Studying emerging questions will require a combination of new and traditional approaches and tools. The questions require information at spatial scales ranging from meters for process studies to pan-Arctic and beyond to link high-latitude change to large-scale systems throughout the Northern Hemisphere. Understanding interactions among changing oceans, terrestrial systems, hydrologic processes, atmospheric dynamics, and social and economic systems will necessitate a broad suite of measurements and observations, obtained at regular time intervals and consistently over decades.
As detailed in this chapter, standard techniques that work in other regions often have deficiencies when applied to the Arctic. For example, remotely sensed data suffer from a lack of appropriate validation data and a need for calibration to Arctic conditions. Social indicators often lack specific relevance to the Arctic. Long-term observations, networks of field-based measurements, and remote sensing techniques are needed to understand and quantify the effects of a changing climate and also to inform and validate modeling efforts. Chronic shortages of appropriate data make it difficult to develop model parameters and to validate model results.
The sections of this chapter describe in more detail various ways research capability can be increased to help address the existing and emerging questions. Many of these improvements will require long-term planning, and all stem from the fundamentally collaborative nature of Arctic research. In keeping with the committee’s Statement of Task, we do not suggest specific actions to be taken but instead raise key topics for consideration by funding agencies and others as they consider how best to address the questions discussed in Chapter 3, as well as how best to continue and improve the strong record of Arctic research described in Chapter 2.
Effective Arctic research is international and national, interdisciplinary and disciplinary, applied and basic, private and public. Cooperation between and among many individuals, institutions, businesses, agencies, and countries will help to maximize investments in research, synthesis, outreach, and infrastructure (Box 4.1).
BOX 4.1 THE CASE FOR ENHANCED COOPERATION
Reflecting its growing awareness of the Arctic as a security concern, the Department of Defense recently highlighted the need for cooperation in the Arctic. The 2011 Fleet Arctic Operations Game Report notes, “As risk increased due to extreme climatic conditions and increased operating and support distances, there was a corresponding increase in the need for specialized information and capabilities. As this trend increased, the required information and capabilities became less available in the U.S. Navy and planners were forced to look elsewhere for the capabilities needed to execute their mission tasking. At the low end of the scale, these could be found inside DoD [Department of Defense], but eventually planners needed to rely on industry, international partners, or the whole of U.S. government. This further reiterates that sustainability in Arctic operations is significantly dependent on strong relationships with international, regional and local partners in government and industry. Mechanisms that strengthen these ties should be prioritized in future planning” (Gray et al., 2011).
This figure illustrates that reliability and sustainability are linked to specialized information and capabilities that are currently enabled by strong relationships. SOURCE: Gray et al. (2011).
Since the Arctic Research Policy Act of 1984, an interagency Arctic Research Plan has been developed every 5 years. The Arctic Research Plan for FY2013-2017, released in February 2013, outlines interagency federal initiatives to better understand and predict Arctic environmental change. Following up on this plan was the first ever U.S. National Strategy for the Arctic Region, released in May 2013, calling for each agency to develop a coordinated strategy or implementation plan.
This alignment of effort within and between U.S. agencies, coordinated by IARPC, could have significant implications for the future of Arctic research if there is a concomitant investment in cross-agency sharing of research and infrastructure. The ongoing Study of Environmental Arctic Change (SEARCH) program and the Arctic Science, Engineering, and Education for Sustainability (SEES) competition run by NSF with cooperation from numerous other agencies are examples of what can be done when agencies decide to co-fund initiatives. Nonetheless, there is still a need for commitments to make the most of opportunities for joint studies across agencies. This is especially important when the missions of different agencies result in complementary work, for example, in synthesizing findings from different research projects so those findings can be applied to meet the needs of various stakeholders. Some synthesis activities have taken place or are under way, but they are often ad hoc efforts made after the majority of research is done.
Cooperation across levels of government is as important as interagency cooperation. It does exist in some forms, such as the North Slope Science Initiative in Alaska, which involves the federal government, state government, and local (North Slope Borough) government and aims to increase collaboration on monitoring, inventory, and research related to development activities. More can be done, however, to coordinate data collection, share costs, and develop a common basis of understanding regarding key issues affecting the Arctic.
Looking beyond the United States, understanding the Arctic is inherently global in nature. The circumpolar North spans the eight nations that constitute the Arctic Council and draws interest from dozens of other countries. Furthermore, changes in the Arctic have global implications. Existing and emerging research questions are often multidimensional across international domains. Arctic research and our ability to act
on our knowledge benefit from cooperation with those who share an interest in Arctic matters.
One of the most influential developments in scientific discovery in recent decades is the internationalization of science. This is in part a result of the vast improvement in international communication. But it is at least equally a consequence of the nature of key scientific questions, which increasingly view the Earth as a system, within which understanding requires a global perspective. Documented trends in international scientific mobility (Van Noorden, 2012) reflect the increased national diversity of the scientific community and emphasize the benefits of cross-fertilization of ideas and methodologies as we move toward a multicultural and interdisciplinary scientific world.
Much Arctic research is undertaken by U.S. researchers outside of U.S. territories and by researchers from non-Arctic countries. A variety of formal and informal arrangements exist by which researchers and agencies cooperate with their counterparts in other countries, including the International Arctic Science Committee (IASC) and its associated bodies, the Arctic Council and its working groups, the International Arctic Social Sciences Association (IASSA), and the Association of Polar Early Career Scientists (APECS). These collaborations help place findings from the U.S. Arctic in a wider context and provide a way to learn from experience elsewhere when it comes to applying science to management, regulation, and governance.
The International Polar Year demonstrated the tremendous value in international cooperation for Arctic research (e.g., NRC, 2012a). Far more was accomplished collaboratively than could have been done by any one country, regardless of Arctic research expenditures. Research under the Arctic Council similarly illustrates what can be accomplished by working together. The scientific community is looking forward to the new Belmont Forum Arctic Collaborative Research Action (CRA) focused on Arctic observing and Arctic sustainability science. The new Scientific Cooperation Task Force (SCTF) of the Arctic Council, co-chaired by Russia, Sweden, and the United States, is a promising step in the right direction. The SCTF will report to ministers in 2015 on ways to improve scientific research cooperation among the eight Arctic States.
There is a great deal of interest in cooperation among individual researchers, among agencies, and among countries engaged in Arctic research. But more could be done to collaboratively address existing and emerging Arctic research questions in a time of rapid change and rapidly expanding human presence. A potential method for fostering international collaboration beyond the level of individual researchers is to explore opportunities for U.S. projects (e.g., SEARCH) to work with international projects (e.g., ACCESS, ICE-ARC). The FY2013-2017 Arctic Research Plan recognizes this with
references to the necessity for international partnerships to meet research goals, for example, “Successful implementation of this five-year research plan will require close coordination among … international partners” (Executive Office of the President, 2013, p. 6).
Improved collaboration is needed on both the funding of research that crosses borders (see “Investing in Research” section later in this chapter) and the logistics of doing international research. Arctic research frequently entails complex logistical arrangements, often international in scope, with long lead times to obtain permission to access remote field sites. But the necessity for international collaboration extends well beyond logistics. Access to the necessary analytical tools and remotely sensed imagery commonly requires international cooperation. Because of the geographically remote nature of much of the Arctic, specialized research platforms and instruments are often necessary to advance regional knowledge and understanding. These needs range from detailed in situ observations to satellite observations and from year-round manned field stations to research vessels. U.S. infrastructure in this regard is finite; international coordination of infrastructure and cost sharing is essential to take advantage of available observing platforms (e.g., ships, aircraft, fixed offshore platforms, coastal research stations). At present, individual projects have the responsibility to navigate these complex issues. A higher-level effort to streamline this process would greatly facilitate research, and the community is looking forward to the findings from the Arctic Council’s SCTF on this issue. Coordination that extends beyond national and international organizations to active participation with the private sector is more likely to result in beneficial new insights. The scientific community also needs to be assured that there are data repositories where data in support of published research can be permanently archived in a format accessible across the international community, and to the public at large (see “Managing and Sharing Information” section later in this chapter).
Interdisciplinary cooperation leads to improved understanding of the complex interactions within and among the physical, biological, and social domains of the Arctic. Researchers often need time to learn to connect the theories, concepts, and language of one discipline to those of another, and for research teams to build a collective understanding of the phenomena they are studying. Interdisciplinary collaboration, however, is often difficult to initiate and can be difficult to sustain without specific allocation of funding for such research. Yet it is in the connections between research domains that many emerging questions lie. Our ability to tackle these with vigor and
success requires considering how interdisciplinary research is encouraged and supported and what can be done to foster greater efforts in this area. A more strategic approach, with suitable direction from IARPC, will allow us to reap more benefit from our Arctic research efforts and expenditures (see “Funding Comprehensive Systems and Synthesis Research” section later in this chapter).
Also of substantial importance is the question of intersectoral cooperation, including public-private partnerships. The private sector sponsors a great deal of Arctic research, often related to the prospects for, and the effects of, industrial activity. Too often, such research is questioned or dismissed amid perceptions of bias due to funding source, but it is shortsighted to ignore the data and findings that come from private-sector research. It is similarly shortsighted to keep most of this research proprietary. Findings of commercial value naturally belong to those who paid for them. But data concerning basic conditions or research that helps illuminate particular processes or changes is valuable for all, and the greater dissemination and use of such data and research can also help provide quality control, reducing the likelihood and perception of bias. Some efforts have begun in this direction, and after evaluation, effective efforts could be promoted and emulated (see “Partnerships with Industry” later in this chapter).
Looking ahead, we need to explore the use of social media as cooperative sources of information as well as cooperative tools to inform decision making. As recommended in the International Study of Arctic Change report, Responding to Arctic Environmental Change, we need “development of an interactive, widely accessible, stakeholder engagement tool that can be used to develop new research priorities and research questions” (Murray et al., 2012, p. 15). Establishment of issue trackers helps identify concerns emerging from communities. Social networking can then help with collecting knowledge through restructuring expert attention to bring in needed expertise and collaborators for problem solving (e.g., Nielsen, 2011). Regarding responses, social networking can encourage contributions—through crowdsourcing, fostering local experimentation, disseminating knowledge and best practices, and supporting implementation elsewhere—thus spreading innovation among communities, agencies, and industry. Through these cooperative processes, social media can foster grassroots approaches to proactive management of Arctic change. Might social media also help with the knotty problem of making scientific products more useful for stakeholders?
FIGURE 4.1 The Sea Ice Outlook from June 2013. The intent of the Outlook is to summarize all available data rather than issue predictions. SOURCE: ARCUS.
The Sea Ice Outlook along with the Sea Ice for Walrus programs are powerful examples. The SEARCH Sea Ice Outlook (Figure 4.1) synthesizes and publicly posts community estimates of the current state and expected minimum of sea ice. The Sea Ice for Walrus Outlook is a weekly report on sea ice conditions for subsistence hunters, coastal communities, and other interested members of the public. The Canadian Polar Commission recently launched the Polar Knowledge App, intended to expand public access to polar information.1 In addition, some science blogs are interpreting scientific studies for a lay public and providing broader context.
Science depends on data. Individual projects generate data specific to their questions and hypotheses, but the interpretation of results usually relies on comparison of those results with data from longer periods or over larger areas, to place them in context. In many cases, this means data from long-term observations or monitoring—without which our ability to detect change, constrain models, and analyze the significance of research findings—is greatly diminished, if not lost entirely.
A major challenge facing society is to ascertain, comprehend, and forecast rates and patterns of change across the Arctic that arise from physical, biological, or human causes. Society can address this challenge through an understanding of the resiliency and vulnerability of the Arctic system. Resiliency is the capacity of a system to withstand disturbances to its structure, function, and feedbacks (Folke et al., 2004; Walker et al., 2004), whereas vulnerability describes the extent to which a system is harmed by exposure or sensitivity to stressor(s) and by constraints on its adaptive capacity in response to the stressor (Turner et al., 2003). When designed to characterize resiliency and vulnerability, monitoring, the long-term and systematic measurement of appropriate system characteristics is essential in meeting this societal challenge.
When suitably constructed, monitoring systems serve a variety of purposes for a variety of stakeholders. On one hand, long-term observations enable quantification of the natural variability, over a range of temporal and spatial scales, of complex “noisy” systems. Once the “noise” is defined and quantified, long-term observations enable detection of gradual, systematic changes. On the other hand, because of the nonlinear character of many systems, a carefully developed monitoring scheme may detect abrupt and/or unanticipated changes (e.g., detecting what we don’t know we don’t know). In this capacity, long-term observations serve as part of an early warning system (e.g., NRC, 2013), which then allows for a choice of responses. These responses will vary depending upon the nature of the change, but they could include collecting focused measurements designed to better understand the emerging phenomenon; development or initiation of mitigating procedures, if deemed feasible; or, in the event of a potential catastrophe, appropriate emergency responses. Long-term observations also provide the temporal-spatial context in which shorter-duration, hypothesis-driven process studies can be undertaken. In this context it allows researchers to determine whether the processes under consideration occurred under typical or atypical condi-
tions. This was, for example, a key ingredient of the U.S. GLOBEC program2 in which short-term process studies were embedded within the framework of a monitoring program.
Monitoring is a synergistic component in modeling and hypothesis development. It provides datasets necessary for the evaluation and development of models and/or suggests investigations needed to improve model parameterizations and/or processes. Models provide an integrated approach to understanding system behavior and can be used to modify the monitoring program as necessary. Models also augment monitoring efforts by suggesting how unsampled system components may be evolving. Monitoring and model results both contribute to the construction of hypotheses on how the system or parts of it operate.
Much of our recognition and understanding of the dramatic changes occurring in the Arctic has emerged from long-term observations. For example, routine measurements revealed the dramatic warming of the Arctic atmosphere and the accelerating decline in sea ice; both are consistent with some of the earliest model predictions of climate response to greenhouse gas warming (Manabe and Stouffer, 1980). Another example is the systematic approach adopted by the Arctic and Bonanza Creek Long-Term Ecological Research (LTER)3 programs conducted in the tundra and boreal forest biomes of Alaska, respectively. Although independently initiated, these LTERs are established along a latitudinal and ecological gradient and each attempts to understand the resiliency and vulnerability of the respective biome to a warming climate. Both LTERs have been in existence for at least 25 years and involve myriad interdisciplinary process studies and modeling activities. Although different investigators are involved in each, there are consistent efforts to compare and contrast the results across biomes.
One important finding from the integration of plot-scale long-term studies of vegetation dynamics, fire cycles, and their links to climate in the Bonanza Creek LTER (Van Cleve and Vierech, 1981; Van Cleve et al., 1983) with broader-scale measurements of a series of wildfire-disturbed boreal forests of interior Alaska is the likely shift in some Alaskan boreal forests from a spruce-dominated to a broadleaf-dominated landscape due to increased burn severity (Figure 4.2). This transition to more high-severity wildfires is occurring in conjunction with thawing of permafrost and the decomposition of previously frozen organic carbon in boreal forest soils. Through large-scale manipulation experiments at the Arctic LTER at Toolik Lake, researchers have found that response to heating soil, shading, or altering soil moisture is slow, with responses delayed until 9 or 10 years post initiation of the treatment (Hobbie and Kling, 2014).
FIGURE 4.2 Conceptual model showing the shift in resilience cycle from a coniferous-dominated (left) to a hardwood-dominated system (right) triggered by an increase in fire severity. High organic matter thickness following low-severity fires in black spruce allows for the regeneration of slow-growing woody plants, inhibits hardwood regeneration, and results in rapid reestablishment of a thick moss layer that insulates the soil and permits the return of permafrost. High-severity fires remove thick organic layers and allow for the rapid establishment of hardwoods, which store large amounts of C and N in aboveground biomass, and create conditions (high litter quality and warm soils) that accelerate forest floor decomposition rates. SOURCE: Adapted from Johnstone et al. (2010).
These experiments are designed to explore future effects of continuing climate change, but at an accelerated rate. The LTER observations and experiments predict increased productivity and biomass of grasses and shrubs by the end of this century and an eventual shift from tundra to boreal forest, with great disruption of fish and wildlife habitats (Hobbie and Kling, 2014). Whole-ecosystem experiments conducted at the Arctic LTER near Toolik Lake, which have continued for more than two decades, have provided valuable insight into aboveground production and biomass in moist tussock tundra. They have demonstrated that the vegetation response to marked climate warming is relatively small when compared to annual variation. Linking these longitudinal studies at the LTERs with shorter-term but broader-scale studies offers opportunity to improve understanding of the changing Arctic and boreal landscape.
As outlined above, the guiding principles behind a monitoring effort seem logical, but the design of a monitoring program in a system as complex and diverse as the Arctic
is far from obvious. A number of questions arise immediately. What is the purpose of a particular monitoring activity? How is it integrated into other monitoring efforts, including those in other regions and/or disciplines? Where should the long-term observations occur? How long should the program continue? What are the specific variables to be measured, at what rate, and over what time and space scales should these be sampled? What measurement techniques (including calibration and algorithms used in interpreting the data) should be used? Who should perform the measurements? Who should pay for it? Who evaluates the utility of the measurements? Who interprets and synthesizes results? How do we ensure that the results of individual efforts are blended into a coherent picture of the emerging Arctic that is of use to stakeholders and society? Although this committee recognizes the importance of these questions, it cannot provide definitive answers, and rather suggests that the following issues be considered.
Involvement of northern communities is an important component of monitoring efforts in the Arctic. This includes not only the use of traditional knowledge but also the involvement of local residents in data collection (e.g., Alessa et al., 2013; Huntington et al., 2011). We expect that a carefully developed approach that involves local residents would provide numerous benefits (see “Growing Human Capacity” section). Local involvement can enhance cross-cultural communication, including ideas about research strategy and interpretation; provide an important degree of ownership by local residents in the measurements being made; stimulate the involvement of decision makers (Danielsen et al., 2010) and schoolchildren within these communities, enhance seasonal coverage; and facilitate overall logistics.
Successful monitoring programs address linkages between different parts of a system (e.g., Alessa et al., 2009; Liu et al., 2013). The Arctic spans a broad latitudinal range that encompasses a number of physical, biological, and social systems. The acquisition of societal data—demographic, infrastructure, health, economic—is essential for many purposes. Thus there are many national and more localized efforts to collect such data, from national censuses to local surveys. The results of these programs are widely used in social science and other research, but they have drawbacks. Some, like the U.S. Census, are conducted at 10-year intervals, providing only coarse temporal resolution. In other cases, different jurisdictions collect information on different aspects of a topic, such as subsistence harvest production versus participation in hunting and fishing. The indicators that are documented are usually chosen for purposes other than scientific research and rarely with the specific context of the Arctic in mind (e.g., AHDR, 2004; Baffrey and Huntington, 2010). The Survey of Living Conditions in the
Arctic (SLiCA4) has attempted to remedy this shortcoming by developing indicators of specific relevance to Arctic societies and their needs, but it cannot gather all that is needed, leaving many gaps in our ability to connect societal trends with each other or with biophysical processes. The Arctic Social Indicators project, which follows up on the activities of the Arctic Human Development Report (AHDR, 2004), offers ideas for indicators of Arctic human development. Other measures of societal factors include adaptive capacity indicators, which could be further developed for the Arctic to allow systematic assessment of adapting to change and allow communities and decision makers to weigh trade-offs in adaptation investments (e.g., Fussel, 2009). Efforts such as these, although limited, can yield lessons about the challenges of collecting societal data.
Monitoring efforts that address the physical and biological systems of the Arctic include observations of the atmosphere and cryosphere and their interactions with the boreal forests and the tundra biomes in the terrestrial realm and the broad continental shelves and subbasins of the marine environment. Each evolves and processes energy and materials in distinctive ways, subject to external forcing. Each also communicates with other systems through energy and material exchanges along a variety of pathways. For example, the marine and terrestrial environments are linked to one another through species migrations, river systems, changing glacial landscapes, and ocean currents. Some of the results from the Bonanza Creek LTER illustrate how addressing linkages within a monitoring program could be considered. That research indicates an increase in carbon export into Arctic river networks as a result of the degradation of permafrost and fire disturbances (Kicklighter et al., 2013). It is also apparent that rivers are the primary pathway by which mercury is entering the Arctic Ocean (Fisher et al., 2012) and that riverine mercury concentrations are likely to increase because of an increase in soil disturbances (Fisher et al., 2012; Leitch et al., 2007). This has implications for the Arctic Ocean’s carbon and suspended sediment cycles, trace metal budgets, and the Arctic trophic system. An appropriately designed Arctic monitoring system would include measurements of state variables and rates of critical processes within each system and energy and material fluxes along the pathways linking each to the other.
Within the marine environment, a similar ecological/latitudinal gradient monitoring approach is evolving in the Bering, Chukchi, and Beaufort Seas under the auspices of the Distributed Biological Observatory (DBO)5 program (Grebmeier et al., 2010). The DBO program is an international effort involving collection of data by Canadian,
Chinese, Korean, Japanese, Russian, and U.S. scientists, coordinated through the international Pacific Arctic Group6 and, within the United States, through the IARPC DBO Interagency task team. As conceived, the DBO is a holistic approach to track and understand the effects of changing oceanographic and sea ice conditions on the marine ecosystem. Until recently, biophysical sampling has occurred at several shelf biological hotspots from research vessels-of-opportunity that transit the region. The biological sampling, which samples water column and benthic organisms, seabirds, and marine mammals to evaluate species composition, biomass, and the size and condition of key organisms, also includes standard physical oceanographic and nutrient measurements. The shipboard sampling is largely limited to the open-water season but is supplemented by satellite measurements and data from oceanographic moorings (two of the DBO sites have biophysical mooring arrays, and two sites have only physical mooring arrays). However, at present many of the moorings are temporary components of limited-duration process studies, under national or international auspices, being undertaken in the region. Although the DBO program provides an emerging opportunity for assessing biophysical changes over western Arctic shelves, a more concerted effort to coordinate and systematize the sampling over seasonal and interannual scales will be necessary. As a result of western Arctic DBO activities, the Norwegian government is proposing a similar DBO project in the marine waters surrounding Svalbard.
The sampling strategy (duration, sampling rate, spatial extent, locations) of a particular monitoring effort will vary, depending upon the process or variable of interest. There will be a need to measure key system attributes at multi-decadal time scales at relevant rates and obvious locations. Other monitoring efforts need to be adaptive, taking into consideration results that emerge from retrospective (including paleoclimatic) studies, models, and other observations. These may suggest a hypothesis-based observation approach, perhaps of shorter duration (3 to 5 years) with a specific focus. If the results are found to address a critical need, then the sampling may transition into a longer-term effort. An adaptive monitoring effort also allows for the findings of an intensive process study to adjust monitoring activities. Statistical approaches or data assimilation models can aid in devising optimal sampling strategies. However, it is almost certain that resources will be inadequate to execute an optimal sampling strategy for many relevant variables. Here again, data assimilation models might clarify the trade-offs in designing options for sub-optimal (from a statistical perspective) sampling designs. Periodic evaluation can be used to determine whether the monitoring efforts need to be modified, augmented, or suspended.
The breadth and complexity of the Arctic system requires that long-term observations be a shared undertaking, involving international partners and coordinated efforts by government agencies, industry, communities, and scientists. We recognize the difficulties inherent in such coordination, given the different mission of each potential partner. Nevertheless, many or some of the core variables comprising a monitoring program will ideally meet disparate missions. One coordinating approach to consider is a national committee composed of various stakeholders and scientists. Such a committee’s charge would be to: (1) enhance coordination among monitoring activities at both the national and international level; (2) seek opportunities to increase sampling efficiencies and organize responses to “surprises”; (3) address the various needs of the diverse suite of stakeholders that benefit from long-term observations; (4) assist in prioritizing these needs among stakeholders; and (5) communicate monitoring activities and results to policy makers and stakeholders in a coherent manner. Such a committee could be organized by an existing entity like IARPC.
Just as science depends on data, scientific progress depends on access to data. As Arctic research expands, and as datasets grow rapidly in an era of information technology, keeping track of what has already been recorded or accomplished is increasingly difficult. Current efforts to coordinate data management and access are commendable, but much remains to be done. Further progress is likely to depend upon concerted and coordinated efforts rather than reliance primarily on individual researchers or funding programs.
Arctic science has a history of large and interdisciplinary programs, so there is some precedent for successful management of complex datasets. The need for interdisciplinary and intersectoral management is not limited to the Arctic, and there is an opportunity for the Arctic research community to become a leader in developing a culture of data management and sharing. Strategies for achieving the greater cooperation necessary for such a culture were addressed earlier in this chapter, and specific suggestions for managing and sharing information are presented in this section.
We now understand the Arctic to be a tightly coupled, integrated system, where changes in one component will reverberate through the system, initiating a cascade of impacts in other components of the system (Roberts et al., 2010). Understanding
and quantifying these system interconnections is possible only through simultaneous analyses of extensive and often numerous complex datasets from disparate sources. As scientific urgency drives our research endeavors to collect more kinds of observations more frequently and at more numerous sites, we are compelled to develop new techniques to analyze the resulting massive datasets (Pundsack et al., 2013). Moreover, the recognized value of well-documented data for application in new and different analyses places utmost priority upon data preservation, stewardship, and access by multiple stakeholders. While placing great responsibility upon individual scientists and agencies, this realization also elevates the collective responsibility of all engaged in Arctic research to strive to garner the greatest value from our investments in observations and monitoring. The recently published U.S. Arctic Research Plan (Executive Office of the President, 2013) has charged all agencies to “demonstrate new and updated cyberinfrastructure tools to enhance data integration and application and identify opportunities for sharing of technology and tools among interagency partners” (p. 21).
To meet these pressing needs for more efficient utilization of our data resources, it is imperative to establish interoperable data management systems that are adequate for academic needs and to assess progress against agency/collaboration goals. Developments in the field of informatics could yield important lessons for managing large amounts of data and creating interoperable systems. Our present system of data submission by researchers and curation by institutions often results in gaps in data awareness, distribution, and quality of metadata. An additional challenge for data management remains that of achieving interoperability of biophysical and socioeconomic data, as well as determining how to integrate traditional ecological knowledge. Integrating data management and quality control into network design aids in overcoming such deficiencies. Currently, tremendous amounts of work are required by researchers who compile data from various sources. Prescribed formats to be used by all agencies, with structured data submission, archiving, and delivery, would greatly enhance efficiency of analyses by the broader community. One solution would be to create an interagency data management committee (possibly through IARPC) to coordinate structure and dissemination protocols. Such a committee could identify high-priority data sets and identify responsible agencies to support data collection. Additionally, advances in curation technology will make integration of diverse datasets easier and analysis of disparate data streams seamless.
Many advances in Arctic science have resulted from broad-scale synthesis of relevant data streams. These advanced analyses have been made possible by technological
advancements in computing power and search capabilities. However, we can foresee even greater advances on the horizon with the advent of data archiving and harvesting techniques. Data curation has long been recognized as an essential function of operational agencies, but it has only recently been acknowledged as an individual responsibility of every investigator. Moving forward with every scientist accepting a commitment to preserve and share his or her data will greatly enhance our capabilities. To realize the utopian community of data sharing, it may be necessary to encourage data submission by requiring a portion of each grant be dedicated to data curation. Concurrently, we need to establish a robust method of documenting and crediting data sharing through a formal citation protocol. Also, such magnanimity of data sharing has not always been the standard; support will be necessary to secure older, stranded datasets and rescue those high-quality observations that may provide essential clues to past rates of change.
The service provided by formal data centers is clearly imperative, but it is quite difficult to secure funding to support such centers. Critical components of our research infrastructure are agency-supported data centers, which are mission- and discipline-specific, yet interconnected and transparent in terms of data accessibility. Reliable computer systems for storing, accessing, and assessing the quality of data are the crucial backbone of institutional repositories. Compatible architecture using a shared cloud environment as a computer platform would greatly enhance data sharing and transparent accessibility.
Real-time monitoring networks are indispensable for detecting and documenting change, providing validation for model simulations, and elucidating the quantitative relationship among related processes (see “Maintaining Long-term Observations” section earlier in this chapter). It is essential that we sustain a commitment to maintain monitoring networks for the long term, but it is also important that we establish a more seamless flow of data from the observations, through quality checking or quality control, into a permanent long-term archive. The flow of data from our observing networks into permanent archives can be disrupted or delayed, limiting our capacity for analyses and syntheses.
A similar challenge arises when working with the traditional knowledge and local observations of Arctic residents. Field scientists have long valued the knowledge and wisdom of local residents. Roald Amundsen spent 1903 to 1905 in what is now known
as Gjoa Haven, Nunavut, Canada, collecting magnetic measurements and learning from Inuit (Amundsen, 1908). These lessons in Arctic survival gave him the knowledge required to complete the trek to the South Pole in 1911. The collective experience of local observers and knowledge passed from one generation to the next reveal evidence of the changing climate and environmental and ecosystem responses to those changes, but this information source has not been fully utilized for its potential value for either inquiry-based science or as model validation data (Huntington, 2000; Huntington, 2011). It is incumbent upon the Arctic research community to more fully engage local residents as partners and collaborators to ensure that the changes observed today are correctly positioned in historical context and that projections of future change connect environmental and social responses. Such an effort would help address the problem of things we think we don’t know, as described in Chapter 2. The Exchange for Local Observations and Knowledge in the Arctic (ELOKA) is among those working to address this challenge (Pulsifer et al., 2012).
Data centers also need to serve a dual mission of archive and synthesis and be capable of integrating individual projects, real-time data streams, traditional knowledge, and “big data” that are now accessible through a myriad of data-mining techniques. We are presently limited in our ability to achieve major scientific advances because of technological limitations in our capacity to efficiently synthesize and analyze big data. The field of bioinformatics, the science of creating an understanding of complex biological systems by leveraging large datasets and computing power, is a mature field. Geoinformatics, using similar techniques in Earth science applications, is by contrast relatively nascent. The big data necessary for such endeavors are emerging from existing sensor networks and geophysical observatories currently placed in the Arctic, with more planned for the future. Such big data processing capability enhances our capacity for integration, synthesis, assimilation, and assessment and lends promise to sweeping advancements in climate, ecosystem, and socioeconomic science. The culture of data sharing and a strong set of data management standards are crucial for the burgeoning field of geoinformatics and deserve high priority.
The goal of an Arctic cyberinfrastructure (CI) is to provide freely and openly accessible quality-controlled datasets to a variety of users, including the public, management agencies, industrial users, educators, and scientists. To achieve this goal, computing infrastructure needs the capability to integrate data from diverse sampling platforms (e.g., autonomous sensors collecting time-series data, process-oriented but relatively short-lived field programs, and traditional ecological knowledge) interactively into a coherent architecture. Ideally such a system would permit users to:
- analyze and model Arctic processes;
- develop and test hypotheses;
- adjust measurement strategies to allow for adaptive sampling;
- facilitate responses to environmental events;
- enhance predictive capabilities on both short and long time scales; and
- contribute to the maintenance and reliability of the measurement systems.
At a minimum, the Arctic CI requires data preservation and access as has been performed traditionally by centrally managed data archives that ingest and serve metadata and data. More advanced data centers such as the Advanced Cooperative Arctic Data and Information Service (ACADIS) and the National Center for Atmospheric Research (NCAR) Earth Observing Laboratory (EOL) also provide software and advice on metadata and data submission and facilitate data searches, access, formatting, and visualization. ACADIS is a joint effort by the National Snow and Ice Data Center (NSIDC), the University Corporation for Atmospheric Research (UCAR), UNIDATA, and NCAR that was established to provide data archival, preservation, and access for projects funded by NSF’s Arctic Science Program, including the Arctic Observing Network (AON). ACADIS also links to the EOL holdings and the data archive of the NSIDC. In addition, ACADIS is presently hosting the PacMARS data archive. PacMARS is attempting to link, under one data archive umbrella, the large number of marine-related datasets (including those funded by agencies other than NSF) from the northern Bering, Chukchi, and Beaufort seas.
Data sources from outside the U.S. academic research community (including those of international scientists and governments, U.S. state and federal resource managers, industry, and the military) will also need to be integrated. SAON was established for this purpose with a goal to “support and strengthen the development of multinational engagement for sustained and coordinated pan-Arctic observing and data sharing systems that serve societal needs, particularly related to environmental, social, economic and cultural issues” (SAON, 2011). The challenges of sustaining international observing networks have impeded success in promoting open access to data among various national data archives. Additional international partnerships and agreements are necessary to promote truly transparent data access, which will open up new avenues of research and application from a variety of stakeholders. For example, the Department of Defense’s first ever Arctic Strategy document stated, “DoD will also collaborate with international partners to employ, acquire, share, or develop the means required to improve sensing, data collection and fusion, analysis, and information-sharing to enhance domain awareness appropriately in the Arctic” (DOD, 2013, p. 9).
The existing Arctic CI facilities allow the achievement of the listed goals to varying degrees. However, as Arctic observing and modeling programs become more interdis-
ciplinary and more comprehensive networks of autonomous measurements evolve, a more sophisticated CI system is desirable (Pundsack et al., 2013). Such a system might follow the design criteria and incorporate the various elements of the developing CI components of NSF’s Ocean Observing Initiative (OOI) and NOAA’s Integrated Ocean Observing System (IOOS). Both programs ingest and serve data in real time from a large number of autonomous sensors. To take full advantage of such autonomous systems, we need to simultaneously improve our communications capability to enable access to sensor networks in extremely remote locations. Presently, lack of infrastructure and high-power requirements of some communication packages place insurmountable limitations on remote monitoring capabilities. As outlined by Chave et al. (2009), the OOI system includes the capability for operator-to-machine and machine-to-machine control of data collection and analysis, enables model interaction with data acquisition processes, supports virtual collaborations of observing system resources among a variety of uses, and provides some degree of automation in the planning and execution of observing system components. In addition, the OOI CI acts as an operating system that provides the messaging, governance, and service frameworks for the system. Meisinger et al. (2009) suggest that this architecture take advantage of the cloud computing environment, which facilitates scalability and flexibility. Scalability addresses users’ requirements that may encompass a broad range of time and space scales and information types. Flexibility allows for the incorporation of technological developments in distributed networks, sensor technologies, models, and computing. These developments are well under way, and the lessons learned from these activities are likely to prove valuable in guiding improvements to an Arctic CI.
Many gains have been made outside the Arctic science realm that could be brought to bear on problems related to the Arctic. From this, we may find a wealth of what we think we don’t know.
Visualization technology is highly developed in the computer gaming industry, both in hardware and software, and such technology can be applied to scientific use. Additionally, visual analysis in industry has become highly advanced, for example the seismic visualization capability of the oil industry. Leveraging advances like these for the use of Arctic scientists and stakeholders could result in significant gains at modest cost. Many users of data have a need for quick, easy visualization. Steps in this direction have been taken internationally through the Arctic Monitoring and Assessment
FIGURE 4.3 “Arctic Collaborative Environment (ACE) Joint Capability Technology Demonstration (JCTD) is an Internet-based, open-access, Arctic-focused, environmental research and decision support system that integrates data from existing remote sensing assets with products from existing and new environmental models to provide monitoring, analysis, and visualization based on earth observation data and modeling. With an initial focus on the Arctic region, researchers, students, search-and-rescue operators, native hunters, etc. can draw from the open-access data.” SOURCE: https://ace.arsc.edu/workspace.
Program of the Arctic Council,7 the Arctic Portal,8 World Wildlife Fund,9 Conservation of Arctic Flora and Fauna,10 and nationally through NOAA’s Arctic Environmental Response Management Application,11 NOAA’s Earth Systems Research Laboratory,12 and the emerging Arctic Collaborative Environment13 (Figure 4.3).
Once databases have the right data in terms of space (e.g., include downscaled model results), time (e.g., are in real time when possible), and utility (e.g., are useful for both basic research and decision support), the visualization challenge posed here is to gen-
erate or determine a system that can adapt to differing data formats, dimensions, and other factors as well as to generate products responsive to the spatial and temporal requirements and formats needed by various user communities. Further, the ability to generate quantitative information becomes important. Methods of analysis (such as differencing, statistical tools, and more complex numerical analyses) are integral needs of such visualization packages. Success depends upon an increase in the types and range of visualization data (e.g., completing multi-beam surveys in the Arctic Ocean, improved access to satellite visualizations, development of real-time interactive visualizations so that sensor activation can be based on automated visual analyses).
In addition to visualization technology, the gaming industry has produced hardware that has been co-opted into the scientific community. See, for example, the applications of the Microsoft Kinect (Mankoff and Russo, 2013). Another example is the integration of Graphics Processing Units into massively parallel computing architectures.
Miniaturization of data logging and wireless technologies including video is a technology transfer issue. For example, modern smartphones contain vanishingly small cameras that are of increasing quality for both still images and video. In addition, the wireless capability of these devices is impressive. The only (significant) deficiency is robustness. These devices are already being investigated for acquisition and control applications.
Digital photogrammetry from traditional aircraft is an underutilized resource. For example, NASA’s Operation IceBridge has flown numerous missions over the Arctic for the past several years, primarily covering targets on land ice and sea ice (e.g., Studinger et al., 2010). Each flight carries the Digital Mapping System nadir-viewing camera; from this camera there is sufficient overlap on images to allow stereo photogrammetry.
It’s getting harder and harder to find a proper block of ice to sustain one of these stations.
—Viktor Bovarsky, former polar explorer
It is critically important to establish and maintain consistent networks of measurements and robust infrastructure to detect Arctic changes. There is a general lack of in situ infrastructure across the Arctic, including both mobile and fixed observing systems. Some long-term observatories are being discontinued and some satellite systems are now retired, creating a gap in observing.
Observations need to be comparable across individual sites, allowing for network-wide analyses and integration. Often there is a need for rapid response. Observations
need to be carried through autumn, winter, and spring, not only in the convenient summer season. There is a need for in situ observations along the coast and below the sea surface as well as coastal observing, because most remote systems (i.e., satellites) have low resolution in coastal zones and no data are collected below the sea surface. Finally, we need to leverage connections with industry.
Mobile platforms are important for monitoring physical, biological, and chemical oceanographic changes in the Arctic Ocean. Mobile platforms include floats (e.g., Argo), autonomous underwater vehicles (AUVs), ocean gliders, remotely operated vehicles (ROVs), and larger platforms such as ships. Recent advances in miniaturizing sensors have also enabled the use of marine mammals as mobile platforms (e.g., the ocean tracking network), which could be extended to smaller animals in other environments as well.
AUVs, such as buoyancy-driven ocean “gliders,” propeller-driven AUVs, and Wave Gliders®14 have substantial potential for environmental monitoring, ocean process studies, and inspection of industrial facilities in the Arctic Ocean and its adjoining shelves. Each of these vehicles can collect high-resolution data that may be transmitted in near real time. Mission protocols can either be preprogrammed or adjusted at sea to permit adaptive sampling. These vehicles operate differently from one another and can be used independently or collaboratively. Both gliders and AUVs have been applied extensively in open-water settings and now increasingly in the ice-free waters of the Arctic (e.g., Shroyer and Plueddemann, 2012; Timmermans and Winsor, 2013). All of these vehicles come in a range of sizes and capabilities. Deployment and recovery of the smaller vehicles can be done by hand from small vessels (including skiffs) and/or through the ice, but larger vehicles require mechanical aids (hence larger vessels or ice camps).
AUVs are well suited for mapping missions because their navigational ability is more precise than gliders, especially if guided by transponders. However, their endurance is limited to hours to days because their propulsion systems consume considerably more power than gliders. Under-ice AUV operations have a long history (e.g., Francois
and Nodland, 1972) with recent applications including under-ice mapping (Wadhams, 2012), seafloor exploration (Kunz et al., 2009), bathymetric mapping (Crees et al., 2010), and coastal hydrography (Plueddemann et al., 2012).
Gliders move vertically by adjusting buoyancy and use wings and a rudder to control horizontal motion. They have relatively long endurance (weeks to months) and can carry a diversity of sensor packages, although these are limited by size, weight, and power consumption. Under-ice glider operations are a more recent development (Curry et al., 2014). Wave Gliders ride the ocean surface and harness wave energy for their propulsion and solar power for recharging their communications and sensor systems. Wave Gliders have been used in mid- and low latitudes, but their performance at high latitudes has yet to be evaluated.
Gliders and AUVs can incorporate a variety of sensors, although the sensor configuration (and subsequent mission) may be limited by the size of the vehicle and the power required for the sensor configuration. Nevertheless, gliders and AUVs easily support standard oceanographic sensors (e.g., Conductivity, Temperature, Depth [CTD] instruments, optics), and AUVs can also incorporate Acoustic Doppler Current Profilers (ADCPs) and side-scan and/or ice-profiling sonars. Each vehicle has the potential to incorporate passive acoustic recorders for marine mammal detection. As new ocean sensors evolve, many of these are likely to be easily adaptable to one or more of these vehicles. Sensor packages for Wave Gliders are more limited, given their size and their propulsion mechanism, which limits the depth to which sensors can be deployed. Nevertheless, Wave Gliders could be useful in sampling the uppermost 5 to 10 m in ice-free conditions during the summer months.
There are several hurdles to overcome to expand the use of gliders and AUVs in the Arctic. For example, gliders have difficulties navigating under ice, although the under-ice navigation approach of Curry et al. (2014) is promising. Those approaches will be further refined as outlined in the Office of Naval Research’s Marginal Ice Zone Program (Lee et al., 2012). In some regions of the Arctic, swift currents may result in glider loss or prevent the glider from conducting or completing its mission. Depending on the capacity of its buoyancy engine, strongly stratified waters (associated with ice melt and/or river outflows) may prevent the glider from surfacing. Larger buoyancy engines such as those used in the Exocetus Coastal Glider (Imlach and Mahr, 2012) could overcome the impediments associated with swift currents and stratification. Through-ice glider deployments and recoveries also deserve further exploration. Necessary glider improvements include incorporating inertial and acoustic navigation systems and a glider propulsion mechanism that would be used intermittently to enable gliders to navigate precisely to an ice hole for recovery.
A variety of short-duration, attended AUV deployments under ice have been demonstrated. Extended, unattended operations beneath the ice in the high Arctic will require substantial new developments for navigating, providing power, and communications. Such developments would include an autonomous on-ice power and communication system that drifts with the ice and incorporates a through-ice docking port by which the AUV can recharge its batteries, transfer data to the surface, and receive new mission guidelines. It would also require the distribution of an acoustic transponder network (drifting with the ice or fixed on moorings or on the ocean floor) and acoustic modems for passing the position of drifting beacons to the vehicle. Improvements in decision-making software for docking and for choosing the appropriate set of transponders by which to navigate are also needed. An alternative docking scenario may be feasible in the event that offshore hydrocarbon development occurs and subsea pipelines extend onshore. It may be possible to incorporate fixed AUV docking ports and communication and power cables with the pipeline.
Although these are formidable hurdles, many of the necessary elements are currently being developed. A specific challenge is to merge these capabilities into an integrated system for use in the Arctic. Substantial advancement is anticipated over the next 3 to 5 years driven by scientific research as well as interest in seafloor mapping and subsea resources. For example, the OOI is addressing unattended AUV power and recharge systems, data storage and communications, and two-way command and control issues. A prudent course of action would be to allow successful resolution of these issues by the AUV community, while simultaneously planning how to adapt AUVs for the unique conditions of the Arctic. It is nevertheless conceivable that such a system may be feasible and applicable to the Arctic within the next 10 years.
In addition to autonomous vehicles, a variety of drifting sensor platforms (buoys) has been developed for Arctic Ocean applications. These buoys are either installed into and drift with the ice or drift in the ocean below the ice. These include ice mass balance (IMB) buoys (Jackson et al., 2013; Richter-Menge et al., 2006), designed to determine rates of ice and snow accretion and ablation; autonomous ocean flux buoys (AOFB) that measure the turbulent fluxes of heat, salt, and momentum between the upper ocean mixed layer and the ice; and ice-tethered profilers (ITP; Krishfield et al., 2008) that sample the upper ocean hydrography and, depending on configuration, a variety of other parameters including fluorescence, irradiance, oxygen, and velocity (from within ~5 m of the ice to 250 to 800 meters, depending upon application). The IMBs, AOFBs, and ITPs use Global Positioning Satellites (GPS) for positioning and transmit data via Iridium. Polar profiling floats (PPF) are analogous to the profiling floats used in the Argo float program. Specifically, the floats drift at a fixed depth but periodically rise to the surface, profiling the temperature and salinity structure of the
water column. Once at the surface they transmit the data via satellite, receive a GPS fix, and then descend again. PPFs do not break through the ice, but they will surface if open water is present and then transmit their data and obtain a GPS fix. For periods of extended under-ice operations, the PPFs use fixed sound sources for geopositioning but store their data until they reach open water. Although most of these devices have been developed for ice and ocean physics applications, it is feasible that other sensors can be adapted to these as well.
Argo floats currently span all oceans except the Arctic, where access to sea surface communications is limited. Enabling them to be used in the Arctic Ocean would greatly advance our understanding of physical changes within this ocean’s deeper water masses. A technology proposed by Sagen et al. (2011) would enable this technology to be deployed in the Arctic Ocean by the installation of a basin-wide undersea navigation and communication system.
Numerous reports have discussed the continued needs for ships capable of working in the Arctic Ocean (e.g., Dunbar et al., 2012; NRC, 2003; NRC, 2007; NRC, 2011; USCG, 2013; U.S. Navy, 2009). All have identified research questions that can be suitably addressed only with the access provided by research, icebreaker, and drilling ships (rather than autonomous or remote instrumentation). Sustained use of ships is also envisioned for deployment/recovery of stationary or mobile installations equipped with autonomous samplers (e.g., moorings, AUVs/gliders). With the diminished summer sea ice extent, and the new availability of the ice-capable research vessel R/V Sikuliaq, as well as other non-ice-capable research vessels, access to a larger portion of the Arctic Ocean during ice-free months can be achieved using the assets at hand.
However, access to some regions of the Arctic will still require the use of a medium or heavy icebreaker. A number of emerging research questions in the Arctic can be addressed only through shipboard access during all times of the year. This can be achieved by expanding the capabilities of ice-capable ships and icebreakers to deploy and support traditional and new equipment, instrumentation, and technologies in ice-covered seas. Research questions pertaining to oceanic gateway, sea surface temperatures, long-term climate excursions, gas hydrates, oceanic-crust architecture, and tectonic as well as magmatic evolution of the Arctic Ocean Basin require access to deep drilling capability with riser and blowout preventer systems. Drilling of the seafloor could be accomplished through management of ship and sea ice movements using both a moon pool and sophisticated ship-handling technology. Advanced ice
clearing capabilities are also necessary for deployment of AUVs and ROVs in sea ice. UAVs will also be increasingly utilized in the Arctic, and research vessels and icebreakers need to be capable of supporting the deployment of UAVs.
Present U.S. icebreaker capability for medium-to-heavy ice is minimal. The USCGC Healy, a medium icebreaker with a primary mission of science (Figure 4.4), is at mid-life (commissioned in 2000) and will need to be replaced, under normal ship life length, in ~15 to 20 years. Furthermore, the Healy crew is rotated approximately every 2 years, diminishing institutional memory and science experience in the operation of the ship. Retaining crews for longer periods of time would improve the operational capacity of the Healy, resulting in more efficient use of science resources. The heavy icebreaker USCGC Polar Star has recently returned to service after extensive refurbishment and will primarily serve national security interests in the Arctic and McMurdo Station in Antarctica. The science mission requirements for a new polar icebreaker were recently updated at the request of the National Science Foundation (Dunbar et al., 2012). That report identified the need for a medium icebreaker research vessel to address current and future research questions while being reasonably economical to operate (in lieu of a heavy icebreaker). Still, it is important to identify a means to increase heavy ice-
FIGURE 4.4 Scientists obtain samples on the sea ice during a cruise to the northern Chukchi Sea using the USCGC Healy (background). SOURCE: Steve Roberts.
breaking capability, either through new construction or by leasing a vessel that can be used either for science or to provide escort services for a less ice-capable research icebreaker.
It is also important to retain and increase access to non-icebreaking research vessels in the Arctic through increased funding for Arctic research, increased coordination of research activities to maximize use of available assets, greater use of private sector assets (including research vessels as well as platforms of opportunity), and the development of large, multi-investigator, multidisciplinary research programs and by operating research icebreaking assets as efficiently as possible. At present, research vessel time is available primarily on the USCGC Healy or non-UNOLS vessels. Because of the downsizing of the UNOLS fleet, availability of platforms in the U.S. research fleet for Arctic work is inadequate. Additionally, the perception that inclusion of ship time in research proposals diminishes the likelihood of funding has driven a decline in the number of proposals requesting ship time across all oceanographic disciplines (UNOLS, 2013).
A range of stationary marine platforms already exists. Some types are used routinely (e.g., moorings), whereas others are relatively rare or absent in the Arctic, although commonly deployed in other oceanic regions (e.g., shore-based installations, cabled marine observatories). There is still much room for improvement in the capabilities and deployment of both stationary (sea-floor deployment) and semi-stationary (sea-ice deployment) platforms from which to monitor a range of ocean and atmosphere characteristics over all seasons. The platform types include bottom-moored and ice-tethered profilers equipped with a range of physical, meteorological, biological, and chemical sensors; free-floating and ice embedded buoys; cabled marine observatories; and shoreline instrumentation such as tide gauges, meteorological packages, and coastal ocean dynamics applications radar (CODAR) in remote locations. It is also desirable to network data collected from these remote installations into a common location.
In addition to these marine-based fixed observatories, there already exist many terrestrial observatories that need to be sustained in order to address critical Arctic research questions. An example is Summit Station in central Greenland, where atmospheric and snow chemistry measurements have been made for decades, making it an important node in the network of Arctic climate observatories. Similarly, Toolik Field Station in Alaska provides an important observational platform.
Efforts to combine these in situ observations with local community and traditional knowledge, so that local residents’ priorities with respect to climate change can be monitored and assessed, are critical. It is also important to integrate local and community-based observing into operational and research activities. We need to empower local residents to monitor their own environments and assist in the coordination of these community-based monitoring observations (Pulsifer et al., 2014). These locally based observing platforms require strong partnerships between communities and scientists to capture the knowledge of community members. One valuable aspect of these observatories is the ability to place current observations in a local historical context. Local involvement is discussed further in the section entitled “Community Engagement” later in this chapter.
Satellite and airborne observations provide the largest spatial coverage of the Arctic and have proven to be important tools for detecting the impacts of climate change. For example, satellite remote sensing data have allowed the quantification of sea ice loss and the mass loss of ice sheets that contribute to sea-level rise; surface temperature changes; atmospheric changes; shrub expansion northward; changing wetlands and lakes on the north slope; and coastal shoreline changes. Remote sensing makes it possible to scale what is observed on the ground at plot scales up to landscape scales for improved broadscale understanding of patterns of change and for extrapolating that knowledge to grid cells for modeling. Satellite remote sensing has and will continue to play a major role in monitoring and detection of change in the Arctic.
Arctic conditions present many challenges to the interpretation of satellite remote sensing data. The Arctic is characterized by low solar illumination, low vegetation biomass, low primary productivity, perennial snow and sea ice, prolonged darkness, persistent low clouds, and frequent temperature inversions, all of which severely limit radiometer accuracy and monitoring capabilities. Much progress has been made in recent decades in remote sensing applications, but many obstacles remain in retrieving useful information from high latitudes. For example, some satellite systems fly in orbits that simply do not provide Arctic coverage. In addition, many remote sensing products and calibration algorithms are developed for temperate or tropical systems and thus may be inappropriate for the Arctic. The standard atmospheric correction algorithms such as those used by the Landsat Ecosystem Disturbance Adaptive Process-
ing System (LEDAPS) do not work well in the tundra because of changing solar angle variation across the scene. In addition, standard image products from sensors such as NASA’s Moderate Resolution Imaging Spectroradiometer (MODIS) or the Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) are developed primarily for temperate or tropical systems; the MODIS net and gross primary-productivity products overestimate low-productivity sites such as tundra ecosystems (Turner et al., 2006). Cloud detection over snow- and ice-covered scenes also remains a challenge for imagers and sounders, and the frequent temperature inversions over Arctic regions are problematic for retrieving vertical profiles. At an Arctic remote sensing workshop in October 2013, participants cited the lack of calibration of remote sensing products to the Arctic as the number one current concern for effectively observing changes in the Arctic.15
For airborne and satellite remote sensing collections, field data are important for training and validation; these data require collection over an area representative of the spatial resolution or minimum mapping unit of the remote sensing platform. In this regard, distributed measurements may be collected across a somewhat “homogeneous” area and averaged to relate to the image observation resolution. For example, for mapping vegetation cover to a minimum mapping unit of 0.2 ha, field characterization data require collection in a representative 40 m x 50 m area. For coarse resolution satellite platforms (1 km or greater) typically the observed landscape is not homogeneous for a sufficient number of coincident field measurements to be made. In this case, an intermediate remote sensing product (~30 m resolution) may be employed, where the field data are used to train or validate the intermediate product across a range of “homogeneous” cover types within the coarse resolution cell of the targeted sensor (Liang et al., 2002), and then these intermediate data are upscaled to the coarse resolution sensor. In some cases this intermediary step is not an option and a network of field measurements is necessary across the resolution cell (e.g., soil moisture from passive microwave; Jackson et al., 2010). The types of data that are necessary for addressing the existing and emerging questions raised in Chapter 3 and that have the potential to become available from satellite sensors include the following:
Lake depth (bathymetry), precipitation, and evapotranspiration. Remote sensing has been/is being used to map where the water is (MODIS for lakes, AMSRE for fractional wetness), but characterization of the impact of climate on surface water and hydrology cannot be attained without information on lake depth and other hydrology parameters (e.g., precipitation, evapotranspiration). NASA’s Global Precipitation Measure-
ment (GPM) mission is scheduled for 2014 launch and will address the precipitation needs.
Sea ice and land ice thickness. These data have been successfully retrieved from NASA’s Ice, cloud, and land Elevation Satellite/Geoscience Laser Altimeter System (ICESat/GLAS). Together with aerial surveys from the IceBridge program, a continuous record of ice and snow thickness estimates is being collected and will be extended by ICESat2 (scheduled for launch in 2015). These data will assist in addressing emerging questions related to changes in ocean/ice/atmosphere energy exchange, ocean currents, and sea level rise.
Snow depth on ice surfaces. The ICESat/GLAS and other laser altimeters are able to estimate the thickness of snow and ice as a freeboard measurement. The laser altimeters reflect from the snow-air interface; therefore, to obtain snow depth on ice, radar altimetry, which reflects from the snow-ice interface, is also needed. Others have used passive microwave data to estimate snow depth. More consistent measurements are needed with better spatial coverage. Snow information is essential to answer questions related to surface energy exchange and for sea ice thickness.
Permafrost, soil moisture, active layer depth, and soil organic carbon stocks. Satellites are beginning to provide estimates of changes in high-latitude vegetation, freeze-thaw processes, soil moisture, and burn severity. However, these are limited by calibration of the systems and algorithms developed for temperate systems. For example, standard burn-severity mapping algorithms do not work well in the tundra, but scientists are developing algorithms specific for the Arctic (e.g., Loboda et al., 2013). Also, many of these systems have coarse spatial resolution (e.g., Soil Moisture Active Passive [SMAP] has 3 to 9 km resolution). Although SMAP will provide data on freeze-thaw processes and soil moisture, its relative utility for defining active-layer depth is uncertain. The resolution of SMAP is also still too coarse to define landscape heterogeneity in conditions influencing permafrost and soil organic carbon distributions (ideally, resolution needs to be closer to 30 m than 3 km). Synthetic Aperture Radar (SAR) satellite systems are of high spatial resolution (~30 m) and widely used for ice monitoring, but they are underutilized in the Arctic for land applications. Use of SAR and/or Interferometric Synthetic Aperture Radar (InSAR) techniques for soil moisture and active layer retrieval, assessment of carbon stocks, permafrost deformation, and other needs in the Arctic have been demonstrated to have great potential but require further research and development for widespread application. Changes in the Arctic terrestrial ecosystem will be assessed in the Next-Generation Ecosystem Experiment (NGEE): Arctic Landscapes project, in which data from satellite-based laser altimeters will be combined with
biogeochemical models. Monitoring and changes in the high northern latitudes will also be assessed in NASA’s Arctic Boreal Vulnerability Experiment (ABoVE), a major field campaign scheduled to begin in 2015 and run for 5 to 7 years over Western Canada and Alaska. For a more detailed discussion of remote sensing tools for understanding permafrost, see (NRC, 2014a).
Atmospheric boundary layer. The strong near-surface inversions under a frequently overcast sky cover present a particularly difficult challenge to satellite sounding systems, yet knowledge of boundary-layer stratification is essential for determining surface-atmosphere exchanges. Higher spectral resolution systems such as NASA’s Atmospheric Infrared Radiation Sounder (AIRS) combined with the Advanced Microwave Sounding Unit (AMSU) have the potential to provide more accurate retrievals of profiles and cloud information in the critical Arctic boundary layer.
Cloud properties. Estimates of cloud optical thickness, phase, and base height—particularly over ice- and snow-covered surfaces—require additional detail and accuracy. Improved retrievals may be possible from AIRS/AMSU, MISR, GLAS/ICESat, ICESat-2/ATLAS, and instruments in NASA’s A-Train constellation. Cloud information is essential for determining the surface energy balance and atmospheric chemical processes.
Sea ice motion. A near-real-time, high-resolution product is needed for assimilation into dynamical models to provide more accurate sea ice predictions. Coverage for such a product is a challenge, particularly for optical systems, and may require a constellation of satellites. The new Sentinel-1 SAR satellite mission will provide high repeat coverage of the Arctic allowing more frequent information on sea ice, including ice motion.
Repeatable landcover mapping techniques at high spatial resolution. High-resolution (30 m or less) circumpolar land-cover maps are needed as baseline, to detect changes and to aid modeling. Current maps are either geographically limited, are of low spatial resolution, or lack accuracy because ground control is limited. The National Geospatial-Intelligence Agency (NGA) high resolution database may be useful for this purpose, as may multi-sensor approaches that include Landsat, SAR, LIDAR, hyperspectral observations, and other satellite data sources.
Digital Elevation Models, ground surface height and geodetic control. Arctic land areas (including ice-covered) currently have poor-resolution Digital Elevation Models (DEMs; 60 to 90 m). High-resolution DEMs are
FIGURE 4.5 Alaska Mapped digital elevation model. SOURCE: alaskamapped.org.
necessary for improved modeling, geospatial analysis, and remote sensing analysis. In Alaska, the State Digital Mapping Initiative is a program using airborne interferometric SAR to produce high-resolution DEMs and imagery (e.g., SPOT) to produce ortho images for mapping16 (Figure 4.5). The NGA provides access to data at no charge to civilian agencies. For example, 2.5 million scenes over the Arctic and Antarctic of commercial submeter imagery have been collected by NGA and are currently being used by the Polar Geospatial Center in Minnesota to create DEMs at 2 to10 m resolution for portions of Antarctica. Such data could be used to map the pan-Arctic.
Measurements relative to a stable datum would enable measurement of seasonal variations of surface elevation dynamics and long-term subsidence associated with degradation of ground ice. This could possibly be incorporated with NOAA’s National Geodetic Survey (NGS) of the national Continuously Operating Reference Station (CORS) network. These are highly accurate Global Navigation Satellite System (GNSS, formerly GPS) receivers. Installation of a small subset of foundation CORS in the Arctic
is needed to supplement the network. The foundation CORS will improve the accuracy of the International Terrestrial Reference Frame.
The Gravity for the Redefinition of the American Vertical Datum, or GRAV-D, is a program initiated by NGS to redefine the vertical datum of the United States. NGS is prioritizing collection of airborne gravity data in Alaska. This is the most cost-effective way to establish geodetic control in these areas and will allow the increase of elevation measurement accuracy from 1 meter (or worse) to 2 centimeters. Less than 10 percent of Alaska has contemporary shoreline data, and less than 1 percent is mapped annually. This program needs to be expanded to the pan-Arctic.
Improved weather and sea condition forecasts. New observing technologies (in situ, airborne, and satellite) could help to fill existing gaps in meteorological and oceanographic datasets and improve weather forecasts. Beyond approximately 60 to 72 hours, forecasts of weather and sea conditions lack sufficient detail. The resolution of the observational fields that support both meteorological and oceanographic modeling exacerbates this discrepancy (see Chapter 3, Emerging Question M4.)
In addition, there are some infrastructure needs to aid in sharing and dissemination of imagery and sensor data. An autonomous network to uplink and disseminate multi-sensor information about sea ice and other Arctic data is needed.
There is also a need to improve access to satellite imagery, including access to foreign satellite observations and commercial data.
Unmanned Aerial Vehicles
In gathering community input for this study, a frequently identified technology that would facilitate Arctic research was unmanned aerial vehicles (UAVs), or drone aircraft. The Arctic is a remote and challenging region to conduct research. In addition to extreme weather/sea conditions and transportation obstacles during much of the year, the Arctic consists of large expanses of sparsely populated areas with limited access that combine to make environmental observations difficult at best. As a result, aircraft ranging from balloons to transport airplanes have long been an important tool for the collection of observations on the physical, chemical, and biological systems of the Arctic.
Although manned aircraft have the capacity to afford access to broad and remote areas of the Arctic, this access is not without significant peril. With extremely limited infrastructure for emergency alternatives or rescue in the case of failure, manned aerial operations are rightly approached with caution. In recent years, reduced tolerance for
risk on the part of investigative agencies and the private sector have increasingly restricted aerial access to remote areas and limited the scope and scale of data acquired.
During the late 1990s and early 2000s the rapid development and utilization of UAVs by the military provided the possibility of new capabilities for aerial operations in remote areas with challenging flying conditions. The UAV industry now includes options ranging from small hand-launched line-of-sight operated craft to large airframes that are capable of extensive periods aloft and long-distance operation.
Emerging UAV capability has the potential to greatly expand and extend our ability to collect information in the challenging and remote conditions of the Arctic. To date the use of UAVs in U.S. airspace, including the Arctic onshore and offshore, has been somewhat limited as the Federal Aviation Administration (FAA) works to maintain the safety of airspace and resolve the potential for interactions between manned and unmanned air traffic. In addition to obtaining certification for the specific aircraft to be used, operators of UAVs are required to obtain a certificate of authorization (COA) from the FAA that establishes the airspace and operating parameters under which the vehicle can be operated. Generally, the airspace available to UAV operation has been limited to designated areas of controlled airspace, as in military reserves or testing ranges. COAs have not been broadly available to the private sector and have been limited to governmental entities with aviation responsibilities, including a handful of universities with established aviation research programs. In the relatively rare cases where private-sector use of a UAV has been possible, it has been through the establishment of a relationship with and sponsorship by one of the governmental bodies or universities. Acquisition of COAs over the last 5 years has required as much as 10 to 12 months from the initiation of the process. More recently, processing times have trended toward 6 months.
Despite these obstacles, the use of UAVs for data acquisition in the Arctic has been advancing. In 2008, a UAV was tested for the purpose of making observations of marine mammals in the Arctic by being launched and recovered from a vessel at sea. In 2012, small UAVs assisted an icebreaker in its effort to provide access for the delivery of fuel to the village of Nome, Alaska. In 2013, experimental UAVs were successfully launched and operated from controlled airspace near Oliktok Point, Alaska, and were tested successfully in the Chukchi Sea.
On December 30, 2013, the FAA announced an initiative to greatly increase the level of access to experimental use of UAVs.17 Though this initiative will not immediately provide access to the national airspace for commercial and civil purposes, the program
will generate data and information related to safe operation of UAVs. Six investigative entities have been selected to operate UAV test sites. These include University of Alaska, State of Nevada, New York’s Griffiss International Airport, North Dakota Department of Commerce, Texas A&M University, and Virginia Polytechnic Institute and State University (Virginia Tech). Through this program, the agency has set in motion a process that will result in the establishment of operational standards and capacities within the coming years.
Observing platforms are effective only when equipped with sensors that measure critical variables. It is particularly important to measure parameters that describe feedbacks among system components (e.g., albedo and ocean temperature).
Improvements in sensor technology would (1) increase the numbers and types of autonomous measurements, particularly biological and chemical characteristics, (2) miniaturize sensors and sensor vehicles, (3) increase data transfer capabilities from remote installations to the laboratory, (4) enable deployment of sensors that can collect high-quality data during all seasons (including winter), and (5) decrease sensor power consumption.
Examples of new sensor types and technologies that need improvement for Arctic deployment include:
- Underwater, airborne, and terrestrial still and video cameras;
- Chemical sensors for nutrients, pH, pCO2, CH4, and other dissolved gases;
- Bottom-pressure recorders for tides, storm surges, and tsunamis;
- Sensors to measure sea ice thickness;
- Sensors for identifying organisms using molecular techniques; and
- Telemetry instruments (low-power, small, inexpensive, fast).
Integrated suites of new instruments would allow sensors to be programmed for event detection, responses to seasonal changes, or alterations of data capture rates based on ecosystem processes. Integrative technologies use smart sensors that can react to external communication. A network of smart sensors could be autonomously coordinated over a wide range of platforms, for example, among fixed, ocean drifting, and autonomous underwater and unmanned aerial vehicles.
Accurate and reliable monitoring of key variables in remote locations and under harsh environmental conditions requires development of robust and inexpensive new sensor technology to provide the density of measurements needed to validate
spatially distributed models. It is important to ensure that instrumentation to be deployed for operation at remote field sites has passed a thorough pre-deployment testing process, including environmental testing, and has been developed to enable module-level serviceability and remote calibration. It may be necessary to adopt more formal approaches such as those practiced by industry and other agencies for testing and evaluation of new systems and technologies and to formalize the assessment of technological readiness of new equipment and processes. Sensors need to be easy to use and install, autonomous, and with remote data transfer to cover vast parts of the Arctic where no data currently exist. Maximizing the value of independent sensor data distributed across a wide geographic area in a range of terrains (oceans, land, coast, continental ice, and sea ice), requires robust data capture, archiving, access, visualization, and integration. Sensor data collection is an area of increasing innovation. For example, most cars and smartphones are now miniature weather stations. Most new cars have temperature sensors, and windshield-wiper speed can be a crude measure of precipitation rate (NRC, 2009). In the data-sparse Arctic, accessing data from these sources could make a large contribution, and cars and smartphones provide an example of how we need to be open to new and unusual methods of data collection. New and emerging sensor data can be fused with visual sensors data (e.g., acoustics, video imagery, photogrammetry, satellite imagery) to yield data products that can enable profoundly new insights about this rapidly changing region.
Additionally, at present there are many important components of the Arctic system that are under-measured due to logistical or technical constraints (e.g., Executive Office of the President, 2013). These include:
- Coordinated measurements of full energy and mass budgets on scales that resolve seasonality and synoptic variability, including development of new methods to measure radiation fluxes and monitor upper ocean heat and mass balance changes while integrating over spatiotemporal variability;
- Long-term observations of key outlet glaciers and tidewater glaciers;
- Monitoring of the biological and physical state of the Arctic environment in concert with quantitative measurements of human interactions with the environment;
- Assessing the effects of clouds and atmospheric constituents on surface radiation balance; and
- Quantifying the impact of terrestrial warming and permafrost thawing on the carbon cycle.
All of the technologies—existing and envisioned, mobile and fixed—for remote measurement of changes in Arctic systems require some source of energy, and power is still a limiting factor in many cases. In addition, the large quantities of data generated by these remote instruments and systems will need robust and inexpensive telemetry systems for transmission of data. Preparing for the transmission of big data is necessary as we move into the most intensive observational period the Arctic has ever seen, including high-bandwidth observations such as real-time video feeds.
There are several excellent examples of solutions to the remote power problem already in existence. For smaller power requirements, the Ch2MHill polar power website18 has been funded by NSF to be a clearinghouse for information on polar power systems in remote environments. For lower power requirements, UNAVCO has developed a small (5-W continuous power) system based on photovoltaic (PV) panels and an optional wind turbine.19 For the larger power requirements, such as for a shore-based High Frequency Radar, Statscewich and Weingartner (2012) developed a Remote Power Module (RPM), integrating PV, wind turbines, and a diesel generator, along with batteries for storage and the required control and switching circuitry. At the largest scale of operation, for example, Summit Station at the center of the GrIS or Toolik Field Station in Alaska, diesel generators are still needed to produce the necessary 80 to 170 kW.
Two key challenges remain in developing systems for future research questions:
- Developing cleaner solutions for the large-power-requirement stations.
- Distributing power from where it can be generated cleanly to where it is needed.
More robust and affordable clean energy sources and improved energy storage systems are essential to meet the data collection and transmission needs discussed elsewhere in this chapter. This is evidenced, particularly at Summit Station, Greenland. Ironically, one of the most pristine sites in the Polar Regions, a location used largely for its clean atmospheric conditions, is powered primarily by a diesel generator running
continuously. Major enhancements to the value of Summit as a facility could be realized by effectively replacing the diesel generator power with renewable, clean energy sources. It is likely that the technology for overcoming this challenge already exists, and the major impediment is cost.
Many locations in the Arctic are ideal for using renewable energy to generate power, and distributing that power is a key way to realize the benefits of such conditions. Related to the idea of power distribution hubs is the idea of using power where it is generated and moving the products of that power (perhaps manufactured goods, or energy-dense material such as hydrogen), as opposed to moving the energy itself.
Another idea is reducing the energy consumption of the instruments themselves, many of which were designed for laboratories where power is not an issue. These instruments are often now deployed in remote locations, where power consumption is one of the biggest limiting factors. Moving forward, large gains may be made by focusing effort on designing instruments to consume less power, as an alternative to developing higher-output power systems.
Broadband communication systems are vital for research activities (e.g., delivery of sensor network data and environmental monitoring) in the Arctic, are central to northern communities’ ability to adapt to climate change, and are important for monitoring and managing the expected increase in economic and industrial activity in the Arctic region. For example, it is well recognized that a robust and reliable high-bandwidth network is essential for fisheries management, weather forecasting, energy exploration and production, search and rescue, and expanding ship traffic. Broadband communication would also contribute to a paradigm shift in education and telemedicine in the Arctic region.
The coverage of geostationary satellites, which provide a robust marine communication system, is limited to approximately 70 degrees north. An example of technology that could provide communications is being proposed by Canada. The Government of Canada is currently developing a polar communications and weather mission (PCW) with international collaborations that currently includes Denmark, Finland, Norway, Sweden, and the United States (Figure 4.6). The proposed mission comprises two satellites operating in highly elliptical orbits with a weather payload (spectroradiometer), space weather instruments, and Ka- and X-band telecommunications.
FIGURE 4.6 This image illustrates the areas of interest for the polar communications and weather mission (PCW) in the Arctic. SOURCE: Canadian Space Agency.
Partially in response to the 2011 Arctic Communications Infrastructure Assessment Report,20 commercial endeavors have been proposed to install a high-bandwidth telecommunications cable from London to Tokyo through the Northwest Passage and along the Alaska coast. The proposal includes thirteen spur cables that would connect to Arctic Ocean coastal communities in Alaska and Canada.
This committee cannot and does not endorse any specific proposal, but because of the urgent need for communications in the Arctic as well as the challenges and resources involved, it would be prudent to pursue a partnership model including other Arctic nations and industry to enable the implementation of these technologies.
Computational approaches to understanding the Arctic system remain central to developing capacity in understanding mechanisms, diagnosing change, ensuring safe field operations, and improving climate change projections. In all of these aspects, the Arctic presents unique challenges. For example, large biases in simulations of the Arctic climate by global climate system models, particularly at high elevations, over ice sheets, and in the marginal sea ice zone, illustrate the fact that modeling capability in this region lags behind that in lower latitudes. Some of these challenges can be ascribed to limitations in our observational capacity. Some problems can be understood as biases originating from inadequately understood processes in lower latitudes. However, in most respects, we face a combination of sparse and noisy data with inadequate understanding of Arctic processes for the purposes of simulation (Kattsov et al., 2010). Further, the difficulties described above in maintaining robust, continuous, high-quality, distributed observations increase our reliance on models of all kinds as tools for understanding the Arctic.
At present, the capability to reproduce observed Arctic amplification and project its effects into the coming decades continues to elude us. This is manifest in the biases in integrative signals such as regional and temporal variability on a range of scales in the atmosphere, sea ice, ocean, and land (e.g., Notz et al., 2013; Stroeve et al., 2012). Specific challenges include the simulation of critical processes, including, for example, the interaction between liquid- and ice-phase microphysics (Klein et al., 2009), precipitation amount and phase (de Boer et al., 2014), glacial melt (Irvine-Fynn et al., 2014), sea ice albedo (Karlsson and Svensson, 2013) and soil freeze/thaw dynamics (Rawlins et al., 2013). These challenges present opportunities for detailed analysis of field observations in concert with targeted simulation (e.g., single-column models, cloud-resolving models, sea ice models, watershed models) that enhance our understanding of these key processes (e.g., Luo et al., 2008; Morrison et al., 2005). The benefits to climate model improvement arising from coordinated field programs (e.g., DOE Atmospheric Radiation Measurement [ARM], the Surface Heat Budget of the Arctic [SHEBA] program) that include the measurement of key parameters for simulation cannot be overstated. Atmospheric reanalyses (e.g., Dee et al., 2011; Onogi et al., 2007; Saha et al., 2010) are an important tool for a range of Arctic research activities, including applications as
diverse as detection of climate change, impacts assessment, and component model development. However, in the context of both data scarcity and model bias, the ability of data assimilation techniques to provide a resource for these activities is limited. Even the current generation of reanalysis products reveals large inter-model differences, particularly in surface meteorology, clouds, and radiation (Jakobson et al., 2012). Quality operational weather forecasts are critical for safe operations in the Arctic. Generally these models are adapted from national operational weather prediction models of Arctic nations, but research has demonstrated that these models require substantial modification to reduce bias (e.g., Bromwich et al., 2009; Schroder et al., 2011). Enhancement of the reanalysis process (including specialized Arctic regional reanalyses) and operational weather prediction will rely on the continuing improvement in understanding Arctic atmospheric processes and their interactions with other Arctic systems.
The ongoing development of limited-area climate system models in the Arctic represents a critical gap in our modeling infrastructure (Proshutinsky et al., 2008). These models allow the testing of our simulation understanding in a framework that has high spatial resolution, uses Arctic-specific physical representations, and ensures that lower-latitude biases are minimized. Although this approach enjoyed considerable advances in earlier decades (e.g., Dethloff et al., 1996; Lynch et al., 1995), development slowed until recently (e.g., Cassano et al., 2011; Dorn et al., 2009; Glisan et al., 2013). These models provide an important platform for testing approaches prior to implementation in global models, as well as providing additional infrastructure for impacts assessment, downscaling, and field campaign support.
Building the operational capacity necessary to address emerging research questions requires a mix of approaches, including partnering to leverage resources. With increased accessibility comes increased activity on the part of tourism, shipping, oil and gas, and other extractive industries. Many of these industries operate extensive investigative and infrastructure development programs. Frequently, the information needs for industry have much in common with the needs of regulatory agencies and curiosity-driven science. When industry operates in remote locations, it also tends to establish or create infrastructure to support safe operations, including housing, transportation, communications, and crisis response capabilities (e.g., search and rescue). Establishing partnerships with these organizations could allow for collection of information that would, in turn, facilitate robust decision making and extend capacities for scientific investigations in the Arctic.
There are many ways in which collaborations with industry can generate mutual benefits and synergies with the science community. At the most basic level, instrumentation of existing industry installations (i.e., ships, platforms, and facilities) operating in the Arctic can allow for collection of data. Industry is often open to allowing investigators to utilize logistical assets, provided that the investigative work is consistent with the mission of such assets and can be conducted in full compliance with industry standards. The private sector is also beginning to lead funding for scientific investigation in the Arctic (see “Investing in Research” later in this chapter). Although a portion of these funded studies is directly operated by, or on behalf of, industry, opportunities exist to co-fund investigative efforts through matching funds or the inclusion of industry in such programs as the National Ocean Partnership Program.
Industry-funded science can also be a rich source of information that could be more effectively tapped by the scientific or regulatory communities. Recognition of the utility of scientific information as a business driver is increasing the extent and quality of industry investment and willingness to participate in greater public–private-sector collaboration. Whereas industry science may be focused on specific impacts-related questions or project-specific areas, data from these studies can inform a broad array of research inquiries. Measures that increase transparency and inclusion in the planning and implementation of industry studies, the peer review and validation of results and reports, and broad sharing and utilization of industry data, all increase the value of this science both to the scientific community and to industry itself.
Examples of effective public–private collaboration on Arctic science are increasing. An excellent example of utilizing industry assets as observation platforms is the Smart Ocean Smart Industries program under the World Ocean Council (WOC). Through this program the WOC, which is an international, cross-sectoral industry leadership alliance, works with the scientific community to identify data needs and mechanisms through which these data may be collected either directly by vessel crews or through the deployment of instrumentation onto industry assets. NOAA also operates the Volunteer Observing Ship21 program for collecting a standard set of weather observations daily from more than 1,000 ships and platforms globally for incorporation in weather forecasting models.
A 2010 agreement on data sharing between three international oil companies (Shell, ConocoPhillips, and Statoil) and NOAA has made the results of a nearly $100 million investment in data on the U.S. Arctic offshore available to the agency and, more broadly, to the scientific community. Under this agreement, data from meteorology/oceanog-
raphy observing buoys are served directly to the National Data Buoy Center and are utilized to improve forecasting in the Arctic. Data from integrated ecological studies and monitoring programs are made available through the Alaska Ocean Observing System.22
Investigators frequently establish ad hoc public-private collaborations by soliciting matching funds, or by combining privately funded opportunities with publicly funded initiatives. Such informal pooling of funding can increase the scope and utility of publicly funded projects by accommodating the utilization of a larger, more capable vessel or adding scientists to the program. Formal public-private collaborations are becoming more common as both communities find new strategies for co-planning investigative efforts and for co-funding research.
An essential element of ensuring that the nation has sufficient research capacity is an adequate supply of people with a unique combination of the necessary skills and knowledge. Arctic questions span many disciplines across the natural and social sciences and thus require some researchers who work at the intersections, crossing and connecting fields and collaborating across international boundaries. Also, research capacity in the Arctic is particularly important because climate change and its impacts are occurring at an accelerated rate. Thus, our capacity to observe and conduct research to understand the observations, and develop appropriate response strategies, needs to keep pace. Building human research capacity includes both training of the next generation and engagement and professional development of the existing community so that we are better prepared to address current and future challenges.
Human research capacity building was a major component of the International Polar Year (IPY). The National Academy of Sciences study on Lessons and Legacies of the International Polar Year (IPY) 2007-2008 showed that there were measurable increases in the number of scientists conducting polar research (NRC, 2012a). This increase was attributed not only to the climate-change-driven need for more polar researchers but also to IPY’s efforts that enabled international research teams to closely coordinate their activities. Two specific human-capacity-building activities deemed successful during IPY were the Association of Polar Early Career Scientists (APECS) and the growth in student participation in the University of the Arctic.
The APECS coordination office is currently funded by three Norwegian organizations.
Other organizations that work with APECS, formed to support early career scientists in specific disciplines, include:
- Permafrost Young Researcher Network (PYRN)
- Young Earth Scientists (YES) Network
- ArcticNet Student Association (ASA)
- Young European Associated Researchers (YEAR)
- Young Earth System Scientists (YESS)
- World Association of Young Scientists (WAYS)
- European Geography Association for students and young geographers (EGEA)
Increased support and funding agency incentives for U.S. young scientists to engage in APECS’s activities would contribute to growing Arctic research capacity.
The University of the Arctic has a range of programs distributed among and coordinated with member higher education institutions that enable building of Arctic human research capacity with important emphasis on the recruitment and involvement of Arctic peoples. As of 2013, the United States had the lowest student involvement in their northern engagement program. Supporting U.S. students (including recruits from northern communities) in the University of the Arctic has the potential to increase human capacity through their established and well-recognized programs. Another key aspect of human capacity building is training young scientists, particularly social scientists, in the linguistic and cultural competency skills for working across the Arctic. Training centers in other parts of the Arctic could serve as models for North America.
Other IPY human-capacity-building successes were related to funding agency incentives for researchers to incorporate northern community engagement in research and as public outreach. Some of these success stories included expansion from academic outreach to include informal education venues (e.g., museums, science fairs, online broadcasts). Continuing funding agency mechanisms that encourage these activities would provide young Arctic residents an opportunity to see research career opportunities directly linked to the future of their own communities.
Arctic residents have played important roles in research for over a century, and their involvement continues to increase. From providing logistical support and safety in the field, to offering insights from generations of observations and experience, Arctic peoples have a great deal to offer. They also have a great deal to gain from sound sci-
entific research, which can address many challenges of rapid environmental and social change in the region. Effective research partnerships have led to major advances in marine mammalogy (e.g., Noongwook et al., 2007; Thewissen et al., 2011) and meteorology (e.g., Weatherhead et al., 2010), the emergence of traditional knowledge as an important topic of study (e.g., Huntington, 2011), and an increase in the number of scientists and scholars who come from Arctic communities. Arctic researchers, similarly, are increasingly interested in making connections with Arctic residents to incorporate traditional knowledge and observations and also to share the results of their work (Figure 4.7).
These trends are encouraging, and yet the Arctic research community has only begun to tap the potential for involving Arctic residents as well as citizen science practitioners who do not live in the Arctic but are still interested in Arctic topics. Arctic residents are alone in observing their environment throughout the entire year, year after year. They all have a lifetime of knowledge from their own observations as well
FIGURE 4.7 Warren Matumeak (left) and Andy Mahoney (right) discussing sea ice conditions near Barrow while examining a satellite image. SOURCE: Henry Huntington.
as what has been passed down by older relatives, a chain that extends back countless generations in indigenous communities. Few of these contemporary and traditional observations and insights are recorded or made available to others, leaving many potential connections unrealized. The power of entraining large numbers of people in addressing research questions or data analyses (e.g., crowd sourcing) has yet to be applied to Arctic research to any substantial degree. There are promising developments in all these areas (e.g., Alessa et al., 2013), but the wider application of successful approaches has not yet occurred.
Three areas are particularly ripe for further attention to increase meaningful engagement of Arctic communities. First, communities themselves need to determine how they want to be engaged. The research burden on Arctic residents can be high, for example, as in being interviewed again and again in the course of different studies with similar objectives. The return of scientific information back to the communities is not always effective. And communities are not always involved in all phases of research, reducing the value of their participation as well as their ownership and/or partnership. At the same time, few individual research projects have the resources to address all aspects of community interest and opportunity, creating a need for other mechanisms to support community engagement on the community’s own terms.
Second, the infrastructure to support community engagement is only now being developed on a larger scale than that of individual projects or, in a few cases, regions of the Arctic. Such infrastructure includes data management, to capture and make available the results of community efforts, as well as communication procedures that can help researchers connect with communities as they plan, conduct, and disseminate the results of their research. Ad hoc approaches have worked for some projects and individuals, but many opportunities have also been missed, especially for building beyond the activities of a single project. The same principle applies to enhancing the capacity of communities to engage in Arctic research. Various Alaska Native organizations have played important roles in this regard, but greater continuity of effort and connections among projects and practitioners can yield even better results.
Third, there has simply been too little experience to date with the various approaches that have been and can be used, limiting the utility of an evaluation of what works and what doesn’t. More needs to be done, engaging more communities on more topics, to build up a better body of practice and experience, from which relevant lessons can be drawn. More experience will also help community aspirations and capacity grow and mature, likely creating greater demand for community engagement along with a greater sophistication in how to make use of research activities and results.
Research requires funding. Funding involves making decisions, which includes considering what is needed, what is likely to work, and what trade-offs are entailed. Most Arctic research funding in the United States comes from government agencies, ranging from studies intended to address the needs of regulatory and other decisions, to curiosity-driven research within broad areas of scientific interest. Additional research, typically addressing specific needs or goals, is funded by the private sector, including industry as well as philanthropic groups. Decisions about what is funded therefore occur at many levels in many places. Nonetheless, some general patterns are evident, and society’s ability to address emerging research questions in the Arctic is closely tied to the way research funding is organized.
Evaluating the strengths and drawbacks of current funding mechanisms for Arctic science in the United States is beyond the scope of this report. Instead, we draw attention to certain features of research funding and suggest a closer look at whether the current approach is optimal for addressing society’s needs. We focus our discussion in six areas: comprehensive systems and synthesis research, funding non-steady-state research, social sciences and human capacity, stakeholder-initiated research, international funding cooperation, and long-term observations. We consider cooperation among countries, among agencies, across disciplines, and with the private sector.
Research is often proposed in response to a request for proposals and then carried out over a 3- to 5-year time frame. Successful research may lead to subsequent projects that build on the results from the initial project, but there is no guarantee of further funding. Most projects are proposed and run independently, only rarely with support for coordination with related initiatives. This system provides flexibility, in that funding streams are committed for a relatively short period and in that researchers have the ability to pursue topics they deem important and, often, to adjust their research as circumstances and preliminary findings warrant. At the same time, implementation of full programs and deep engagement with and the ability to explore the wider connections or ramifications of a particular topic are often limited within a 5-year project. Similarly, the ability to coordinate and cooperate across projects may be curtailed by time as well as by the demands of producing individual project results and then the competitive aspects of seeking further funding.
These drawbacks are especially apparent when trying to grapple with a comprehensive view of the Arctic, encompassing its myriad components, each with its own complexity. The challenges of “systems” research and interdisciplinary collaborations are well known. How those challenges can be overcome is less apparent, but continuity, coordination, and leadership are likely to play major roles. Other funding approaches are used in other countries, and some innovative approaches have been tried in the United States in recent years. For example, long-term projects under the leadership of scientists with strong records of accomplishment and collaboration have been funded elsewhere. The part of the Bering Sea Project (Wiese et al., 2012) that was funded by the North Pacific Research Board was organized as a single project with one principal investigator, rather than as a collection of individual projects, in order to emphasize interdisciplinary collaboration and a high degree of integration of ecosystem understanding. Integrated and cross-disciplinary proposals could also be developed through the National Science Foundation’s new option for program managers to handle proposals through an “Ideas Lab” model.23 A request for participation in the Ideas Lab is announced. Interested participants are invited to submit an application that outlines their ideas on a specific Ideas Lab topic. Selected participants will attend an interactive, multi-day program of collaborative discussion to construct new ideas and approaches. Subsets of teams will then submit full integrated proposals. Another way to integrate projects is to announce at the outset that the intent is to support a balanced suite and also support a coordinating office, as NSF did with the Climate Change Education Partnership program.
Synthesis activities, similarly, are often challenging in that they lack the allure of new field research. In some cases, the rationale for investing in synthesis is not readily articulated before the synthesis activity has started but only emerges from the interactions of those involved and the interpretation of the various streams of data and insight that are to be connected in the course of the synthesis. Some examples exist, such as efforts under the Outer Continental Shelf Environmental Assessment Program in the 1970s and 1980s, synthesis workshops undertaken by NSF’s Arctic System Science Program (e.g., Overpeck et al., 2005), the NSF’s and the North Pacific Research Board’s Bering Sea Project (Wiese et al., 2012), NSF’s Arctic Freshwater Integration project,24 and recent efforts for U.S. Arctic waters (e.g., the Pacific Marine Arctic Regional Synthesis [PacMARS] and the Synthesis of Arctic Research [SOAR] programs), but these are the exceptions rather than the norm.
Because of the funding structures and norms, there is currently an imbalance, with
most research initiated by individuals and small groups, and few resources directed toward larger-scale synthetic thinking and study. Other countries have different ways of handling synthesis research, including making large-scale and longer-term investments. Some invest in training of reviewers, so that they are better able to handle interdisciplinary and integrative proposals. The extent to which various approaches work and the trade-offs that they entail (e.g., opportunities for young researchers vs. continuation for established researchers) require careful evaluation to determine whether they do in fact produce a better comprehensive understanding of the research area in question, and at what cost. If so, then new funding approaches could be considered by U.S. agencies in light of their specific missions for Arctic research, to ensure the maximum benefit for society from its investment.
Understanding an Arctic in transition may require greater risk on the part of funding agencies and a greater acceptance of uncertainty on the part of reviewers to make headway against an uncertain future. Funding non-steady-state research will be necessary to better understand the dynamics of thresholds, resilience, and transformation in a rapidly changing Arctic. Obtaining funding for research into steady-state processes can sometimes be more straightforward than funding non-steady-state research, as steady-state proposals can provide convincing evidence of feasibility. However, given the potential for nonlinear change, tipping points, and emergent properties, it is important to ensure that investigations of emerging, non-steady-state research questions are funded as well, even if that means greater willingness on behalf of the funding agencies to take risks. Alternative approaches to proposal review and decision making could be utilized, along with locally inspired social-ecological experiments.
In titling this report The Arctic in the Anthropocene, the committee intended to draw attention to the central role of humans in the emerging research questions. There are pressing needs for social science research as identified in Chapter 3 and for recognition of the role people play in research infrastructure, discussed earlier in this chapter.
Support for the social sciences, including economic, behavioral, and decision research, has lagged behind that of the natural sciences. As we attempt to prepare ourselves, our communities, and our country for a more rapidly changing future (IPCC, 2014),
investments in social science are more critical than ever. Many of the questions we have identified in this report have at least some connection with the social sciences (Figure 3.18b).
In addition to conducting the research, ultimately it is people who are central in enhancing cooperation and coordination, sustaining long-term observations, managing and sharing information, building and maintaining operational capacity, and providing the capacity to meet the challenges. The committee heard from many in the community who had stepped in to fill gaps but were not supported in doing so and were stretched thin in responding to multiple demands forced by the rapid pace of change. To do this, people have to be engaged, trained, retrained, and supported so that we have the requisite expertise, provide for follow-through in research infrastructure, operations, and administration, and can rapidly respond to new ideas and fresh perspectives.
Critical questions are emerging from stakeholders, including decision makers and communities, that are not traditional participants in federal research (things we think we don’t know). There is not currently a consensus within the research community that this type of research is important, so it is less likely to rise to the top during proposal reviews and funding decisions—what we know we need to know will often take precedence over what we think we don’t know.
An evaluation of how current funding mechanisms affect the ability of nontraditional research organizations to participate in Arctic research is needed (see also the “Intersectoral” subsection under “Enhancing Cooperation” and the “Growing Human Capacity” section earlier in this chapter). Approaches used by other agencies, regions, and countries are worth considering applying to the Arctic.
A major barrier to international collaboration is the nature of the present framework for funding basic research. International collaborations can by stymied by failure to obtain funding approval from agencies in more than one country. Most nations have a national funding organization that is constrained by unique rules and guidelines that rarely accommodate multinational proposals. This somewhat arbitrary limitation impedes true international collaboration. Peer review of proposals also lacks consistent guidelines internationally, and proposal target dates are not synchronized. There are
few official channels (e.g., Belmont Forum25) through which program managers can communicate internationally to set common research goals. Removing these barriers to efficient international collaboration requires long-term, sustained commitments from national funding agencies, as well as the development of policies that serve the interests of both national funding agencies and the scientific community. An Arctic activity is forthcoming from the Belmont Forum, which is a welcome first step, but a long-term sustained program supporting international collaboration would yield many additional benefits.
Global leaders are beginning to recognize the importance of cooperation in the Arctic. For example, in August 2013, the Russian news agency ITAR-TASS reported that:
“Japan believes there is a strong need to conduct continuous monitoring and research in the Arctic, in particular, in connection with global climate change,” Hakubun Shimomura [minister of education, culture, sports, science and technology] continued. “In view of the fact that Russia is a country to which the largest territory in the Arctic belongs, we consider cooperation with it as absolutely necessary. In particular, we need to work together in the sphere of creating monitoring stations in the Arctic, the use of the icebreaker fleet, exchange of experts and the general expansion of research in this sphere.” The minister said that a regular meeting of the Japanese-Russian Joint Commission on Scientific and Technological Cooperation will be held in Tokyo this September. “It will exactly discuss further prospects for the development of interaction and cooperation between the two countries in this part of the world. …We plan to put forward a concrete proposal on Arctic research cooperation, in particular, with regard to cooperation in the sphere of observation and personnel exchange,” said the minister.
Change can be detected only by observations over time. The precision by which change can be measured depends on the consistency, frequency, and breadth of those observations. At present, there are relatively few consistent, frequent, spatially extensive datasets for the Arctic. Instead, we have a smattering of ad hoc stations, incomplete time series, and varying methods. The “Undetermined Arctic” section in
25 The Belmont Forum was established to overcome some funding challenges by advancing international collaboration in research through joint announcement of targeted programs: “(1) strengthening engagement between the research funding agencies and the academic research community as represented by ICSU and (2) improving coordination of early phase engagement on GCR strategies and priorities in order to improve co-design, co-alignment, and co-funding of major research programs (http://www.igfagcr.org/index.php/challenge).” “The Forum requires each Collaborative Research Action to address the Belmont Challenge:To deliver knowledge needed for action to avoid and adapt to detrimental environmental change including extreme hazardous events. Belmont further requires consideration of human and natural systems in each proposal, and a minimum of three nations involved in each project (http://www.climate-cryosphere.org/news/clic-news/521-update-on-international-research-funding-from-the-belmont-forum).”
Chapter 3 addressed the rationale for better long-term observations. Here we address the implications for funding.
Consistent, system-wide observations over time require sustained support. Long-term funding commitments, however, are rare. Furthermore, the payoff from long-term observations is typically time-delayed, making it easy to justify spending money on relatively short-term research efforts that produce results in a few years rather than over the course of decades. The result on the funding side is a patchwork of efforts that have little coordination and thus exhibit little synergy, in that the monitoring of one component in one location readily lends itself neither to detecting the connections between that component and other parts of the system nor to evaluating the relationship among trends observed in different locations. Complicating matters in the Arctic is the fact that processes interconnect across national borders, requiring cooperative, long-term international observations.
One alternative is the development of a coordinated program of long-term observations, designed not from individual interest or based on what proposal happened to get funding but, rather, from a vision of understanding the system as a whole, and with a sustained commitment to funding. Such an approach is the idea behind the international Sustaining Arctic Observing Networks (SAON) initiative and other efforts such as the Circumpolar Biodiversity Monitoring Program. Though meritorious, these efforts are still largely a collection of ad hoc efforts, with funding dependent on those responsible for each separate component of the overall network.
Our ability to detect change and to determine what new features of the Arctic system are emerging is thus compromised and will remain so until there is a lasting commitment to long-term observations. Because agency interests will always be focused on specific missions or mandates, we need to explore how to put in place a network backbone that provides continuity as well as disciplinary and regional breadth. This network would serve to explore promising scientific approaches and generate new findings while keeping track of key variables and indicators of change. Other activities, such as more focused agency programs, would benefit because they could plug into this network.