Grand Challenges in Earth Surface Processes
A number of overarching intellectual questions regarding natural and anthropogenic processes are fundamental to understanding the form, composition, and evolution of our planet’s surface. From among these questions, the committee has identified nine grand research challenges, each of which is introduced in the form of a primary research question with associated opportunities for further investigation. The grand challenges are exemplary, but they cannot provide a full assessment of all current research in Earth surface processes. Focus was given, instead, to integrating a range of topics that are representative of the breadth and depth of research in the field. The nine challenges are supported by information gathered from the community of Earth surface scientists1 during this study (Appendix B).
WHAT DOES OUR PLANET’S PAST TELL US ABOUT ITS FUTURE?
One of the most remarkable aspects of the Earth surface system is the extent to which it records its own evolution. The visible landscape incorporates a wealth of information about its past, and the archive of sediments and sedimentary rock extends the record to very early in our planet’s history (Figure 2.1). The Earth’s record keeping has given us, among other things, a detailed picture of the evolution of the atmosphere, oceans, tectonics, life, and the surface system itself. Both scientific curiosity and societal demand drive us to read this fragmentary and often puzzling archive in an attempt to understand how the surface environment in which we exist has evolved through time and therefore how it may change
in the future. Our growing ability to quantify processes specific to the near-surface environment also lets us reconstruct Earth’s past in increasing detail including aspects of its geochemistry, biotic processes, topography, and particle and solute fluxes. The information stored in landscapes and sediments provides a kind of “time telescope” that allows us to view what is in effect a temporal sequence of alternative Earths—recognizably our planet, yet often surprising and unfamiliar.
In addition, the history of landscapes strongly influences their present state and future evolution. In the Earth’s northerly regions, the current landscape pattern was largely created by glacial processes that peaked around 18,000 years ago, and present-day debates about agricultural best practices and practical water quality standards require a sophisticated understanding of how the landscape is evolving in response to the deglaciation. Likewise, coastal erosion is still influenced by uplift and subsidence in response to shifts in ice loading over Holocene time. The fate of global deltas on which hundreds of millions of people depend in turn depends on the delicate interplay of subsidence and sedimentation developed over geologic time. In upland regions, the general importance of tectonic history is obvious; more subtle are the possible effects of variations in uplift rate and climate in influencing the balance of sediment storage and release to downstream river systems. Soils are among the clearest examples of the influence of past time—the soils that support global agriculture represent the integrated effects of tens of thousands of years of biogeochemical processes.
The evolutionary road from the prebiotic Earth and an atmosphere devoid of oxygen to the human-dominated conditions of today has been well documented in the popular
press: for example, extreme variations in global ice cover from nearly ice-free to (some believe) near-complete cover; meteorite impacts; sea-level variations of hundreds of meters; supercontinent assembly and breakup; mass extinctions; the growth and decay of mountain ranges; and the rich variety of life forms that preceded human existence. The sedimentary record of Earth’s colorful past is a composite of self-recording landscapes. Beyond the drama and fascination of Earth history, the record in landscapes and sediments gives us an archive of natural experiments, performed at full planetary length and time scales, from which to reconstruct planetary dynamics. This deep-time record of landscape evolution also opens an avenue to explore the connections between the Earth’s interior and its surface boundary. The morphology and properties of Earth’s surface constitute fundamental, directly observable attributes that help constrain the long-term behavior of the Earth’s convecting interior and the tectonic plates that gradually move across its surface. Landscapes on other worlds also provide a chance to test and expand our understanding of the Earth’s surface processes by applying them to materials and basic surface conditions (e.g., gravity, atmospheric pressure) very different from the ones with which we are familiar.
Practical reasons also exist for studying the record of surface evolution. Nearly every aspect of the present form, composition, and function of Earth’s surface reflects its evolution over geologic time; thus, the surface environment—including the services it provides and the hazards it poses—cannot be understood outside the context of its history. Second, the subsurface heterogeneity that controls the availability of resources such as water, hydrocarbons, and minerals is in effect a three-dimensional tapestry of fossilized surface dynamics created via crustal subsidence and burial of paleo-landscapes in sedimentary basins. Lastly, the history of surface evolution in landscapes and sedimentary basins provides us with a rich archive from which to extract information on extreme events, natural variability in space and time, and how the surface responds to change. This archive is sometimes the only means we have to observe critical thresholds and other forms of nonlinear response to our modifications of the Planet’s surface systems, such as those induced by human activity.
The past as the context for the current environment. Although “sustainability” has been defined in different ways, at its core it connotes management with a long, perhaps indefinite, time horizon. The further ahead we would like to forecast, the further back must we look to understand how the system arrived at its current state and how it is likely to evolve. For example, consider two contrasting scenarios for beach retreat, an issue that has consumed billions of dollars in the United States. One is a case of local erosion caused by interruption of longshore sediment flow by jetties. In the other, the retreat is the result of long-term land sinking caused by adjustment from the last glaciation. Although the short-term responses to both cases might appear similar, their contexts and hence strategies for sustainable management may be entirely different.
Scenarios such as this, in which long-term processes strongly influence the modern environment, can be found across the Earth’s surface. Centennial-scale or longer variations in seismicity and climate may control landslide frequency in steeplands, altering the supply of sediment to river systems, with a broad spectrum of environmental consequences. Over much of the Northern Hemisphere the strong imprint of glacial history persists in the distribution of soils, surface biota, river courses, and groundwater. The extraordinary fertility of the breadbasket of the United States is related to soils developed on sediments deposited in the wake of glacial retreat; now human activities are changing the physical structure and geochemistry of these soils and locally stripping them rapidly from the landscape (Box 1.1; see also Sections 2.4, 2.8, and 2.9).
As long as we thought we could simply control the landscape for the convenience of humankind, temporal trends such as these were primarily of academic interest. Now momentum is growing around the world to move away from this type of “hard” engineering to a more sustainable approach based on working with natural tendencies rather than against them. This cannot be done unless we know what the natural tendencies are—a task that requires a new synthesis of process understanding with research to reconstruct past history quantitatively.
Subsurface architecture. Buried surface features such as channels and beaches serve as conduits and reservoirs for water as well as for oil and natural gas (Figure 2.2). In addition, the new field of carbon sequestration focuses on attempts to remove carbon dioxide (CO2) from the atmosphere and store it indefinitely in subsurface reservoirs—often ones from which hydrocarbons have been extracted. Again, this cannot be done safely and sustainably without detailed knowledge of the subsurface plumbing system.
The petroleum industry spends billions of dollars creating extremely detailed images of the subsurface architecture created as basin subsidence and sediment input build depositional sequences in three dimensions (Figure 2.3). These images are used to characterize hydrocarbon reservoirs, but as records of geomorphic evolution, their potential value for understanding surface dynamics is enormous and has hardly been tapped.
Four billion years of natural experiments. Given the complexity of the surface environment, with its interwoven and highly nonlinear physical, geochemical, and biotic systems, it is not surprising that theoretical methods for predicting its evolution are only in their early stages. In addition to hypotheses, we must learn how the Earth’s surface works through field observation and experimentation. Intentional scientific experiments in the field are common, but necessarily involve short length and time scales; reduced-scale laboratory experiments are proving to be useful, but they cannot capture all aspects of surface dynamics. Humankind seems determined to carry out full-scale unplanned “experiments,” through activities such as changing land use, reconfiguring and regimenting natural landforms, and
altering the climate. We will learn from these experimental results even as we try to predict and perhaps influence them. Yet the increasing influence of humans at global scales impels us to make better use of the roughly 4-billion-year record of natural experiments that have already occurred on Earth. The record of natural experiments includes extreme events such as meteor impacts and rapid climate changes, as well as states of the Earth—for example, one that is nearly ice-free—that are quite different from the one we know. Although we have much to learn, we do know that the world of direct human experience represents only a small fraction of our planet’s highly varied history and of its potential future. Now as we put increasing pressure on environmental services, urbanize hazard-prone landscapes, and become predominant geologic agents ourselves, understanding the full range of planetary behavior has become crucial. Numerous other examples of insights gained from studies of surface evolution on a range of time scales are included in other sections of this report.
The archive of Earth history records a far richer range of states and behaviors of the landscape than we can observe directly in the tiny slice of time we occupy today. As such, sedimentary and other landscape records can provide critical information about how the landscape could change in the future. Three primary research opportunities to understand Earth’s surface and its future through examination of its past include (1) quantitative reconstruction of landscape records, (2) application of dating and imaging tools to understand landscape history, and (3) linking studies of the surface and subsurface.
Interpretation of landscape records by Earth surface scientists to this point has been mostly qualitative. The crucial next step is to accelerate the process of learning to read these often fragmentary records quantitatively, in order to understand long-term dynamics and reconstruct extreme events and variability and to provide the results in a form that can be used by decision makers. Recent examples of reconstruction include the spectacular findings from ice-core analysis showing that climate can change dramatically (see also Section 2.7)—over time scales measured in years—and from quantitative estimates of storm and earthquake frequency from sediment studies. These inspiring results represent a first step to unlocking the surface and sedimentary archive. One way to focus this type of reconstruction work is to seek “distant mirrors”2: times in Earth’s past with high potential to shed light on problems of prediction facing us today, such as accelerated climate change. Initiatives to study the Paleocene-Eocene Thermal Maximum (approximately 55 million years ago), an ancient warm interval potentially analogous to an anthropogenically warmed Earth, provide a good example of this approach.
New imaging methods allow us to study the present morphology of the landscape in unprecedented detail (Box 2.1; see also Sections 2.2, 2.3, and 2.6; Appendix C). Equally importantly, new radiometric and other surface-dating methods (Box 1.2; see Sections 2.2 and 2.3; Appendix C) allow us to measure rates of surface change and add temporal evolution to our snapshot of the current state of the landscape. These tools are precisely those needed to quantify landscape evolution and to reconstruct the path that has led to the present state. In tectonically active areas, analysis that includes data from these tools can, for instance, extend seismic records to time spans beyond those of human records (see also Section 2.3). In any area, reading geomorphic history quantitatively will allow us to measure fluctuations in key environmental quantities, such as fluxes of sediment and nutrients through Holocene time (the last 11,700 years of Earth history), and to constrain the magnitudes of extreme events (see also Section 2.7). These essential data provide the natural context for the observed acceleration of anthropogenic changes to the surface environment.
Digital Landscapes: Documenting Topography at the Scale of Transport and Erosion Processes
Airborne laser mapping began only in the mid 1990s, and the technology has improved rapidly since then. In 2003 the National Science Foundation (NSF) supported the founding of the National Center for Airborne Laser Mapping (NCALM) to provide research-grade airborne laser swath data to the national research community, to advance the technology, and to provide education and training for students. By January 2009, more than 60 projects covering some 13,000 km2 had been flown. For the GeoEarthScope program, all of the San Andreas Fault System (~3,000 km2) was surveyed, and those data (plus other fault zones flown by NCALM) are freely available via the GEON OpenTopography Portal (http://www.geongrid.org/index.php/topography/opentopo). NCALM makes available the data for all its surveys through its web page (http://calm.geo.berkeley.edu/ncalm/ddc.html). By the end of its first 10 years, NCALM will have flown about 80 seed projects in support of graduate research, providing key data and contributing to a new generation of researchers advancing the field of Earth surface processes through use of high-resolution topography. In 2008, NSF supported an international workshop on High-Resolution Topographic Data and Earth Surface Processes to explore how high-resolution topographic data can advance understanding of Earth surface processes (Merritts et al., 2009). Nearly half of the more than 60 participants were graduate students or postdoctoral researchers who had received a doctorate within the past three years.
Connecting Surface and Subsurface
For too long the closely related sciences of geomorphology and sedimentary geology have been pursued separately. The confluence of the need to understand and predict the evolution of the surface environment and the availability of large volumes of high-resolution information on the subsurface archive—much of it from the oil and gas industry—provides a new level of motivation to bring these two fields together both to improve prediction in the subsurface and to use the archive to understand how the surface system works. Concerted effort to understand the links between the Earth’s surface and its deep interior also provide an opportunity to meet this challenge (see also Section 2.3). Another key element is to develop observations and methods that link processes across the range of time scales from the present day to deep time. Laboratory experiments that, in effect, speed up time are one way of approaching this. Another bridge to deep time is the set of natural climate experiments constituting the current “ice house” (glacially influenced) period. These experiments are being investigated intensively for their value in understanding climate dynamics and testing numerical climate models.
HOW DO GEOPATTERNS ON EARTH’S SURFACE ARISE AND WHAT DO THEY TELL US ABOUT PROCESSES?
A glance out an airplane window on a clear day is enough to remind us of the remarkable capacity of landscape processes to create spatial patterns. These geopatterns comprise a diversity of scales and forms, and most show a fascinating mix of order and disorder. Familiar examples include the treelike, branching patterns of stream networks that create erosional and depositional landscapes; river channels, with their ornate meanders and braids; sand dunes; glacial valleys and landforms; deltas; barrier islands; and the zones and fabrics of soils (Figures 2.4-2.6). Physical landscape patterns are often closely associated with biotic ones, ranging from the variation in forest type with upland elevation to riparian ecosystems tied to stream channels to the exquisite control of marsh vegetation by small changes in land elevation and wetting frequency. In inhabited landscapes, we notice the human footprint, visible as clearly “unnatural” patterns (Figure 2.7; see also Section 2.8). These include cultivated areas with simple geometric boundaries quite unlike the intricate patterns of natural landscapes, as well as cities and towns that may exhibit locally regular spatial structures.
New kinds of surface geopatterns appear as we explore landscapes that are new or are on unfamiliar scales. For example, advances in sonar and other underwater imaging techniques have enabled us to visualize underwater landscapes as if from an airplane, revealing spatial patterns on the seafloor that often appear to be scaled-up cousins to their terrestrial counterparts (Figure 2.8).
Microscopic imaging has also revealed patterns of mineral dissolution at micron scales that are similar to those developed on the scale of landscapes. Measurement of molecular biological signatures has revealed the details of spatial patterns, recognized for decades, in the distribution of biota as a function of depth and position in soils and sediments (see also Section 2.4). Satellite images of other bodies in the solar system also reveal geopatterns that show familiar forms developed in unusual materials and under conditions different from those on Earth.
Most of the natural geopatterns we see are self-organized—they emerge spontaneously from local interactions as opposed to being imposed by some outside influence. Patterns that arise from the internal dynamics of landscapes are autogenic, a term that encompasses both spatial and temporal variation. Most geopatterns, regardless of scale, are also dynamic—they develop over time and in many cases remain dynamic even when statistically in steady state. Natural geopatterns are also resilient. A kind of “geomorphic natural selection” weeds out unstable patterns and favors those that are resilient in the face of natural forces and fluctuations (see also Section 2.7). Where do these patterns come from? What do they tell us? How can we use them? As old as these questions are, new observational and analytical methods are now available to yield deeper answers to them.
Pattern formation and self-organization have emerged as interdisciplinary sciences in their own right. Fractal geometry, the study of structures and patterns with non-integer dimensions, arose from an attempt to measure the length of the coastline of Britain. Fractals and their relatives continue to facilitate our understanding of the spatial structure of landscapes. The emergence of complex systems and pattern formation as research areas in physics, mathematics, and other sciences has brought new attention to geopatterns from these communities.
Consider the evenly spaced valleys in locations as disparate as the humid Appalachian Mountains of Pennsylvania, the Coast Ranges of semiarid central California, and the steep flanks of arid Death Valley, southern California (Figure 2.9). How did these valleys form, and
what processes control their spacing? Valley spacing appears to be a fundamental, emergent signal that tells us something about the types and rates of erosional and depositional processes that shape the landscape, despite varied climatic settings and geological foundations. A successful theory of landscape evolution must be able to explain such fundamental signals and their variability in space and time. Prior to the availability of light detection and ranging (lidar), however, measuring valley spacing accurately was difficult because of vegetation cover and the relatively low resolution of topographic data (Box 2.1).
Landscape geopatterns provide a template for a broad spectrum of processes on and around the land surface. The subdiscipline of landscape ecology is devoted to the study of spatial patterns in ecosystems, including ties to those in the subsurface (see also Section 2.6). Landslides and river floods, with their economic and human costs, are strongly conditioned by the self-organized geometry of drainage basins and the hydraulic characteristics of near-surface rocks and soils. Intensively patterned soils (vertical profiles and spatial catenas of soil properties), as well, reflect geochemical and physical segregation strongly conditioned by microbial processes (see Section 2.4). Overall, the distribution and intensity of a range of landscape processes, such as erosion and deposition or nutrient uptake and release, are tied to spatial structures in ways that we are only beginning to recognize and understand.
That geopatterns are found throughout the landscape is reason enough to want to know how they arise and what they mean, but there are also pragmatic reasons for studying them. Geopatterns record information about present and past conditions and structure the behavior and response to changes in landscapes and their linked ecosystems. The locations of landslide hazards and river and coastal erosion, for example, are closely tied to spatial patterns in the landscape. New evidence suggests that this is also true of less obvious processes, such as soil biochemistry and nutrient cycling. As we strive to manage natural systems better—to work with nature rather than against it—we ought to understand how natural systems organize themselves and how the resultant patterns may contribute to their resilience (see also Section 2.7).
New observational methods and tools have improved our ability to measure familiar geopatterns, revealed new geopatterns on scales and in places that previously could not be observed (including other planets) and, in some cases, have allowed us to measure rates of pattern evolution. Recent breakthroughs in computational spatial analysis, for instance, using geographic information systems (GIS) to study connectivity, distance-decay functions, shape analysis, edge roughness, and association between patterns show great promise. The need also exists to build stronger linkages between spatial analysis and GIS in the study of Earth surface processes. In addition, advanced tools in modeling and statistical analysis provide powerful new ways to quantify and extract information from the spatial structures of landscapes (see Appendix C). Modeling changes in channel morphology from upstream to downstream can also yield information to understand what controls the rates of change. Below, major opportunities for research in geopatterns are summarized under three broad questions.
How Do Local Interactions Give Rise to Extensive, Organized Landscape Patterns?
The study of pattern formation has already developed into a subdiscipline of its own. Approaches to the origin of landscape patterns span the full range of methods: from classical perturbation and normal-mode analyses to a variety of numerical models including generic evolution equations, coupled systems of partial differential equations, and cellular (for example, coupled map lattice) models. The combination of these new approaches, inexpensive computing, and more rigorous quantitative training of researchers has improved our capabilities for modeling surface geopatterns. Models can now generate plausible representations of most types of landscape patterns, often from relatively simple interactions. However, only in a few cases do we have what might be termed a “standard model”; many of the models proposed so far have enough free parameters to make them difficult to test. Research opportunities include advancing model development, with an eye toward making testable predictions, and field and/or experimental campaigns to discover the critical interactions and to test and constrain models.
What Does Spatial Organization Tell Us About Underlying Processes?
Modeling geopatterns leads naturally to the question of what we can infer from them about underlying processes. For example, we associate tributary (collection) channel patterns with net erosion and distributary (dispersive) patterns with net deposition (Box 2.2). The rapid development of new observational tools such as lidar (Box 2.1) and high-resolution, three-dimensional seismic reflection (Appendix C) provides unusual opportunities to answer new questions. Are aspects of tributary network patterns sensitive to rate and spatial distribution of uplift? Can distributary and shoreline patterns provide quantitative constraints on the relative importance of rivers, waves, and tides in shaping coastlines? What can soil geopatterns tell us about chemical fluxes? Answering these questions would require measurement of local physical, geochemical, and biotic processes and attributes in addition to extant and preserved topography. Thus, relating spatial variation of smaller-scale (hard-to-measure) quantities to larger landscape patterns would be important to reveal underlying explanations. Exploration of these variations using field and laboratory studies is rapidly expanding. An exciting development has been the discovery of process “hot spots”—areas where a high level of activity concentrated in a small location can be identified from relatively simple morphologic measures. Another promising approach is to use topographically based estimates to provide field scientists with a set of reference values for key local variables that serve as a starting template for observation. Both efforts are in their infancy but point to the enormous potential for understanding the Earth surface environment through discovering the quantitative connections between surface patterns and processes.
How Can We Use Landscape Patterns to Improve Prediction?
We are currently faced with the problem of forecasting the response of the Earth surface system to changes that include climate, land use, and sea level. The scale and complexity of the system require that we use every means at our disposal to get traction on the problem,
realizing that in most cases our predictions, like weather forecasts, will include estimates of uncertainty, reflecting the inherently stochastic nature of the underlying processes. Thus, as we gain a better understanding of the origin and dynamics of patterns on Earth’s surface, we must also learn to use them for prediction. For example, a major effort is under way in hydrology to use statistical approaches for the prediction of variables such as streamflow in ungauged basins for which only topography and rainfall data are available. This effort reflects the difficulties in modeling the rainfall-runoff process mechanistically from first principles (see also Section 2.5). Exploiting the pattern of the drainage basin is a key to this effort.
The successful use of patterns for prediction requires that two issues be addressed. First, because any persistent natural pattern must be resilient to some extent in the face of disturbance, the origin and limits of this resilience to change must be addressed (see also Section 2.7). For example, as surface systems are altered by a range of human impacts (Section 2.8), will naturally organized structures be distorted or disrupted and, if so, at what thresholds, and how can these thresholds help us to predict future change? Additionally, prediction of landscape patterns requires expansion into the subsurface (see also Section 2.1). The traditional question of how two-dimensional surface patterns are recorded as three-dimensional subsurface structures to infer three-dimensional rock properties is still important for understanding the flow of subsurface fluids. New opportunities arise from understanding how “fossilized” surface patterns can help us to better understand and predict surface evolution in the present.
HOW DO LANDSCAPES INFLUENCE AND RECORD CLIMATE AND TECTONICS?
Tectonic plates are hard and dense, and evolve slowly; by comparison, the atmosphere is fluid, thin, and quick to change. One might think that their interaction would be limited to things such as the well-known effects of terrain on local climate. In fact, one of the major advances in the Earth sciences of the last two decades has been to show that the connections between the climate and tectonic systems are far deeper and more subtle than this. The issues include such interdisciplinary concerns as climate variability, the evolution of landscapes and the biological communities (including humans) they host, and the deformation that builds mountains and basins. These diverse processes interact with one another on time scales ranging from millions of years to individual storm and flood events that last hours or weeks, and seismically triggered mass wasting that lasts for seconds but affects the landscape for years to millennia.
Rugged mountain topography reflects the most intense interplay between tectonic uplift and erosional removal of rock material (Figure 2.10). Rates and patterns of deformation in active convergent mountain belts are strongly influenced by the breakdown of rocks and their
removal by hillslope collapse, river incision, gullying, and other erosion processes whose actions occur over relatively small scales but cumulatively control the erosion of entire mountains.
The rapid erosion in response to snowmelt and intense rainfall in rugged terrain and the deposition of sediment in neighboring alluvial lowlands in response to flooding are familiar occurrences to those living in mountainous areas. Fault movement, earthquakes, and erosion are also familiar to those living in active tectonic areas. Less well-known is the recent discovery that the long-term rate and pattern of erosion influence the rate and spatial distribution of tectonic motions, both at Earth’s surface and in the interior. This recent insight fosters an entirely new set of research questions that can bring together climate scientists, geophysicists, structural geologists, and Earth surface scientists in an effort to
understand how processes acting at Earth’s surface both influence, and are influenced by, processes within the solid Earth.
The study of Earth’s deep interior and much of traditional Earth science cannot be decoupled from study of Earth’s surface dynamics. Moreover, the interplay of climate and tectonics produces landforms and deposits that encode the history of climatic and tectonic conditions—a record of essential scientific and societal value. Fundamentally, surface dynamics influences tectonics because topography dictates, in part, the state of stress within the Earth’s interior. As a consequence, erosional and depositional modification of topography may produce a direct deformational response. Numerical and physical experiments have shown that a concentration of erosion in one area may trigger the inception of new faults or reactivate old ones. Deposition can also suppress deformation on frontal thrusts and raise base level, thereby reducing erosion rates in the interior of the range. Because rates and patterns of erosion may be climatically controlled, deformation within Earth’s lithosphere may also be climatically controlled. Rapid uplift of rocks along New Zealand’s Alpine fault, for example, may be ascribed largely to intense precipitation (in excess of 10 meters per year) driven by orographic lifting of moist trade winds impinging on the western flank of the Southern Alps (Figure 2.11; see also Figure 1.2). If precipitation and erosion were distributed more evenly across this mountain range, the rate of vertical motion along the Alpine fault would be a fraction of what we observe today. Biota may also directly or indirectly affect deep-Earth processes and tectonics (Section 2.6) either by altering precipitation patterns (and thereby erosion rates) or by influencing rates of weathering and surface erosion directly.
Some of the most intriguing research questions in the interaction of climate and tectonics center on the relative sensitivity and rates of the numerous feedback mechanisms among climate, topography, ecosystems, physical and chemical denudation, and the deformation of rocks in active, convergent mountain belts. Complex feedback mechanisms involve, for example, orographic enhancement of precipitation, evolution of the temperature field within the deforming interior of the Earth, and the strength and viscosity of rocks. Rapid erosion can significantly increase geothermal gradients, causing elevated temperatures within the crust. These elevated temperatures can weaken rocks, focusing and accelerating deformation, which may in turn increase topographic relief and thus increase chemical and physical erosion rates through a variety of mechanisms, forming a positive feedback loop. Away from active orogenic zones, the lowlands also provide an interesting record of climatic and tectonic events (Box 2.3).
Chemical weathering and biogeochemical cycles influence, and are influenced by, biological processes and physical erosion. Recent findings show that (up to some limit) the faster the physical erosion rates, the faster are the chemical erosion rates. Chemical erosion, in this sense, is loss of mass due to chemical dissolution. This coupling to physical erosion rates establishes an important link to rates of tectonic deformation and
Lowland Landscapes and the Record of Climate and Tectonics
The Earth’s lowlands constitute an area that includes most of the modern landscape and the most heavily populated regions of the continents. In these regions, erosion rates have slowed over hundreds of millions years as landscapes have been reduced to generally low elevations with long hillslopes, mantled with relatively deep soils, saprolites, and sedimentary covers. This record of weathering and sedimentation on hillslopes and in lakes and caves across eastern North America, Amazonia, Australia, and Africa is being rapidly unraveled with a combination of geochemical studies; cosmogenic isotopes; detailed geochronology with K-Ar, 40Ar/39Ar, 18O profiles, and paleomagnetism, and modeling of the simultaneous evolution of the chemistry and physics of regolith evolution and transport. The work reveals a rich history of climatic and tectonic events and of recent human occupation. In some parts of these lowland environments, erosion rates by water and wind are high because of the occurrence of sparsely vegetated, erodible sedimentary covers left behind by glaciers and meltwater rivers during glaciation. In others, the low rate of erosion leaves rivers starved of sediment and flowing on bedrock. In regions that have been affected by human colonization and intensive land use during the past several centuries, some of the lowland rivers have been transformed from sediment starved to alluviated (see also Section 2.9). Understanding how these patterns are being created by tectonism, climate change, and sea-level variation over the Late Cenozoic era and the degree to which they have been disrupted by human intervention poses scientific questions that link geological studies of landscape evolution with the need for predictions of the future of landscapes under human and other natural influences (see also Section 2.8) .
These continental-scale lowlands also are drained by the world’s largest rivers, the vast alluvial lowlands and deltas which host hundreds of millions of people who depend on both the life-sustaining benefits of the floodplain soil and water resources and the environmental hazards that arise from the rivers’ hydrologic and geomorphic behavior. Although the generally quiescent tectonic framework of these cratonic regions is the basis for their low topography, low rates of erosion, and thick soil and sedimentary covers, they are nevertheless subject to subtle flexure and faulting that have important consequences for the largest rivers and deltas where small alterations of elevation and gradient affect sediment transport, river courses, and the evolution of deltaic wetlands. Subtle influences of this kind, along with the aforementioned human disruptions pose refractory problems for the management of sustainable valley-floor and coastal regions in most of the major river valleys of the world. Interdisciplinary studies of Earth surface processes provide a geographically and temporally extensive perspective on these problems, which currently are being tackled by local engineering strategies alone. This perspective provides a critical context in which to more fully examine, understand, and address those problems.
establishes a critical negative feedback in the global climate system. Chemical weathering of silicate rocks directly influences global climate by consuming atmospheric CO2 (see also Section 2.4). Thus, the correlation of chemical weathering with physical erosion establishes one potential mechanism by which tectonics may influence global climate: rapid rock uplift enhances physical erosion, leading to greater chemical weathering and
a drawdown of atmospheric CO2 resulting in a cooler climate. The exact causes and strength of this linkage remain topics of debate. Other linkages between Earth surface processes and global climate that require study involve the sequestration of carbon in the sediments of foreland basins and deltas, the potential for reestablishing carbon storage in soils in the vast cratonic regions that are managed by humans for agriculture and forestry, and the potential for decreases in the carbon reservoir in boreal regions if warming of the soils combined with melt-driven hydrologic and geomorphic disruption causes drainage and oxidation, and methane release.
The linkages among surface processes, tectonics, climate, and landscapes operate at geologic time scales (tens of thousands to millions of years), but have practical implications because they influence the distribution of active faults and their rates of displacement. The record of paleo-earthquakes and deformation patterns preserved in landforms and surficial deposits has been recognized as an essential complement to monitoring active deformation using repeat geodetic surveys, global positioning systems and seismometer networks, and satellite imagery (for example, interferometric synthetic aperture radar [InSAR]) (see Appendix C). These records are particularly useful because historical and instrumental records are too short to fully gauge the seismic hazard potential of earthquakes with great destructive potential and recurrence intervals from 100 to 10,000 years. Research is also being conducted to understand and use the continuous changes in the pattern of strain accommodation along individual faults. Commonly observed differences between slip rates determined over geologic and decadal time scales are driving investigation into the mechanics and history of earthquake clustering, fault growth, and interactions between nearby faults. The only information available for this history is that recorded in landforms and sedimentary deposits (see also Section 2.1).
The availability of new tools and the development of new approaches to the study of this important record of fault activity have opened exciting new opportunities. Airborne and ground-based lidar instruments (Box 2.1; Appendix C) now allow rapid mapping of topography at a spatial resolution and spatial coverage not even imagined a decade ago, greatly facilitating quantitative study of deformed landscapes. Much of this work focuses on the use of slowly eroding landscape features as passive strain markers. High-resolution topographic data are also unlocking a complementary record of tectonic displacements and climatic conditions preserved in the form of actively eroding landscapes (Figure 2.12). New methods of geochronology, including cosmogenic nuclides (Box 1.2), optically stimulated luminescence (OSL), and probabilistic analyses of carbon-14 (14C) data, also afford unprecedented resolution of the age of deposits and landforms, the timing of events, and the rates of deformation and erosion (see also Appendix C).
Several particularly promising lines of research concern the interaction of climate and tectonics. A well-developed understanding exists of the behavior, over geologic time scales, of model systems characterized by simple rules for rock deformation, steady plate
convergence, subduction of mantle lithosphere, and surface erosion, which is assumed to vary with mean annual rainfall and regional topographic gradient. Although observations are generally consistent with these simple model predictions, we have not yet quantitatively tested the predictions in natural systems—testing these models remains an important research objective and opportunity. In addition, the next generation of models will have to account more completely for the dynamics of the lithosphere and mantle, the complex rheologies of rocks, oblique convergence, the role of sedimentation on the flanks of mountain ranges, and a more quantitative and complete representation of climate relative to its interactions with topography, ecology, and erosional processes and rates.
Four research opportunities emerge as particularly promising toward advancing our understanding of the linkages among climate, surface dynamics, and tectonics: (1) quantification of the role of climate in surface processes; (2) influence of mountain building and surface processes on climate; (3) sedimentation and mountain building (see also Section 2.1); and (4) interactions of surface processes, climate, tectonics, and mantle dynamics.
Quantification of the Role of Climate in Surface Processes
Climate is expected to exert a powerful influence on landforms, weathering, and erosion. However, much remains unknown regarding the causes and effects of these influences, and in some instances, data fail to demonstrate the expected dependences. Thus, accurate, quantitative predictions about the relationships among climate, weathering, topography, erosion, and the rate of sediment delivery to depositional basins are difficult to make. This knowledge gap persists in part because of the complex interplay among the relevant processes and the erosion rate, but it also reflects a dearth of data collected systematically to investigate these relationships. Thus, a high-priority research question remains: How does climate modulate the relationship between topography and erosion rate?
The assumption that erosion rate increases with precipitation (or runoff ) is used in most models linking climate, tectonics, and erosion, yet this assumption is not well supported with field data or theory. The relationship between climate and erosion transcends long-term geologic and geodynamic questions. Understanding this relationship is essential to (1) understand the role of climate and climate variability in the form and function of Earth’s surface in general; (2) interpret the intensity of tectonic activity and, therefore, seismic hazard; and (3) predict Earth surface response to climate change.
The availability of new tools (for example, OSL, cosmogenic nuclides, isotopic indicators of material sources, flow paths, and residence times; e.g., Box 1.2) creates an opportunity to develop millennial-scale datasets appropriate to quantify the modulation of weathering and erosion by climate. As highlighted in Section 2.1, the sedimentological and isotopic records of paleoclimatic conditions and particulate and chemical fluxes into depositional basins are also valuable information resources to address this research opportunity. Ultimately these datasets can be used to guide and test explicit theories for how climate influences erosion, transport, and deposition processes (see also Section 2.5).
Influence of Mountain Building and Surface Processes on Climate
Surface processes influence local and global climate in numerous but poorly understood ways through their influence on topography, land cover (ecosystems, albedo), soil composition, and soil moisture content. At the largest scale, the rise of mountains alters large-scale atmospheric circulation and its complex interactions with the oceans; mountain belts are both barriers and deflectors and influence the pattern of heat exchange among the Earth, the oceans, and the atmosphere. In addition to affecting circulation patterns and driving orographic precipitation, the rise of major mountain belts can increase climate variability of the atmosphere and oceans over thousands to tens of thousands of kilometers downstream, both through lee-side planetary wave responses (relatively well understood) and through the effects of rugged topography, and complex patterns of albedo and surface energy and
water fluxes (less well understood). The influence of land surface-atmosphere interactions (energy and moisture exchanges modulated by topography, biological communities, and soil hydrology) also operates on a more local scale and across all landscapes (not just in the mountains). In addition, oceanic feedback to the atmosphere plays an essential role in climatic response to changes in topography, land cover, and orbital forcing. Ocean feedbacks to changes in large-scale landscapes and climate are also central in determining the impact of sea-level rise on coastal landscapes and understanding geological and paleoclimate records.
Although widely discussed, these feedback mechanisms between landscapes, the ecosystems they host, and climate remain poorly quantified. Indeed, although seen in some models, it has proven difficult to establish from observations whether a significant net feedback of the land surface on climate exists. New modeling tools such as coupled global and regional dynamic vegetation and climate models developed in the past decade, datasets, and statistical analysis methods show promise for resolving this issue. A recent progress report on the U.S. Climate Change Science Program (NRC, 2007b) concludes that inadequate progress has been made on the problem of the potential feedbacks among land-use and land-cover changes, ecosystems, and climate. Most research to date has focused on Earth surface and ecosystem response to climate change. An important opportunity and a challenge lie in the study of the coevolution of climate and Earth surface conditions. Two examples include the carbon budget of permafrost regions and the climate impact of dust aerosols. As areas of permafrost thaw in response to global warming, vast stores of carbon and methane may be released into the atmosphere, establishing a positive feedback (see also Sections 2.4 and 2.7). Dust aerosols have a significant impact on radiative forcing of the climate and on the health of humans and animals. Dust aerosol concentrations depend on soil composition, soil moisture, topography, and land cover in the source region, which in turn depend on current and past physical, chemical, and biological processes over dryland. Dryland ecology is especially vulnerable for projected climatic drying and increasing land use through activities such as grazing. Degradation of dryland ecosystems could further reinforce climatic drying and increase dust aerosols globally.
Sedimentation and Mountain Building
Sedimentation on the flanks of actively deforming mountain ranges significantly influences the rate and pattern of active faulting and erosion rates. Few models, however, have evaluated the role of sedimentation in the evolution of mountain ranges. Insight about this interaction will come in part from analysis of the sedimentary record itself. Sedimentation may generate an extensive record of the history of erosion rates and deformation patterns that can be studied in outcrop and, importantly, imaged seismically in the subsurface (see also Section 2.1). In recent decades, research into the interaction of climate and tectonics has focused almost exclusively on the erosion and exhumation record; new tools have recently
Thermochronometer techniques rely on radioisotopic dating to constrain the thermal histories of rocks and minerals. In the last 5 to 10 years, thermochronometer applications have broadened to include processes at Earth’s surface. This breakthrough in Earth surface process-related applications, or the “thermochronometer renaissance,” is the result of two developments: (1) new thermochronometer techniques and the refinement of existing techniques such as (U-Th)/He and 4He/3He thermochronometry now enable the reconstruction of near-surface thermal histories of crustal rock materials (temperatures less than ~90º C) in ways that were not previously possible; (2) recent applications of detrital thermochronology from sediments provide new perspectives on spatial and temporal variations in catchment erosion.
Thermochronometer techniques are widely used to quantify landscape denudation rates and magnitudes over time scales of ~104 to 108 years. The sensitivity of apatite (U-Th)/He data to shallow crustal depths has attracted new interest because subsurface temperatures at, and cooling ages from, these depths are influenced by the shape of the overlying topography. A distribution of cooling ages collected across a mountain range can then be used to infer how the paleotopography and topographic relief appeared at the time of rock sample cooling in that landscape. Recent development of apatite 4He/3He thermochronometry now allows reconstruction of rock cooling histories between ~80º and 20º C. The added information of the cooling history, rather than just the cooling age from the (U-Th)/He technique, provides powerful constraints on the tempo of near-surface denudation. These techniques and extensions of them can, for example, be used to quantify the timing and magnitude of glacial erosion or river incision in a landscape and to measure rates of weathering processes in stable cratonic settings.
Exciting opportunities for applications of thermochronology in the future include (1) addressing catchment erosion processes with detrital thermochronometer samples from river sediments, moraines, hillslopes, and alluvial fans to quantify the original elevation of a sediment source; the distribution of erosion from different geomorphic processes; catchment-wide denudation rates; forest fire frequency and magnitude; or temporal variations in denudation; (2) application of (U-Th)/He and 4He/3He thermochronology to mineral phases sensitive to very low, near-surface temperature histories to date weathering processes and/or reconstruct climate change; and (3) high-density sampling of bedrock across mountain ranges to reconstruct paleotopography over million-year time scales.
become available that offer an opportunity for new studies of the sedimentary record of mountain building. These tools include cosmogenic nuclides (Box 1.2), low-temperature thermochronology and detrital thermochronology (Box 2.4), and three-dimensional imaging of subsurface sedimentary structures (Appendix C). Three-dimensional imaging of subsurface horizons allows extraction of a time sequence of high-resolution paleo-land surfaces—a direct record of landscape evolution and a resource that has hardly been tapped to date (see also Section 2.1).
Interactions Between Surface Processes, Climate, Tectonics, and Mantle Dynamics
Study of the interactions between climate, topography, erosion, and mantle dynamics is an exciting, novel area of interdisciplinary research involving atmospheric scientists, geodynamicists, geophysicists, and Earth surface scientists. This research is at the frontier of problems in mantle tomography, modeling of mantle convection, lithospheric geodynamics, orographic precipitation, and tectonic, topographic, and climatic controls on erosion and deposition. Geophysicists can now image the current buoyancy structure of the mantle at moderate resolution. This information can be used to seed models of mantle convection that can then be run either forward or backward in time to predict, or retrodict, motion in the mantle over space and time. These models can also be used to generate time series of expected patterns of rock uplift and subsidence associated with mantle convection. An exciting frontier of research lies in the coupling of surface process models to mantle convection models to explore potential feedback mechanisms: Can climate, through the agencies of physical and chemical denudation, sediment transport, and deposition influence convection in the mantle? Equally important will be observational studies to test model predictions and refine models.
Opportunities for Interdisciplinary Collaboration
New opportunities for advancing the theory of weathering and erosion are arising through nascent interactions among atmospheric scientists and Earth surface scientists who traditionally have worked independently. One interdisciplinary research need in the area of interacting landscapes, climate, and tectonics is for a more sophisticated understanding of the controls on the frequency, intensity, and duration of rainfall at the regional scale as a function of local topography, land cover, and global climate. Meeting this need is a major challenge for atmospheric science and is vital to quantitative predictions of the influence of climate on landscapes, ecosystem functioning, erosion, sediment production, transport, and deposition.
At a larger scale, what might be the connections between landscapes and climate due to the topographic and land-cover influences on atmospheric conditions and circulation patterns? How do we build coevolving topography, landscape, and tectonics models? These and other questions will become accessible through direct collaborations among climate scientists and Earth surface process scientists. Insights and tools from several other grand challenges (for example, see Sections 2.1, 2.4, 2.5, and 2.6) will also be required to do so.
Recent research progress sets the stage for a transformation in scientific understanding of the impact of climate and climate change on the Earth’s surface. Hydrologists are making significant progress on predicting soil moisture patterns and runoff in response to known rainfall inputs. Collaboration between hydrologists and other Earth surface scientists is beginning to shed light on the coevolution of topography, hydrology, and biological com-
munities (see Section 2.6). Independently, atmospheric scientists and ecologists are making significant progress on quantifying the influence of land surface-atmosphere interactions (energy and moisture exchanges modulated by topography and biological communities) on regional climate (NRC, 2007b). However, further progress is required on studies of the feedback mechanisms between, and the coevolution of, the Earth’s surface, its ecosystems, and climate. In addition, developing a theoretical understanding of spatial and temporal patterns of rainfall and how these patterns are controlled by regional climate, topography, and land cover remains a formidable challenge. For example, although climate models can predict differences in mean annual precipitation between the Cascades, the Rocky Mountains, and the Himalayas, neither sufficient data on the differences in storm rainfall statistics between these sites nor theories or models capable of explaining them exist. Computational initiatives in regional climate modeling, wind and wave energy fields, orographic precipitation, and landscape evolution (for example, the Community Surface Dynamics Modeling System [CSDMS]3) have begun to address some of these issues. Orographic and land surface effects on rainfall are particularly challenging problems, as is predicting seasonal dynamics of rainfall in mountain regions; these effects are important but traditionally have received little attention in climate science.
HOW DOES THE BIOGEOCHEMICAL REACTOR OF THE EARTH’S SURFACE RESPOND TO AND SHAPE LANDSCAPES FROM LOCAL TO GLOBAL SCALES?
If Earth’s surface consisted solely of outcropping bedrock, the surface area available to anchor and nourish life would be a small fraction of what we see today. Instead, in the near-surface environment, rocks disaggregate into particles and react to form new minerals during the formation of soils, which retain some of the solutes for time scales that support plant growth and other biogeochemical processes. The weathering and eroding landscape varies both chemically and physically over space with strong patterns that reflect topography, lithology, biota, and climate; these changes occur over time in ways that we cannot yet predict quantitatively. Importantly, such bedrock weathering processes contribute to landscape evolution, influence biogeochemical fluxes, and impact regional climate. As landscapes evolve, biota play an active role in retaining some of the soluble elements, serving to anchor existing soil on hillsides and to accelerate soil formation. The breakdown of bedrock—a major factor in Earth surface processes—is among the least understood of the important geological processes.
The transformation of rock into small and large particles is driven by stresses associated with tectonics, chemical reactions, topography, salt and ice growth, mineral swelling, biotic activity, and thermal fluctuations. This transformation process can occur in an instant,
such as when a tree, rooted into underlying rock, suddenly falls over, pulling up roots and underlying rock. The destruction can also be slow as waters percolate gradually through the bedrock and progressively leach mobile elements to create openings. A comprehensive theory of how these processes interact and how the rates of chemical weathering and soil formation are related to environmental conditions is lacking. The near-surface fracturing of rock and the transformation of rock into particles are key steps in converting mechanically strong, low-permeability bedrock into an erodible, hydrologically active material. Particles derived from rock become the soil covers of hillslopes and the sediments of rivers, floodplains, deltas, and beaches. Thus, the process involved in transforming rock into soil influences all aspects of Earth’s surface dynamics.
Soil formation is not, however, only of academic interest. Our food comes from plants grown in soil. The rapid rate of soil erosion due to land use relative to the slow rate of transformation of rock into soil endangers soil resources worldwide (Box 1.1). The fate of soils, the base of agriculture, is of great concern.
The chemical weathering of rocks and soils affects climate, the chemistry of groundwater and rivers, the strength of rocks and erodibility of landscapes, the availability of nutrients in soils, the fate of anthropogenic contaminants, and the properties of the ecosystems that cover Earth’s surface. Even the long-term evolution of the atmosphere is driven by these biogeochemical processes, coupling the fluid envelopes to the solid Earth in feedback loops that, in turn, govern climate. Soil, Earth, and life scientists of all backgrounds are crossing disciplinary boundaries to develop a more integrated picture of Earth’s surface down to and including altered bedrock. This new aspect of research on Earth surface processes emphasizes the biogeochemical cycles of all the elements, at all depths, at human-impacted and pristine localities, over all relevant time scales.
Perhaps the most significant of these biogeochemical reactions involves carbon. Throughout the history of Earth, CO2 emitted to the atmosphere has combined with water to return to Earth as carbonic acid (H2CO3) in precipitation. This acid dissolves minerals and then travels in rivers to the sea as bicarbonate, where it precipitates with cations as sediment. Hence, this chemical reaction removes CO2 from the atmosphere and, on geologic time scales, plays a major role in mediating atmospheric levels of CO2 and consequently climate. This series of chemical processes links volcanic activity (releasing CO2 to the atmosphere), mountain building and chemical weathering (more fresh rock and faster CO2 removal from the atmosphere), and climate (temperature linked to concentration of CO2).
These processes can proceed in the absence of life, but on Earth, biotic processes strongly interact with the carbon cycle. The litter and decomposed organic matter of soils contains about four times more carbon than in the entire terrestrial biosphere (Sposito, 2008). Litter and soil carbon can be sequestered through accumulation or burial for thousands of years only to be rapidly released back to the atmosphere through disturbances
such as agriculture, fire, or thawing of permafrost. Hence, the soil reservoir and vegetation influence short-term carbon dynamics. Considerable research is under way on the dynamics of this soil reservoir of carbon, particularly in the light of how it will change with changing climate and land use. It is a challenging problem that requires understanding the interactions of such factors as vegetation, insect infestation, hydrology, fire, deforestation, agricultural practices, and feedbacks with global warming. Current models disagree on whether soils will be a source or a sink of carbon over the next 100 years, and such soil carbon behavior will be affected drastically by land management practices. The inevitable release of methane associated with the thawing of frozen ground in northern latitudes is also of special concern (see Section 2.7).
Just as we cannot accurately model the carbon cycle, we similarly do not have the tools to project the fluxes of the major nutrients nitrogen and phosphorus as a function of climate, land use, lithology, or other variables. Little research has focused on the effects of different crops on the nitrogen-to-phosphorus ratios in runoff. The implications of widespread increases in nitrogen fluxes in precipitation to landscapes far removed from the source of the nitrogen are also unknown. Furthermore, biogeochemical cycling within the Earth’s surface is strongly affected by rapid invasions of species as well as slower ecosystem responses to changes in climate or land use. How do ecosystem changes, often driven by climate change, interact with nutrient availability?
Of further interest is the fact that more than 30 bioessential elements are also extracted from solid Earth materials by biota, often biogeochemically coupled with the cycling of carbon, nitrogen, and phosphorus. For example, molybdenum acts as the most important catalytic center in fixing gaseous nitrogen into the nitrogen species used by biota. Other micronutrients play important roles in the health of ecosystems and in human beings in general. It has been suggested that an increase in the availability of micronutrients might improve human health globally.
Understanding the biogeochemical cycling of elements needed by biota in the present and the past will allow better understanding of how much human intervention through landcover change will affect the natural balance of biogeochemical cycles (see also Sections 2.6 and 2.8). New technological advances that allow measurement of the isotopic signature of many of these biogeochemical cycles are revolutionizing our understanding of how metal nutrients are cycled by biota. Changes in metal content of soils and sediments over geologic time are documented in the isotopic signatures of elements. Studies of the mobility of minor and trace elements in the modern and ancient Critical Zone may illuminate responses to environmental and climate change.
The prediction of biogeochemical evolution of water from its arrival as precipitation striking the canopy of vegetation through its reaction with soil and rock to its discharge via rivers to the sea is one of the great challenges in Earth surface processes (see also Chapter 1).
Critical Zone Observatories and Critical Zone Exploration Network
The Critical Zone Observatories (CZOs) are natural laboratories built around watersheds or groups of watersheds that are investigated by interdisciplinary teams of National Science Foundation (NSF)-funded ecologists, geochemists, geologists, geomorphologists, hydrologists, and soil scientists using field, laboratory, and modeling approaches. Based on the initiatives put forward in a document “Frontiers in Exploration of the Critical Zone” (Brantley et al., 2006) as well as discussions with the Consortium of Universities for Advancement of Hydrologic Science, Inc. (CUAHSI),a the first three CZOs were funded from the Division of Earth Sciences (EAR) within the NSF Geosciences Directorate. These three CZOs received five-year grants beginning in 2007 to investigate chemical, physical, and biological processes within the natural laboratories. The funding is being used to emplace instrumentation, collect data, and use models to answer forefront questions at the crossroads of hydrology, geomorphology, soil science, geophysics, ecology, and geochemistry. Implicit to the goals of the CZO program is recognition of the contribution that Earth scientists make toward understanding the Critical Zone.
The three CZOs represent sites that vary in some attributes; for example, two sites lie on granitic material (one in the Sierra foothills of California and one in Colorado) and one on shale (in Pennsylvania). Furthermore, each of the three CZOs is designed around a different set of questions: the Southern Sierra CZO is studying how a change from snow- to rain-dominated climate is impacting hydrology and biogeochemistry in the CZO; the Boulder Creek CZO is studying the influence of differing erosion histories on Critical Zone architecture and function; the Susquehanna-Shale Hills CZO is measuring water, energy, and solute budgets from the water table through the atmospheric boundary layer with particular focus on rates of regolith formation. Three new CZOs have been funded in 2009.
Although each CZO is different, integrative modeling efforts are going on at both the intra-site and the cross-site levels. For example, at the Sierra CZO, models are addressing how the snow-rain transition affects the Critical Zone across an elevation gradient. Similarly, researchers at all three CZOs are using cosmogenic nuclides and models to quantify the rates and mechanisms of the bedrock-to-regolith transformation. As part of this latter initiative, data from the CZOs are being compared to data from seed sites within the Critical Zone Exploration Network (CZEN) as well as to the six satellite sites associated with the Pennsylvania CZO to investigate how variations in temperature and precipitation affect regolith formation
Aspects of this challenge have been addressed—though as yet incompletely. The Critical Zone Observatories (CZOs) will add considerably to this knowledge (Box 2.5).
Water chemistry can be monitored intensively, but the chemical and material properties of the soil and bedrock that influence this chemistry remain technically challenging to document, especially at larger spatial scales. Models for evolution of soil chemistry have been developed but are limited by our lack of knowledge with respect to (1) the kinetics of chemical reactions in nature; (2) fundamental thermodynamic equilibrium constants;
on shale and granite. Projects that span all three CZOs also use fluorescence spectroscopy to characterize dissolved organic matter in surface waters and new instrumentation to quantify water fluxes through efforts facilitated by CUAHSI.
To coordinate the CZOs, the group holds an annual all-CZO meeting. In addition, occasional workshops are held. For example, a small workshop promoted by the Boulder CZO led to the ongoing effort to characterize dissolved organic matter in the CZOs with monthly samples. In addition, a National Steering Committee has been constituted with members from hydrology, geomorphology, and geochemistry drawn from academia and government agencies. The CZOs encourage outside collaborators to use the facilities with funding available in existing NSF programs. Such collaborations are initiated by contacting the principal investigator of each CZO to use instrumentation and data. To foster collaborations, each CZO seeks to establish infrastructure, data sharing, and models that attract researchers in areas of interest. For example, the rich sets of hydrological, geochemical, geomorphological, and ecological data that are collected at each CZO generally do not yet include many of the new isotopic tracers although use of these methods is anticipated in future research at the observatories. Furthermore, the rapid growth in molecular biological and hydrological sensor techniques also represents an opportunity for geobiologists and hydrologists to build on the CZO efforts. CZO investigators are developing mechanisms to share data for the sites online. In addition, Critical Zone scientists in Europe, Australia, and China are developing collaborative projects with CZO scientists as observatories are created abroad.
The CZOs were established separate from the Long Term Ecological Research (LTER) Network and the emerging National Ecological Observatory Network (NEON) in recognition that questions in Earth science related to the Critical Zone are often best addressed at sites chosen to study specific Earth surface processes. LTER sites provide detailed geochemical and biological observations, but often provide less focused geochemical, hydrological, and geophysical data compared to the CZOs (see also Chapter 3). Other proposed networks are under discussion through efforts of the initiative known as the WATERS Network (WATer and Environmental Research Systems Network), currently pursued by the science community and the Engineering; Geosciences; and the Social, Behavioral, and Economic Sciences Directorates in partnership at NSF.
(3) the influence of biota; (4) the coupling between chemical and physical processes; and (5) the reactivities of materials ranging from fractured bedrock to nanoparticles.
Numerical models have also been developed that couple hydrology to simplified chemical reactions. However, such models have not yielded mechanistic (quantitative and process-based) predictions of first-order observations such as the discharge-dependent concentration of specific elements in river runoff—an achievable goal. The problem also remains of predicting the pathways and residence times of water in the subsurface en route to the channel and the chemical evolution of stream water as it interacts with its bed,
floodplain, and biota. Figure 2.13 summarizes the various scales of these processes. Water in a stream (2.13A) integrates waters that have traveled through soil and bedrock (2.13B) dissolving and exchanging elements with the evolving soil and weathered bedrock. Within the soil and rock, mineralogical changes document the effects of chemical reactions at the scale of hand samples, mineral grains (2.13C and D), and the mineral interface (2.13E).
These chemical and physical processes produce a weathered residue that is typically strongly structured vertically across the landscape. Soils of varying thickness and properties typically overlie weathered bedrock. Soils are distinguished by physical and chemical differentiation into distinct layers or horizons—all soil classification systems are based on such horizons. In environments of little physical erosion (typically flat surfaces), soils progressively develop and the changes can be used to assign relative ages of landscapes. On sloping landscapes where soils are eroded and transported, soil properties may become essentially time independent. Greater understanding of soil-forming processes is needed to predict the vertical structure of soil and the time evolution of its properties. Such understanding may even prove useful for planetary studies. For example, some researchers now believe that Mars passed through an early, relatively wet phase of clay formation followed by drying and salt accumulation. Future missions to Mars will examine soil profiles for clues about Mars climate history and the prospect of past life.
The development of cosmogenic nuclide methods (Box 1.2) for calculating lowering rates (denudation) of landscapes has led to new observations relating chemical and physical erosion to climate and tectonics. Early results have been surprising and have profound implications. In many cases, researchers have detected little of the expected climatic (temperature and precipitation) influence on chemical weathering rates. Instead several studies have found that rates of chemical erosion (weathering) vary directly with the rate of physical erosion: the greater the physical erosion rate, the greater is the chemical erosion rate. This finding is also detected in data from river monitoring of sediment and dissolved load. One inference is that more rapid physical erosion removes weathered detritus, exposing fresher surfaces for faster chemical reactions, and that this influence is stronger than that of temperature and water flux. Since physical erosion rates are observed to increase with uplift rate, these findings have important implications for modeling the coevolution of tectonics, topography, and climate (see also Section 2.3). These intriguing surveys are now generating investigations to produce mechanistic models.
The transformation of rock to erodible debris and the shedding of mass through dissolution by subsurface flow influence many processes. Advances in monitoring technologies, geochemical analyses and models, shallow subsurface geophysics, geobiology, nanogeoscience, and rock mechanics will help move the field of Earth surface processes from correlation to explanation and from mapping to prediction.
WHAT ARE THE TRANSPORT LAWS THAT GOVERN THE EVOLUTION OF THE EARTH’S SURFACE?
Do the shapes and organized patterns of landscapes reveal the processes that formed them (Figure 2.14)? Are particular landscapes more likely to change rapidly as the Earth warms and as human activity expands across Earth’s surface? Complementing the research challenges posed in other sections of this chapter for these kinds of fundamental questions (see also, for example, Sections 2.1 and 2.8) is the need for a mechanistic understanding of processes that link climate, hydrology, geology, biota, land use, topography, and erosion rates. To tackle this challenge we need to discover, quantify, test, and apply laws that define the rates of processes shaping the Earth’s surface.
All fields of study search for rules or laws that relate cause and effect, action and reaction, or process and form. In the sciences we typically seek mathematical expressions that can be tested and applied widely. In the physical sciences these commonly are the constitutive relationships that form the foundation of the prediction of deformation, flow, and transport of various materials. Some disciplines have made great strides in this endeavor and can focus, therefore, on the richness of behavior that these laws predict. Such predictions can be as simple as the oscillation of a pendulum or can require linking many laws to explain a complex interacting system, as in global climate modeling or ocean circulation simulations.
A simplified set of mathematical expressions for mass conservation applicable to landscape evolution modeling spotlights the critical need for laws in Earth surface processes. Landscape evolution involves the coupled evolution of two layers (Figure 2.15): a bedrock layer with thickness b, which is the vertical height of the bedrock surface above some datum (usually sea level), overlain by a layer of transportable debris (soil and/or sediment) with vertical thickness h. The elevation of the land surface above the datum is z = b + h. The coupled mass conservation equations are
The evolution of the land surface depends on the thickness change of the bedrock with time (t) above the datum and the thickness change of the soil-sediment mantle. Bedrock thickness (b) is a balance between U, the rate of bedrock uplift (or, if negative, subsidence),
and E, the rate of conversion of bedrock to transportable sediment. The transportable debris (soil or sediment) thickness (h) is a balance between the conversion rate of bedrock to transportable debris and the spatial gradient of the volume flux of transportable debris, qs (a vector quantity). The parameters ρr and ρs are the bulk densities of rock and soil or sediment (Figure 2.15). For simplicity, we neglect chemical processes and assume that conversion of bedrock to sediment is strictly mechanical.
This equation is equally applicable to the practical problem of predicting the production and routing of sediment through watersheds (although the uplift term is usually ignored). In this compact form, the influences of climate, geology, land use, and biota are embedded in the E (bedrock erosion) and qs (sediment transport) terms.
To solve these equations, we need to know or be able to predict or retrodict the rate of uplift or subsidence of the landscape and to have mathematical expressions for bedrock erosion (or conversion of bedrock to soil) and sediment transport and deposition (Figure 2.16; Box 1.1). Comparison with any real landscape also requires that the first mass conservation equation be integrated over some time period; therefore, the documentation of landscape evolution and the study of landscape history are essential for validating attempts to explain landscape evolution quantitatively. Here, the focus is on the erosion and transport processes.
Developing Mechanistic, Field-Tested Mathematical Expressions
At present we lack mechanistic, field-tested mathematical expressions for most of the processes that shape Earth’s surface (Figure 2.16). This gap represents a great opportunity and an enormous need for future research. In the Earth sciences, this research challenge is similar to the need to develop friction laws for faults or temperature- and pressure-dependent viscosity models for mantle convection and glacier flow. The large knowledge gap in Earth surface processes is particularly challenging because it requires dealing with complex materials (for example, soil, organic matter, and bedrock) that deform, flow, and change properties over a wide range of scales of time (seconds to tens of thousands of years) and space (microns to kilometers) as various forces and transport processes (for example, water, wind, ice, and biota) act upon them (see also Sections 2.3 and 2.4). Some models from data collected at field sites are beginning to show the implications of various transport laws on the form of the land surface, taking into account features such as bioturbation and soil thickness variations (Figure 2.17).
The mass conservation equation has no explicit time scale, so the erosion and sediment transport laws could be written to apply either to instantaneous movement or for longer time periods during which short-term variations—such as individual storms, dissolution, or particle motions—are averaged. The latter are particularly appropriate when the landforms of interest, such as drainage basins and mountain ranges, evolve over tens of thousands
to millions of years. Transport laws applicable at these longer time scales, or geomorphic transport laws, should have parameters that are physically meaningful and can be evaluated with field measurements. Ideally, such parameters would allow us to conduct scaled experiments in the laboratory or to apply transport laws in new settings, such as Mars or Titan.
Significant progress has been made recently in developing and applying geomorphic transport laws for soil transport and river incision into bedrock. These laws are used to explore many important issues, including the relationships among tectonics, climate, and
topography; the response time of landscapes to changes in uplift; and the processes that control the heights and gradients of hillslopes. These transport laws now are being revised and expanded to include the effects of climate, grain size, and sediment supply, and discoveries made with these laws are being applied in hydrology, geotechnical engineering, and geophysics.
In contrast to these examples of progress, we still lack geomorphic transport laws for such fundamental processes as landsliding, overland flow erosion, glacial erosion, chemical erosion, wind erosion, and transport and deposition of flocculated mud. No laws exist to explain or predict the erosion of bedrock-dominated landscapes or what controls the size of sediment shed from hillslopes into rivers. The breakdown of bedrock into erodible debris, the first step in hillslope erosion, is also poorly understood. Recent studies suggest that as landscapes erode, they create strongly curved surfaces that generate sufficient force to crack bedrock, making it more likely to be eroded (Figure 2.18). Hence, bedrock strength decay, topography, and erosion might coevolve.
A particular mathematical difficulty arises when models are developed to incorporate physical, chemical, and biological processes because the thermodynamic and kinetic equations describing chemical and biological reactions often are highly nonlinear and operate over vastly different time scales. Furthermore, the rates of geochemical and biochemical reactions appear to differ between laboratory and field settings, and we therefore lack confidence in extrapolating geochemical kinetics laws from one system to another. As a consequence, only a few models incorporate sediment transport, geochemical, and biochemical transformations (see also Section 2.4 for discussion of the role of chemical weathering and connections to biogeochemical cycles).
Although it represents a difficult problem, developing transport laws for these processes is now possible. New dating methods enable us to determine rates of processes and their spatial variation across landscapes. Reactive transport codes prove useful in extracting the laboratory-field discrepancy for extrapolating geochemical kinetics. High-resolution topographic surveys made possible from airborne and ground-based lidar surveys can resolve landforms over large areas at sufficiently fine scale to link process and form mechanistically. NSF has also supported NCALM to provide research-grade topographic data, to advance the technology, and to provide education and training for students (see also Box 2.1). Innovative field instrumentation enables us to monitor processes directly and motivate and test conceptual models (Figure 2.19). Physical modeling also exploits this new instrumentation.
The founding of the NSF-supported CSDMS (Section 2.3) will encourage the development of sediment transport law theory and its incorporation into landscape evolution models. New models exploiting parallel processing and supercomputers may permit retention of fine-scale process laws over applications on large spatial and temporal scales. The
NSF funding of three Critical Zone Observatories (CZOs; Box 2.5), which are dedicated in part to observing and modeling the evolution of soil and weathered bedrock, will advance significantly our understanding of weathering and landscape evolution. Furthermore, advances in various geophysical tools are enabling us to monitor from space and explore the subsurface in entirely new ways (Box 2.6).
Rate Laws Explicitly Accounting for Geology, Climate, Biota, and Land Use
A challenge to all geomorphic law formulations is how to account explicitly for geology (bedrock type), climate, biota, and land use. This list is surprising because
qualitatively, even introductory textbooks recognize specific roles that rock properties, climate conditions, and biotic activity play in Earth surface processes and landscape evolution. We lack, however, metrics that quantitatively link landscape form to these controls. Furthermore, few studies define how specific aspects of bedrock properties, climate, or biota should be included in transport laws. A small number of pioneering models have been proposed that include these major controls. In all parts of this chapter, the development of transport laws is important, especially transport and erosion expressions that apply at shorter time scales. Although not highlighted further here, the role of bedrock in influencing hydrologic, ecologic, and geomorphic processes has not yet been put on a mechanistic footing. Many ecologic and geomorphic processes may require formulations based on short time scales (such as storm events or actions of individual organisms) that then can be applied over long time scales (for landscape evolution modeling) or to land-use analysis. These kinds of processes point to fundamental questions about what characteristics of climate, bedrock, or biota strongly influence processes and how these characteristics can be incorporated into transport laws and modeling. The processes by which bedrock, climate, and biota dictate landforms remain frontier problems in Earth surface processes.
The Importance of Shallow Geophysics
Understanding processes that influence the chemistry, form, and function of landscapes requires quantification of rates of mass transfer and the composition and structure of the Earth’s surface. For example, quantifying the sediment flux down a hillslope and its relation to spatial and temporal variations in vegetation and precipitation is important for improving a hillslope geomorphic transport law and for geochemical models of regolith formation. Equally important is the subsurface structure of landscapes, including the depth to unweathered bedrock, fracture density of the bedrock, and spatial variations in the composition of weathered and unweathered hillslope material. Numerous remote-sensing and ground-based geophysical methods have been developed to quantify these aspects of the near- and subsurface environments (see figure below). Hydrogeophysics, for example, has recently emerged to develop tools to improve subsurface monitoring of bedrock moisture dynamics (Section 2.6). The focus of many shallow surface geophysical studies is on environmental engineering problems, such as contaminant transport in shallow aquifers or satellite-based observations of ground cover. As described in a 2006 CUAHSI review of geophysical instrumentation for watersheds, advancing shallow geophysics tools for the near-surface bedrock environment and making them widely available for research investigations is important (Robinson et al., 2008). The content of this report highlights new and exciting research opportunities for the shallow and remote-sensing geophysics communities to engage in the study of Earth surface processes. The recent formation of a new focus group, Near Surface Geophysics, in the American Geophysical Union attests to some of the new interest in this type of research. Extending these techniques to enable measurement of geochemical and biological properties of natural systems will become increasingly important.
A diverse range of ground and airborne geophysical techniques are applicable to study the evolution of the form, composition, and function of landscapes. Tools commonly used in the environmental engineering communities include ground penetrating radar, high-frequency seismic methods, microgravity, and a range of electrical and magnetic methods (see figure below; also Appendix C). These tools can image variations in subsurface physical properties (such as density, electrical resistivity, and acoustic velocity), all of which relate to variations in soil, ice, or bedrock composition and/or the presence of air or water in pores and fractures. A wide range of ground or airborne remote-sensing techniques is also available, and new techniques are under development. Remote-sensing techniques have proven useful for imaging numerous characteristics relevant to Earth surface processes. Examples include high-resolution imaging of topography, topographic and ice surface change over time, and vegetation; imaging surface velocities of glaciers and mass movements; multispectral imaging of Earth’s entire surface and atmosphere every one to two days; and measurement of spatial and temporal variations in rainfall. Focusing these and other new and developing geophysical techniques toward studying the evolution and operation of Earth’s surface is important.
Interdisciplinary collaboration between Earth surface scientists, geophysicists, and engineers is needed to advance these emerging technologies, provide training, and make such geophysical instruments widely available. The availability of global positioning systems and seismic instruments through the University Navigation Satellite Timing and Ranging (NAVSTAR) Consortium (UNAVCO) and Incorporated Research Institutions for Seismology (IRIS) is one model of what could be done for shallow geophysics. Examples of exciting interdisciplinary research problems in Earth surface processes that can be addressed through geophysical studies include (1) quantifying variations in soil and weathered bedrock thickness and the physical, biological, and
HOW DO ECOSYSTEMS AND LANDSCAPES COEVOLVE?
How different would Earth be if life had never evolved? Will the biotic response to future climate change drive significant changes in landscape processes? Biota break rocks into soil particles, exhaust gases into the atmosphere, and alter runoff chemistry (see also Section 2.4). The movement, growth, addition of organic matter, and dissolution associated with biota make otherwise dense or perennially frozen ground permeable and cause rainfall and snowmelt to enter into the subsurface. Flow of water into the subsurface may reduce flood peaks and later sustain low flows during dry periods. Organisms also weave strength into weak materials, stabilizing hillslopes against erosion, and confining river flow within narrower channels. Life—through digesting, dilating, exhaling, decaying, pushing, and weaving—strongly influences the form and pace of surface erosion and strongly modulates biogeochemical cycling. Hence, life simultaneously affects climate, hydrology, erosion, and topography (Figure 2.20; see also Sections 2.2, 2.3, and 2.4).
Despite our awareness of the role of life in Earth surface processes, we often have only a qualitative appreciation of this connection. As discussed elsewhere in this chapter, however, biota are important and, in several cases, central. Questions such as those posed above call for a mechanistic understanding that can form the foundation for making predictions or performing modeling experiments to explore and discover the richness and importance of these interactions. The incorporation of the biosphere into global climate models, for example, transformed these models, enabling them to address accurately such vital issues as the controls on CO2 buildup. The climate community accomplished this task by model building, monitoring, and large-scale field campaigns to obtain data on controlling mecha-
nisms. A similar effort is needed in the Earth surface processes community in conjunction with the development of some of the research opportunities outlined below.
The quantitative study of the interactions of biotic, physical, and chemical processes in the Earth’s surface, or geobiology, is in its infancy. The most active area of research into the connections of biotic processes with inorganic Earth materials is at the microbial level, and research has focused on discovering the composition and functions of these organisms and their influence on the environment. This effort is a central part of the emerging field of geomicrobiology, and many Earth sciences departments have made recent hires in this area. Microorganisms were the only life form on Earth for most its history, and the influence of microorganisms on the Earth’s geochemical cycles remains profound (see also Section 2.4). Since the earliest stages of Earth’s evolution, methane gases produced by bacteria may have significantly elevated the atmospheric temperature, keeping it warm even when the then-young Sun was 30 percent less luminous. The evolution of some forms of photosynthesis by microorganisms drove up oxygen levels, enabling a vast diversity of life forms to emerge. Interest in the influence of microbes on geochemical cycles has increased with the discovery of water on Mars, the methane lakes of Titan, and numerous extrasolar planets—all of which raise the question of the possibility of life elsewhere. Although research in geobiology has been expanding rapidly, little work has been directed toward the effects of microbial activity on the form, composition, and evolution of the Earth’s surface. Some studies have shown that the surface crust formed by microorganisms in arid landscapes may strongly influence runoff and erosion processes. At present, however, we do not know if microbes affect the shapes of hillslopes or the form of rivers, nor can we quantify how biotic activity influences ecosystem services.
We do know that plants and animals such as trees, worms, gophers, and insects strongly influence surface processes and consequently the form and composition of the surface and ecosystem functioning. In many forested environments, trees root into the underlying bedrock, and if the tree falls, the roots tear the bedrock, converting this rock into soil material, transporting both soil and rock downslope. Through such processes, forests make and move their own soil. Charles Darwin devoted his last book to the action of worms in which he quantified the rate of formation and transport of soil by worms and concluded, “When we behold a wide, turf-covered expanse, we should remember that its smoothness, on which so much of its beauty depends, is mainly due to all the inequalities having been slowly leveled by worms.”4 While soil formation rates and soil transport rates by biota have been quantified in some settings, we lack theory to guide prediction of these process rates and have limited ability to include biota mechanistically in landscape evolution models.
Strong coupling of life and landscapes also occurs in river systems and marshlands.
Recently several research groups have presented field and laboratory data demonstrating that riparian vegetation, while seeking the moisture, nutrients, and light along river valleys, typically adds considerable strength to banks, traps and slows overbank flows, drops clumps of roots and trunks into the river that armor the bank, or can float away and become trapped into jams of debris. Through these actions, vegetation may transform wide channels with multiple bars and flow pathways into single-thread channels or convert the channel into multiple vegetation-lined threads (see also Section 2.2). The vegetation can control the rate of channel shifting and, ultimately, direct the formation of floodplains. So significant is large woody debris in channels to creating habitat that millions of dollars are spent annually adding wood to streams to recover lost ecosystem functions due to past wood removal (see also Section 2.8). In marshlands, vegetation plays a primary role in controlling fluxes, deposition patterns, and channel development. An understanding of this coupling is crucial to predicting the fate of tidal marshlands during the sea-level rise expected with climate warming and designing successful marshland restoration projects (see also Section 2.9). Great theoretical and practical questions now present themselves with respect to how riverine or marshland ecosystems and their channel morphology and dynamics are coupled.
The study of the coupling of ecosystem and hydrologic processes has recently been labeled ecohydrology and is rapidly expanding as a field of research. This coupling is dramatically played out in the water-limited environments of arid landscapes. Arid zone vegetation is often found to be patchy, and various models to explain this have been proposed that argue for codevelopment of vegetation patches, runoff and infiltration, soil moisture, and erosion. Closely coupled to the presence of vegetation, fluxes of water in arid regions also directly influence salt contents of soils, which can control soil permeability. Connections to human activity are also involved. Some have proposed that land use has reduced the crust and vegetation cover of arid lands, leading to more dust erosion, and some of that sediment is deposited in mountain glaciers (as in the Rocky Mountains) leading to lower albedo and greater melting. Only a few recent papers have argued for the use of the word ecogeomorphology, but the idea of expanding these studies to include landscape evolution is established and research is expanding rapidly. One example of such research is the exploration of the coupled ecologic, hydrologic, and geomorphic evolution of hillslopes in landscapes where biologic communities are strongly influenced by hillslope orientation (aspect) and hillslopes display a distinct topographic asymmetry. Such research at the interface of hydrology, geomorphology, ecology, and geochemistry is at its inception, but shows great promise.
The coupling of life and landscapes may seem least direct when considering entire ecosystems but, in fact, may be most crucial to understanding the large-scale evolution of topography. As land masses collide and elevate rocks, life may be more than a passive passenger as mountains build and are eroded. Biotic processes directly affect the form and rate of erosion. Bedrock breakdown, soil development, and hillslope erosion—all driven or
mediated by biota—strongly influence the size and composition of sediment entering rivers (see also Section 2.4). Sediment load and particle size impose constraints on river slopes, and steeper channels will tend to lead to higher mountains. Vegetation extracts moisture from the ground and exhausts it into the atmosphere, cooling the air, and nourishing clouds that later return the moisture to the ground as rain and snow. This exchange process affects the global heat balance—hence atmospheric circulation and the spatial pattern of precipitation. This pattern in turn may direct both chemical and physical erosion rates and, on mountains, affect the spatial pattern of unloading. The pattern of erosional unloading directs the evolution of mountain systems, and by these connections, biota may influence the shape, height, symmetry, and surface chemical composition of mountain ranges. To explore quantitatively the evolution of landscapes and climate (see also Section 2.3), the effects of biota need to be specifically and mechanistically included in models (Figure 2.21).
Recent developments indicate that we now are ready to make significant advances in understanding the coevolution of landscapes, life, and ecosystems. In addition to the emergence of the fields of geobiology, ecohydrology, and ecogeomorphology, new efforts are under way in field observatories that explicitly address linkages between biota, Earth surface
National Center for Earth-surface Dynamics
NCED is an NSF-funded Science and Technology Center (STC) created to catalyze development of an integrated, predictive science to examine processes that shape the Earth’s surface and, specifically, to predict the coupled dynamics and evolution of landscapes and their ecosystems (http://www.nced.umn.edu/). NCED research is organized into three integrated projects: (1) Desktop watersheds seeks to exploit the spatial structure imposed by tributary channel networks, expressed by high-resolution topography, to provide static and dynamic predictions of local physical, geochemical, and ecosystem properties. (2) Subsurface architecture uses information from modern systems, experiments, and stratigraphic records to develop a predictive understanding of delta evolution and apply this understanding to delta restoration. The work will also improve prediction of variations in porosity and permeability that control the flow and accumulation of water, oil, and gas in the subsurface. (3) Stream restoration addresses the scientific basis for this multibillion-dollar activity in the United States through a combination of research and training developed in coordination with agency, industry, and academic partners including the U.S. Geological Survey, U.S. Army Corps of Engineers, Environmental Protection Agency, U.S. Department of Agriculture, U.S. Bureau of Reclamation, and Bureau of Land Management. The goal is to move restoration practice away from reliance on analogy to an analytical, process-based approach.
Knowledge transfer, education, and diversity are all integrated into NCED’s research programs. Knowledge transfer includes exchange and engagement with the broader research community via workshops, working groups, a visitor’s program, short courses, and postdoctoral researchers. NCED’s education program uses the familiarity and aesthetic appeal of landscapes to engage a broad spectrum of learners in NCED science. The centerpiece of the NCED education program has been collaboration between the Science Museum of Minnesota (SMM) and NCED’s academic institutions and applied partners. One outcome has been the successful EarthScapes exhibit (http://www.nced.umn.edu/Earthscapes.html) at SMM. A major new traveling exhibit initiative, H2O: Water = Life, developed by SMM and the American Museum of Natural History in New York, along with three STCs, is also helping the public learn about Earth surface processes (http://www.amnh.org/exhibitions/water/). The NCED-SMM collaboration led to the NSF-funded Future Earth Initiative (http://www.smm.org/exhibitservices/history/futureEarth/), which will serve as a center for informal education activities on human influence on the environment. NCED’s diversity program addresses the mismatch between the current spectrum of participants in environmental science and the U.S. population overall. Working with Ojibwe tribal elders, NCED has developed environmental camps that use innovative, culturally sensitive programming to excite Ojibwe children about environmental sciences and encourage them to excel in school and pursue science-related careers. NCED also has a vigorous recruiting program for minority participants in its research program, and minority participation increased from 8 percent when the program started to 17 percent in 2008.
processes, and landforms. The National Center for Earth-surface Dynamics (NCED) was established in 2002 to advance a predictive Earth surface dynamics science and specifically seeks to apply this predictive capability to ecosystem and land-use management (Box 2.7). Three CZOs were initiated by NSF (see Box 2.5) to quantify and predict the interactive processes occurring in the zone between the vegetation canopy and the underlying groundwater table (with emphasis on chemical weathering, biogeochemical cycling, soil formation, and hydrology), and the CZO network is growing to provide field sites for study along environmental gradients. Explanation of weathering and soil formation patterns, biogeochemical cycling, vegetation-mediated hydrologic dynamics, and landscape evolution at these observatories will require collaborations across many disciplines. Twenty-six Long Term Ecological Research (LTER) network sites also conduct ecological research that integrates ecologic with geologic, hydrologic, and atmospheric sciences (see also Section 2.4 and Chapter 4). Research questions at several LTERs specifically focus on the biological-physical feedbacks that shape landscapes at decadal or longer time scales. The proposed National Ecological Observatory Network (NEON) has identified 20 potential sites (about 225 km2 each) across the United States to monitor ecosystems and their drivers for the next 30 years. In addition, as many as 60 shorter-term monitoring sites may be established. Although the coupling to Earth surface processes and landscape form, composition, and evolution is not explicitly a goal of NEON, the data gathered will provide a wealth of information for such endeavors. An important opportunity for critical interdisciplinary advances may be missed, however, if biologists focus on NEON sites, Earth scientists focus on CZOs, and hydrologists develop their own observatories. In May 2008, the Meeting of Young Researchers in Earth Sciences (MYRES) held an international conference on the “Dynamic Interactions of Life and Its Landscape” attended by more than 60 participants (see also Preface). These current graduate students and recent doctoral degree recipients reported on laboratory, field, and numerical modeling studies that are beginning to lay the foundation for this emerging field.
WHAT CONTROLS LANDSCAPE RESILIENCE TO CHANGE?
Geological and modern records of sediments and landforms indicate that Earth’s surface changes under the influence of external drivers such as climate change, tectonics, volcanism, and human activities, as well as the internal dynamics of erosion and deposition acting over long periods of time (Figure 2.22). Some changes are gradual; others are abrupt or rapid, ranging over time scales from the seconds in which a landslide occurs to the millennia over which glaciers move across a mountain range and dramatically change the landscape. As used in this report, “abrupt” change refers to a response that is nonlinear and much faster than the forcing, while “rapid” change refers to a response that can be linearly related to forcing.
The shape of landforms and their rates of evolution fluctuate within ranges reflecting the stochastic nature of the processes and materials that drive their operation. A singlethread meandering river channel will fluctuate in width, depth, or sinuosity as it is perturbed by floods, by fluctuations in sediment supply, and by its own bend-lengthening and bend-cutoff processes. Channel cutoffs may occur and form lakes in a floodplain, yet the meandering channel can retain a remarkably similar geometry for centuries, even as the channel migrates across its floodplain.
Under other circumstances, changes in external drivers and internal dynamics of landforms intersect to impose a new set of processes on the landscape. If riparian and floodplain vegetation is thinned by drought, the floodplain may become erodible by overbank flows, the bank strength may be reduced, and a multithread, less sinuous channel may develop with frequent avulsions, a steeper gradient, higher sediment transport rate, and a higher rate of channel and floodplain change. Such a transition is often labeled “channel instability,” but it represents an enduring change in the probability distributions of channel forms and operations, under an altered set of processes, driven ultimately by a change in climate and vegetation in this example.
Landscapes generally fluctuate within some predictable range in response to variation in external drivers and interaction among internal processes as long as the processes that
alter it remain below some level of intensity. When conditions change with sufficient magnitude and duration, landscapes may become altered beyond the range within which they can recover. The driving agents have then exceeded the landscape’s resilience to change, resulting in change to another form with a different intensity of erosion, sedimentation, or dominant morphogenetic process. Through such changes, the forms of landscapes can themselves experience sudden changes, such as from extensive and smooth, aggradational features (alluvial fans and valley fills) to sharp-edged, incising forms (steep escarpments with landslide scars, dense valley networks, or gullies) (Box 2.8). The material-transporting systems of Earth’s surface can also be intensified or even reorganized so that, for example, a river-dominated coastline with a copious sediment supply can have its sediment regime transformed into one of net sediment loss when a delta lobe switches position or river sediment is impounded by dams. The resilience of surface features to change may be defined (in either observed or conceptual terms) by the magnitude of the disturbance that they can absorb before changing form or behavior.
Progress has been made in the past two decades in developing numerical models that simulate erosion and sediment transport processes, driven by climate and land-use changes. The models predict spatial and temporal patterns of landform evolution, sediment supply to river channel networks, and the conditions required for switching between periods of sediment accumulation and removal. More recently, emphasis has been placed on applying these models to real river basins in regions for which it is possible to provide time control on rates of erosion and sequences of sedimentation. The models are driven by long-term weather records. Some of the models simulate how vegetation cover is also driven by climate changes. The model results can then be compared to sedimentary records of climate change to understand, for example, the range and magnitude of conditions that a landscape has experienced over time.
This section highlights the state of scientific knowledge related to rapid and abrupt change on Earth’s surface and why some landscapes undergo a “change of state” whereby their form may change dramatically or different morphogenetic processes may become dominant. The section also highlights the needs and opportunities for investigating controls on the resilience of landscapes to change (see also Section 2.9).
Some areas of Earth’s surface are more vulnerable than others to changes in state. Polar, glacial, and periglacial regions are currently nearing, or are in, a change of state and are predicted to continue to do so with persistent global warming. Paleoclimate records show temperature swings between glacial and interglacial periods that are several times stronger in the polar regions than in the tropics because of greater sensitivity to orbital change and snow and ice feedbacks. Changes in the state of the landscape also pertain to areas in which concentrations of human population impose increasing pressures on Earth’s surface environments (see also Section 2.8). The goals and challenges of research in Earth surface processes include identifying thresholds of change, understanding the environmental processes most
The Future of Landslide Prediction
The future of landslide hazard research lies in making predictions about when, where, and how big the landslides will be, and in developing procedures that convey warnings and risk to the public so that loss of life and damage to infrastructure or ecosystems are reduced. Such effort is critical, and particularly important in anticipating the effects of climate change and land use in areas prone to slope instability (see figures below). Land use, especially road construction and vegetation clearing (including industrial timber harvesting), has led to great increases in landslide frequency. The underlying causes for this are well understood and landslide theory and empiricism can be used to guide management decisions.
Landslides occur in sizes that vary from a sand pile avalanche of a few centimeters to entire landscapes involving many cubic kilometers of earth material. They can move slowly, creeping along in wet years. They can remain stable for decades or millennia, then collapse suddenly and travel at meters per second. Entirely new landslides may occur where no previous evidence of landsliding exits. The risk and management response differs with scale and speed. This variability of scale and speed, while challenging to anticipate, is not random. Types of landslides generally take on certain patterns that can be described statistically (see Section 2.2).
At present much can be done to delineate where landslide hazards are greatest by mapping historical occurrences of landslides and quantifying correlations between occurrence and probable controlling factors (e.g., hillslope gradient, geology, and vegetation). These correlations can be applied across large areas to identify where landslides are likely to occur in the future. Such correlations have improved in the past 20 years with the availability of digital spatial information for attributes such as topography, geology, vegetation, rainfall, soil depth, and other factors. High-resolution topographic data obtained from airborne laser mapping greatly increase our ability to identify and delineate landslides and to predict their future occurrence. Hazard maps based on correlations may have limited ability to predict future events under changing climate and land use, hence fundamental research on landslide mechanics remains essential.
Landslides commonly are caused by elevated water content in the ground, which reduces the strength of materials. Landslides also are caused by ground shaking during earthquakes, and can contribute significantly to the loss of life and destruction of infrastructure. Rainstorms, rapid snow melt, ground ice melting, and simply wetter years can lead to elevated water content and slope destabilization. Various international research groups are exploring the use of precipitation forecasts (from global or regional climate models) to predict when an area may experience landslides. These forecasts are often developed at rather coarse scales and require methods to predict local precipitation. Once the forecast is made, it needs to be translated into landslide likelihood. The simplest approach has been to use empirical relationships, typically relying on analyses of the rainfall intensity and duration at which landslides occur. These predictions can be made more specific by using landslide hazard maps to identify more sensitive parts of the landscape where such landslides are most likely to occur. Real-time monitoring of precipitation, water content, and even ground deformation is being done in some areas to add further information. Advances in fine scale (in space and time) precipitation forecasts and real-time monitoring, coupled with landslide potential mapping, have led to landslide warning systems being tested in various countries.
vulnerable to change, understanding the mechanisms that make some landscapes resilient to change, and anticipating and investigating options for mitigating or even reversing the effects of change.
Studying the impact of abrupt changes on Earth’s surface requires looking into geologic records of their occurrence. As discussed in Section 2.1, studies of Earth surface response to past changes foreshadow potential future change and clarify the evolutionary trends that produced today’s landscapes. Documenting the occurrence and impact of rapid or abrupt climate change events on landscapes is particularly challenging in that it requires high-resolution paleoclimate reconstructions to complement stratigraphic records, as well as documentation of Earth surface response encoded in the physical, chemical, and biotic properties of current landscapes. While new high-resolution marine and ice-core records are continually being collected for paleoclimate research and other purposes (Box 2.9), less work has been done on terrestrial (paleosol) records, which can be more sensitive to geographical and latitudinal variations in climate.
In the coming decades, significant advances will be required in the measurement of past and present rapid landscape change (via field and remote-sensing studies) and in improved predictive models. The tightly coupled response of landscapes to changes in climate, tectonics, biota, and human activity at multiple spatial and temporal scales requires increased emphasis on interdisciplinary research to take advantage of emerging opportunities and tackle the challenges highlighted below.
Disappearing Glaciers and Glaciated Landscapes
Over the past several decades, mountain glaciers and continental ice sheets have retreated rapidly. Accelerated melting of glaciers has contributed to sea-level rise, increased numbers of glacial lakes, glacial-flood outbursts, altered river flows, and changed flooding patterns. Over millennial to million-year time scales, glaciers are powerful geomorphic agents as glacially carved and glacially deposited landscapes attest. Several challenges, highlighted below, face the scientific communities studying ice sheets (for example, Greenland and Antarctica) and mountain glaciers.
Current models of continental ice-sheet dynamics match overall ice-sheet trends observed during the twentieth century, but do not predict the observed rapid marginal thinning currently taking place. Such a deficiency implies that understanding of the mechanisms controlling ice dynamics, particularly with regard to mechanisms that accelerate ice flow as temperature approaches the threshold for melting, is inadequate. Current models could, therefore, be inaccurate in predicting future loss of ice sheets and long-term rates of glacial erosion. This problem is a leading cause of the uncertainties in projected future global sea-level rise; the supply of buoyant freshwater to the North Atlantic; and the resulting risk of abrupt climate change due to altered ocean circulation patterns. The deficiency of
the ice dynamics models is attributed to inadequate understanding, and thus lack of treatment, of subglacial hydrology, basal sliding, subglacial sediment deformation, and ocean-ice interactions at ice-sheet margins (Iverson, 2008). Future efforts to improve coupled ocean-atmosphere-ice sheet models will incorporate the processes summarized above. Parallel efforts in satellite-, ground-, and marine-based studies are also needed to provide data for model evaluation. Additional quantification of chemical and physical processes in periglacial environments could also improve predictability of changing chemical fluxes of carbon and methane as well as physical changes that have impacts on societal infrastructure.
More robust predictions are also needed to evaluate mountain glacier contributions to sea-level rise and how climate change will influence mountain glacier interactions with the hydrologic cycle and alpine natural hazards. Central to these scientific challenges is a need for an improved quantitative understanding of the long-term evolution of alpine landscapes (Owen et al., 2008). The main tasks confronting the community include quantifying processes controlling glacier motion and erosion (such as glacial hydrology and sliding); clarifying past spatial-temporal patterns of mountain glacier fluctuation in relationship to regional climate and ocean-atmospheric circulation; continued development and testing of numerical mountain glacier models and their effects on mountain landscapes over repeated glacial-interglacial cycles; characterizing the role of climate variability on glacier variations; and predicting the distribution, size, and nature of glaciers in the future.
At high latitudes and in high-altitude mountainous environments the regolith and underlying bedrock are permanently frozen except for a seasonally thawed “active layer” near the surface. Global warming is rapidly changing permafrost environments and has contributed to an increase in the active layer thickness resulting in rapid changes in periglacial landscapes and landforms. Permafrost thawing has changed patterns of soil organic carbon and nitrogen content, vegetation, biogeochemical processes, rates of permafrost creep, and mass movements. The impacts of these changes on the form, function, and evolution of periglacial environments and the atmosphere are not well known.
Frozen and unfrozen water is disseminated throughout the regolith, while in other parts ice is highly segregated into layers and “frost wedges,” several meters thick. The low temperatures and evaporation rates at these latitudes maintain the regolith in a moist condition, so that some of the carbon fixed by photosynthesis each year accumulates in soil. The amount of this accumulation has varied since deglaciation, and a considerable amount of field research and modeling has recently been invested in reconstructing the postglacial history of accumulation of this vast carbon reservoir, which currently stores about 14 percent of soil carbon.
The carbon reservoir emits both methane and CO2 to the atmosphere, directly through
the soil surface and by concentrated bubbling through lakes. The future stability of this carbon reservoir in the presence of global warming and whether or not the carbon reservoir will be vented as methane or as CO2 are major uncertainties in climate projections (see Section 2.4). Thus, modeling the history and future of soil carbon storage and linking these permafrost models to climate models is important research and requires more information about the distribution and age of permafrost and its relationship to groundwater and lake hydrology across northern regions. Concern about the future of the permafrost and its carbon reservoir has been raised by the recognition that permafrost in interior Alaska has warmed by approximately 1.5ºC since the 1980s and that temperatures in boreholes on the North Slope of Alaska rose by 2-4ºC during the past century. In addition to being driven directly by the atmospheric radiation budget, permafrost is also thawing around lakes as they warm and along coasts as a result of wave erosion of frozen but unconsolidated sediments as sea level rises and as marginal sea ice melts and no longer buffers wave action.
The landscapes of the northern permafrost regions have largely been created by sediments left behind from glacial and fluvioglacial processes. In areas where the permafrost is continuous or at least widespread, warming is leading to greater seasonal thawing and drainage of groundwater, thereby causing the creation or expansion of lakes as the regolith collapses upon melting. In areas of discontinuous and sporadic permafrost, remote-sensing studies have revealed a decrease in lake area due to the thawing of their substrates and breaching of new outlets. The fate of lakes in subarctic regions has an important influence on the evasion of methane and thus potentially on atmospheric chemistry and the radiation budget. Quantifying the chemical and physical processes active in permafrost environments relies on continued development of physically based, process-oriented models as well as mapping and characterization of permafrost depth and distribution from multiple geophysical, geochemical, and field-based techniques.
Coastlines in Transition
Coastal areas are the dynamic interface between land and ocean. Coastal areas worldwide are experiencing eustatic sea-level rise that is exacerbated in many areas by subsidence of already low-lying land surfaces. The current situation is alarming because of the number of people living in lowlands within 150 kilometers of the coast; this number has been projected to increase from 3.6 billion in 1995 to 6.4 billion, or 75 percent of the world’s population, by 2025.5
Existing estimates of land-sea elevation changes have been made without considering how the effects of subsidence of coastal land areas, altered sediment transport and deposition, or changes in the biosphere may interact with eustatic sea-level rise, which itself varies
around the globe because of the combination of isostatic, tidal, and rotational effects. The combined effects of these processes, in addition to climate change, may amount to several meters of relative land-sea level change over the coming century.
Many coastal zones are especially sensitive to sea-level rise because, with low relief, small vertical changes in sea level affect extensive areas. As sea level rises, low-lying areas are subject to tidal inundation, flooding by storm surge, and increased wave energy due to greater water depth. These processes alter the relationships between sediment deposition and erosion by altering the productivity, distribution, density, and types of vegetation and animals, as well as the physics of the coastal system. Tidal marshes are obvious examples of coastal landscapes in which ecological and physical processes are tightly coupled. Marsh vegetation (type, density, productivity) depends on the relationship between surface elevation and tidal range, while marsh vegetation influences water velocity, sedimentation, and erosion of the marsh surface. How large do changes in hydroperiod or wave energy need to be to alter vegetation density to such an extent that sedimentation and erosion from marsh surfaces are affected? How does increasing tidal inundation due to sea-level rise alter the allocation of plant photosynthate (chemical product of photosynthesis) to roots, thereby changing the rate of marsh elevation increase? How does deposition of wrack (marine vegetation) carried into the marsh by storms affect the composition of the plant community? In some cases, coastal retreat is accelerated; in other places, little change in the coastal landscape is observed. What combination of rate process changes cause tidal marshes to be converted to mudflats or subtidal zones? How will sea-level rise influence the morphodynamics of the large lowland river deltas around the world and of river-dominated coastal zones? Could such effects lead to rapid wetland drowning and abrupt landward shifts in shoreline? Other consequences of sea-level rise include salinization of aquifers and estuaries, and damage to coastal infrastructure and economies, particularly fisheries and agricultural production.
A major challenge facing Earth surface process science is to couple biological and ecological processes with physical processes to produce predictive models of coastal landscapes that consider the biological effects on flow and sediment transport and to understand how coastal ecosystems are altered as function of morphology, flow, and sediment transport. Without such coupling it will be impossible to approximate the behavior of these systems. The Coastal Zone is an area in which interaction among hydrological, geological, and ecological scientists is especially crucial (see also Section 2.6).
Understanding Limits of Landscape Resilience to Change
The conditions that lead to landform transformation are difficult to predict, and few state changes have been well monitored. Comparison of neighboring landscapes affected by differing climate, tectonic, and volcanic events, or alterations of vegetation by anthropogenic
disturbances, provide “natural experiments” that can be exploited through field and modeling studies of the changes. River channels and coastal sedimentary landforms tend to be more sensitive than hillslopes or glaciers to such alterations, and provide some of the most convenient examples for study on the time scale of instrumental and isotopic records.
Characterizing climate in terms that are relevant for process studies at Earth’s surface is also needed. Whereas many mechanistic explanations of these processes are event-based (for example, a rainstorm or windstorm of finite duration, intensity, and extent), tests of climate models are based mostly on mean climatology. The ability of these models to represent extreme weather events, which have strong impacts on landforms, is not clear. The outputs of most climate and weather models are provided on spatial scales that are too coarse to resolve rain rate at the scale important to landforms and are not organized in a manner useful to understanding landforms, despite the introduction of concepts such as geomorphically relevant climatic statistics more than a half-century ago (see also Section 2.3). High-resolution regional climate models (spatial resolutions from a few kilometers to a few meters), coupled with the surface hydrology and vegetation dynamics and ultrahigh-resolution global climate models (spatial resolutions of a few tens of kilometers), are being developed. Progress in incorporating this type of resolution could support the development and testing of predictive, mechanistic descriptors of land surface response to climate change.
HOW WILL EARTH’S SURFACE EVOLVE IN THE “ANTHROPOCENE”?
Humans have changed the face of Earth throughout history and pre-history. Through agriculture and urban development, humans have altered land cover, promoting soil erosion (Box 1.1) and interfering with hydrologic and biologic processes (Figure 2.23). Dams and levees have been built for flood control, water supply, and hydroelectric power, while interrupting sediment movement along rivers and causing loss of habitat and biodiversity. Chemicals have entered water and soil through industrial and agricultural practices. Human activity has also been linked to our warming climate over the past several decades, and this warming, in turn, affects Earth’s surface processes. Humans have changed the surface of the Earth more drastically over the past 50 years than at any time in history. These effects may increase with a growing global population that is projected by 2050 to surpass 9 billion and possibly reach 10.5 billion inhabitants (United Nations, 2009). The environmental impacts of human population growth and the accompanying resource consumption and disposal have become so pervasive that the term “Anthropocene” has emerged in the scientific literature to signify a new geologic era (Crutzen and Stoermer, 2000), with the contemporary global environment dominated by human activity (Steffen et al., 2007; Zalasiewicz et al., 2008). Clearly, even identifying “natural” landscapes on the planet is now difficult (see Box 2.10).
What must we do in order to understand, predict, and respond to rapidly changing
landscapes that are increasingly altered by humans? This question is among the most pressing challenges of our time, and its scientific component falls squarely within the purview of Earth surface scientists. We now need to build conceptual and numerical models that account explicitly for human-process interactions on Earth’s surface, even if such models may have inherent limits relative to our desire for quantification and prediction. Models should strive to define the limits of predictability and the degree to which certain problems are unpredictable. In addition, because theories for Earth’s surface systems are developed largely for natural landscapes and because such landscapes are increasingly rare, we also need new theories based on enhanced understanding of human-landscape interactions. Such knowledge is important in developing the tools needed for adaptive management and for guiding decision making, especially in the face of uncertainty and change. Because much of landscape restoration targets a return to “natural” conditions, clearer understanding of the human modification of
Earth’s surface would also enable Earth surface scientists to inform resource managers when this goal is possible or sustainable at reasonable expense (see also Section 2.9).
Significant advances have been made over the past decades in defining the extent of human impacts on Earth’s surface. Rapid development in tools, particularly lidar and satellite remote sensing as outlined elsewhere in this chapter, has also provided unparalleled opportunity to acquire new data for examining the significance of human alteration of landscapes (see Box 2.4; Appendix C). These data can facilitate development of predictive frameworks for human-landscape systems if knowledge across disciplines also can be integrated with new models and approaches. Recent workshops supported by NSF have facilitated collaboration between geomorphologists, geochemists, and ecologists working on the dynamics of Earth’s surface that include human impacts. They have also enabled the establishment of common areas of research in which natural and social scientists are beginning to integrate social and environmental analysis (e.g., Vajjhala et al., 2007; Nagel et al., 2009). In this context, the following paragraphs highlight three central areas of research that are poised for substantial progress over the coming years and identify the challenges to, and high-priority needs for, making advances possible.
Long-Term Legacy of Human Activity
The dominance of humans in shaping some modern Earth surface environments is clear (Figure 1.3). Over decadal and centennial scales, the impacts of human alteration of Earth’s surfaces, especially changes in land use, will likely exceed those of climate change that has captured global attention. Yet, the expected magnitude and ramifications of these impacts are not well understood or quantified. For example, how do we separate the effects of human influences from those of other, natural processes? How close are human-impacted landscapes to thresholds for ecosystem collapse or the onset of rapid erosion or mass wasting, especially under global climate change (see also Section 2.7)? How do we predict multiple and/or compound effects of human activities? Which regions are particularly vulnerable to change? What might be the impacts of these changes on human society?
Understanding the legacies of human impacts on landscapes remains an urgent research need especially because the information would provide society with options for adaptation. Meeting this need requires accelerated field and remote-sensing studies to identify the key characteristics of landscapes affected, in addition to using historical records, survey maps, and data from paleoenvironmental studies for documenting and gauging landscape change over long periods. Emphasis is needed in sensitive environments prone to rapid change (see also Section 2.7). These include coastal and urban areas where human populations are concentrated, as well as polar, mountain, and arid regions where increasing human impacts pose pressing issues of resource use. Developing predictive models of system response to anthropogenic changes also requires reliable
The Illusion of “Natural” Landscapes
Many landscapes around the world reflect a long history of human presence. This presence is obvious in some cases where rivers, for instance, have been channelized, in constructed environments where hillsides have been intensively terraced (Figure 2.23), or where water-intensive crops have been grown through irrigation in normally arid zones. In other areas, human influences are not as apparent because the initial alterations occurred centuries or even millennia ago (see figures below). For example, catchments of the Pacific Northwest region of the United States contained abundant large wood at the time of European settlement, but so much wood has been removed to reduce flood hazards and improve navigation in rivers that people now tend to perceive wood in rivers as “unnatural” and requiring restoration. In the eastern United States, studies have also shown that before European settlement, small anabranching channels within extensive vegetated wetlands accumulated little sediment but stored substantial organic carbon. However, tens of thousands of water-powered mill dams built in the seventeenth to nineteenth centuries raised the regional base levels of stream channels, inducing sedimentation behind the dams and burying the wetlands. Subsequent breaching of the dams caused channel incision and accelerated bank erosion that resulted in the modern incised meandering streams. The concept that these modern, meandering channel forms represent “natural”, or ideal, conditions has guided theory development in fluvial geomorphology for decades, but has only recently been proven incorrect with investigations of the seventeenth to nineteenth century human influence in these areas. Returning landscapes to their original, natural conditions underlies much of the practice of stream restoration (see Box 2.13). Deciphering what is truly “natural” on an Earth that is not static presents challenging scientific, sociocultural, and policy questions for Earth surface scientists (see also Section 2.9).
quantitative data about human impacts on processes and process rates. Our capability to collect large datasets through surveys (for example, with airborne laser swath mapping [ALSM]) or field monitoring (using networks of wireless devices) will improve our ability to address the long-term human impacts on landscapes over the various time scales involved; however, significant challenges remain in data analysis and interpretation. A need exists for increased collaboration between Earth surface scientists and geospatial scientists with expertise in the application of existing and emerging geospatial and remote sensing technologies. Lack of understanding of the human responses to anthropogenic landscape change also limits development of the integrated models needed by society. This limitation points to the need to strengthen interactions between the natural and human sciences.
Complex Interactions within Anthropogenic Landscapes
Interactions in human-impacted landscapes have been examined recently within a framework for complex environmental systems (Pfirman and the AC-ERE, 2003). This framework recognizes the interconnectedness among Earth’s surface components, the nonlinearity of the relationships between them, the historical conditions and inheritance that relate to local settings, and the range of spatial and temporal scales for examining these systems. For example, although the installation and operation of dams provide many economic benefits, these structures also induce physical, biological, and chemical interactions in rivers that may not have been anticipated (Box 2.11). These interactions often intertwine with social processes, as in the case of the Klamath River of California and Oregon, where dikes and levees have increased water supplies since 1905 but adversely impacted threatened and endangered fish (NRC, 2008). Although management actions undertaken during the dry year of 2001 protected the basin’s declining fish population, they reduced the water available in the system and triggered further physical and biological reactions, as well as social conflicts related to agriculture.
Although singular impacts and responses are known in many cases, multiple stressors, their impacts, and consequent interactions make human-influenced landscapes difficult to model or predict. How do social processes influence these interactions? How will the mutual interactions adjust to global climate change? Which environmental processes are most vulnerable to change? How will these changes affect the sustainability of water and biotic resources? To improve predictive capabilities for these complex interactions, focused interdisciplinary field studies that link multiple processes are needed as well as experimental studies and theoretical modeling (see also Section 2.6). Incorporating decision making and social processes in some of these interactions also requires strengthening ties between the natural and social sciences.
Coupled Human-Landscape Dynamics
Traditionally, Earth and environmental scientists have treated humans as external drivers of change in examining human impacts on the environment. For river landscapes, for example, studies have emphasized adjustments in fluvial forms and processes following deforestation, mining, impoundment by dams, and urban development. Knowledge of how landscapes have changed after human disturbances has been important in formulating theories of landscape change and assisting management. Yet the traditional approach is insufficient to capture the interrelationships and feedbacks that characterize landscapes affected by increasing and multiple human-caused stressors. Examining coupled natural and human systems enables integrated understanding of how humans both influence and are affected by natural patterns and processes (Pfirman and the AC-ERE, 2003).
The reciprocal relationship between natural and human systems may exhibit positive or negative feedbacks. In Kenya, for example, where conversion of forests into cropland has degraded soils and decreased crop yield, greater food insecurity has hastened the conversion of remaining forests to agriculture, illustrating a positive feedback that accelerates environmental degradation. Negative feedbacks often involve modification of human behavior and policy that potentially slows or halts the impact. For example, flow regulation in Cypress Creek, Texas, eliminated the variability in flows and eventually threatened the culturally and aesthetically valuable Bald Cypress tree, prompting local stakeholders to reevaluate and adjust policies for dam releases. Historically, degradation of Earth’s surface by humans has been accompanied by only weak negative feedbacks on human behavior because humans have proven innovative in their ability to endure their own continued alterations to an environment with little regard to its original or “sustainable” function. Weak feedbacks also result where the causes of the impact originate at different scales and faraway places, as in the case of the impacts of global climate change on Arctic residents. Earth surface science plays an important role in public policy by recognizing these diffuse, and often subtle, influences while there is still time to mitigate them. Understanding coupled human-landscape dynamics is among the epochal challenges for predicting the evolution of Earth’s surface in the presence of the growing human population and for developing innovative solutions for environmental management.
The study of coupled natural and human systems has thus far emphasized ecological phenomena, producing an emerging science for social-ecological systems that explicitly incorporates human decisions, cultural institutions, and economic and political systems. Enormous gaps in knowledge remain, however, especially with respect to understanding the impact-feedback loops within geomorphic systems. Social processes influence both the original human impact and the potential responses and feedbacks within landscape systems—in other words, human systems influence whether crops are grown on particular soils or how a city is developed, and human systems need to manage and adapt to the
Constructing and Removing Dams
Of all the human alterations of the Earth’s surface, the construction of dams and the acceleration of erosion through agricultural and other land clearance activity are among those having the most profound effects on river sedimentation and morphology. Nearly two-thirds of the world’s rivers are regulated by dams, levees, and other structures. In the United States alone, approximately 80,000 dams are registered in the National Inventory of Dams, many built in the 1960s. Millions more small impoundments dot the landscape. These structures can collectively store a volume of water equaling almost one year’s runoff. The large ones alter the timing and magnitude of river flows and dissolved and particulate material, as well as the locations and pathways of chemical reactions. The dams also compartmentalize our nation’s rivers, fragment landscapes and aquatic habitats, and create new reservoirs for sediment and carbon. Globally, dams have reduced the flux of sediment reaching the world’s coasts by 1.4 ± 0.3 billion metric tons annually, despite increased quantities of sediment introduced through soil erosion and transported by rivers (2.3 ± 0.6 billion metric tons per year) (see figure below). The diminished transfer of sediment from land by rivers is increasingly important with respect to global denudation and geochemical cycles, the functioning and loss of coastal ecosystems, and the stability and evolution of deltas and depositional environments (see also Box 2.2). Climate change is expected to accelerate and confound these impacts, as melting glaciers (see also Section 2.7) increase the potential need for water storage by dams, and sea-level rise further threatens the viability of coastal living space. Quantifying these potential impacts is hampered by the limited availability of data for sediment loads. Since many reservoirs are now approaching their limits of usefulness because of sedimentation, and as environmental impacts and the associated loss of ecosystem services are increasingly recognized, decommissioning and removing dams and other structures also bring new opportunities for ecosystem restoration (see Section 2.9).
environmental changes, such as accelerated erosion, that may be caused by these activities. Thus, a primary need in understanding the coupled system is the exchange of analytical perspectives between Earth surface scientists and social scientists, including cognitive-behavioral scientists, economists, political scientists, sociologists, and human geographers. Anthropogenic Earth surface changes also need to be quantified in terms of how they, in turn, affect humans directly or indirectly. For example, accelerated soil erosion (Box 1.1) can be quantified in terms of how it affects the water-holding capacity of the soil profile, and thus plant and crop production, and predicted over time scales that can be expected to be valued by humans. These changes can also be translated into the language of ecosystem services, human perceptions and valuation, and the extent to which predictable thresholds exist that would trigger feedback responses to accelerate or decelerate the original impact. Because policy and institutional processes are the key instruments that society has for effect-
ing necessary feedbacks, scientific frameworks for coupled human-landscape systems would also require linkage to policy mechanisms. Successful transmission of scientific information is required to guide policies for mitigation and adaptation, pointing to the need for research findings to be articulated in forms that are useful for policy making, as well as the identification of pathways for transmission of this information to policy makers.
Science is far from developing a general theory of coupled human-natural systems, even though such a theory may offer the potential to slow or reverse environmental degradation and improve human well-being. Because such a theory would include knowledge of societal perceptions of environmental impacts and the ability and willingness of societies to react to these changes, much focused inductive and empirical work is required to investigate these interacting processes in a range of Earth surface environments, and in societies at various levels of institutional development and types of organization.
HOW CAN EARTH SURFACE SCIENCE CONTRIBUTE TOWARD A SUSTAINABLE EARTH SURFACE?
As human activity became widespread and intensive some 5,000 years ago, the surface of Earth was—as it continues to be—altered in many ways (Section 2.8, Figure 1.3). With increasing scientific understanding of the causes and cumulative, long-term effects of human-induced changes, a consensus has emerged that at least some of these disrupted or degraded landscapes can and should be “restored” or “redesigned.” The meaning of the term “restored” is not agreed on with precision, but it generally refers to a set of strategies to reestablish processes that functioned before the most recent and intense anthropogenic disruption. Where this is not possible, a more limited set of natural processes may be reestablished, or a fuller set may be reestablished at a smaller scale than the original case. Restoration is often incorporated into engineering strategies for making landscapes safer for human habitation (such as coastal barrier reinforcement or flood risk reduction).
Landscape restoration may involve river channels and floodplains (river restoration), entire ensembles of hillslopes and small channels (watershed restoration), or coastal wetlands and waterways. Landscape restoration is typically managed by coalitions of engineers, biologists, and planners, but as restoration activities become more sophisticated they will include a broader range of Earth surface scientists. They will also involve social scientists, planners, and practitioners, as in the development of integrated assessment tools to guide decision making. The design of landforms (in the case of rivers, floodplains, and tidal marshes) or land conditions (in the case of watershed restoration) that will function within a predictable, preferred range has had some success. The real challenges emerge in designing landforms that interact with the desired biological processes and communities required to establish sustainable ecosystem complexity (Box 2.12). In fact, biotic processes are central to the evolution of many landforms (see also Section 2.6). Hence, the design and monitoring of restored landscapes and their ecosystems should both rely on and add to emerging models of interacting biological, physical, and chemical processes in landscape evolution. Furthermore, the long-term cost of a project will be minimized and the likelihood of its sustainability will be maximized if the original design lies within the stable range of landscape fluctuation and if it respects long-term trends in landscape evolution.
To this activity, research in Earth surface processes brings knowledge of how major environmental drivers (such as water regime or sediment supply) scale the size and shape of landforms that are treated as the “reference conditions” for restored features. Thus, specific research challenges arise about controls on the size and geometry of redesigned rivers, tidal creeks and inlets, deltas, and beaches and how they are controlled by the expenditure of flow energy and by sediment supplies. In other cases, where landforms originally existed in sediment-starved environments or were dominated by intense biogenic drivers such as large wood or riparian vegetation, little theory exists at the present time for how such landscapes
might be recreated or what the alternative sustainable states are. In these cases, the need is for extensive field investigations to support new theories of the nature and functioning of biogenically reinforced landscapes that would retain sediment and chemicals and release them slowly into water bodies.
Research in Earth surface processes also focuses on the transport processes that redistribute sediment and chemicals, and interact with biotic processes such as seed dispersal, plant survival, and colonization of aquatic substrates by woody plants, algae, and macrophytes. Attempts to modify the sediment and carbon storage in marshlands and soils require new field studies of the accumulation of organic sediments. Where human effects involve the disruption of sediment supplies, restoration entails reducing excessive sediment supplies from disturbed areas using principles of hydrology, geomorphology, plant community regeneration, soil profile enhancement, and engineering. In other cases, perturbations such as the removal of dams or the augmentation of beaches involve increases in sediment supplies to landforms, and the effects of these perturbations need to be predicted through advances in the study of mixed-grain-size sediment transport that may transform or even engulf the landform itself. The nation’s gradually expanding program of dam removal from rivers (now ~50 projects per year; see also Box 2.11) adds urgency to the need for research into sediment management along rivers. In most dam removals, the major problem to be solved is not the removal of the dam or the restoration of flows, but the management of impounded sediment. Strategies for stabilization, removal and disposal, or release of this sediment involve large costs and ecological and societal consequences, and pose research challenges for Earth surface scientists and engineers involved with sediment transport and channel mechanics (Box 1.1).
Earth surface processes research also illuminates the long-term context of restoration plans. Restoration of landscapes requires designs that will survive long-term conditions such as highly stochastic sediment supplies, regional land-use changes, the relaxation of some landscapes from deglaciation, sediment starvation, crustal deformation, and sea-level change. Studies of these phenomena clarify the limitations on restoration and, especially, the role of time in allowing ecosystem complexity to develop. In other situations, the disruption of flows and sediment supplies or the deepening of coastal waters through land surface subsidence renders restoration to a preferred state with desired biological characteristics impossible on time scales of human interest. Research into the Late Cenozoic evolution (the last several million years) of sedimentary landforms and their ecology and biogeochemistry provides a basis for understanding this long-term context and sustainable environmental range of landscape restoration (see also Section 2.1).
At the same time, restoration projects provide field-scale experiments within which the evolution of landforms or soil profiles can be studied with a resolution and degree of simplicity that are rarely possible in nature. Initial conditions are well known and usually simpler than natural landscapes, and processes occur at a natural scale as opposed to miniaturized laboratory conditions. Often, variables such as sediment supplies to beaches,
Coastal Wetland Restoration
Many tidal marshes have been converted to non-wetland habitat by dredge-and-fill activities, nutrient loading, groundwater extraction, contaminated runoff, fire management, and invasive species. Coastal wetlands have declined worldwide by 0.5-1.5 percent per year over the past few decades. In Louisiana, losses in coastal wetlands have risen to ~60 km2 per year or nearly 18 hectares (~45 acres) per day (Barras et al., 2003). Sea-level rise and crustal subsidence also lead to loss of wetlands, but the destruction of wetlands by direct and indirect human activities is of much greater consequence (see also Box 2.2).
The rate at which coastal wetlands continue to be filled and drained belies their ecological and economic importance as wetlands. Wetlands are among the world’s most productive ecosystems. They provide habitat, food, nurseries, and refuge for fish and shellfish. Wetlands store and cleanse water, prevent coastal erosion, and reduce storm surges. In the 1960s, public awareness of the importance of wetlands led to policy changes intended to slow wetland loss by encouraging wetland mitigation and restoration (e.g., Section 404 of the Clean Water Act in 1972).
Coastal wetlands develop at the interface between land and water. Their very existence results from a tenuous, but enduring mass balance of mineral and organic sediment that is imported, exported, and generated in situ in locations that lie approximately at mean sea level and are simultaneously subject to the eustatic rising of sea level, the warping of Earth’s crust as a result of the last deglaciation, tectonics, and sediment loading (often altered by human activity), and anthropogenic effects such as groundwater extraction. Many coastal wetlands have very low topographic gradients (see figure below), so that differences of only several centimeters in salt marsh elevation strongly influence vegetation patterns and ecological and biogeochemical processes.
Restoration of coastal wetland ecosystems requires the land surface to be sufficiently high in the tidal range that vegetation is not flooded constantly, but also low enough that the energy and sediment supply provided by regular flooding is achieved. Vegetation exerts strong controls on water movement and sediment deposition and erosion that shape the marshes and tidal creek networks. Until recently, only the effects of physical processes on biotic processes were considered in restoration of coastal marshes. This approach focused on contouring sites, planting native species, redirecting sediment dispersal at river mouths, and reintroducing tidal flushing. The outcomes of such efforts, however, were natural vegetation patterns, but not functionally diverse marshes (e.g., Tijuana River Estuary, California; Barn Island Wildlife Management Area, Connecticut).
Reestablishing functional, as well as structural, characteristics of coastal wetlands will require fundamental advances in understanding the coupling of ecological and morphodynamic processes at a range of scales (Fagherazzi et al., 2004). Within individual marshes, emphasis needs to be placed on incorporating
base levels of wetlands, and flows to rivers are manipulated in ways that allow quantitative expectations to be developed (Box 2.13). Observations in the field of the effects of real trees, root zones, material properties, and sediment supplies, for example, cannot easily be reproduced indoors. In other cases, animals are excluded from large areas or small plots, or climate is controlled through watering or soil profile warming and fertilization. Field experiments
with replication are often possible within the confines of restoration projects. Research opportunities include the high-resolution testing of predictions of the stable scale or range of some landforms, as well the trajectory of other features that may not equilibrate. Recently developed geochronological and geophysical tools and geochemical tracers enhance the nature and level of detail of studies that are now possible of both the mechanisms and the
What shape and gradient should a river have in order to behave in a more or less predictable manner consistent with human designs? How much water should flow through it? What plants and animals should live in its waters and on its banks? In the United States, at least $1 billion is spent on river restoration annually, and the number of restoration projects is growing exponentially (Bernhardt et al., 2005). The environmental and economic motives for restoration vary with geographic region and history of environmental impact. Restoration practice has outpaced scientific investigation, in that many of the basic scientific questions relevant to stream restoration have yet to be answered or even investigated. Better understanding of the processes that determine river evolution and response to environmental change, for example, could improve our ability to design sustainable river systems that provide ecosystem services that people desire.
The term “river restoration” begs the question: Restore to what state? “Restoring” a river usually means returning it to a close approximation of its condition prior to a disturbance. Yet understanding a river’s past and present states is often complicated, requiring the use of a broad set of tools and disciplinary approaches that include many Earth surface processes. Rivers are dynamic features that adjust to changes in climate, hydrology, vegetation, and human activities. Increasing sediment input to a river—by logging of neighboring forests, for example—may increase erosion of stream banks. Left unchecked, stream bank erosion can cause the river to become wider and shallower and—in severe cases—to migrate erratically and flood more frequently. Wastewater discharge or runoff from agricultural fields can change the chemical load of rivers as well as their habitability for different species. River restoration projects not only take stock of past and current processes that control the long-term stability of streams, but also attempt to predict what modifications will best achieve the project’s goals.
In addition to the scientific framework, rivers are restored within particular constraints of regulatory goals, economic needs, and public preferences that vary geographically. Projects are guided by a variety of federal, state, and local regulations that oversee land use, water quality, and flood control, for example. Regulations also often interact with economic incentives that drive decision making. In one case, after determining that fish passage upgrades and maintenance costs would exceed the value of Oregon’s power-generating Marmot Dam (see figure below), the dam was decommissioned and removed in 2007. The dam breaching released 100,000 m3 of accumulated gravel and sand into the river downstream within the first 48 hours (Major et al., 2008). Although the full consequence of this added sediment load downstream over the coming decades cannot yet be predicted quantitatively, the decade of research that preceded dam removal made accurate predictions of some of the impacts of sediment release and aided the decision-making process for decommissioning and removing the dam.
rates of change to be expected from managed or reconstructed landscapes (Boxes 1.2, 2.1, 2.4, 2.6; Appendix C). Thus, landscape restoration provides an opportunity both for learning and for service by the Earth surface processes research community.
In addition to restoring and redesigning landscapes, society needs Earth surface scientists to guide decisions regarding the functioning and evolution of the surface, while
recognizing uncertainty and gaps in scientific knowledge. In human-dominated regions, for example, landscape-modifying processes are often simplified by engineering and other forms of resource management, but re-naturalizing the processes is not simple. Complex historical cascades of decisions have often been made about rights to resources of land,
water, and minerals and multiple layers of interests have solidified in the status quo. If this history has resulted in the degradation of resources or ecosystem services, or even the viability of human living space, the complexity of human interests creates resistance to unraveling the damage. This difficulty is compounded by the complexity of the physical, chemical, and biological processes themselves, which makes it difficult to predict the results of changing resource utilization or other societal action. Yet, society needs Earth surface scientists to construct models for assessing the probable effects of various possible restoration actions. The models may initially be crude and may require sequential elaboration and refinement. Learning how to develop such assessment tools in a politically sensitive, useful way requires that natural scientists studying Earth’s surface processes collaborate with social scientists who study all of the other processes that affect the surface, as well as with practitioners in industry, engineers, and planners.
Landscape restoration is a complex, high-priority goal for many researchers, practitioners, policy makers, and the public, but only recently have these different communities begun to examine together the legacy of past restoration efforts in the context of new data from specific regions of the Earth’s surface. Earth surface scientists have vital contributions to offer as restoration activities are carried toward quantitative models and predictions. Just as importantly, Earth surface scientists are in position to guide future decisions through inclusion of understanding of human-process interactions, even while recognizing uncertainty and limits of predictability. Such guidance may enhance the goods and services that Earth’s surface provides to society and thereby produce enormous economic benefits, as well as aid in developing a sustainable living surface for the next generation.