One of the most promising areas to apply new sensing technology to monitor, assess, and respond to pressing environmental challenges is in water resources. The Foresight and Governance Project, in conjunction with the Center for Embedded Networked Sensing (CENS), launched a new white paper, Distributed Sensing Systems for Water Quality Assessment and Management, at an event at the Wilson Center on 23 February 2007. Three of the report's authors, Thomas Harmon, Professor of Environmental Engineering at University of California, Merced, Deborah Estrin, Director of CENS and Professor of Computer Science at University of California, Los Angeles, and Jeffrey Goldman, Director of Program Development at CENS and the report's lead author, presented the paper. Robert Olson, Senior Fellow at the Institute for Alternative Futures, moderated the session.

Sensor networks allow advanced, automated data collection over wide areas, said Dr. Harmon. By linking autonomous sensors together, an observer can monitor large systems quite closely, and in a variety of different ways. For example, a septic system could be monitored for leaks by checking for contaminants downstream, by tracking physical and chemical patterns across the septic field, or even by periodically introducing a marker chemical into the system and following its transmission. Such analysis could be carried out in real time, and leak sources could be identified and even plotted on a map, said Dr. Goldman.

Networked sensing systems can simplify complex problems, such as so-called "non-point source pollution," Dr. Harmon argued. By taking data at many points over a large area and using appropriate models, such a system can cut through the complexity and break the problem into many small, point-source pollution problems, which can be addressed individually.

Modeling is key to the success of networked sensing systems, Dr. Estrin contended. With well-designed models, computers can aggregate and analyze the data collected by such systems and arrive at simple conclusions. For example, Dr. Goldman proposed a farm monitoring system, which could analyze water in the soil and automatically determine irrigation levels for the farm, or alert the farmer if he is using too much or too little fertilizer. By developing and using good models, sensor networks could help managers by collecting and interpreting data on the fly—fast enough for it to be incorporated into quick decisions.

Sensors have a great deal of versatility. As Dr. Estrin explained: while some sensors are meant to be planted and left in place, many sensors are handheld or robotic and can move around, allowing greater control over measurements. Moreover, current sensors can collect data on physical properties (e.g., temperature), they can test for chemical concentrations (e.g., measuring nitrate levels in soil), and some can even check for the presence of biological indicators (e.g., the presence of certain bacteria in water). Generally, Dr. Estrin said, physical and chemical sensors are much more advanced and field-ready, while biological sensors remain limited. Methods for detecting microorganisms typically require lab work and a lot of time, making biological sensors ill-suited for miniaturization.

Sensor networks often depend on different types of sensors working together in multi-scale sensing systems. Stationary and mobile sensors can be used in concert, Dr. Estrin noted—for example, stationary sensors can identify areas where large variations exist, and direct robotic sensors, or researchers, to those locations to take a closer look. Dr. Harmon described the use of proximal indicators, such as increased turbidity for bacterial concentration, to help identify areas that require more attention from researchers.

As sensors grow smaller, more sophisticated, and more rugged, the field applications of networked sensing systems will continue to grow, opined Dr. Goldman. By developing new sensors and engineering appropriate models to interpret the data they gather, sensor networks should become an important tool in water management. Dr. Goldman recommended that to get from here to there—the vast potential that networked sensing systems present—"education and training of future water resource professionals is needed to ensure that a ready supply of people will be available to wield these advanced tools now and in the future." When asked by Robert Olson about longer-term visions of the future, Dr. Estrin replied that she foresees sensor technologies helping us "achieve sustainability in many different dimensions" as they enable "more precise and real-time measures of our impact," which will inform management and decision-making.

The audience included many representatives from government agencies (the Environmental Protection Agency, National Science Foundation, State Department, Department of Defense, U.S. Geological Survey, and U.S. Army Corps of Engineers), as well as members from industry and NGOs.