2

Water Science and Resources Challenges for the Next 25 Years

The United States and the world face critical issues that put increasing pressure on water resources, including growing populations, climate change, extreme weather, aging water-related infrastructure, and demands for food, energy, and industrial production. These issues can create threats to water quantity and quality and result in greater exposure to hydrologic extremes and hazards, affect economic and policy decisions, and exacerbate the balance of tradeoffs between human and ecological water uses and needs. Scientific and technological advances to help quantify, characterize, understand, and predict water resources will be key to address these issues nationally and globally, helping inform and support balanced water use, management approaches, and access to safe water for people and ecosystems.

The Water Mission Area (WMA) tasked the committee to identify the highest-priority water science and resources questions for the United States over the next 25 years. As it considered this quarter-century perspective, the committee spent considerable time deliberating about whether the pressing water issues of today will still be relevant in 25 years or whether an entirely new set of water science and resources questions will emerge. The committee’s consensus is that the fundamental water issues of today, as well as their societal relevance, will only become more significant and far-reaching as time proceeds.

As part of its information gathering, the committee considered the significant body of work represented in previous National Research Council reports such as Toward a Sustainable and Secure Water Future: A Leadership Role for the U.S. Geological Survey (NRC, 2009), Global Change and

Extreme Hydrology (NRC, 2011), Challenges and Opportunities in the Hydrologic Sciences (NRC, 2012a), and Preparing for the Third Decade of the National Water-Quality Assessment Program (NRC, 2012d). The committee also discussed a wider variety of water science and resources issues with experts and stakeholders from government agencies, academia, and nongovernmental organizations.

On the basis of these resources, coupled with its knowledge and additional background research, the committee identified six cross-cutting water challenges:

• Understanding the role of water in the Earth system;
• Quantifying the water cycle;
• Developing integrated modeling;
• Quantifying change in the socio-hydrological system;
• Securing reliable and sustainable water supplies; and
• Understanding and predicting water-related hazards.

A discussion of each of these challenges is presented below. Another theme that was consistently raised by the spectrum of stakeholders who provided input to the committee was the continuing, rapid expansion of technologies to observe, process, and visualize water-related information. Based on consideration of these six water science and resources challenges and the opportunities provided by technological advances, the committee compiled a set of high-priority science questions to address the challenges. These priority questions are presented in the last section of this chapter.

WATER SCIENCE AND RESOURCES CHALLENGES

Understanding the Role of Water in the Earth System

The water cycle is of central importance in Earth system functions and is a primary driver of energy and biogeochemical cycles (e.g., Eagleson, 1986). As water moves through the atmosphere, the lithosphere, and the biosphere, it facilitates many physical, chemical and biological processes and carries with it latent heat, sediment, nutrients, and carbon that regulate the Earth’s climate system from diurnal to geological time scales (Schlesinger and Bernhardt, 2013; Korenaga et al., 2017). The hydrologic cycle has undergone profound changes throughout Earth’s history, responding to changes in the atmosphere (e.g., radiative budget from orbital and greenhouse gas forcing), the lithosphere (e.g., plate tectonics, orogeny, subsidence, weathering), and the biosphere (e.g., evolution and upland colonization of land plants, the emergence of angiosperms). Changes in the hydrologic cycle feed back to the energy and biogeochemical budgets

of the atmosphere, the lithosphere, and the biosphere through complex interactions (Clark et al., 2015; Good et al., 2015). Understanding how the water cycle responds and feeds back to global change in the past and in the near future, and with increasing human pressure on water and other parts of the global system, remains a key challenge in global change research (Haddeland et al., 2014; Kaushal et al., 2014). It also presents opportunities to transform the knowledge and improve the predictive capabilities of the Earth as a system. This fundamental scientific knowledge is needed for the effective management of water resources.

Quantifying the Water Cycle

Effective management of water resources demands knowledge of where and how much water exists in key stocks (e.g., in glaciers, snowpacks, aquifers, lakes, rivers, soils) and how water moves within and among these different stocks (NRC, 2012a). Quantifying the water cycle is exceedingly difficult because the stocks, flows, and residence times of water vary spatially and temporally, as do the relevant parameters that influence their residence time and location (e.g., soil, land cover, aquifer characteristics, topography, dam and reservoir management, agricultural practices). Accurate measurements of the physical and chemical properties of water are needed, as well as an understanding of the biogeochemical transformations that affect water quality (Michalak, 2016). Land-cover and land-use change, infrastructure such as dams and levees, and climate change intersect with and alter the water cycle, adding more complexity to efforts to quantify it.

Developing Integrated Modeling

Models are essential tools for integrating and synthesizing disparate observations, for elucidating complex interactions and testing hypotheses, and for reconstructing past conditions and predicting future trajectories of co-evolving systems. Models can aid in assessing the potential for rapid environmental change and can help harness vast amounts of data in visually compelling ways that can provide information to decision-makers. Historically, hydrologic simulation has focused on individual components of the hydrologic cycle (e.g., groundwater models, surface water models); consequently, an urgent need exists for improved, integrated models of water systems (Davies and Simonovic, 2011). Such models can simulate multiple parts of the hydrologic cycle simultaneously, and build on the explosion of information and advances in data and computation technology. These models can ideally link to other models of the terrestrial system, such as those for land use and land-use change, ecosystems, and the biosphere;

geochemical components such as carbon and nitrogen; and the atmosphere, including weather.

These types of integrated models are needed as society faces mounting pressure over competing demands for limited natural resources and the need to improve predictive capabilities of the state and behavior of the water system, as well as the delivery of that knowledge to users. Integrated modeling for water will need to link groundwater and surface water (Refsgaard et al., 2010); provide measurements of water quantity and quality, land-atmosphere feedbacks, and precipitation; and represent human decisions explicitly (Davies and Simovic, 2011; Jaeger et al., 2017). Such models will need to be constrained by observations (NRC, 2008) and provide outputs that can be tuned to user needs at a range of scales (Garcia et al., 2016).

Given the complexity and range of spatial-temporal scales of water movement through the hydrologic cycle, it may be impractical to integrate all stocks and fluxes in a single model, especially when also integrating complicated, scale-dependent suites of natural and human drivers and feedbacks. The challenge will be to formulate problem-driven conceptual models that identify key drivers and responses at relevant scales and to remove barriers among different disciplines (e.g., vadose zone, surface water, groundwater, water quantity, and specific quality) to develop integrated modeling approaches.

The National Water Model (NWM) is one example of the potential for integrated water models to help support and enhance decisions about flooding, droughts, and other societal risks in real-time (see Box 2.1). This integrated model was recently developed jointly by the National Oceanic and Atmospheric Administration’s (NOAA’s) Office of Water Prediction, the National Center for Atmospheric Research (NCAR), the academic community (i.e., Consortium of Universities for the Advancement of Hydrologic Sciences, Inc.), the U.S. Geological Survey (USGS), the U.S. Army Corps of Engineers (USACE), and other federal agencies.

Quantifying Change in the Socio-Hydrological System

Human activities and actions can affect water resources in a variety of ways. Water infrastructure and its operations (e.g., dams, groundwater wells), coupled with urban and agricultural land-use change, can affect water quantity and quality, while land-use management practices and contaminants from human activities also alter its quality. These activities influence water resources availability and quality across the United States and globally (Gleick, 2003a; Nilsson et al., 2005), and can reduce or enhance flood risks. For example, global changes in the freshwater discharges from land to the oceans are a reflection of alterations to the hydrologic cycle in response to climate change (Syed et al., 2010; Durack et al., 2012)

and to continued human alteration of water flows through ongoing reservoir construction (Chao et al., 2008). Changes in climate that result in warmer temperatures, stronger storms, more droughts, and changes to water chemistry can further alter water quality (Georakakos et al., 2014). These effects, and others such as nutrient pollution leading to harmful algal blooms (including cyanobacterial), can influence the economy by affecting riparian development, ecosystem health (and their services), recreation and tourism, water treatment, and public health and well-being (Repetto, 2012; Michalak, 2016).

Securing Reliable and Sustainable Water Supplies

Society depends on the availability of clean and reliable surface water and groundwater for the provision of safe drinking water, food and energy production, industrial water supply, healthy ecosystem functions, and support of recreational opportunities and tourism (NRC, 2012a). Reliability of water resources, in turn, depends on efficient and effective infrastructure for water delivery and conservation (EPA, 2016). The impact of extreme weather events such as drought and floods, as well as the effects of climate change, represent threats to water resources (Boehlert et al., 2015). In addition, natural and managed water systems are increasingly affected by land and water management decisions (NRC, 2009).

To manage water resources in ways that can meet societal and environmental needs, the amount of water available has to be quantified and the reliability of water sources in the face of pressures from extreme events and failing infrastructure has to be understood (Gleick et al., 2013). Real-time information is also needed to effectively manage these resources to guard against overuse. A sound science base can inform management decisions regarding aquatic ecosystems that have to respond to changing flow regimes and water quality (Vörösmarty et al., 2010; NRC, 2015). In addition, ensuring that environmental flows of water are adequate necessitates interdisciplinary work to quantify ecosystem habitat and processes (Novak et al., 2016). Considering water resources in the context of integrated resources management can bring an integrative understanding of the feedbacks between water and these other resources that will ultimately affect water security (Barthel and Banzhaf, 2015).

Understanding and Predicting Water-Related Hazards

Water-related hazards represent some of the world’s costliest natural disasters in both economic and human terms (see Box 2.2) and are increasingly exacerbated by human activities and global change. Scenarios of climate change predict, for example, that the United States will see an increase in heavy precipitation, regionally varying changes in flood hazard, and increased risk of chronic long-term drought in the western United States (USGCRP, 2017). Understanding these complex interactions, managing the effects of human activities on water-related hazards and extreme events, and communicating the associated risks for lives, livelihoods, and the environment will need science informed by high-quality, long-term data and analysis and high-resolution modeling tools. Impacts from human activities are not limited to climate; for example, widespread use of de-icing salt on roadways in the northern United States has resulted in increased chloride concentrations in freshwaters, which can degrade ecosystems and

compromise drinking water. The highest contamination is associated with urban land use and impervious surfaces (Boutt et al., 2001; Kaushal et al., 2005). Better understanding of local impacts can also be complemented by increased understanding of large-scale feedbacks and the implications for atmospheric and terrestrial systems. Understanding the hydrological impacts of humans on water-related hazards is a continuing challenge (Gleick et al., 2013).

EMERGING TECHNOLOGIES

New technologies will emerge over the next 25 years that will allow for observations that come from a wide variety of sources, are more affordable, offer previously inaccessible location data, provide “fit-for-purpose” (i.e., designed to meet the specific requirements of its intended purposes) tem-

poral and spatial resolution, and deliver measurements of new parameters. Together with the wide adoption of those technologies, there will be a need to develop systems (e.g., hardware, software, management frameworks and protocols) that can rapidly bring together disparate datasets from a wide variety of sources, assess the data for quality, and then process, store, and share them in near real-time ways that are informative and accessible for users. The growing field of hydroinformatics will also be critical for combining the societal issues related to water use and need with large volumes of scientific datasets to provide innovative new insights into water resources.

While it is impossible to foresee all the advances in techniques and equipment that might occur over the next 25 years, great gains in the ability to understand and predict water quantity (surface and subsurface), quality, flow directions and rates, and residence times have come from observations related to satellite-based or airborne platforms, shallow-earth geophysical techniques, miniaturization of sensors, and data acquisition in remote settings. Biogeochemical markers (e.g., viruses, genetic material) used as environmental tracers and environmental DNA (eDNA) are poised to change the landscape of understanding environmental dynamics, including water resources. High throughput genetic sequencing of eDNA samples (e.g., application of metagenomics methods) could revolutionize the detection of pathogens and invasive species and provide important insights into aquatic ecosystem health and the cumulative impacts of multiple stressors. As new technologies gain hold or become established, there will be opportunities for greater analytical precision, accuracy, and lower costs for almost all laboratory methodologies. Real-time data processing, real-time links among data sources (e.g., between satellite observations and land-based sensors) and better analysis and visualization capabilities will increase the value of each sample collected.

Sensors for making ground-based observations have advanced significantly in ways that will allow high-density sensing in the future. These include applications such as the use of ultrasound to determine river stage, radar for measuring stream velocity, and the application of autonomous vehicles (both airborne and ground-based) as platforms for lidar and other remote-sensing applications. Distributed temperature sensing allows measurements to be made over time and space in streams and aquifers (in wells), which can also be used to indirectly measure soil moisture. Improvements are also likely in the quality and resolution of real-time precipitation measurements at locally relevant scales using radar and other remote-sensing tools. Snowpack measurement can be conducted using sensors that are able to determine snow depth, density, grain size, and albedo. Geophysical tools currently in use and in development (e.g., ground-penetrating radar, passive seismic, magnetic resonance imaging, optical and acoustic borehole

imaging) can acquire valuable information about the subsurface including the presence, general composition (i.e., saline versus fresh), and flow of water. There are likely to be leaps in the application of artificial intelligence approaches to maximize autonomous, decentralized systems such as sensor networks.

Chemical measurements by deployable sensors are currently limited to only a basic set of water quality parameters (e.g., temperature, pH, dissolved oxygen). Probes to measure nitrate have been in existence for some time but remain relatively expensive to deploy and maintain. Commonly used water-quality probes require frequent field maintenance, which substantially adds to their costs (Wagner et al., 2006). Sensors that measure light absorbance can indirectly measure other water quality parameters, such as dissolved organic carbon and algae (Wymore et al., 2018). There have been promising developments in sensors that can measure other chemical species by voltammetry and microfluidic lab-on-a-chip technologies (Buffle and Tercier-Waeber, 2005; Cogan et al., 2015; Barton et al., 2016; ter Schiphorst et al., 2018). Other advances, including the miniaturization of instruments such as mass spectrometers, could make them field-deployable over the next 25 years (Snyder et al., 2016). However, issues regarding powering and hardening these sensors have been challenging.

The use of space-based Earth observation platforms supported by the National Aeronautics and Space Administration (NASA), NOAA, USGS, and international agencies have dramatically changed observations of the water cycle (NASEM, 2018a). Current missions that are sensing critical water parameters include the follow-on missions of the Gravity Recovery and Climate Experiment (GRACE-FO),¹ the U.S. Soil Moisture Active Passive (SMAP),² and the Global Precipitation Measurement (GPM)³ satellites. The Surface Water and Ocean Topography (SWOT)⁴ Mission to survey the topography of the Earth’s oceans and terrestrial surface waters is expected to launch in 2021. The Moderate Resolution Imaging Spectroradiometer (MODIS)⁵ and the Visible Infrared Imaging Radiometer Suite (VIIRS)⁶ can measure snow-covered area, snow grain size, and snow albedo. Passive microwave sensors can estimate snow water equivalent over relatively flat, treeless regions, although global space-based measurement of snow water equivalent remains elusive. The application of space-based observations will continue with the advanced sensors developed and launched by NASA

___________________

¹ See https://gracefo.jpl.nasa.gov; accessed September 4, 2018.

² See https://smap.jpl.nasa.gov; accessed September 4, 2018.

³ See https://pmm.nasa.gov/GPM and https://svs.gsfc.nasa.gov/Gallery/GPM.html; accessed September 4, 2018.

⁴ See https://swot.jpl.nasa.gov; accessed September 4, 2018.

⁵ See https://modis.gsfc.nasa.gov; accessed September 4, 2018.

⁶ See https://jointmission.gsfc.nasa.gov/viirs.html; accessed September 4, 2018.

and international partners, as well as rapid deployment of large numbers of private-sector-funded satellite missions.

PRIORITY RESEARCH QUESTIONS

Given the water science and resources challenges and emerging technologies described above, the committee identified 10 science questions that, if addressed, would make the most significant contributions to respond to these and other challenges in the future. The questions below are not presented in any order of priority, as each is of critical importance for science and society. Although USGS could potentially contribute to advancing any of these questions, the committee further refined this set of questions in Chapter 3 to a smaller set representing those that would have the highest potential to advance USGS strategic science and other government priorities.

1. What is the quality and quantity of atmospheric, surface, and subsurface water, and how do these vary spatially and temporally?

Measuring water quantity is fundamental to the management of water resources. Understanding the amount and reliability of water sources is also key to water treatment and energy production and to industries that depend on water to produce their goods. As noted by Vogel et al. (2015), traditional approaches to water resources planning have centered on water quantity, while water quality assessments and the effects of human activity over time are assessed independently. There is growing awareness of the need to consider integrated water quantity and quality from a multidisciplinary perspective, including coupled interactions and human influences on both (Tundisi et al., 2015). Integrated understanding will become more pressing in the coming decades as water security issues become even more urgent. Water stocks, fluxes, and water quality parameters (including a range of natural chemicals, the contaminants introduced by human activities, and the water temperature of streams, rivers, and lakes) need to be measured, analyzed, and modeled. However, new technologies and methods used to measure water quantity and quality need to be fully validated before these data can be used in an interpretative manner.

Groundwater is increasingly being relied on for drinking water, agriculture, and industrial and energy production (e.g., Alley et al., 1999) and integrated studies will increasingly be desired. For example, understanding interactions between surface water and groundwater over time has grown in importance as water withdrawals from both surface and subsurface sources have increased. Because this interface is a highly reactive biogeochemical zone, it influences important functions such as the processing of major elements (e.g., carbon, nitrogen, phosphorous) and the attenuation

of organic contaminants (Berner, 2003; Zeng et al., 2012; Schlesinger and Bernhardt, 2013). Coordinated programs to monitor, assess, and model the individual and joint effects of changes in the use of surface and subsurface waters and of the introduction of contaminants to these waters will continue to be essential. However, making improvements in all aspects of the programs as technologies advance in the coming decades will be a major challenge.

2. How do human activities affect water quantity and quality?

Many human activities affect the quantity of water available over time and space (Khatri and Tyagi, 2014). Effects of water use range in scale from very large—such as the major water diversions associated with the Colorado River Aqueduct or the Central Arizona Project (Zuniga, 2000) that bring water to arid regions of the country—to much smaller and less obvious transfers, such as water lost from leaking infrastructure (e.g., irrigation canals, water mains, and sewers). Human influences can also affect water quantity issues such as flood risk. Urbanization and agricultural land use can enhance rapid runoff from storms and rapidly channel stormwater to streams and rivers. Flood protection works can remove natural attenuation processes, increase flood risk downstream, and increase floodplain inundation levels (Wheater, 2006).

In addition to knowing the volumes available in surface water and groundwater and the flows within and between them, a critical science need is to learn how human activities affect these sources and sinks so that future impacts can be anticipated. Threats to water quantity are regional in nature and typically occur in areas that are both arid and population-dense and in agricultural regions that are heavily reliant on groundwater for irrigation (MacDonald, 2010). Additionally, many regions rely on sufficient winter snowpack in mountainous regions to deliver water throughout the rest of the year and to recharge deep aquifers (Barnett et al., 2005; Mankin et al., 2015); yet, snowpacks are continuing to decline (Allchin and Déry, 2017; Mote et al., 2018). Even humid regions with normally abundant resources are still susceptible to sustained drought and increased demand (e.g., the Delaware River Basin⁷ or New York City [NASEM, 2018b]).

Urban land uses can affect water quality by increasing non-point storm water runoff and point sources of wastewater effluent (NRC, 2009); degrade groundwater quantity and quality through excessive drawdown, induced infiltration, reduced recharge, and multiple contaminant sources; and mobilize sediment. The expansion and intensification of agricultural land use has negatively influenced water quality at local, regional (Hansen et al., 2018), and continental scales (Diaz and Rosenberg, 2008), and also

___________________

⁷New Jersey v. New York, 347 U.S. 995 (1954).

reduced the availability of water. Efforts in the United States to remove low-head dams and to return streams to their natural flows could mobilize stored sediments. The effect of these sediments to downstream environments could be deleterious to benthic ecosystems (e.g., Mbaka and Mwaniki, 2017), but the limited evidence to date show little impact with respect to legacy organic contaminants (Cantwell et al., 2014). Providing new data and science to address these issues is a critical need, given the increasing urbanization of the world and the need to produce more food as the global population expands.

Water is also used to transport human and industrial waste and is an essential component of numerous industrial processes (Peters and Meybeck, 2009). Produced and flowback water as byproducts of unconventional hydrocarbon development have become more prevalent over the past decade, raising water quality and quantity concerns (Vidic et al., 2013; NASEM, 2017). Excess nitrogen, phosphorus, and pesticides from agricultural runoff, organic solvents that contaminate groundwater, microbial contaminants such as E. coli, per- and poly-fluoroalkyl substances (PFASs), and thermal pollution from power plants, municipalities, and reservoir releases all affect water quality in freshwater systems across the nation.

By increasing the export of dissolved nutrients downstream, land uses that disturb soils can fuel harmful algal blooms with cascading implications on environmental and human health (Rabalais et al., 2002, Kovacic et al., 2006). Harmful algal blooms threaten both potable water supplies and aquatic ecosystems alike (NRC, 2011), yet the ability to predict their extent and duration remains uncertain. Changing land use also affects large-scale hydrological cycle functions. Changing urban and rural landscapes, including irrigated agriculture, alter the evaporation feedbacks from land to atmosphere and have the potential to influence long-term propagation of atmospheric moisture and storms (Vörösmarty et al., 2004; Sterling et al., 2013).

Emerging contaminants (e.g., viruses, pharmaceuticals, new pesticides, personal care products, estrogenic compounds, nanomaterials) represent an unquantified potential threat to water quality and public health. Much remains unknown about the fate and transport of these contaminants (NRC, 2001, 2013). Research to minimize or mitigate negative effects of emerging contaminants on both water quantity and quality is critically needed (Rosi-Marshall et al., 2014), as the fate and ecosystem consequences of these compounds remain understudied (Boxall et al., 2012; Rosi-Marshall et al., 2013).

3. How can water accounting be done more effectively and comprehensively to provide data on water availability and use?

Accurately and completely accounting for water—understanding how much water is available, how it changes throughout the year, and how it is allocated—is critical for effective water management. However, current data-collection programs, including both measurement methods and the resulting data, are not sufficient to allow for the comprehensive accounting of water use and availability (Cooley et al., 2013). For example, in the United States, current water-accounting reports generally quantify only water withdrawals and do not distinguish between water that is returned to the hydrologic system or is used for consumption (Kenny et al., 2009). In the United States, USGS has made efforts to quantify some of these water uses, including surface water and groundwater use and trends in water use from 1950 to the present.⁸ Many regions in the world, however, completely lack basic water-use data. Even when data are collected, availability tends to be limited and data quality may be questionable. Establishing consistent efforts for the collection, compilation, analysis, and reporting of comprehensive water-use data is needed to allow effective management of resources and avoid water catastrophes. Data from all countries will be necessary as global interconnections in water resources become ever more prominent.

Countries across the world need to improve water accounting (Gleick, 2003a). In the United States, for example, while there are direct measurements of domestic water withdrawals at the local and state levels, the amount and type of data gathered is generally dependent on need (Dieter et al., 2018). This is particularly true in many rural areas or areas where there is currently an adequate water supply to meet all existing needs including human consumption, irrigation, and power generation (Dieter et al., 2018). In areas where demands are high and water resources are limited, water-use data are generally acquired at or near the end use (e.g., household or industry delivery point) (Maier et al., 2016; Siegrist, 2017). Because “consumptive use” (defined as withdrawn water that is no longer available for immediate use) is not generally quantified, it is difficult to assess how these losses affect existing stocks, especially in areas experiencing water insecurity. Finally, no standard protocol exists to delineate between direct (e.g., bathing, drinking, agriculture) and indirect water use (e.g., used to produce goods, virtual water). Indirect water use can only be estimated, as there is no direct means to measure these stocks and these models are reliant on USGS water-use reports (Blackhurst et al., 2010).

The procurement of adequate data is key to effective accounting systems. Data are needed on national water use, streamflow, groundwater,

___________________

⁸ See https://water.usgs.gov/watuse; accessed September 4, 2018.

water quality, and ecosystem needs in order to accurately account for water resources.

4. How does changing climate affect water quality, quantity, and reliability, as well as water-related hazards and extreme events?

Climate change is intensifying natural perturbations in the water cycle, as warmer temperatures increase evaporation, precipitation, and the water-vapor holding capacity of the atmosphere (Giorgi et al., 2011; Trenberth, 2011). When storms occur under such conditions, they can be more intense and lead to floods; while during hot, dry seasons, these same effects amplify drying of the land surface and contribute to droughts. Under current rates of carbon dioxide forcing, average annual U.S. temperatures are projected to warm by 6°F to 7°F over the next 50 years, with disproportionate warming in mountain regions (Pepin et al., 2015). Glacier retreat and reduced snowpack with earlier snowmelt will change the magnitude and timing of mountain river flows. In northern regions, permafrost thaw will change landscapes and hydrological connectivity, while earlier river ice breakup will change flood risk. More generally, the change to more intense precipitation and faster runoff will likely reduce groundwater recharge and surface-water base flows and climate-driven changes to terrestrial ecosystems, including carbon fertilization and changing fire regimes, will further modify hydrological response.

Water quality is also being altered by climate change. Higher temperatures lead to warmer water in streams and rivers, often negatively affecting cold-water fisheries, facilitating the success of invasive species (e.g., sea lamprey in the Great Lakes), and potentially exacerbating the effects of land-use change (e.g., deforestation) or industrial activities (e.g., water discharges from thermal power plants). Intense rain events, especially those following severe wildfires, increase erosion and sediment transport, which in turn mobilize contaminants. Warmer summers exacerbate harmful algal blooms (including cyanobacteria blooms) in lakes as toxic algae proliferate in warmer waters (Chapra et al., 2017). Rising sea levels can lead to saltwater intrusion into coastal aquifers (Green et al., 2011). Changing water quantity and degrading water quality are already presenting severe challenges to water resources management and are anticipated to accelerate as climate continues to change. Changes in the water cycle in future climate scenarios will not affect all parts of the world in the same way and could have significant and different impacts on a regional scale. New science is needed not only to project expected changes in water resources in the future but also to quantify, analyze, and explain the uncertainty surrounding future projections.

There is a need to understand the likely effects of climate change on extreme events. While enhanced resolution of climate models will improve

the explicit representation of precipitation (Rasmussen et al., 2011), it is also likely that predictive uncertainty associated with scenarios of future climate and their impacts on regional hydrological systems will remain high (Trenberth, 2010). Particular challenges arise for northern environments, where the hydrology is dependent on cold-region processes of snow, ice, and frozen soils and where these regions are rapidly warming with potentially large feedback effects (DeBeer et al., 2016).

Infrastructure design and risk-management practices in the United States and many other countries are based on the assumption of climate and catchment stability (Milly et al., 2008). Design methods for water infrastructure are based on observed flow and precipitation records, which are typically very limited in record length in comparison with the frequencies of extreme events that are of concern (Galloway, 2011). A new paradigm is needed to guide new infrastructure design and manage the increased risks associated with socioeconomic development and accelerating environmental change (Milly et al., 2008). Furthermore, given the inherent uncertainties associated with projections of future climate and land-use impacts on the hydrologic cycle, there is a need to develop a framework for decision-making under uncertainty (Lempert and Schlesinger, 2000; Lempert et al., 2003; Wheater and Gober, 2015).

5. How can long-term water-related risk management be improved?

A 2001 National Research Council report identified grand challenges for environmental sciences (NRC, 2001). One of the challenges was “to predict changes in freshwater resources and the environment caused by floods, droughts, sedimentation, and contamination in a context of growing demand on water resources” (NRC, 2001, p. 31). Given the apparent continuing escalation in water-related disasters (see Box 2.2), this remains an important challenge for water science in the future. The report noted the following:

All of these elements—new data, new integration, new science—will continue to grow apace in the future. It is imperative to continue advances in the ability to make useful multi-decade forecasts that can enable informed management of future risks to water supplies, water quality, and risks from floods and droughts.

Risks from land and water management are often multifaceted, affecting water flows, water quality, and ecosystem health. In addition, vulnerability to hydrological events is driven by societal policy, not simply the physical system. For example, human exposure to risk may be affected by choices such as installing engineering infrastructure for hazard mitigation (e.g., seawalls, levees), providing flood insurance in flood-prone areas (which may encourage people to live in areas at risk of more severe events [Gober and Wheater, 2015]), or using nature-based or green approaches to reduce the risk of flooding (Zimmermann et al., 2016). Improving the scientific understanding of these changes will be driven by more effective integration of physical sciences, social sciences, and engineering. Projections of future risks, however, are likely to continue to be associated with large uncertainty, especially when dealing with the human system. Interdisciplinary, integrated models will be needed to understand and predict the influence of environmental and societal change, as will tools that support decision-making under uncertainty, including vulnerability assessment, robust decision-making, and adaptive management.

6. How does the hydrologic cycle respond to changes in the atmosphere, the lithosphere, and the biosphere through Earth’s history and in the near future? And how do the hydrologic responses feed back to and hence accelerate or dampen the initial changes in the atmosphere, the lithosphere, and the biosphere?

The flow paths and residence times of water (and the resulting water-rock contact time) in the shallow subsurface are highly sensitive to climate and topography, but they also regulate climate through weathering and the long-term carbon cycle and regulate topography through weathering and erosion (Maher and Chamberlin, 2014). Water availability drives plant evolution and adaptation and hence ecosystem structures and their functioning (Stebbins, 1952; Axelrod, 1972); once established, ecosystems alter the water cycle through deep roots, transpiration, and precipitation recycling (Boyce and Lee, 2017) and the carbon cycle through sequestration in the soil (Shevliakova et al., 2013) and organic carbon burial (Berner, 2003).

Over shorter time scales, the water cycle regulates the Earth system through its fundamental role in land-based ecosystem structure and productivity. In the coming decades and centuries, the ability to predict the future climate trajectory rests on the ability to predict the response of land ecosystems to carbon dioxide fertilization and warmer temperatures, which is likely to enhance ecosystem productivity (Bonan and Doney, 2018). Many ecosystems, however, are currently under water stress, and others are predicted to experience increasing water stress. How water availability will shift in space and time in the near future, and how these shifts will enhance

or reduce an ecosystem’s capacity for carbon uptake and climate mitigation, are key questions to be answered. Given the short length of records from historical observations, much information can be gained from analysis of paleo records including tree rings, sediment cores, ice cores, and pollen studies, among others (NRC, 2007).

7. How can short-term forecasting for climate, hydrology, water quality, and associated social systems be improved?

Short-term hydrological forecasting (defined as forecasts over time spans of a few hours to a few weeks) is critical to society, providing immediate predictions of extreme, high-risk hydrologic events such as floods, as well as informing management and infrastructure decisions related to droughts, river flows, groundwater levels, and water management. Such short-term forecasting uses hydrologic models, which can assimilate current observations in real-time (e.g., precipitation, river levels, snow water equivalent, groundwater levels, water quality, water use) and make appropriate allowance for the loss of such information in extreme events (Young, 2002; Cloke and Pappenberger, 2009). In the future, developments in the availability of spatial data can be expected to improve current forecast reliability for river flows and lake and reservoir levels, although these ultimately depend on the accuracy of meteorological forecasts, which are high for short-term forecasts but degrade for seasonal forecasts. In the United States, a new national weather forecasting model is being implemented (Adams, 2016) to augment existing National Weather Service capability (see Box 2.1). There is significant further potential for delivery of improved forecasting, for example of floodplain water levels, urban flooding, and high groundwater conditions. Short-term forecasts that could warn of floods at any location, including the identification of the potential for failure of flood-defense infrastructure, would be a major advance. Global flood and drought warning products are currently being refined (e.g., Alfieri et al., 2013; Sheffield et al., 2014) and will be enhanced as global data products improve. The scope of water-related hazard warning systems can be expected to expand to include short-term (i.e., hours to days) threats from water quality (e.g., travel time of accidental pollution, warning of harmful algal blooms, seasonal forecasts of groundwater levels).

8. How do institutions and governance and institutional resilience impact the quantity and quality of water?

Diverse institutions play roles in the management and regulation of water supply, use, quality, and infrastructure in the context of societal values and preferences. Whether formal or informal, institutions encompass property rights, markets, regulations, policies, and socio-cultural norms

(Jaeger, 2015). They can function at the local level, such as municipal water pricing and conservation measures; at the state level, such as the doctrine of prior appropriation with water rights; and at the federal level, with implementation of regulatory environmental flows for endangered species. Societies develop infrastructure (e.g., dams, canals, pipelines, wells, sewers, treatment plants, reservoirs) to store, manage, deliver, and treat water for human needs.

Institutional resilience (the ability of an institution to adapt successfully to perturbations) is a concept that has been used to describe the ability of water decision-making and management bodies to create flexible solutions to new water resources challenges. In contrast, institutional failures can have devastating impacts on water resources. Instances of lead contamination (e.g., Washington, DC [e.g., Renner, 2004] or Flint, Michigan [e.g., Campbell et al., 2016; Chavez et al., 2017]) demonstrate failures in water-related decision-making. In another example, thousands of gallons of a chemical used to remove impurities from coal spilled into West Virginia’s Elk River in 2014, contaminating drinking water in Charleston, West Virginia, and several municipalities downstream. Institutional failures, both before the spill and during its response, contributed to the disaster (Lukacs et al., 2017). In cases such as the Elk River spill, stronger institutional resilience could lead to more rapid and coordinated efforts by a variety of entities, including federal agencies such as USGS WMA, to characterize and treat water-related contaminants (Whelton et al., 2017).

The challenges of managing water resources in the future are daunting. New approaches that do not simply focus on strictly technical and single-objective solutions will need to be developed. For example, adaptive governance, which explicitly integrates human dimensions and institutional arrangements to develop management goals and plans for achieving those goals, could be better integrated into the decision-making process (Akamani, 2016). To allow for wise use of water resources, the development of management approaches that are practical and flexible in the face of deep uncertainty are needed.

9. How can understanding of the connections between water-related hazards and human health be improved?

Human health and well-being can be affected by water-related hazards in diverse ways (Yusa et al., 2015), including physical injury or death, mental distress, and exposure to pollution and diseases. Floods, for example, can be a direct threat to life and property and can cause stress and consequent health responses. Hurricane Maria, for example, is estimated to be responsible for more than 2,975 deaths in Puerto Rico (GW Milken Institute School of Public Health, 2018) and is expected to cost at least

$139 billion to rebuild property, the energy grid, and the water system, among others.⁹

In addition to loss of life and property, floods are commonly accompanied by contamination from human sewage and have been linked to waterborne disease clusters (Epstein, 2005; McMichael, 2006). As demonstrated by Hurricane Harvey in Houston, Texas, floods may create pollution hazards through inundation of existing or former industrial sites or can mobilize river sediments that contain legacy contaminants from mining or mineral and chemical processing (EPA, 2017). Droughts can affect respiratory and mental health, infectious diseases (whether through water, food, or vectors), and illnesses related to toxin exposures, as well as food and water security (Yusa et al., 2015). Water quality may directly affect health through consumption of unsafe drinking water. For example, dissolved organic carbon, when subject to chlorination, can lead to carcinogenic disinfection byproducts (Pressman et al., 2010).

Substances in groundwater that affect human health and the health of ecosystems originate from a variety of sources (e.g., Moody, 1990). Naturally occurring elements such as arsenic and uranium can enter groundwater at unacceptable levels through geochemical processes (Nordstrom, 2002; Orloff et al., 2004). Elevated levels of nitrates and other nutrients in groundwater is a significant problem in many shallow aquifers (Rivett et al., 2008), and contamination by persistent manmade chemicals such as chlorinated solvents (Moran et al., 2007) is a nationwide concern. In agricultural areas, pesticides are ubiquitous in surface waters and groundwater used as drinking sources (Kolpin et al., 2002; Gilliom, 2007; Moschet et al., 2014). Harmful algal blooms can sicken humans who ingest contaminated seafood and can have long-term health effects including liver disease, cardiovascular disease, developmental defects, and neurobehavioral illness (Paerl and Otten, 2013; Malham et al., 2014; Campos et al., 2015; NSTC, 2016). In addition, recreational waters in close proximity to urban and agricultural areas could see increases in human exposure to pathogens (Soller et al., 2010).

Human health connections to water are multifaceted. Understanding these connections entails data assembly and analysis from sectors that are not typically seen as connected—meteorological forecasting, disease vector mechanisms, pollution control, disaster management, and water supply, among others. Coordination of the required interdisciplinary research in itself represents a formidable challenge, particularly as the impacts of climate change and land-use changes related to a growing global population occur in this century (Pandve, 2010; Khedun and Singh, 2013).

___________________

⁹ See https://www.npr.org/2018/08/09/637230089/puerto-rico-estimates-it-will-cost-139-billion-to-fully-recover-from-hurricane-m; accessed September 4, 2018.

10. How can competing uses for water resources be managed and maintained to sustain healthy communities and ecosystems in a changing world?

At the heart of water resources challenges is the conflict among competing water uses. One of the many such conflicts is between human societies and ecosystems. For example, widespread groundwater depletion for irrigation poses threats to groundwater-fed streams, lakes, and wetlands, as well as their dependent aquatic ecosystems (Winter et al., 1998). The alteration of natural flows disrupts aquatic ecosystems, and the risk of ecological change increases with the magnitude of flow alteration (Poff and Zimmerman, 2010). Water is also necessary to maintain ecosystem function in terrestrial ecosystems such as forests, wetlands, and prairies, especially under changing water regimes (Baron et al., 2002; Grant et al., 2013). Riparian floodplain forests along river and stream corridors rely on seasonal inundation to maintain their structure and function and in turn provide ecosystem services to adjacent freshwaters through shading, bank stability, carbon inputs, and nutrient retention (Richardson et al., 2007). From uplands to lowlands, water science challenges include measuring, monitoring, and understanding water quantity and quality needed to maintain the structure and function of ecosystems and maintaining ecosystem services in the face of environmental change.

An additional set of conflicts arises among competing human water uses, such as water resource allocation for drinking water supply, industry, and agriculture, as well as the management of reservoirs for hydropower, water supply, and flood protection. Conflicts arise at multiple scales, often involve many actors, and may cross jurisdictional boundaries, including among states or international governments. The food-energy-water nexus was conceived to help clarify the interdependence of and tradeoffs between some of these critical components (Beddington, 2009; Perrone and Hornberger, 2014; Scanlon et al., 2017). Increasing pressures of climate change, population growth, and technological shifts pose sustainability and management challenges to the food-energy-water nexus. In particular, as noted by Vaux (2012, p. 145), “the allocation of water between agriculture and the environment is a significant issue globally, nationally and locally and can no longer be taken for granted.” To better understand the complex tradeoffs in the food-energy-water nexus, USGS will need to bring together a broad spectrum of disciplines to provide data, information, tools, and the expert knowledge necessary to manage competing uses of water. A holistic approach will be needed to address the technological and non-technological issues to help decision-makers prioritize their efforts, such as engineering and technological innovations for water allocation and resource management, as well as social issues such as environmental justice and affordable access to high-quality water. There is increasing recognition that the social process of stakeholder engagement with water science is at least as important as the knowledge yielded by the science (Wheater and Gober, 2015).