful for informing society’s adaptation planning decisions may require paleorecords dated accurately enough to determine the duration of the collapse with an uncertainty of half a century or less—which is very roughly the timescale for turnover/ replacement of much common coastal infrastructure. Therefore, we suggest pursuing sediment and ice cores with this scale of chronological accuracy or better. This can, however, be pursued together with other research approaches, described below, that make use of different types of indicators and thus provide independent, complementary information for constraining ice sheet models.

How Much Sea Level Rise Was Caused by Past WAIS Collapse?

To determine how much sea level rise was caused by past WAIS collapse, one must determine the geographical footprint (and thus the total volume of ice lost) during past marine ice sheet collapse. This requires mapping the areal extent of the past collapse region, and use of ice flow models to predict the thickness of the adjacent remaining ice sheet. One promising approach for gaining such information is to employ cosmogenic isotope exposure dating techniques on short bedrock cores taken from beneath the WAIS. In concept, if the ice sheet was removed during the Eemian, some areas of above-sea-level bedrock would have been exposed to cosmic radiation, and as a result, cosmogenic nuclides (e.g., ¹⁰Be and ²⁶Al) would have been produced in the top few meters of rock. It is challenging to detect cosmogenic isotopes stemming from an irradiation of only several thousand years’ duration happening 125,000 years ago, but it should be feasible (J. Stone, personal communication, 2014).

The combination of these sub-ice datasets, along with cosmogenic data from nearby moraines and related glacial deposits that record ice margin positions, will provide a rich opportunity to assess changes in WAIS ice volume over time. Access to these bedrock archives would be facilitated by rapid, agile drilling systems. The Ice Drilling Program Office is constructing an agile sub-ice rock drill for this purpose, which will reach ice depths of 700 m and take rock cores 10 m long, as well as a deep-ice rock coring system (Rapid Access Ice Drill) designed to reach 3,300-m depth with 25-m rock cores.²

_________

²http://icedrill.org/equipment/raid.shtml.

Another valuable line of evidence for evaluating how WAIS could contribute to sea level rise is to improve understanding of the “style” of past ice sheet retreat—for instance, the details of regional variation in ice retreat and interaction with outlet glaciers. Glacial geomorphologic studies that highlight changes in ice elevation and marginal position can shed light on these details, and thus may be another important contributor in guiding the modeling of the present-day ice sheet retreat. Terrestrial deposits that document WAIS recession, especially after the Last Glacial Maximum and throughout the Holocene, may provide context for evaluating WAIS collapse during the preceding (Eemian) interglacial period, as well as into the future. Likewise, reconstructed ice-surface profiles from trimlines and erratics on nunataks in West Antarctica and from moraines alongside outlet glaciers may help document changes in ice thickness as a function of both the timing and spatial pattern of grounding-line retreat, all of which provide critical test points that help refine glaciological models for rapid ice sheet retreat (e.g., Anderson et al., 2014; Conway et al., 1999; Hall et al., 2013).

A well-integrated program of sediment and ice observations yielding multiple archives and datasets can directly inform and help constrain models that are used to predict future melting of the West Antarctic Ice Sheet. Models are crucial for assimilating these datasets with other relevant observations, and making the best possible

estimates of the amount of sea level rise caused by WAIS collapse. These models can be driven by the forcing factors thought to be most critical during periods of past WAIS collapse—in particular, by local ocean temperatures, which can be validated by comparison with the proposed paleorecords (i.e., with existing ocean temperature records that are augmented by the marine sediment coring efforts proposed here). The models could then be used with modern climate-forcing estimates to project future sea level rise rates.

This research goal aligns well with priorities that have been identified by the broader scientific community—for instance, in the SCAR Horizon Scan, in the U.S. Ice Drilling Program’s Long Range Science Plan (IDPO, 2015), and in the recommendations of a 2014 conference on Paleoclimate Records from the Antarctic Margin and Southern Ocean attended by many U.S. scientists.

Programmatic Considerations

NSF/PLR has supported a tremendous amount of highly successful research on the changing ice sheets over the past few decades. The support for this work has generally been divided between four research sections: Integrated Systems, Geology and Geophysics, Oceans and Atmosphere, and Glaciology. Coordination within the research community for the WAIS activities was chaired for over two decades by Dr. Robert Bindschadler and is now chaired by four scientists on a rotating basis. These past efforts have led to many important advances in scientific understanding, but the increased urgency of the issue necessitates a greatly expanded coordination effort among the U.S. research community and at the international level, with strong leadership and support from NSF.

Within NSF/PLR, a new dedicated section or science office may need to be established to oversee such planning efforts, and to manage the ongoing fieldwork and modeling activities, the development of essential technologies, and the expansion of collaborative activities with other federal agencies and other countries. Some of the major research components outlined above could each be supported as university-based centers, perhaps modeled after the NSF’s Centers of Excellence funding platform. Beyond traditional collaborations, programs should be set up to support graduate student and postdoctoral researcher exchanges between disciplines, and cross-training in field and modeling methods.

The questions raised here span multiple science disciplines, which in many cases have traditionally had little direct interaction. Component i represents a new generation of multidisciplinary studies that bring together those studying Antarctica’s ice, ocean,

atmosphere, and climate—in a common quest to better understand the processes that currently control rates of melting and ice sheet flow. Component ii will link ice core, marine, and terrestrial research on the speed, mechanisms and extent of past WAIS collapse events. It will bring together Antarctic ice core, glacial modeling, earth science, and marine communities in order to develop integrated insights and the greatest scientific return on investment.

While there are real distinctions in the research strategies and geographical focus of the two components of this initiative, they cannot be isolated efforts. Ongoing interaction and integration among these different research communities will yield the innovations that can elevate the Changing Ice Initiative above and beyond the USAP West Antarctic research that has been carried out to date. For instance, this initiative would encourage those studying current atmospheric and oceanic dynamics to seek new insights by working closely with those collecting paleoclimate records; and it would foster the evaluation of modern observational data in the context of climate records extending back over decades, centuries, and millennia.

The initiative outlined here is a large and complex effort that will stretch the current planning and management capabilities of both NSF and the research community at large. Detailed science and implementation plans should be further developed through broad community workshops with international participation. Thereafter, the annual WAIS community meetings that have taken place in recent years should continue, but should entrain a wider array of disciplines than have historically participated in such gatherings.

Strategic Priority II
How Do Antarctic Biota Evolve and Adapt to the Changing Environment? Decoding the Genomic and Transcriptomic Bases of Biological Adaptation and Response Across Antarctic Organisms and Ecosystems

Background Context and Motivation

Antarctica and the immense encircling Southern Ocean encompass uniquely isolated ecosystems of chronically frigid, extreme conditions. This isolation and cooling began over 30 million years ago (DeConto and Pollard, 2003; Zachos et al., 2001), punctuated with periodic climatic transitions up to the present day (Naish et al., 2009; Naish et al., 2001). For many Antarctic species, especially macrofauna, the remoteness of the continent and the insulating Antarctic Circumpolar Current effectively impeded dispersal in north-south directions. Species confined within the Antarctic region have had to

evolve continually to adapt to changing environmental challenges. This continual evolution has led to the distinctive Antarctic biota of today.

Unlike other landmasses, Antarctica has no terrestrial vertebrates, and land invertebrates and plants are rare. In the Southern Ocean, some key taxonomic groups are low in diversity while other groups have proliferated. Plants, invertebrates, fishes, marine birds, and mammals that have succeeded in colonizing the harsh Antarctic environments are endemic, and microbial “extremophiles” dwell in ice, soil, rocks, frozen mountain lakes, and subglacial environments under thick ice sheets. Antarctica provides an invaluable vast natural laboratory for understanding species evolution and functional specializations driven by extreme and changing environments through its geological history.

The tempo of Antarctic environmental changes is being hastened in modern human times by global climate change and increasing commercial fisheries in the Southern Ocean. There remains much to understand about how cold-adapted and specialized Antarctic species will respond and keep pace with these changes, and how Antarctic ecosystems may be altered. Warming ocean temperatures, which in turn affect oxygen availability, and increasing ocean acidification both present major challenges to cold-adapted marine species (Somero, 2010). The same warming ocean that threatens to collapse Antarctic ice sheets (see Strategic Priority I) could also bring about radical ecological changes for marine biota in sea ice and coastal ecosystems. In the western Antarctic Peninsula, for instance, some regions are experiencing ecosystem changes associated with declines in sea ice, changing precipitation, and melting glaciers (Schofield et al., 2010). This includes a large decline in Adélie penguins, concomitant with declining biomass of their major prey, Antarctic krill— attributable to climate-related loss of sea ice that supports krill proliferation (Fraser et al., 2013; Trivelpiece et al., 2011).

Climate change could also open up the continent and Southern Ocean to invasive species. For instance, because of warming water temperatures, cold-intolerant crab species long thought to be excluded from cold Antarctic continental shelf waters (Aronson and Blake, 2001; Thatje et al., 2005), have recently been found in the Palmer Deep basin of the western Antarctic Peninsula shelf (Smith et al., 2012). Continued invasion of these top benthic predators into shallower shelf waters may significantly alter biodiversity and ecosystem structure in the coming decade.

These clear signs of an environment in flux, with cascading effects on Antarctic biota, point to a compelling urgency to understand how vulnerable or resilient Antarctic species and ecosystems might be, both to natural and anthropogenic selection pressures. There is likewise an urgent need to establish baseline references for Antarctic

species and ecosystems, as this provides the foundation for detecting and understanding future impacts of climate change and human activities. To help gauge the fate of these ecosystems requires ascertaining the diversity of life forms that have succeeded in the harsh Antarctic environment and better understanding of both adaptation over evolutionary timescales and adaptive potential (i.e., phenotypic plasticity and adaptability) to contemporary environmental change (Nicotra et al., 2015).

The question of how life has adapted to survive and exploit extreme Antarctic and Southern Ocean environments has been widely studied at different levels of biological organization, ranging from the molecular level to physiology and ecology for microbial species (e.g., Bakermans et al., 2014; Deming, 2002; Dolhi et al., 2013; Grzymski et al., 2006; Smith et al., 1994) and more complex organisms (e.g., Bakermans et al., 2014; Devries, 1971; Dolhi et al., 2013; Fields and Somero, 1998; Hill et al., 2011; Kawarasaki et al., 2014; Magalhães et al., 2012; Thatje et al., 2008). For example, it was discovered that the microbial ecosystem in the dark, freezing subglacial Lake Whillans is sustained by autotrophic bacteria that can derive energy from oxidizing inorganic compounds; and these autotrophs and their products serve as food that sustains heteotrophic bacteria (Christner et al., 2014). And research revealed that cold-blooded fishes that abundantly colonized the Southern Ocean have survived through evolutionary innovation of antifreeze proteins and adaptive changes in various functional proteins (Cheng and Detrich, 2007; Pörtner et al., 2007).

In contrast, a frontier that remains little explored in the studies of Antarctic life and evolution is the fundamental and comprehensive core of genomic information encoded within species. Decoding this genomic information across Antarctic ecosystems will richly inform us about the breadth of biological diversity that evolved in this extreme environment, and about the evolutionary capacity governing species ability to continue to evolve. In addition, decoding transcriptomes and metatranscriptomes can provide information about the functional capability and plasticity of these cold-specialized species to deal with continued environmental challenges.

Numerous recent technological innovations now make it possible to rapidly advance investigations of the genetic and functional bases of Antarctic life, evolution, and adaptive potential. For instance, massively parallel sequencing has made genome and transcriptome sequencing highly feasible, and the supporting infrastructure is now commonplace across research institutions and universities. Decoding small microbial genomes can be accomplished rapidly. Sequencing and analyses of larger, more complex invertebrate and vertebrate genomes requires progressively greater effort, but is still within reasonable timescales of a few years.

New methods such as reduced-representation genome sequencing can be used to assess genetic variations among individuals of populations of the same species, which helps elucidate species resiliency to habitat change. Species biodiversity and interactions among species can now be assessed using high-throughput sequencing methods, as an alternative to the traditional, logistically demanding “catch and identify” biodiversity surveys. Environmental DNA (eDNA),comprising metagenomes and/or fragmented genomic DNA from microorganisms and higher taxonomic groups (in the form of genetic material within bits of metabolic wastes or tissues shed to the environment), can be recovered from sediments, ice, or water samples. They can then be sequenced and analyzed to reveal past and present species diversity (Kelley et al., 2014; Thomsen and Willerslev, 2015; Thomsen et al., 2012). One can even deduce predator-prey relationships for higher taxonomic organisms by sequencing DNA recovered from animal gut contents or scats (Boyer et al., 2013; Murray et al., 2011).

Given these tremendous advancements in technologies and methodologies, the field is poised to make countless new discoveries. Three particular areas where great progress could be made are described below.

1. (i)   Antarctic Biodiversity Across Time and Ecosystems as an Indication of Evolutionary Success and Potential.

DNA sequence in a genome specifies the unique identity of an organism. Genome and metagenome sequencing of extant and fossilized macrofaunal organisms, viral and microbiota (bacteria, archaea, protists) species and communities (in seawater, ice, subglacial lakes, soil or rock, and in ancient pollens or other forms of ancient DNA), will greatly expand understanding of the species and biodiversity that uniquely succeeded in colonizing Antarctica through geological timescales.

Environmental microbes often cannot be grown in the laboratory, but information about the genetic diversity of microbes in the environment can be gained through direct metagenome sequencing of samples of seawater, ice, and soil, etc. Information from metagenome sequencing of entire species assemblages of microbial communities can be analyzed using bioinformatics and molecular phylogenetics, to reveal community genetic and taxonomic diversity. Currently advancing technologies to sequence single-cell genomes will enable the capture of microbial genomes that are truly representative of the Antarctic microbiome (Rinke et al., 2013). Genomic characteristics that allow one to assess taxonomic identity or affinity will help resolve enduring questions about the origin and extent of endemism of Antarctic microbiota.

Decoding the genomes of taxonomic groups that have undergone adaptive radiation and proliferated in the harsh Antarctic environment will help us understand the genetic and genomic factors that have shaped Antarctica’s distinctive faunal composition (see, e.g., Figure 3.3). Whole-genome assessments of the genetic diversity within geographically distinct populations of a species can provide important information about how resilient or vulnerable these species may be to drastic environmental change.

1. (ii)   Species Functional Response to the Changing Antarctic Environment as an Indication of Their Phenotypic Plasticity

Combining genomic sequence information with transcriptomic (and metagenomic/ metatranscriptomic) sequences can provide a powerful foundation for understanding the diversity of Antarctic species and species assemblages. To understand how organisms might respond to rapid environmental changes, one must elucidate the cellular and metabolic pathways that govern a species’ or community’s ability to cope with changes, through their functional response to contemporary and/or future conditions. A species’ cellular and metabolic functions are encoded in the protein-coding genes and DNA and RNA regulatory elements in the genome. Systemwide transcriptome

FIGURE 3.3 A giant Antarctic toothfish being released by a research diver back into McMurdo Sound. The toothfish and the family of Antarctic notothenioid fishes to which it belongs are “swimming libraries” of cold-adapted and cold-specialized genes. SOURCE: Paul Cziko and Chi-Hing Christina Cheng.

and metatranscriptome sequences provide global profiles of gene transcripts, from which encoded functions can be inferred. Sequencing native transcriptomes can help us understand the functional phenotypes forged by the selection pressures of extreme polar conditions (Bilyk and Cheng, 2013; Dilly et al., 2015; Huth and Place, 2013). Sequencing native metatranscriptomes can elucidate functional ecology, which is particularly useful for microbial assemblages (Chan et al., 2013).

The extent to which these functional capacities limit or allow species to deal with environmental stressors can be tested under experimental conditions and assessed via transcript sequencing and analyses. But designing well-formulated experiments requires better knowledge of the environmental conditions and variability that Antarctic life faces in its natural setting. One gains this knowledge by monitoring basic environmental parameters such as temperature, soil moisture, ocean salinity, and pH. The collection and sharing of such data can be aided by existing LTER programs, and perhaps by a Ross Sea “Research Coordination Network” (as suggested earlier in this chapter).

There is increasing use of high-throughput sequencing of transcriptomes to understand Antarctic species response to changing environments (Bilyk and Cheng, 2014; Enzor and Place, 2014; Teets et al., 2012). But the value of these transcriptomes is limited by the lack of whole-genome sequences of Antarctic species. These whole-genome data (from which one derives transcript sequences) are needed in order to identify and quantify responding gene transcripts and spliced variants, and to determine how genetic variations between related species affect response capability.

For microbiotic systems, metagenome and metatranscriptome sequencing and analyses are at incipient stages and yet have already provided insights into soil and rock microbial diversity and community functional ecology in McMurdo Dry Valleys (Chan et al., 2013). Microorganisms play important roles in elemental cycling within polar ecosystems, including in the Southern Ocean, which is estimated to account for up to half of the annual oceanic uptake of anthropogenic CO₂ from the atmosphere (Arrigo et al., 2008; Gruber et al., 2009), and which (through vertical mixing) is thought to supply enough nutrients to fertilize three-quarters of the biological production in the global ocean north of 30°S (Sarmiento et al., 2004). Robust estimates of microbial abundance and diversity are critical to understanding elemental cycling in the Southern Ocean, and genomic tools are the only means to map this microbial diversity and to identify the microbial species (phytoplankton, bacteria, viruses) that play key roles in mediating elemental fluxes. Many insights can be gained by using genomic profiles to generate microbial community inventories, and by combining this with studies of the functional analysis of gene expression, to characterize microbial responses over a range of environmental conditions.

1. (iii)   Evolutionary Cold Adaptation/Specialization and Future Evolutionary and Adaptive Potential

DNA sequence in the genome and genome structure both provide historical records of a species’ genetic and genome structural changes over time. Decoding the genome sequences and structures of Antarctic species whose evolution was shaped by extreme polar conditions and comparing them with decoded sequences and structures of related non-Antarctic temperate species provide a powerful basis to assess how Antarctic species evolved and adapted over geological timescales, and to assess their evolutionary potential to adapt to future environmental changes. Coupling this assessment of evolutionary potential with assessment of phenotypic plasticity (through comparisons of functional response, as informed by transcriptomes and experimental testing) allows detection of Antarctic-specific genetic, genomic, and functional innovations and gains and losses that underlie cold adaptation and specialization. These comparisons may also elucidate how much genomic and functional “reengineering” might be necessary and possible for a particular Antarctic species, in order to attain more temperate characteristics that would allow it to survive in a changing climate.

Even more definitive assessments of the adaptive potential of Antarctic species to survive a warmer world can be achieved by studying species that were of Antarctic origin but later colonized temperate non-Antarctic regions during the lineage’s evolutionary history. Comparative genomic analyses of these related Antarctic/non-Antarctic species allow one to see the genomic and functional retooling required for adapting to the more temperate environments. Estimating the rate of evolutionary genetic change can indicate whether Antarctic species can evolve at a sufficient pace to remain viable in the face of future environmental change. These comparative analyses can yield powerful insights into past and present cold-adaptive/specialized processes as well as future adaptive potential.

The Proposed Effort

The value of decoding genomes of key Antarctic organisms in order to understand evolutionary adaptations and ecological success was recognized over a decade ago (NRC, 2003b). More recently, both international and U.S. community-scale assessments (Kennicutt et al., 2014b; NRC, 2011a) have reiterated these same research themes as a priority for Antarctic science. Our community engagement discussions likewise revealed widespread interest to advance the understanding of past adaptation and future adaptive capacity by leveraging powerful molecular and genomic approaches and technologies.

An Antarctic Genomics Initiative, inclusive of genomes and transcriptomes of individual species and species assemblages, will involve biologists working on diverse species in a coordinated pulse of activity—with a shared goal of decoding the genomic and functional bases of organismal adaptation in a changing environment. Genomic sequencing could encompass ancient DNA, viruses, bacteria, and complex eukaryote species from major Antarctic habitats, including ice sheets, soils, outcropping rocks, surface and subglacial lakes and streams, the ocean, and sea ice. A priority focus can be given to keystone species and organisms/communities that are fundamentally important for addressing questions about Antarctic adaptation in the past and in the future. As a necessary adjunct to the genomic analysis in different Antarctic habitats, this initiative should also advance understanding of the environmental properties and variability driving species responses.

High-throughput sequencing can be performed on various platforms appropriate for the complexity of the specific samples. This effort could lend itself well to collaboration across universities and research institutes, as well as cooperation and shared funding support between NSF and other federal agencies (e.g., DOE’s Joint Genome Institute, Los Alamos National Laboratory, and National Institutes of Health Sequencing Centers).

We suggest that NSF could implement this initiative in a set of calls for proposals designed to encourage interplay among lab-based genomic analyses, field-based environmental investigations, and collection of biological samples and environmental physical data. These calls can be concomitant or phased, and could be broken out roughly as follows:

• Sequencing of existing samples. Many investigators already have substantial collections of materials from key taxa or biological samples in frozen storage from prior Antarctic field research programs. These could be readily used for genome sequencing. Thus an initial component of this initiative could center on sequencing genomes or metagenomes of the existing species/samples that are well suited for providing answers to the types of genome-enabled inquiries described above.
• Acquisition of new samples. The second component, which could begin in parallel, is the acquisition of additional species or samples for sequencing, to allow genomics-enabled inquiries that existing samples do not cover. New field acquisitions such as a series of seasonal environmental DNA samples can address dynamic microbial or planktonic fauna changes that affect the food web. Genetic materials also can be retrieved from archival materials such as ice, sediment, or rock cores. Surveys of environmental properties and

processes would accompany these collections, to characterize the conditions that contribute to genomic diversity and functional ecology.

• Field experimentation. The third component supports new field experimentation, and testing of hypotheses that are developed through the genomic data analyses carried out in the first two components of sequencing and analyses work.

Critical to the success of this initiative is concurrent support for bioinformatics advancements, to aid in assembling and annotating the genomes to be analyzed. It will be particularly useful to support collaboration among bioinformaticists and biologists mutually interested in certain taxonomic groups or environmental metagenomes, to continually add to and refine draft genomes/metagenomes for accuracy and completeness. This will help avoid the pitfall suffered by many genome sequencing projects, where the draft genome remains fragmented and unimproved, thus undermining the utility of the genomic data for the community.

This initiative should include NSF-sponsored workshops targeted at fostering communication among different disciplines, engaging early-career researchers, and training community members in new bioinformatics tools. Community-wide discussions will ensure that the most relevant candidate species and ecosystems are selected for analysis, that appropriate sequencing approaches are used, and that data analysis and sharing standards are developed.

As noted above, this initiative could be based in part on sequencing genomes from suitable biological samples already available. There could also be opportunities to leverage existing terrestrial and marine research projects to collect new biological samples and related metadata (as long as careful attention is paid to sampling protocols). At a time of budgetary constraints and ever more costly field deployments, this initiative can be an effective way to advance Antarctic biological research without taxing already overstretched budgets and field logistics.

Despite the United States being the global leader in the invention and continued innovation of high-throughput sequencing technologies, U.S. Antarctic biology still remains at the cusp of the genomic revolution. Thus far, only one Antarctic metazoan genome (the small genome of the Antarctic midge) has been decoded by a U.S. team (see Box 3.7). Meanwhile, significant efforts have been and are being led by other countries. Advancing U.S. leadership in this field requires a concerted and well-executed initiative, which will generate a trove of high-quality genome sequence data from a range of Antarctic species and samples of important scientific value.

Well-assembled and annotated genomic and transcriptomic sequences of strategically chosen species and samples will catapult the abilities of biologists now and for gener-

ations to come to understand the genomewide basis of evolutionary adaptation and specialization that allowed some species to flourish in the harsh Antarctic and Southern Ocean environment. It will also provide the genomic framework for assessing the ability of Antarctic organisms to evolve and adapt to continuing environmental change. Studying the cold-adapted proteins and molecular adaptations of these organisms may even offer promising opportunities for discovering low-temperature industrial and medical applications.

Antarctica and the Southern Ocean encompass some of the most extreme and pristine regions on Earth. The biome in this vast region has been shaped by over 30 million years of isolation and chronic cold. Scientists have only begun to understand the impacts of climate change and human activities in the region on cold-attuned Antarctic biota and ecosystems. Such impacts will continue in unforeseeable ways in the coming years, and thus urgently need careful investigation. This initiative would advance fundamental understanding of how Antarctic life has succeeded in the changing Antarctic environment through time, of how extant biota and ecosystems continue to respond to current changing conditions, and of how their viability may change in the future.

Strategic Priority III
How Did the Universe Begin, and What Are the Underlying Physical Laws That Govern Its Evolution and Ultimate Fate? A Next-Generation Cosmic Microwave Background Program

Background Context and Motivation

The cosmic microwave background (CMB) is the fossil light from the early universe of nearly 14 billion years ago. Measurements of the CMB have revealed the seeds of all structure existing today in the universe. But, how were the seeds produced? Data indicate that the seeds resulted from quantum fluctuations on the smallest scales being stretched to astronomical scales during the first moments of the universe, in a process of accelerated expansion known as inflation. If the universe did indeed start with inflation, what can one learn about the physics of this process, and how can one use that information to constrain physics at high energies?

The theory of inflation provides a compelling framework to understand CMB measurements; it led to predictions of several properties of the CMB which were subsequently confirmed by observations. A key test remains, however—detection of the imprint on the CMB of gravitational waves generated during inflation, when the universe was less than 10^–34 seconds old and only about 10^–29 m in size (many orders of magnitude smaller than a single proton). Such a detection would provide a spectacular confirmation of the inflationary origin of our universe. In addition, it would open a window on physics at energy scales many orders of magnitude greater than could ever be probed with particle accelerator laboratory research. It would provide evidence of the long-sought, but so far elusive, quantum nature of gravity.

Measurements of the CMB have already provided remarkable insights into the makeup of the universe, determining the relative fractions of ordinary matter, dark matter, and dark energy, as well as the presence of the cosmic neutrino background. Neutrinos are the second most abundant particles in the universe (second only to photons—particles of light), and a cosmic neutrino background was released in the first second of the universe. Trillions of these neutrinos pass through your body each second. Their nonzero masses indicate the existence of new physics beyond our current standard model of particle physics. Fully understanding the physics of neutrinos will extend our knowledge of the basic workings of nature and may offer important insights into a process known as baryogenesis, which produced the asymmetry (imbalance) between baryons and antibaryons in the very early universe and therefore the origin of all the ordinary matter in existence (including us).

A next-generation ground-based CMB experimental program, referred to as CMB Stage IV (CMB-S4), is being developed to provide definitive measurements of the early universe, detecting inflationary gravitational waves or at least setting limits that rigorously rule out classes of models (e.g., so-called large-field inflationary models).The next generation of CMB measurements would also offer enough sensitivity to make precise determinations of the number and type of neutrino species, as well as the sum of their masses. The CMB-S4 program builds on the successful CMB programsusing telescopes at the South Pole and the high Atacama Plateau in Chile, and possibly will add a new site in the Northern Hemisphere to allow observations of the full sky. The program will be designed to be highly complementary to the continuation of the successful Long Duration Balloon CMB programs, or even a future satellite mission. Future balloon and satellite measurements can provide broader-frequency coveragefor constraining foregrounds as well as wide-sky coverage for pursuing the largestangular scales.

The precision and accuracy envisioned for the next-generation measurements will help unlock the story of the universe’s evolution that as of yet remains hidden in the CMB. During its 14-billion-year journey across the universe, the electromagnetic and gravitational interactions of the CMB photons with intervening structures impart subtle spectral distortions and spatial correlations in the CMB. Untangling these signals will provide insights into compelling questions such as: How and when did the first stars and galaxies in the universe form? What are the properties of the mysterious dark energy that is causing the present-day expansion of the universe to accelerate? What is the ultimate fate of the universe? These experiments address fundamental questions about our origins and the workings of the natural world, questions that have been asked for millennia. In striving to answer these questions, scientists have discovered phenomena that cannot be explained with current understanding of physics and that thus demand further investigation.

CMB science has been highlighted in several reports, including astronomy and astrophysics decadal surveys (NRC, 2010) and most recently in the report of the Particle Physics Project Prioritization Panel (P5), Building for Discovery: Strategic Plan for U.S. Particle Physics in the Global Context (Particle Physics Project Prioritization Panel, 2014). The P5 report included the CMB-S4 project under all funding scenarios considered and recommended to “support CMB experiments as part of the core particle physics program,” stating that “the multidisciplinary nature of the science warrants continued multiagency support.”

Despite its highly technical nature, this research has generated a great deal of popular interest, as people are naturally curious about questions relating to the origin, makeup,

evolution, and ultimate fate of the universe. Tremendous excitement was generated by the apparent BICEP2 detection of B-mode polarization induced by inflationary gravitational waves. Unfortunately, that result has now been shown to be just an upper limit rather than a confirmed detection, but the hunt continues with more enthusiasm and interest than ever.

One cannot predict the societal benefits that could eventually result from the pursuit of new physics. The last time there was such clear evidence for new physics was roughly 100 years ago, and the result was the quantum revolution leading to lasers and electronics that have transformed human society. It has been said that there are two types of science—applied and not yet applied. It is critical to invest in both. The investment in applied science for CMB research is the development of new detector technology that could have applications as diverse as x-ray and optical astronomy, sensors for advanced light sources, medical imaging, and homeland security.

In this rapidly progressing field of research, the CMB-S4 project is the next logical step, and possibly the last major step for the ground-based CMB observational program. The U.S. research community is ready to move ahead and the technology development is in place. Taking the next step now will ensure U.S. leadership and continued return on NSF’s investment in CMB research.

The Proposed Effort

The entire CMB-S4 program will likely be on the order of 10 telescopes of two types, small apertures similar in size to BICEP and larger apertures, possibly 5 to 8 m in diameter. The family of telescopes will be located at the South Pole, Chile, and possibly a new northern site to optimize the science return. The 10-m South Pole Telescope would likely remain. (See Figure 3.4.) The deployment will likely stretch over 3 to 4 years, and operations will continue for roughly 5 years after deployment is complete. Of all these sites, the atmospheric observing conditions at the South Pole are superior and would allow the most sensitive observations; while the other sites allow more sky coverage.

An infrastructure upgrade will be required for CMB-S4 at the South Pole. The scope of the upgrade will be equivalent to the renovation or replacement of the current Martin A Pomerantz Observatory—which may soon become unavailable because it is being buried by snow. The power requirements for CMB-S4 at the South Pole, driven primarily by the helium cryocoolers, are similar to those of the current program.

For efficient operation of the CMB-S4 telescopes at the South Pole, it is highly desirable to improve the 24/7 Internet/voice access and increase the data transmission. The current transmission of CMB data from the South Pole is around 120 GB/day. It

FIGURE 3.4 A view of the BICEP telescope and the 10-m South Pole Telescope located at the Amundsen Scott South Pole research station. SOURCE: Steffen Richter.

would be desirable to increase the transmission to roughly 1 TB/day. Lower transmission rates would require considerably more on-site analysis, with physical transport of the backed-up data media during the austral summer seasons. Although possible, this would severely diminish the science output by not allowing the full power of the CMB-S4 science team to work on real-time data inspection and data analysis.

This next-generation CMB initiative will be an interagency effort. In addition to three NSF divisions involved (PLR, Physics, and Astronomical Sciences, with PLR leading management of South Pole activities), the DOE Office of Science has also expressed interest in being involved, particularly in the scaling-up of detector arrays and in handling the computation required for the large datasets. NASA leads the ballooning and space-based CMB studies and there are nascent community discussions of planning a possible future NASA satellite designed to complement the ground-based effort. The project is expected to include international partners. The current international partners of the ongoing CMB projects would naturally be expected to participate in CMB-S4, including the United Kingdom, Canada, Germany, Japan, Chile, and others.

Regarding the anticipated management needs, a Joint Oversight Office with members from all participating agencies and NSF divisions will be necessary. It will be highly beneficial to have a dedicated NSF program officer to work with the CMB-S4 project and interface with other agencies.

SELECTING AND SUPPORTING THE RECOMMENDED PRIORITIES

The Committee’s task was to help NSF make some hard choices, which means that many compelling research topics raised in the community engagement discussions are not listed among the recommended priorities. We did not feel it would be constructive to list the merits and limitations of all the various, wide-ranging research ideas suggested. Rather we simply note that the ideas that were not selected did not fare as well in meeting some or all of the evaluation criteria, as compared to the recommended priorities. And emphasis was placed on activities for which NSF/PLR in particular is best suited to lead.

One unique example to note is the IceCube-Gen2/PINGU proposal (discussed in Box 2.4). Although the Committee acknowledges that this project has the potential to lead to exciting new scientific advances, we were cognizant of the fact that construction of the original IceCube facilities had major impacts on the USAP logistical system, and we had concerns that high demands on USAP logistical support likely required for the IceCube-Gen2 construction would be incompatible with our recommendations to expand logistical support for activities for West Antarctic research.

This report outlines only general-level guidance for what the recommended initiatives should look like, with expectations that the more fine-grained implementation details will be shaped by NSF and by the relevant research communities. These planning processes will also allow robust cost estimates to be determined for each of the initiatives. On a very general level, we surmise that the Antarctic Genomics Initiative and the Cosmic Microwave Background Initiative can be supported within the existing PLR core program budgets (for Antarctic Organisms and Ecosystems, and Antarctic Astrophysics and Geospace Sciences, respectively), without significantly impeding the ability of those programs to continue support for traditional PI-driven proposals. And for both initiatives, NSF and the research community should consider whether establishment of a science and/or technology center could be an appropriate vehicle for supporting and advancing the work.

The scale is significantly larger for the proposed Changing Ice Initiative. If carried out at the scope and duration envisioned, this effort may require resources equivalent to much of the research budget of the following PLR Antarctic Science core programs—Earth Sciences, Glaciology, Ocean and Atmospheric Sciences, and Integrated System Science. And yet we would not want to see the proposed initiative overwhelm the capacity of these programs to provide ongoing support for traditional PI-driven proposals—an outcome that runs contrary to the overall strategic vision proposed at the start of this chapter.

This apparent conflict may be ameliorated in part by leveraging new and expanded partnerships between PLR and other divisions of NSF, other federal agency programs, and other countries’ Antarctic research programs. But more importantly, the Committee fully believes that because of the urgency and potential magnitude of the threats that ice sheet collapse poses for human society, this issue must be looked at through a different lens—beyond the constraints of what current budgets dictate is feasible. This research will ultimately provide critical guidance on when, where, and how sea level may occur, and it will thus help society make the needed adaptation investments in a reasonably informed matter. Although the cost of this research is large relative to current PLR/ANT program budgets, it is tiny relative to the projected costs of adaptation to and damage from sea level rise (see Box 3.3). We thus suggest there is strong rationale for NSF to seek to significantly augment the funding available to support this research.

Moreover, it is worth noting the potential co-benefits to USAP science that could result from the recommendations to bolster support for deep-field work offshore and near, under, and on the ice shelves. This enhanced logistical capacity could yield a wealth of new opportunities for other types of smaller (PI-driven) investigations that would otherwise lack the critical mass to justify logistical support. This may even include opportunities for disciplines that are not directly related to the initiative itself—for instance, opportunities for ecology researchers to join research cruises and coastal expeditions to collect samples that would otherwise be impossible to obtain.

There are inevitably difficult trade-offs that must be made in any effort to truly set priorities among research needs. However, the Committee has tried to provide a framework that balances the desire to make major leaps forward on a few critical research topics with the need to preserve flexibility for NSF to continue responding to new creative ideas arising from across the research community.