Written by George Wyeth, Environmental Law Institute
Reliable data on environmental conditions is the foundation of pollution control in the U.S. Both the Clean Air Act and the Clean Water Act use the same fundamental approach: set standards for levels of pollution, measure air quality and water quality, and use that data to determine what regions or water bodies fail to meet the standards and require an aggressive response. Under these laws, the federal Environmental Protection Agency, and environmental agencies in the states, have developed sophisticated systems for monitoring air and water quality.
Citizen science (some prefer to call it community science) plays an increasingly significant role in this arrangement, potentially altering longstanding practices. In water pollution control programs, this has been true for over two decades; some states have had non-governmental volunteers helping to take samples since the 1990’s. In the air program, the potential for the public to contribute data has only come along more recently, with the emergence of new technology and low-cost sensors, and agencies have responded cautiously.
Thus, at present the experiences in the air and water programs are very different. The question is whether the two models will eventually converge.
The use of volunteers to help assess water quality was a matter of necessity for the states, who are required by law to submit reports every two years to the federal government on the state of their rivers, lakes and streams. Monitoring all these water bodies far exceeded their capacity, so they reached out to local groups and individuals for help. Sometimes the data generated by these volunteers could be used in the reports; sometimes it was not reliable enough for that purpose, but could help the state prioritize its own monitoring.
A key to the success of volunteer water monitoring was the fact that the testing techniques were within the capabilities of nonspecialists, with a basic level of training and oversight. Water samples are collected by hand, using relatively inexpensive tools. In some cases, results can be determined in the field, and in others the samples are sent to a laboratory. Therefore, monitoring by state officials and by volunteers was essentially the same process. To ensure data quality appropriate for regulatory use, agencies have provided training, guidelines, approved research designs, and other forms of assistance to the volunteer groups.
Air quality monitoring works much differently. Government agencies use a network of very high quality, and costly, monitors spread across the country. Based on the data the network provides, the EPA and state governments determine where pollution exceeds acceptable levels, and develop plans for reducing emissions in those areas. Such systems are prohibitively expensive for most environmental or community organizations. As a result, air pollution monitoring has remained exclusively in the hands of governments (or private companies).
Over the past ten years, however, new technologies for measuring air quality with much cheaper sensors have emerged. This creates the possibility for non-governmental groups, such as community organizations, to assess air quality on their own. However, data generated in this way has not been used by agencies to a great extent, for a number of reasons:
- The low-cost sensors are not as precise or as reliable as agency monitors, which undergo extensive review before being approved for regulatory purposes. Knowledge of the quality of data from these sensors is uneven, as technology changes frequently.
- Generating actionable data also requires a scientifically rigorous research design. The fact that data is being collected by non-professionals also makes agencies cautious about using it.
- Private air quality monitoring has been done by advocacy groups. Agencies often are wary of relying on groups they perceive as having an agenda.
- Communities facing serious air pollution problems often view agencies with distrust, making collaboration difficult.
As a result, agencies generally do not consider community-generated data acceptable for their purposes. There are occasional exceptions – for example, air samples taken by a neighborhood group in Tonawanda, New York, persuaded prosecutors to conduct their own investigation which led to criminal charges against a polluting facility. However, this remains the exception much more than the rule.
Technical concerns about using data from unconventional sources are legitimate. However, such data is potentially very valuable in its own way. In particular, community data gathering can provide information on air quality at a much finer geographic scale than the agency network is capable of. It can also be an avenue for improving relations between agencies and communities. Therefore, agencies could benefit if they were to find ways of using that data as much as possible.
The experience in water programs suggests that agencies can address their concerns about data quality by working with the public in a number of ways:
- Identifying a range of ways data can be used, and the type of data required for each use.
- Testing new sensors, and describing the uses they may be appropriate for.
- Providing direct training, equipment, guidance on research designs and proper data gathering protocols.
- Taking into account the contribution to data quality of crowdsourcing; data from a large number of low-cost sensors will likely be more reliable than a test of any one sensor would indicate.
Agencies can think more creatively about how such data could be used even if it does not meet regulatory standards. Such data could be used, for example, to help in locating the agency’s sensors to assess local air quality, or to help enforcement officials identify potential violations.
Non-agency data can also shed much light on pollution in urban neighborhoods. The environmental concerns of overburdened, low-income communities are especially important at this particular time in history. Data from the community can inform planning by measuring conditions on a more granular level than agencies are generally capable of. Agencies can find ways of using such data even taking into account its higher degree of uncertainty.
These steps would require some work (and funding), but the experience in the water program suggests that it is worth the effort. In this way, a system can be set up for the use of non-governmental data, with clear guidance to communities and other external groups to ensure that their expectations are realistic and that their time and effort are well spent. The result should be beneficial for all concerned.
The views expressed are those of the author and do not represent those of ELI; ELI is provided for identification only.