Critical Infrastructure for Ocean Research and Societal Needs in 2030 (2011)

One of the committee’s primary tasks was to “identify major research questions anticipated to be at the forefront of ocean science in 2030 based on national and international assessments, input from the worldwide scientific community, and ongoing research planning activities” (see Box 1.1). In response to this charge, a range of recent government plans, task force documents, research planning assessments, disciplinary reports, and primary literature (e.g., NSF, 2001; USCOP, 2004; JSOST, 2007; CEQ, 2010) were reviewed by the committee. From these documents, and from information gathering sessions with experts in ocean science and policy, the committee identified 32 compelling science questions that are anticipated to be at the forefront of ocean science in 2030, ranging from broad global challenges that require both interdisciplinary and multidisciplinary research to regional, local, or discipline-specific topics. These questions are clearly relevant for 2010 but are not simple issues that will result in solutions with a few more years of intensive effort. Instead, they reflect challenging scientific problems that will likely take decades to solve, especially if only limited resources are available.

The act of defining research questions that will still be relevant in 2030 has many challenges. Almost certainly, new discoveries and technological advances will alter the research landscape, redefining or even providing answers for some questions. It is nearly impossible to anticipate the nature of such transformational discoveries and even more difficult to pose questions that anticipate their impacts. Instead, the committee (guided by the planning assessments cited above) focused on questions that are likely not only to still be relevant, but even more pressing in 2030. For example, nearly 20 years ago Policy Implications of Greenhouse Warming (NRC, 1992) posed a series of research issues associated with geoengineering schemes as potential avenues to mitigate climate change. The past few years have seen numerous workshops and reports devoted to developing geoengineering research agendas as a response to climate change (e.g., IPCC, 2005; The Royal Society, 2009). The science has certainly advanced over the two decades between these reports, but compelling science questions on the viability and impacts of these options remain.

Although such a list of questions can never be exhaustive, the committee feels these are comprehensive enough to capture the major infrastructure needs for 2030. As discussed in Chapter 1, these questions are organized within the context of four overarching societal drivers: enabling stewardship of the environment, protecting life and property, promoting sustainable economic vitality, and increasing fundamental scientific understanding. These drivers are similar to critical themes identified in Charting the Course of Ocean Science in the United States for the Next Decade: An Ocean Research Priorities Plan and Implementation Strategy (listed in Chapter 1). This chapter also aligns with several priority objectives of the National Ocean Policy (CEQ, 2010; E.O. 13547), discussed in greater detail within specific science questions.

In the next 20 years, significant anthropogenic environmental impacts are very likely, given the magnitude of the growing world population¹ (De Souza et al., 2003; Rockstrom et al., 2009). However, increased understanding of the ocean’s physical, chemical, and biological responses, particularly in the context of anthropogenic forcing factors (e.g., climate change, resource extraction and utilization, waste production and nutrient pollution), has potential to limit many adverse impacts.

Human activities, from fishing to energy extraction, are having impacts on all regions of the ocean, from estuaries to the deep ocean. However, perhaps the most significant and striking impacts are found in coastal and polar regions. The coastal zone, an area vulnerable to multiple stressors, is of particular societal and environmental significance. Although

it comprises only about 8 percent of Earth’s surface, this area supports more than 25 percent of total global primary production and yields nearly 90 percent of present world fisheries production (Ryther, 1969; Sherman, 1994). Oceanrelated activities and industries provided over 2.3 million jobs in 2004.² About 35 percent of the world’s population currently lives within 100 km of a shoreline (Nicholls and Small, 2002); this number is projected to grow to 75 percent in a few decades (Vitousek et al., 1997). Over two-thirds of the world’s largest cities, with populations greater than 1.6 million, are located in coastal areas. These are often in the vicinity of estuaries or coastal wetlands, accounting for more than 50 percent of wetland loss (Walker, 1990; Anderson and Magleby, 1997). Coastal governance issues (e.g., coordination and support of ocean and coastal management; coastal and marine spatial planning³) are currently at the forefront of both public attention and national priorities (CEQ, 2010); and this is not expected to decrease by 2030.

The polar regions will almost certainly also be of profound importance in the next 20 years, as noted by inclusion in the National Ocean Policy (NOP) objectives (CEQ, 2010; E.O.13547). Although they do not have significant populations in numbers, they are presently subjected to rapid environmental changes (e.g., warming, sea ice reduction, changes in freshwater fluxes) that may have great impacts for commercial activity, including resource extraction and transportation. These also require special considerations when discussing ocean infrastructure needs. The following 13 questions were chosen to encompass a broad range of issues regarding environmental stewardship from the poles to the equator.

The trapping of heat by anthropogenic greenhouse gases is likely to lead to sea level rise on a wide range of spatial and temporal scales (NRC, 2010b). As so many people live and work near sea level, sea level study and prediction will continue to be a topic of active research in the coming decades. In 2007, the Intergovernmental Panel on Climate Change (IPCC) estimated sea level rises between 0.18 and 0.6 m by 2100 (IPCC, 2007). More recent estimates that take into account ice melt on Greenland and western Antarctica increase these estimates to between 0.8 and 2.0 m (Pfeffer et al., 2008). Increased heat in the ocean-atmosphere system causes sea level rise in two ways: (1) a warmer ocean is less dense, and thus has more volume even if its mass remains constant; (2) melting of ice on land adds mass to the ocean, raising sea level (Nicholls and Cazenave, 2010). Even if these fundamental effects were perfectly understood and predicted, there would still be issues related to regional sea level rise that depend strongly on local conditions (Milne et al., 2009), including subsidence, tides, and storm activity. Tides and storms contribute to local inundation, so the most damaging effects of a higher sea level will likely be felt more frequently. Seasonal effects could be significant, as runoff contributes to flooding in areas of high precipitation. For low-lying coastal communities, sea level rise will be a threat to societal infrastructure (e.g., streets, buildings, sewage, drinking water supplies, gas, electricity [Nicholls and Cazenave, 2010]). Ports and naval facilities, in particular, will need to address the impact of sea level rise and changing dynamics of coastal erosion and sedimentation in order to maintain effective operations. Also of concern are more than 200 existing marine laboratories that currently provide support for a wide range of ocean research and education activities (Sebens, 2009), which will have to adapt to coastline changes as a result of rising sea level. On regional and global scales, ocean temperature and therefore sea level will continue to change in response to natural, interannual modes of climate variability such as the El Niño-Southern Oscillation (ENSO), and many of these changes will be irreversible over both short and long time scales (Solomon et al., 2009).

Major changes have and will continue to take place in the world’s ocean (e.g., changes in temperature, stratification, circulation, oxygen distributions, trace metals inputs, and pH) (e.g., Sarmiento et al., 2004; Doney et al., 2009; Reid et al., 2009; Keeling et al., 2010; Steinacher et al., 2010). These changes all have direct and indirect impacts on ecosystem processes, including limitation of primary production by nutrients, shifts in the major phytoplankton groups that dominate open ocean waters, and changes in zooplankton behavior and distributions (Reid et al., 2009). Global trends in primary productivity have been linked to changes in surface temperature and mixed layer dynamics (Behrenfeld et al., 2006; Martinez et al., 2009; Chavez et al., 2011). While some of the basin-scale trends are correlated with natural oscillatory cycles (e.g., the North Atlantic Oscillation, Pacific Decadal Oscillation), the exact mechanisms that force changes in ecosystem productivity are still uncertain (Martinez et al., 2009). Indeed, a recent study concludes that long (~40 years) records of persistent, high-quality, global-scale data are needed to separate decadal oscillations from climate effects on ocean productivity (Henson et al., 2009; Chavez et al., 2011).

Modulation of the surface ocean ecosystem’s composition, stock, and productivity influences the biological pump that functions to transport atmospheric carbon dioxide (CO₂) incorporated into organic carbon into the deep ocean

(Sarmiento and Gruber, 2006). However, the link between surface productivity, fluxes to depth, and the rate at which this material degrades in the ocean interior is currently not well understood and quantified (Burd et al., 2010). The challenge of understanding the ocean’s role in the global carbon cycle and its response to a changing environment requires an expanded scale of observation in both space and time (K.S. Johnson et al., 2009; Chavez et al., 2011). Global-scale observations of phytoplankton stock, functional group distributions, and productivity are currently constrained by, and limited to, remotely sensed ocean color, which senses only the near-surface conditions of the ocean. New observational strategies are needed to study and understand the link between phytoplankton productivity, carbon export, and climate forcing.

Interactions between climatic forcing and anthropogenic changes in greenhouse gas concentrations affect global ocean circulation, which in turn influences global climate (e.g., Broecker, 1997; Clark et al., 2002; Sutton and Hodsen, 2005). These interactions will have an impact on ecosystem dynamics. Changes in species composition, species distribution, or trophic interactions, which can be caused by shifts in the geographic range of ecosystem components, may result in alterations of ecosystem resilience and productivity (Pereira et al., 2010). The degree of genetic connectivity and species-specific life history characteristics mediate the resiliency of populations and communities and the ability to recover from both human and natural sources of disturbance. Studies of the mechanisms of genetic connectivity (both passive transport of gametes or early life stages and active movements of older individuals) are needed to identify the space and time scales of biological and physical processes that link populations and communities, and to identify factors that enhance or limit gene flow and dispersal.

Community response to disturbance is also determined by patterns of species interactions. For example, disturbance of corals or other habitat-forming organisms may have a larger impact on the community than a similar magnitude of disturbance to other taxa (Sebens, 1994). Similarly, the removal of important predators in some ecosystems has been shown to significantly alter abundances in different trophic levels (e.g., their prey, their prey’s prey, other predators of their prey [Wootten, 1994; Estes and Duggins, 1995]). Ecosystem-based management approaches, such as that advocated by the NOP (CEQ, 2010), are presently being developed in part to address these issues.

Disturbances to species composition and distribution include invasive species, which can displace native species, change community structure and food webs, alter fundamental processes such as nutrient cycling and sedimentation, and are a major threat to marine biodiversity (Carlton and Geller, 1993; Molnar et al., 2008). Invasive species have transformed marine habitats around the world, caused human disease, and led to significant ecological and economic damage (Takahashi et al., 2008). Many marine species have been transported to nonnative areas by ship ballast water or hulls. By 2030, it is predicted that commercial shipping will be able to exploit seasonal ice-free Arctic shipping routes (e.g., Wilson et al., 2004; Stroeve et al., 2008); this may exacerbate the movement of invasive species and have other impacts (e.g., vessel whale strikes). The foundations for a quantitative global assessment of the impacts of invaders and their routes of introduction will likely be in place by 2030, but additional information will be needed to develop large-scale strategies necessary to prevent future introductions while adapting to existing invaders.

The combination of large-scale biogeographical shifts, changes in local community structure caused by ocean warming and acidification, and impacts from invasive species will have far-reaching consequences for marine biodiversity, ecosystem structure, and population dynamics. Yet many of the current changes and their impacts remain unreported, for lack of comprehensive global marine ecosystem monitoring efforts. In order to provide effective stewardship of the marine environment, infrastructure that can quantify ecosystem changes and manage human activities in response is a need for 2030.

Marine biogeochemistry and ecosystems are likely to be affected by the chemical changes related to increasing dissolved CO₂ in the ocean, as well as the attendant ocean acidification (Feely et al., 2004; Fabry et al., 2008; NRC, 2010d). Lower carbonate saturation states are apt to lead to less calcification, diminishing alkalinity removal from the surface ocean into the deep ocean. Over thousands of years, lower carbonate saturation will lessen the sedimentation of calcium carbonate (CaCO₃), altering the carbonate compensation depth (where dissolution equals supply) and lysocline depth (where seafloor carbonate dissolution begins and accelerates as a function of increasing depth). The response of biological productivity to the diverse factors affected by ocean acidification is likely to alter the global ocean nutrient distribution.

Phytoplankton may respond directly to increased dissolved CO₂ through faster carbon uptake when other factors are not limiting (Riebesell, 2004). Many phytoplankton and zooplankton species are sensitive to other chemical changes associated with decreasing pH (e.g., trace metal speciation changes, which affects the bioavailability of essential metals such as iron or zinc, and the toxicity of other elements such as copper and arsenic; Shi et al., 2010). However, understanding how these complex ecosystems respond to ocean acidifica-

tion is extremely limited. Laboratory experiments and field observations suggest that calcifying organisms and communities (e.g., planktonic foraminifera, coral reefs, and oyster reefs) can be affected by present ocean acidification levels and will be strongly disturbed by doubled atmospheric CO₂ (e.g., Anthony et al., 2008; De’ath et al., 2009). The impact of these disturbances on community food webs, however, is unknown. Of particular concern is the ability of corals to respond to increased ocean acidity, because they form habitat for many ecologically and commercially important species (Hoegh-Guldberg et al., 2007). There are also direct chemical responses to ocean acidification. Decreased pH would affect both organic and inorganic chemical speciation of trace metals that form strong oxyhydroxide complexes such as iron, aluminum, and thorium, and alter the kinetics of reactions. Extreme shifts in pH (comparable to that expected in the 22nd century [Caldeira and Wickett, 2003]) could affect the stable redox state by altering the uptake ratios of elements and their subsequent recycling from biological debris. However, it has also been suggested that, with the exception of calcification, other major biogeochemical cycles will not be affected by ocean acidification (Joint et al., 2011).

Climate change is likely to influence the distribution of chemical elements through ocean circulation and temperature, biogeochemical responses to the physical climate, and alterations in weathering and transport of key nutrients. A warmer climate will tend to stabilize upper ocean stratification, diminish vertical mixing, and reduce the upward flux of nutrients and productivity (e.g., Reid et al., 2009; Sarmento et al., 2010). However, an altered climate is also likely to affect wind patterns and hence the positions and strengths of currents, upwelling zones, and the timing of seasonal transitions; these changes are more difficult to predict without very high resolution coupled ocean-atmosphere models and data to force and constrain them. All other things staying constant, warmer surface water will contain less oxygen, leading to lower oxygen at depth. Reduced oxygen will lead to the expansion of denitrification zones and a long-term (thousands of years) reduction in the oceanic nitrate inventory, although this could be offset by high anthropogenic fixed nitrogen emissions (Keeling et al., 2010). Changes in winds and continental climate could alter the flux of dust and atmospheric aerosols into the ocean, influencing the distribution of high-nutrient, low-chlorophyll regions (Jickells et al., 2005). Additionally, climate-induced changes in temperature, salinity, and pH will affect mineral solubility (e.g., CaCO₃) and trace element speciation (Reid et al., 2009). The effect of climate change and anthropogenic emissions in continental settings will alter weathering and transport by rivers, with potentially large consequences for the coastal ocean, and in the longer term (many thousands of years), for the entire ocean.

Carbon has a vital role for supporting all life on Earth. Dissolved organic carbon (DOC) in the ocean is one of the largest pools of fixed carbon on the planet, approximately equal to the amount of CO₂ in the atmosphere (Hedges, 2002). Fluxes of organic carbon may be expected to change markedly in a warmer climate (Riebesell et al., 2009). This is of significance because fixed carbon can be converted into sugars during photosynthesis and is then usable by heterotrophs. The amount of carbon residing in this pool is thought to have changed by two to three orders of magnitude over geologic times scales (Rothman et al., 2003). In the modern ocean, DOC exported to the ocean interior contributes about 20 percent of the global ocean biological pump (Hansell and Carlson, 2002). This export occurs largely by overturning circulation, which is likely to be altered because of future changes in ocean stratification; the DOC fields and export in a more stratified ocean will be considerably different than what is observed today. Much of the DOC in the ocean has resisted qualitative and quantitative analysis, as the microbial processes that control its composition and abundance are enigmatic (Hedges et al., 2000). Turnover in some components is extremely fast, while much of the material has an apparent ¹⁴C age of thousands of years (Druffel et al., 1992). Thus, models of global carbon cycling are limited by knowledge of the time scales for DOC cycling in the ocean. The processes that regulate interactions of this material with the microbial ecosystem are just beginning to be understood. Emerging new analytical tools provide scientists with the capability to directly probe composition and rates of change of a broad spectrum of components in the DOC pool (Mopper et al., 2007). These capabilities can be linked with environmental genomics and studies of protein structure and expression to greatly expand the understanding and predictive capabilities regarding the vast pool of DOC in a changing climate (Kujawinski, 2011).

The ocean’s capacity to transport, store, and exchange huge amounts of heat with the atmosphere has a profound effect on the climate system—both natural and anthropogenic. Natural climate variability orchestrates large changes in weather and climate over much of the globe on interannual and longer time scales (Joyce, 2001; Visbeck et al., 2001; Trenberth et al., 2002; Kerr, 2005). One such example is ENSO, a recurring change in the distribution of heat on the equator that involves weakened upwelling in the eastern Pacific and attendant warming (Philander, 1990). Impacts of ENSO are felt in fisheries off Peru, western U.S. coastal

waters, precipitation across North America, hurricanes striking the southeastern United States, and sometimes in globalscale atmospheric conditions (McPhaden, 1999). Basin-scale changes in sea surface temperature (SST) of the subtropical North Pacific have a dominant mode (the Pacific Decadal Oscillation), with known effects on precipitation (Davis, 1976; Mantua et al., 1997). Ocean warming due to anthropogenic climate change involves both trapping of heat by greenhouse gases and its redistribution. A complete knowledge of the ocean’s energy balance, as well as the redistribution of heat by ocean currents, is fundamental to understanding the climate system’s response to natural and anthropogenic drivers. The ocean’s boundary currents, especially those on the western sides of basins, are key to poleward heat transport (Bryden and Imawaki, 2001). In turn, it is expected that the increased heat and freshwater added to the ocean will affect the stratification, currents, and ocean conveyor belt. Future research surrounding this question is likely to focus on sustained observations, analysis of changes as they occur, and improved modeling for prediction.

Alteration in the global water cycle is a crucial issue for civilization. The ocean is the main reservoir of free water on the planet, containing 97 percent of Earth’s water (Baumgartner and Reichel, 1975). It accounts for 86 percent of global evaporation and 78 percent of global precipitation (Schmitt, 1995, 2008) and is central to regulating the water cycle. Because the vapor pressure of water is an exponentially increasing function of temperature, alterations in the water cycle can be expected and have already been documented with climate change (e.g., Curry et al., 2003; Boyer et al., 2006; Yu, 2007). Water evaporates more readily from a warmer ocean, so an intensification of the water cycle and changes in the distribution of salinity are expected with anthropogenic warming. Cloud-climate feedbacks, which will remain a major research challenge, are an important element of understanding changes to the global water cycle. Freshening of the high-latitude ocean through increasing input of freshwater from melting will increase ocean stratification (e.g., Schmitt, 2008), suppressing mixing and greatly affecting nutrient supplies and ocean ecosystems. Increased stratification could also slow down the ocean conveyor belt, which will affect the large-scale flux of freshwater, heat, and carbon dioxide in the ocean (e.g., Yashayaev and Clarke, 2008). Ocean salinity feeds back on the circulation and mixing (Schmitt, 2008) and thus has influence on ecosystems and future climate states. In addition, distributions of SST are good predictors of rainfall on land (Schmitt, 2008). Large changes in drought and flood patterns will affect both ecosystems and societal infrastructure.

Changes in coastal boundaries include both gradual and abrupt alterations of the shoreline, wetlands, and seafloor. These can be natural changes such as erosion or deposition, subsidence, faulting, and storm or tidal surges, or they can be consequences of human activities. Anthropogenic changes to coastal boundaries include creation of artificial boundaries (e.g., breakwaters, jetties), modifications to wetlands and rivers (e.g., infilling, channelization, subsidence due to oil and gas activity, damming and reduced sediment supply), and potential impacts from climate change (e.g., sea level rise and loss of coastline; e.g., Nicholls and Cazenave, 2010). Physical and geochemical fluxes across the coastal boundaries include, but are not limited to, significant air-sea interactions, riverine and groundwater inputs to the ocean, and saltwater intrusion to the coastal zone. These processes occur at a wide range of scales, from the submeter scale to many kilometers. Included in this range is the submeso-scale, where variability is spatially intermittent with highly energetic regions depending on proximity of varying water masses and currents. Understanding physical processes at the submesoscale promises improved prediction of chemical and biological distribution at coastal boundaries. Meanwhile, time scales for dynamically important coastal processes also span orders of several magnitudes, from seconds to months or even years, and effects can accumulate over time.

Coastal regions throughout the nation and world are simultaneously affected by a number of significant stressors. Human activity (e.g., agriculture, sewage treatment, runoff) alters both the concentration and composition of nutrients entering marine systems (Peierls et al., 1991; Howarth et al., 1996). Excessive amounts of nitrogen and phosphorus are entering streams and rivers, eventually reaching estuarine and coastal waters and causing eutrophication, which can result in harmful algal blooms and episodes of hypoxia (Anderson et al., 2002). Chemical pollutants can severely affect the biology of marine organisms. A variety of commercially important species bioaccumulate toxic pollutants, while other species’ reproductive traits are impacted by estrogenic chemicals from human activity (Morel et al., 1998; Vos et al., 2000). Coastal development and recreational activities have led to habitat loss and degradation for many species including fish, marine mammals, and seabirds, particularly in coral reef and sea grass communities. Commercial and recreational fishing affect coastal ecosystems, both through the removal of target species and the unintentional bycatch of other organisms (Stevens et al., 2000; Pauly et al., 2002). Marine shipping is introducing many nonnative species to coastal areas (Ruiz et al., 2000). On top of these near-term

ecosystem stressors, communities will also have to respond to potential changes in temperature, acidity, and ultraviolet exposure due to climate change (Halpern et al., 2008). The cumulative effects of these various stressors will likely affect ecosystems in complex ways that cannot be predicted by sim­ply adding the effects of each individual component (Crain et al., 2008). This highlights the importance of efforts that develop ecosystem-based monitoring and management tools for marine resources but also shows the inherent challenges involved in effectively implementing these tools.

One of the most dramatic signs of rapid change in polar regions is the observed decrease in sea ice cover in the Arctic Ocean; between 1979 and 2009 the annual minimum extent of Arctic sea ice cover decreased at a rate of ~11 percent per decade (Stroeve et al., 2008). These changes in the extent and concentration of sea ice can alter the seasonal distribu­tions, geographic ranges, patterns of migration, nutritional status, reproductive success, and ultimately the abundance and stock structure of several fish, marine mammals, and sea­bird species (e.g., Tynan and DeMaster, 1997). Furthermore, because the albedo (surface reflectivity) of snow and ice is several times that of ocean water, loss of sea ice increases the amount of solar radiation that is absorbed by the Arctic Ocean, warming the surface waters and creating a positive feedback cycle that causes even more sea ice to melt and thus amplifying warming trends. Along the West Antarctic Peninsula, midwinter surface atmospheric temperatures have increased by 6°C (5.4 times the global average) during the past half century, 87 percent of the glaciers are in retreat, and the concentration of winter sea ice has decreased (Ducklow et al., 2007, and references therein). Heat from the ocean is implicated as a major driver for the deglaciation, as increased supply of heat associated with Upper Circumpolar Deep Wa­ter flux is believed to be associated with strengthening winds over the Southern Ocean. The increased heat is itself partly a consequence of anthropogenic activity (greenhouse gas emissions and/or ozone depletion). Atmosphere-ocean-ice interplay at the West Antarctic Peninsula results in a positive feedback that amplifies and sustains atmospheric warming.

Rapid climate changes in polar regions are triggering pronounced shifts and reorganizations in regional ecosys­tems and biogeochemical cycles (Moline et al., 2008). While large ecosystem changes have been detected (e.g., shifts from marine mammal to pelagic fish [Grebmeier et al., 2006]), linking shifts in the physical system to biological changes remains difficult; however, overcoming this gap is a critical step in establishing any level of predictive skill (Schofield et al., 2010a). The complexity of marine ecosystems, com­bined with chronic undersampling, limits the understanding of how a shifting ocean will affect regional and local marine food webs.

With the future expansion of commercial activities in coastal waters and the ocean, ocean sciences must be prepared to address accidents and spills. The U.S. Coast Guard’s National Response Center reports that there were over 34,000 spill incidents of all types in 2010.⁴ Perhaps of greatest concern are oil spills. In the 25-year period of 1974-1997, there were 742 oil tanker spills worldwide that released more than 1,000 barrels (136 metric tons) of oil each (NRC, 2007b). In U.S. waters more than 70,000 barrels (~9,800 metric tons) of oil or refined petroleum products are spilled every year on average (NRC, 2003b). In April 2010, the explosion and sinking of BP’s Gulf of Mexico Deepwater Horizon oil platform resulted in an unprecedented disaster, with 60,000 barrels (~8,200 metric tons) of oil per day issu­ing from the deepwater well for 87 days (National Commis­sion on the BP Deepwater Horizon Oil Spill and Offshore Drilling, 2011).

Spill responses include deployment of mechanical containment and recovery systems (e.g., booms, skimmers) or nonmechanical methods (e.g., surface burning, oil dis­persants). Dispersants act to reduce break up and dilute the oil by mixing it into the ocean. The biological and physical processes that determine the ultimate fate of dispersed oil, and its potential toxicity to the marine environment, are poorly understood (National Commission on the BP Deep­water Horizon Oil Spill and Offshore Drilling, 2011). Of particular concern is the fate of dispersed oil in areas with high suspended solids, as it is unknown how chemically dispersed oil interacts with suspended sediments, over both short and long terms, compared to naturally dispersed oil (NRC, 2005).

The danger of possible oil spills in the Arctic will be an issue of future interest. Decreasing Arctic sea ice (Stroeve et al., 2008) as a result of climate change will attract greater amounts of commercial shipping and oil and gas exploration. An Arctic oil spill is likely to be much more difficult to con­tain than in other regions: spills in or under ice-covered areas would be harder to track, would require different techniques than those in open water, and harsh, remote conditions would increase the difficulty of getting spill recovery assets in place.

A related issue is the existence of more than 8,500 sunk­en vessels worldwide (Michel et al., 2005), three-quarters of which date back to World War II (Hamer, 2010). These shipwrecks could harbor between 2.5 and 20 million tons of oil (Michel et al., 2005), as well as hazardous chemicals and munitions. The lower estimate of oil contained within

these shipwrecks is at least twice as much as the Deepwater Horizon spill (Hamer, 2010).

Geoengineering can be classified as deliberate actions that modify environmental processes in order to mitigate other environmental impacts that result from human activities (The Royal Society, 2009) and are generally considered global in scope (NRC, 1992). Many projects presently being discussed focus on storing CO₂ in the ocean, either by (1) pumping liquid CO₂ into the deep ocean or into the subseafloor, (2) enhancing weathering reactions of CO₂ with carbonate or silicate minerals and storing the products in the ocean, or (3) accelerating natural mechanisms of ocean carbon uptake by seeding the ocean with nutrients, thus removing CO₂ from the atmosphere. Direct injection of CO₂ into the deep ocean will decrease acidification of surface waters but will exacerbate the problem at depth as a result of CO₂ leaks or natural seepage back to the ocean floor (Caldeira and Wicket, 2003; Blackford et al., 2009). Early experiments on deep sea communities suggest that they may be more sensitive to changes in pH due to increased CO₂ than shallow water communities (Barry et al., 2004). In addition, elevated dissolved CO₂ concentrations may impose a physiological strain on marine animals, especially in hypoxic regions, which are likely to expand as the ocean absorbs anthropogenic CO₂ or it is injected into the ocean as part of geoengineering projects (Peltzer and Brewer, 2008; Brewer and Peltzer, 2009). Enhanced weathering reactions avoid the major pH changes (and ensuing acidification) associated with storing CO₂ directly in the ocean, but are potentially expensive and require extensive mining of source materials (The Royal Society, 2009). Perhaps the most discussed nutrient addition project is iron fertilization (Cullen and Boyd, 2008), which follows the principle that growth rates and biomass accumulation by phytoplankton are limited by the availability of iron in as much as 40 percent of the world ocean (Moore et al., 2002). If iron could be added to these deficient areas (via ship or other platform), it would increase plankton growth rates and perhaps increase removal of carbon dioxide from the atmosphere (Coale et al., 1996). These types of experiments would result in a deliberate modification of marine ecosystems, which could shift many open ocean areas from a low-biomass, low-primary productivity condition to moderate productivity (similar to the coastal ocean). It is difficult to predict the impacts of this activity with certainty, but concerns have been raised about the formation of low-oxygen areas and harmful plankton blooms (Cullen and Boyd, 2008) and the potential limited impact of fixing carbon in the deep ocean (e.g., de Baar et al., 2005). Other geoengineering projects at more modest scales have also been discussed, such as reoxygenation of the Baltic Sea to reduce phosphorus and decrease seasonal eutrophication (Stigebrandt and Gustafsson, 2007).

The protection of life and property is a compelling societal objective. The research that supports this objective continues to focus on predicting and mitigating natural hazards associated with the solid earth (e.g., earthquakes and volcanoes) and weather (e.g., severe storms and drought). In addition, several new areas have become more prominent either because of recent catastrophic events (e.g., tsunamis) or growing concerns related to climate change and variability (e.g., sea level rise and ocean acidification). The prediction and mitigation of adverse human health outcomes has emerged as a major area of research related to climate change science. Societal concerns, combined with the potential for significant advancement in prediction and mitigation, are likely to drive interest in these six research areas well beyond 2030.

The effects of offshore earthquakes can be monumental, whether direct (e.g., ground shaking and rupture) or indirect (e.g., triggering a tsunami). Many of the largest earthquakes in the world occur offshore.⁵ In the United States, major off-shore seismic hazards span the west coast from California to Alaska, including the offshore component of the San Andreas Fault as well as the Cascadia and Alaska subduction zones. Although paleoearthquake data can constrain occurrence intervals, earthquakes still cannot be predicted. In the next two decades, there is likely to be progress on this front, as earthquake early warning methods that detect the beginning of a large fault rupture based on initial portions of the primary (compressional) waves have recently been developed (Allen et al., 2009, and references therein). This allows for a warning to be issued before the arrival of larger, more damaging secondary (shear) waves. As observations that are collected close to earthquake epicenters provide extra information that can strengthen early event warning systems (McGuire, 2008; Yamada and Mori, 2009), instrumenting offshore seismic hazards could improve prediction and detection of potentially damaging offshore earthquakes.

The majority of seafloor volcanism occurs along the global mid-ocean ridge spreading network, within ocean island arc environments, and at hot spots. Shallow, large volcanoes common in arc, back-arc, and hotspot environments can present significant environmental hazards. One historic example of this was the 1883 Krakatou eruption in Indonesia, which changed global climate through its

eruption of ash and gases (The Krakatoa Committee of the Royal Society, 1888; Self and Rampino, 1981) and led to a tsunami that killed 36,000 people living around the Sunda Strait (Kious and Tilling, 1996). Understanding of how strain accumulates in the seafloor, the spatial and temporal evolution of crustal movement, and the migration and release of magma and volatile elements is critical to developing predictive models of volcanic eruptions, and possibly lessening their impacts.

Tsunamis can result from earthquakes, submarine and aerial landslides, volcanic eruptions, and in rare instances meteorite impacts that rapidly displace large volumes of water in the ocean. Generally, damaging tsunamis are caused by earthquakes greater than magnitude 7 (NRC, 2011b). However, catastrophic submarine landslides caused both by volcanic eruptions, large-scale collapse of volcanic islands, earthquakes, or other slope instabilities can also lead to tsunami generation; historic mega-tsunamis reaching 365 m above sea level have been related to flank collapse (Moore et al., 1989, 1994; Clague et al., 2002; McMurtry et al., 2004; Pérez-Torrado et al., 2006). Tsunami waves can be centimeters to tens of meters tall, last over a period of several hours, and cause flooding of low-lying areas, greatly affecting coastal communities. The December 2004 Sumatran earthquake and resulting tsunami fundamentally changed the global perception of tsunami threat, with the loss of more than 200,000 lives and billions of dollars in property damage (Schiermeier and Witze, 2009). The ability to predict the initiation of tsunami waves remains as elusive as the ability to predict earthquakes and landslides; however, once a tsunami-generating event has occurred, the arrival time of the first waves can be predicted for a given site within a few minutes. This is the same timeframe in which a “near-field” tsunami (one that originates near an at-risk community) can occur. Tsunami models have also performed reasonably well in forecasting tsunami wave heights since the installation of an open ocean sea level observing network; however, near real time wave height forecasts are only available with considerable delay on the order of a fraction of an hour or more (NRC, 2011b). While efforts to create a global warning system and educate at-risk communities have expanded significantly since the 2004 tsunami, the next two decades are likely to see increased population growth and property development along the coast. Maintaining the tsunami warning systems, and educating this population about high-risk, low-probability events like tsunamis will remain a challenge (NRC, 2011b).

Hurricanes and other severe weather events have the largest economic impact of any natural hazards (Kunkel et al., 1999). Prediction of hurricane and tropical storm paths has improved; however, progress is still needed in the prediction of the intensity of such severe weather (NRC, 2010f). While climate variability and change may influence severe weather, this remains an area of active research (Bader et al., 2008; NRC, 2010b). According to the IPCC’s Fourth Assessment Report (Pauchiri and Reisinger, 2007), it is more likely than not that there is a human contribution to the observed trend of hurricane intensification since the 1970s. Although there is increasing certainty of the link between climate change and more intense hurricanes and tropical storms, the effects of climate change on their frequency remains unclear (Bender et al., 2010; Knutson et al., 2010). ENSO events in the Pacific, which occur every 4-7 years, tend to suppress hurricane activity in the Atlantic, particularly inhibiting the formation of major hurricanes (Category 3 or higher). Climate change also has potential to impact the distribution, frequency, and intensity of other forms of severe weather (e.g., coastal flooding), with great impacts on coastal populations. Increased storm frequency and severity will also increase risks to all maritime operations. Ports, ships, and offshore structure (e.g., oil platforms and wind farms) will need to be designed and engineered to withstand extreme conditions and to ensure crew safety and environmental protection. As demonstrated by the 2005 devastation of the Gulf Coast by Hurricane Katrina, especially in the context of a changing climate, hurricane prediction and mitigation of impacts will remain a top priority for ocean and atmospheric science.

Sea ice collisions create pressure ridges that rise several meters above sea level and descend tens of meters below the air-sea interface (Williams et al., 1975; Wadhams and Horne, 1980; Wadhams, 1988), posing a collision hazard to ships transporting personnel and materials within the Arctic and Southern oceans, as well as the North Atlantic, Bering Sea, and Great Lakes. Climate change has led to significant thinning of ice shelves at both poles (Pritchard et al., 2009), causing ice shelf collapse in both the Antarctic (Scambos et al., 2009) and Arctic (Copland et al., 2007) that release hazardous chunks of ice into the Southern and Arctic oceans. Declining sea ice cover, as noted in the Arctic (NOAA Arctic Report Card, 2010) also has implications for sea level rise (Shepherd et al., 2010). Since 1979, satellites have monitored the changing extent of ice in and around the polar seas (Zwally et al., 2002; Stroeve et al., 2007, and references

therein), but cloud cover limits the ability of satellites to precisely map the distribution of sea ice and icebergs, and existing models cannot accurately predict where ice will be found. In addition to posing collision hazards, large icebergs have grounded in the shoals off McMurdo Station (Robinson et al., 2010), hampering efforts to resupply that important Antarctic scientific station. Along the Arctic and sub-Arctic coastlines, the reduced span of shore-fast ice leads to greater exposure to storm surges; as a result, many shorelines are eroding rapidly with attendant loss of societal infrastructure to the native communities that live there (ACIA, 2004).

Humans are significantly altering the coastal environment, with many actions that have potential to negatively affect human health. There is a growing need to identify the source, transport, fate, and impact of chemicals in common use by industry, agriculture, and households that are eventually discharged into coastal waters. Anthropogenic activity has changed the concentration and composition of nutrients entering marine systems (Peierls et al., 1991; Howarth et al., 1996), leading to degraded coastal water quality. Increased nutrients lead to greater growth of phytoplankton and macroalgal biomass, which heightens turbidity, depletes oxygen, decreases marine biodiversity, and alters ecosystem structure and function (NRC, 2000a; USCOP, 2004). This has been linked to increased frequency and intensity of harmful algal blooms around the world (Hallagraeff, 1993; Pearl, 1997; Anderson et al., 2002; Babin et al., 2008). Harmful algal blooms can lead to devastating fish and mammal kills, and can sicken and even kill humans (Anderson, 1994; Glibert et al., 2005). Another form of pollution, sewage discharge in coastal waters, can lead to increased levels of pathogens and viruses, which can be unsafe both for human exposure and for a variety of marine life (e.g., Goyal et al., 1984; Lipp et al., 2001). The production and use of traditional (e.g., PCBs [polychlorinated biphenyls], heavy metals) and emerging contaminants is also likely to continue into the future. Many emerging contaminants, including compounds such as flame retardants, insect repellents, pharmaceuticals (e.g., steroids, hormones, antibiotics, analgesics), and domestic waste (e.g., detergents, fragrances, caffeine) persist in the environment, accumulate in tissues, and can be toxic to humans and aquatic life. Others interfere with hormone systems governing reproduction and growth (Morel et al., 1998; Vos et al., 2000).

Broad-scale shifts in the ocean-climate system are likely to affect human health patterns. ENSO is associated with changes in precipitation patterns across the globe with major implications for human health and welfare (Glantz et al., 1991). For example, ENSO significantly increases the flood frequency for coastal California (Andrews et al., 2004), while other regions are affected by more severe droughts (Philander, 1990). Changes in these precipitation patterns have been linked to epidemics of malaria on the Indian subcontinent and South America (Bouma and van der Kaay, 1996; Bouma and Dye, 1997). In East Africa, Rift Valley Fever⁶ (a viral zoonosis) epidemics have coincided with unusually high rainfall associated with ENSO-related Pacific and Indian Ocean SST anomalies (Linthicum et al., 1999).

In and around the Arctic Ocean, climate-related changes are having diverse effects on human populations and their physical and physiological well-being. Arctic environmental change has already adversely affected critical ecosystems that many native communities are dependent on for their livelihoods. The decreasing time available to use shore-fast ice as a platform (Druckenmiller et al., 2009) in combination with a general decrease in sea ice extent (Stroeve et al., 2007) has resulted in shorter seasons for subsistence hunters to find bears, walruses, and seals, which are staples of many indigenous diets (ACIA, 2004). The net result of these factors, in combination with other societal forcing functions, is a migration of some indigenous populations out of Arctic communities. Beyond these examples of direct effects, indirect impacts such as changes in ecosystem health or sea level are discussed throughout this chapter.

The United States, with over 12,000 miles of coastline,⁷ has strong economic ties to the ocean. Traditional uses, such as oil and gas extraction, fisheries, and recreation, are still large components of the ocean economy. Other activities, including aquaculture, wind power, and marine hydrokinetic resources, are likely to become much more important in the next two decades. Scientific research that identifies oceanic resources in the broader context of impacts that might be incurred through utilization will promote this societal objective. Sustainability of these resources for future generations is of great importance, as is minimizing adverse impacts on the marine environment. The next three questions examine these important future issues.

The ocean and inland waters provide about 20 percent of the protein supply for a growing human population (UNFAO, 2009). Overexploitation of fisheries stocks and unsustainable fishing practices have created significant threats to marine biodiversity (Myers and Worm, 2003; Pauly et al., 2003; Worm et al., 2006, 2009) and to food

security in some parts of the world. Global wild fishery catches leveled off in the 1980s (Pauly et al., 2002) and some experts fear large-scale extinctions of commercially important species (Dulvy et al., 2003). Since the 1980s, perperson seafood production has kept pace with population growth only because of the growth of aquaculture production. Even with better management of wild fish stocks, aquaculture is expected to play an increasingly important role in future global seafood supply (UNFAO, 2009). Both wild capture fisheries and aquaculture production have the potential to create significant impacts on ocean systems. Trawling and other benthic fisheries can severely impact communities through the destruction of seafloor habitat (Thrush and Dayton, 2002). Overfishing of predatory species can fundamentally alter food webs, which has the potential both to impede recovery efforts for the stocks and to lead to jellyfish blooms that further affect fisheries (Scheffer et al., 2005; Purcell et al., 2007). Characteristics of deep sea fish (e.g., slow growth rates and maturation, long life, and low birth rates [Devine et al., 2006]) make them susceptible to overfishing, although the full impacts of removing these deep-sea species from the food web are not yet well known (Koslow et al., 2000). Aquaculture is responsible for the introduction of a variety of nonnative species, and the animal waste products from some operations are a significant source of water pollution (Wu, 1995; Ruiz et al., 2000). In addition, many aquaculture programs involve the farming of carnivorous species that rely on fishmeal and fish oil (NRC, 2011a), increasing total fishing pressure in other fisheries (Naylor et al., 2000). Research into potential methods for sustainably managing fisheries, such as monitoring the status of fish stocks and their role in ecosystems, creating accurate catch limits, and establishing marine protected areas, will be critical to ensure future food production from the ocean. Equally important is the goal of maintaining ocean biodiversity, which may be difficult to achieve while also maximizing fisheries (Brander, 2010). Management strategies that enable both sustainable fisheries and biodiversity conservation are needed and will require improved environmental and fisheries data resources and substantially better modeling capabilities. The use of ocean space for farming finfish, shellfish, and algae (Goldburg et al., 2001; NRC, 2010c) will also need to be balanced against competing energy, national security, and recreational needs.

For the foreseeable future, traditional oil and natural gas extraction will continue to fill a significant proportion of U.S. energy needs (e.g., Musial and Butterfield, 2004; Greene et al., 2007). The U.S. outer continental shelf is a major focal point for energy industries, accounting for an estimated 30 percent of national oil production and 11 percent of natural gas production in 2009.⁸ In recent years, there has been increasing oil production in deep waters (greater than 1,000 ft), especially in the Gulf of Mexico (USCOP, 2004). The scope of energy extraction is likely to continue to incorporate deeper waters, as well as smaller reservoirs and additional, alternative sources.

One such source is methane hydrate, an ice-like substance formed from a combination of gas and water at high pressures and low temperatures. Burning methane produces less carbon dioxide than other fossil fuel combustion, and its abundance in U.S. continental margins and permafrost could provide greater energy security for the United States (NRC, 2004a, 2010e). Although most methane hydrate is found at low concentrations and is not currently economically viable, more concentrated methane hydrate accumulations (found in deepwater marine and Arctic sands [Boswell and Collett, 2006]) could be likely targets for a future economic resource. However, potential degassing of methane hydrate at atmospheric conditions is a technical challenge for recovery and could affect the global carbon cycle.

There is also international interest in mining seafloor massive sulfide deposits that contain economically valuable minerals (Hoagland et al., 2010). In some cases, such as oil and gas production, resource utilization in the ocean is driven by the difficulty of satisfying demand with economically accessible terrestrial resources.

Commercial activity in the ocean is growing and may possibly become an important part of the U.S. energy portfolio, especially the unique opportunities to harness renewable energy. These include installations of wind farms in coastal environments, development of marine hydrokinetic power (from waves, tides, ocean currents, and ocean thermal gradients), and siting of solar collectors on a large scale. Renewable energy activities, like offshore wind farms and marine hydrokinetic systems, exploit unique properties of the ocean; in this case, higher wind speeds that occur over the ocean as compared to land or strong wave energy or tidal currents at certain locations (e.g., Bay of Fundy, Hudson River). In each case, the economic viability of these sources will be enhanced by matching optimal environmental conditions with appropriate energy infrastructure design. Each of these uses will have some associated environmental and societal impacts in addition to their significant economic value: habitat disturbance or destruction, injury or fatalities for birds and marine organisms, aesthetic concerns, and changes to indigenous cultures. Comprehensive coastal and marine spatial planning (such as that outlined in the NOP [CEQ, 2010]) will be needed to manage these and other compet-

ing activities in the ocean. The optimization of renewable energy production while minimizing impact represents an important, emerging area of research.

Beyond their societal objectives, investigating the science research questions posed in the previous sections will, in turn, contribute to increases in fundamental understanding of the ocean and its relationship to the Earth system. Fundamental research, even if not directly applicable to a problem of societal relevance, has considerable merit in its own right. It has a long history of producing discoveries that advance scientific understanding, many of which eventually lead to an increased ability for stewardship of the environment, protection of life and property, and promotion of sustainable economic vitality. An essential component is understanding-driven research, which provides a foundation to increase current knowledge of the ocean in order to improve predictive capability. There is also a compelling need for human exploration, both to understand how Earth functions and to unravel the many remaining mysteries on the nature of physical, biological, chemical, and geological processes that occur and interact. These 10 fundamental questions range widely in scope and scale, from entire Earth system processes to individual organisms.

Understanding of Earth’s interior is critical to a range of societal issues, including earthquake detection and hazard assessment; the development of volcano and tsunami warning systems; the role and effect of fluids in Earth’s crust; energy and mineral resource exploration; and even nuclear test monitoring and treaty verification (Forsyth et al., 2009). The past four decades of geophysical research have established that heat from Earth’s deep interior powers convection in its liquid outer core, generating a planetary magnetic field, and that heat in the solid mantle drives plate tectonics. Mantle convection also regulates the chemical composition of the surface layers, drives the exchange of materials between the planetary surface and its deep interior, and produces chemical fluxes into the ocean and atmosphere. Although it is known that the mantle and core are in constant convective motion, their motion can neither be precisely described nor confidently calculated with respect to past differences (NRC, 2001, 2008b). Patterns of convection are poorly understood, although there may be internal boundaries resulting from chemical differentiation within the mantle, with mineralogical phase changes controlled by pressure and temperature (Forsyth et al., 2009). Finally, although plate tectonic theory explains many surface features of the planet, it is not currently understood why Earth has plates or what the relationships might be between plate tectonics and Earth’s abundant water, continents, and life. Further study of Earth’s interior can help determine what the surface environment was like in the past and predict what it might become in the future.

Earth history contains a rich and diverse record of climate change, operating across a broad spectrum of time scales (Ruddiman, 2010). Given the evidence for significant anthropogenic influence on the climate system, better understanding of the rates at which climate changes and the climate system’s sensitivity to various factors have become extraordinarily important to society. For example, sea level was more than 120 m lower than present at the peak of the last glacial period about 20,000 years ago (Church et al., 2008), then rose between 19,000 and 7,000 years ago (Lambeck and Chappell, 2001). From approximately 2,000 years ago to about 1900, sea level changed very little (Lambeck et al., 2004), but anthropogenic increases of greenhouse gas concentrations are now causing sea level rise. The rate and magnitude of future climate change is closely tied to the expected impacts both on human and natural systems, and many of the changes may be largely irreversible within millennial time scales (Solomon et al., 2009). Therefore, an increasing ability to predict future climate change and to better understand uncertainties in climate prediction is also an urgent societal need. Clarifying possible rates of climate change is critical to understanding potential resiliency of global marine ecosystems. High-resolution oceanic and terrestrial paleoclimate records help assess rates and magnitude of past climate change in the context of Earth’s surface, atmospheric composition, and variations in solar input, and may provide analogues for predicted future change. At the same time, describing changes in the modern climate system, while focusing on areas of greatest uncertainty in current climate processes, is likely to improve predictive skill and increase understanding of complex interplay of processes within the climate system.

Interactions between the ocean and atmosphere are complex and multilayered: the atmosphere imparts momentum to the ocean; precipitation and evaporation affect ocean salinity; heat and gas exchange between the atmosphere and ocean; dust containing nutrients and toxins is deposited from the atmosphere to the ocean; cloud condensation nuclei are injected into the atmosphere from the ocean; sunlight is attenuated by the atmosphere and infrared radiation emitted from the ocean is trapped by the atmosphere. These interactions directly and indirectly affect physical, chemical,

and biological processes in the ocean. The atmosphere is driven by SST, so knowledge of the upper ocean is needed for prediction of climate and weather, including improved hurricane prediction. Ocean circulation is largely driven by winds, so accurate knowledge of wind stress is essential to the specification and prediction of currents at all scales. In addition, wind-driven ocean waves modulate fluxes of many properties (e.g., gas exchange). All of these fluxes are essentially turbulent, requiring parameterization to relate them on larger-scale, easier-to-measure quantities, and to be represented in models. Ocean-atmosphere interactions also drive the coupled biogeochemical system. In polar regions, sea ice acts as a porous layer between the ocean and atmosphere, as well as a source of gas fluxes, even in winter. The surface ocean microlayer regulates particle and gas exchange into the overlying atmosphere. Both the microlayer and seawater below it produce and concentrate organic compounds that are potentially ejected into the air; however, limited measurements of the resulting aerosols’ organic compositions constrain current understanding and modeling.

Observation, theoretical understanding, and parameterization of mixing are essential to climate prediction. The ocean is a global turbulent fluid, with length scales ranging from ocean basins to molecules and time scales from seconds to centuries or longer. All of these scales interact, so that mixing processes occurring at small scales end up affecting global circulation. However, mixing processes must be parameterized in ocean climate models, as they occur at scales too small to be directly simulated, given the resolution of present-day models. The ocean is also a highly anisotropic fluid, with vertical gradients much stronger than horizontal gradients. Maintenance of the vertical stratification requires mixing to balance the upwelling that occurs through the deep ocean (Munk, 1966). Ocean observations of turbulent mixing have established that there is relatively small diffusivity in the interior ocean (Gregg, 1989), while substantial mixing is found in surface and bottom boundary layers and in regions of flow over rough topography (Davis et al., 1981; Polzin et al., 1997; Sanford and Lien, 1999). Sources of energy for mixing are dominated by the wind and the tide. Tidal mixing has received a great deal of attention because the time scale is predictable and amenable to observation (Rudnick et al., 2003). Wind is an important energy source (Wunsch and Ferrari, 2004), whether it occurs directly through the surface mixed layer or indirectly through mesoscale eddies spun off of major ocean currents. The study and parameterization of horizontal stirring of the ocean, the submesoscale (scales of kilometers), and subgridscale processes for ocean models (Fox-Kemper and Ferrari, 2008) are central to climate predication and likely to remain important. One of the properties of turbulence is intermittence, so that variability is concentrated in space and time, as at ocean fronts or under storms. The study and prediction of such events is an ongoing research question.

The oceanic crust is the largest fractured aquifer system on the planet. Fluid circulation through the crust and overlying sediments generates enormous chemical fluxes in the ocean, profoundly alters the composition of basement rock, and supports a potentially vast subseafloor biosphere (Fisher, 2005; Fisher and Wheat, 2010). Hydrothermal flow significantly influences the thermal, mechanical, and chemical state of subducting tectonic plates, impacting seismicity and volcanism (e.g., Gill, 1981; Peacock and Wang, 1999). High-temperature hydrothermal fluids are produced at mid-ocean ridge vent systems, but low-temperature fluid circulation occurs in approximately half of the seafloor crust (Parsons and Sclater, 1977; Fisher and Von Herzen, 2005), accounting for two to three orders of magnitude more seawater circulation than the mid-ocean ridge (Davis and Elderfield, 2004). Off-axis fluids within the seafloor can be transported laterally tens of kilometers (Fisher et al., 2003), which has implications for microbial connectivity and movement in the subseafloor. Despite its importance for global-scale processes in the ocean, the subsurface fluid flow system is undersampled and its biogeochemical impacts are not yet well resolved.

Ocean sciences will continue to play a leading role in understanding the fundamental, unresolved questions of how Earth’s life began and has evolved over time. The late-1970s discovery of submarine hydrothermal vents fueled by undersea volcanoes (Spiess et al., 1980) led to the hypothesis that life may have originated in these hot spring systems (Corliss et al., 1981). Since that time, more than 200 active vent systems have been discovered, and studies of these environments have profoundly changed thinking on where and under what conditions life exists on Earth (e.g., Wilcock et al., 2004; Kelley et al., 2005; Martin et al., 2008). The field of astrobiology uses limits from vent environments (e.g., temperature, pressure, salinity) to guide the search for life on other planets (Woodruff and Baross, 2007). It also uses advancements in molecular sciences and in experimental analyses focused on abiotic synthesis of organic compounds, which have driven new hypotheses about the conditions necessary for life and its evolution (e.g., Shock et al., 1990; Cody et al., 2004). Beyond the vents, the most extensive populations of microscopic life may exist in vast, largely unexplored areas of oceanic crust and sediment away from active mid-ocean ridge systems. Using ocean drilling core

samples, microbial activity has been documented at depths of over 500 m beneath the seafloor (Parkes et al., 1994), and the total amount of carbon associated with subseafloor bacteria and archaea exceeds that of any other ecosystem on Earth (Gold, 1992; Whitman et al., 1998). Genetic and functional diversity of the deep ocean biosphere, conditions under which organisms can live and thrive, and their contributions to oceanic carbon and other biogeochemical cycles are just beginning to be explored.

The past two decades have revealed staggering molecular, biochemical, and species diversity in the ocean, a complexity that is reshaping views of the structure of oceanic food webs (e.g., Delong and Karl, 2005). Studies in the 1970s underestimated the number of microorganisms by three orders of magnitude; marine viruses (arguably the most abundant and diverse form of life on Earth) were only appreciated in the 1980s (e.g., Fuhrman, 1999). The Census of Marine Life,⁹ which operated from 2000 to 2010, detected numerous new megafauna species, and mass sequencing of microbes in the oligotrophic Sargasso Sea revealed more than a million protein-encoding genes and discovered a large number of new genes (Venter et al., 2004). Despite this progress, existing biodiversity has still not been quantified nor have robust abundance estimates been achieved, and it is not well understood how physical, chemical, and biological factors maintain diversity. Interpreting microorganism complexity is further complicated by potentially high rates of lateral gene transfer and mutation, which suggest high rates of molecular evolution in the ocean (Frigaard et al., 2006; Oliver et al., 2008). How this molecular evolution translates into the innovation of new species and biochemical pathways is an open question. Oceanographers are drawing upon rapid advances in technology from the medical sciences to perform techniques such as genome sequencing, quantification of protein structure and expression, and metabolite analysis in order to address marine problems. It is anticipated that the resulting data sets will help scientists more accurately study the evolution, biochemistry, physiology, and diversity of marine organisms.

Understanding ocean ecosystem dynamics and predicting changes over time requires knowledge of species diversity, distribution, and abundance throughout the ocean (Pereira et al., 2010). Although species living in the ocean’s upper reaches are relatively well known, far less can be said about species in the bathypelagic (1,000 to 4,000 m) and abyssopelagic (4,000 to 7,000 m) zones (Robison, 2004, 2009; Heino et al., 2010; Wiebe et al., 2010). This lack of knowledge is even more notable since the bathypelagic zone accounts for 60 percent of the ocean’s volume, making it the largest marine habitat on Earth. Comprehensive understanding of deep-sea biodiversity has eluded oceanographers because of the fragility, rarity, small size, and/or systematic complexity of many taxa, as well as the difficulty in sampling the more mobile larger invertebrates and fish (Sutton et al., 2008, 2010). For many groups, there are long-standing and unresolved questions of species identification, systematic relationships, genetic diversity and structure, and biogeography (e.g., G. Johnson et al., 2009).

The global geographic scale of the investigations required to address these issues, as well as the three-dimensional complexity of the world ocean, make complete knowledge of marine deep-sea biodiversity and ecosystem dynamics challenging. New technologies and advanced sensors will play a significant role in developing fundamental knowledge about deep-sea ecosystems. However, physical access to remote regions of the ocean and the deep-sea interior will still be required to discover and observe the organisms living in these areas, and to understand their interactions and dynamics.

Marine organisms have extraordinary abilities to sense and respond to their surroundings and, in many cases, to actively communicate within or between species. These sensory processes underlie many observed spatial and temporal patterns that cannot be explained by ocean physics or chemistry. Basic sensory systems used to perceive environmental conditions and communicate within and between species include vision (e.g., light vs. dark, complex colors, patterns, shapes, movements), hearing (acoustic signals of varying wavelength and sound patterns), chemosensory (waterborne “smell,” surface-bound “taste” compounds), and somatosensory (e.g., physical contact, temperature, body position, pain). These mediate all fundamental biological and ecological processes, spanning from reproduction to habitat selection and predator-prey interactions. For example, sea anemones perceive chemical signals from other anemones that have been wounded by predators and respond with a characteristic defensive contraction (Howe and Sheikh, 1975); packs of dolphins utilize sonar to coordinate swimming behavior, aggregating prey in small spatial zones to minimize grazing effort (Benoit-Bird et al., 2004); and bioluminescence in the deep ocean has been hypothesized to help increase reproductive success and/or provide protection against predators (e.g., Haddock et al., 2010). Communication is also found commonly in microbial systems, where quorum-sensing bacteria produce and release chemical-

signal molecules that increase in concentration as a function of cell density to stimulate gene expression of neighboring bacteria. Communication can regulate a diverse array of physiological activities (e.g., symbiosis, virulence, competence, conjugation, antibiotic production, motility, sporulation, biofilm formation; Miller and Bassler, 2001). Because these sensory mediated processes are central to evolutionary life histories, population dynamics, and community ecology, a more complete understanding will be central to predicting marine organisms’ responses to various ocean environmental changes in the future and for developing sustainable ecosystem-based management strategies.

Earth’s carrying capacity, or maximum number of organisms that can be supported without undergoing environmental degradation,¹⁰ is dynamic and ultimately finite. With a human population that is projected to exceed 9 billion by 2050 (UN, 2009), people have become the dominant consumer of most of the world’s major ecosystems (Rees, 2003). However, the human population needs more than ecosystem products; there are many ecological goods and services provided by nature that are essential for human sustainability (Costanza et al., 1997). These fall into three categories: renewable natural capital (e.g., species, ecosystems), replenishable natural capital (e.g., oxygenated air, freshwater), and nonrenewable natural capital (e.g., fossil fuels, minerals; Rees, 1996). The current human population is living beyond sustainable means provided by renewable and replenishable natural capital and is sustainable only by use of nonrenewable resources (Daily and Ehrlich, 1992). For example, industrialized fisheries, which are calculated to dramatically reduce community biomass in less than two decades (Myers and Worm, 2003), represent a domain in which carrying capacity issues are already clear and may turn some renewable resources into nonrenewable capital. Nutrient pollution related to terrestrial agriculture and ocean aquaculture also affects carrying capacity, because they are implicated in the development of oxygen minimum zones and hypoxia (e.g., Turner and Rabalais, 1994; Diaz and Rosenberg, 2008) and raise concerns about coastal pollution (e.g., Costa-Pierce, 1996), respectively. Oil and gas production and mineral extraction, other nonrenewable resources on time scales of human development, can have significant impacts on the ocean environment. More research is needed into the ocean’s contributions to human carrying capacity, especially with regards to oxygen production, climate moderation, carbon removal from the atmosphere, and production of food and mineral resources.