Contaminants are widespread in the 21st century environment, posing threats to human life and wellbeing and to the integrity of natural ecosystems. Even remote areas, such as the Arctic, are polluted with toxic materials such as mercury and chlorinated organic chemicals, despite the fact that the human activities generating them often are many thousands of miles away. Contaminated water bodies are a special concern, with major problems documented in large rivers (e.g., the Mississippi and Hudson), coastal waters (e.g., Gulf of Mexico, Chesapeake Bay), numerous ground-water aquifers, and many lakes of all sizes. Despite improvements in wastewater treatment, many surface-water bodies continue to be impaired by nutrient enrichment, and pathogenic microorganisms still are a common problem in the nation's waters, impairing recreation and threatening drinking water supplies. Organic chemicals and heavy metals pose risks to human health and to aquatic organisms.

To address large-scale water problems of the 21st century, we need to improve our understanding of the entire water cycle and associated biogeochemical processes and how they are influenced by human activities. Four critical deficiencies exist in our current abilities: • Basic data about water-related processes at the needed spatial and temporal resolution. • The means to integrate data from different scales, media and sources (observations, experiments, simulations). • Sufficiently accurate modeling and decision-support tools to predict underlying processes and forecast effects of different management strategies. • The understanding of fundamental processes needed to transfer knowledge and predictions across spatial and temporal scales—from the scale of measurements to the scale of a desired management action. For example, we need better ways to relate contaminant loadings to different types and levels of human activities, as well as better ways to predict transport processes and degradation mechanisms of contaminants and more reliable tools to assess the risks they pose to the environment and humans (Ref. NRC Report "Envisioning the Agenda for Water Resources Research in the Twenty-First Century", NAS Press, 2001). Because pollutants move between air, water, and land, we need to understand the interplay between these media and how efforts to control pollutants in one compartment affect environmental quality in other media.

The Engineering and Geosciences Directorates at the National Science Foundation (NSF) are collaborating in planning a national-scale environmental observatory network focused on water research and education. WATERS (WATer Environmental Research System) Network is an outgrowth of earlier initiatives of the directorates: CLEANER (Collaborative Large-scale Engineering Analysis Network for Environmental Research) and Hydrologic Observatories. Development of WATERS Network is being accomplished by scientists and engineers in the academic community who recognize the need for such network to enable better understanding of human-dominated water-environments, their stressors, and the links between them.

The goal of WATERS Network is to transform the ways the environmental engineering and hydrologic science communities perform research on large-scale, human-stressed, water systems through a collaborative analysis network, using high performance sensing, modeling and cyberinfrastructure tools. The initiative has the potential to transform engineering and science education by engaging the academic community in large-scale, complex real-world problems and by providing easier access to water information, modeling tools, and databases currently spread among many widely-distributed resources and not accessible in an integrated manner. The strategic intent is to create a system where theorists, experimentalists, and computational experts collaborate on significant water resources issues, identifying and resolving knowledge gaps related to them. The collaboration will operate using a framework analogous to adaptive management, in which research strategies are improved over time based on the knowledge gained by studying how environmental systems respond to specific actions. Modeling will be the central component for analysis, knowledge synthesis, and design of further experimentation. Modeling tools will include systems analysis and life cycle assessment models that incorporate consideration of economics, uncertainty and risk in decision-making.

Scientists and engineers associated with the development of WATERS Network within the academic community have established a series of grand scientific challenges as targets for the network to address. Key examples include the following: • How do we better detect, predict, and manage the effects of human activities and natural perturbations on the quantity, distribution and quality of water in near-real time? • What natural and human factors control the patterns and variability of water cycle processes at scales from local to continental? • Is there a universal theory of Continental Water Dynamics that accounts for patterns and variability, and their evolution over time? The integrated system of distributed facilities and researchers will support collection of critical environmental data with advanced sensor array systems and in situ instrumentation, facilitate data mining and aggregation, and provide analytical tools for data visualization, data analysis and predictive modeling of large-scale, dynamic water-cycle processes. The cyber-enabled facilities will facilitate participation from the broad engineering and science community, including educators, students, practitioners, and public-sector organizations, who will have access to data, models, and software. Ultimately, the improvements in scientific understanding produced by the network will result in more effective management approaches for water, based on enhanced observations, experimentation, modeling, engineering analysis, and design.

WATERS Network thus is envisioned to consist of four components: (a) interacting field sites networked through cyberinfrastructure; (b) groups of investigators studying landscapes affected by human activities, including agricultural and urbanized regions; (c) centralized support facilities including specialized personnel and analytical and experimental facilities, and (d) an analysis network with common modeling platforms and analysis protocols that will serve the community for collaborative investigations. A critical design principle for the network of field sites will be to provide a system that will enable multi-scale, dynamic predictive modeling for water, sediment, water quality (flux, flow paths, rates), including: near-real-time assimilation of data, feedback for observatory design, and point- to national-scale prediction. Thus, field sites will be nested over a range of scales from plot and sub-catchments to large river basins. Aside from scale, master design variables for locating sites will include climate (arid versus humid), geomorphology (e.g., coastal versus inland sites), and land-use/cover.

Recent progress in planning for WATERS Network can be summarized as: • A project office established in August 2005 has organized a community planning effort, including preparation of draft reports on a science plan, cyberinfrastructure, sensors, organizational structure, role of social sciences in the network, and educational plans. • A team of scientists and engineers from the environmental engineering and hydrologic sciences academic community is collaborating on the conceptual network design—to be completed in early spring 2007. • A workshop on modeling needs for environmental observatories (EOs) in May 2006 enhanced collaboration among NSF's several EO initiatives; a workshop on the role of social science research/education in EOs is in the planning stage. • NSF's Environmental Engineering and Technology (EET) and Hydrologic Science (HS) programs jointly funded 11 new “test-bed” projects to gain field experience with EO development and operation (2006). In addition, three projects funded by the solicitation “Cyberinfrastructure for Environmental Observatories: Prototypes” focus on cyberinfrastructure for water research, and an ongoing project to develop a Hydrologic Information System (HIS) has succeeded in developing capabilities to extract and merge water data from various databases, along with a work-flow system to streamline analysis of data from diverse sources. • The National Research Council issued a report (May 2006) supporting the concept of WATERS Network and providing advice on science questions and implementation issues.