Part of the Knowing Our Waters Project
Previous installments of the Knowing Our Waters storytelling series have focused on the work of volunteer water monitoring groups across the Marcellus Shale that are concerned about the impacts of oil and gas drilling in their communities. These stories have run the gamut, from tales of frackwater seeping into acid mine drainage sites, waste flows traveling from drilling operations in Pennsylvania into New York State landfills, schools using monitoring for environmental education, and sporting enthusiasts tracking the quality of high quality trout streams.
When reflecting on the geographical distribution of these efforts, one of the big questions is whether or not monitoring programs are more concentrated in areas most heavily impacted by unconventional oil and gas drilling, or if there are other explanations for why and where volunteers pick up the tools of monitoring and enter their watersheds. Furthermore, what is the relative distribution of volunteer monitoring versus governmental monitoring in these highly impacted watersheds?
In this article, guest author, Dr. Abby J. Kinchy, Associate Professor of Science and Technology Studies at Rensselaer Polytechnic Institute, explains her investigation into these questions, and highlights some surprising findings made by her research team in the Watershed Knowledge Mapping Project. An extended version of these findings can be found in a recently published paper in the journal Environment and Planning C: Government and Policy.
Kirk Jalbert, Manager of Community-Based Research & Engagement
Mapping Spatial and Social Inequalities of Water Monitoring
By Abby J. Kinchy, Associate Professor of Science and Technology Studies, Rensselaer Polytechnic Institute
Maps by Matt Kelso, Manager of Data & Technology
Are governmental and non-governmental efforts to monitor water quality reaching the watersheds that are most heavily impacted by Marcellus Shale gas drilling? My colleagues and I in the Watershed Knowledge Mapping Project at Rensselaer Polytechnic Institute (Kirk Jalbert, Sarah Parks, and Simona Perry) wanted to find out.
With a grant from the National Science Foundation, we surveyed watershed protection organizations across Pennsylvania and New York to find out where volunteers and nonprofit organizations — here referred to as civil society organizations (CSOs) — were monitoring streams for the impacts of fracking. We also sifted through publicly-available government databases on water quality, in order to determine where, and how frequently, state, federal, and other government agencies were sampling surface water quality. A significant portion of this data was also used in generating the Knowing Our Waters master map of volunteer monitored watersheds in the Marcellus Shale seen below.
What we found surprised us. CSOs were monitoring an impressively large proportion of watersheds in Pennsylvania – and many in the Marcellus Shale region of New York, as well. Government agencies were also monitoring surface water in more watersheds than we assumed, given the criticism of those agencies that we often heard from watershed advocates (a table describing each of the government monitoring programs included in this study is at the end of this article). State governments report to the EPA about the status of water quality, and the US Geological Survey samples water quality nationwide. In addition, multi-state agencies such as the Susquehanna River Basin Commission (SRBC) and the Delaware River Basin Commission (DRBC) have water monitoring programs.
Yet, despite all of this water monitoring activity, in some of the watersheds with the highest density of gas shale gas wells, there is little or no water monitoring by CSOs or government agencies. We refer to these unmonitored places as “spatial knowledge gaps.” There are many reasons why we should be concerned about these gaps in gathering data about our watersheds. Lack of knowledge obscures the dangers of industrial development, leading to public uncertainty, faulty conclusions about environmental impacts, and obstacles to effective governance (Frickel, Campanella, et al., 2009; Rocchini et al., 2011).
Gaps in water quality knowledge are a longstanding problem in the United States, as we found by examining published critiques of the nation’s water monitoring capacity since the passage of the Clean Water Act (CWA) in 1972. The CWA requires states to monitor water quality and report their findings to the federal government. One of these critiques, written in 1993, reflected on the first two decades of the CWA by stating, “scientists still cannot reliably answer the most basic questions about national water quality,” such as “What fraction of the nation’s surface water and groundwater fails to meet water-quality standards for toxic substances and conventional pollutants?” The author suggested that the reason for this was that the EPA’s reporting requirements have “not produced the kind of baseline and comparative database needed to evaluate past water-quality trends and to develop future strategies.” At the time of this article only 36% of river miles in the US were assessed due to insufficient resources for water monitoring (Knopman and Smith, 1993, page 19).
Note: it is important to explain that mapmakers represent watersheds at many different scales. The map of nongovernmental monitoring groups above represents watersheds at the “HUC 12” scale. This is a designation used by mapmakers for extremely local watersheds (e.g. small tributaries to larger streams). In the remaining maps and analysis below we used the “HUC 10” scale, which are larger in area. HUC 10s contain HUC 12 watersheds nested within them (as tributaries). Our reason for shifting to a HUC 10 level of analysis was that, in a few instances, local governmental monitoring data was not available at the granularity of HUC 12.
Knowledge Gaps and Social Inequality
In our view, it is important to view knowledge gaps not just as a science or public policy problem, but also a manifestation of social inequality. As we note in the journal article, uneven distribution of public and private water monitoring efforts and resources place some communities at a disadvantage compared to others when responding to environmental threats:
Gaps in monitoring indicate a lack of local capacity to identify and generate evidence of water quality problems. Places where streams are not monitored are vulnerable because they lack the infrastructure and personnel to detect water pollutants that may not be visible to the unaided eye, and because they do not have baseline data about water quality that would enable them to demonstrate that a change has occurred as a consequence of a new industrial activity (like fracking).
Our analysis revealed significant gaps in government monitoring. Only 53% of watersheds in Pennsylvania are continuously or frequently monitored by state, federal, or multistate government agencies. Only 22 watersheds (out of 332 watersheds in the state) have more than 3 monitoring locations per 100 square miles.
We then focused our analysis on the 25 watersheds that had the greatest density of shale gas wells, as of February 2013. A section of our article interprets these findings:
On average, there are 162 shale gas wells in each of these 25 watersheds, which is a density of about one well per square mile. There is a clear difference in government knowledge investments between the two gas drilling regions of the state, with a relative absence of government monitoring in the southwest. Of the 25 watersheds with the highest density of shale gas wells, seven are in southwestern Pennsylvania. Strikingly, five of the seven southwestern watersheds on this list have no frequent or continuous government monitoring. In comparison, the SRBC’s water monitoring program that specifically targets watersheds in gas drilling areas makes northeastern Pennsylvania a more highly monitored area.
As this Knowing Our Waters series highlights, the volunteers and environmental advocacy groups that make up the CSO community have frequently stepped up to fill in the gaps in government water monitoring. “I guess I see ourselves as doing a lot of what the state should be doing, which is just monitoring and making sure that problems are identified and addressed in an efficient manner,” one volunteer explained to us in an interview. In our article, we summarize the impressive contributions of these CSOs:
By 2012, CSOs were monitoring Marcellus-related drilling impacts in 56% of watersheds in the Marcellus Shale region of Pennsylvania—a greater proportion of watersheds than is monitored by state, federal, and multistate agencies combined. Even more impressively, CSOs are monitoring at twice the density of government monitoring locations—over 2 per 100 square miles. Of the 25 watersheds with the highest density of shale gas wells, 12 are monitored by CSOs—primarily in the southwestern watersheds that have little or no government monitoring. It is evident from the analysis of these 25 watersheds that CSOs are filling in spatial knowledge gaps in the southwestern part of the state. CSOs also make notable contributions throughout the rest of the Marcellus Shale region of Pennsylvania. Remarkably, CSOs are the sole source of continuous or frequent monitoring data in 60 (22%) of the watersheds in the Marcellus Shale region.
However, CSOs are not usually coordinated with one another or with government agencies. The absence of coordination creates the possibility that efforts will be duplicated in some places, while other places will be neglected. One goal of our analysis was to find out if this was the case in Pennsylvania. This bar graph represents the variation in government and CSO monitoring efforts across the 25 watersheds that have the highest density of shale gas wells.
The map below shows all watersheds that are monitored by CSOs (according to the original Watershed Knowledge Mapping dataset). The map also highlights remaining knowledge gaps: watersheds where neither government agencies nor CSOs are monitoring on a frequent basis.
Map of watersheds that are still not monitored on a continuous or frequent basis by any organization or agency (based on FracTracker’s 2015 data on CSO monitoring and Watershed Knowledge Mapping project’s 2013 data on governmental monitoring).
Relationships to Environmental Justice
A key finding of our study is that “Despite the considerable efforts by government agencies and especially by CSOs, 25% of watersheds in the Marcellus Shale region of Pennsylvania are still [as of 2013] not monitored on a continuous or frequent basis by any organization, either government or CSO.” Thankfully, these gaps are shrinking as more CSOs and local governments get involved in monitoring, as reflected in the FracTracker map above.
Why do these gaps exist? Is there any pattern to the distribution of the gaps, or do they occur by random chance? The next step in our analysis was to explore some hypotheses about which kinds of communities lack water quality knowledge. In our mapping efforts, we did not find any obvious relationships between knowledge gaps and demographic characteristics like rates of poverty or racial and ethnic makeup of communities living in particular watersheds. However, population density appears to play some role.
…[C]ompared to other watersheds with shale gas wells, the unmonitored watersheds are in areas of lower population density (on average 112 people per square mile, compared to an average of 156 people per square mile in the counties surrounding watersheds with water monitoring).
We then wanted to examine the prevalence of water monitoring activity in areas with high concentrations of poverty and/or minority communities. Government agencies regard these regions as “environmental justice areas” (EJ areas) because of past tendencies to concentrate risky industrial activities near these vulnerable populations. Are watersheds in these areas more closely monitored, because of environmental justice concerns? Or do they lack monitoring resources?
Our analysis found no evidence that government run water monitoring programs were less prevalent in EJ areas. In fact, when compared to other census tracts we found that EJ areas were less likely to intersect with unmonitored watersheds. As we noted in our paper, “Only 27% of EJ areas intersect with an unmonitored watershed, compared to 54% of other census tracts.” In other words, watersheds that overlap with EJ areas are, more often than not, monitored by a government agency.
How do these numbers compare to CSO water monitoring organizations? Are citizen scientists also monitoring the watersheds of EJ areas? Indeed, we found that watersheds that overlap with EJ areas are more likely than not to be monitored by a CSO—75% of EJ areas intersect with a watershed monitored by a CSO (compared to 42% of other census tracts). In our paper, we suggest some reasons for the distribution of water monitoring efforts that we observed. In particular, we focus on the historical land uses of different regions.
[I]t appears that both government agencies and CSOs have often chosen to establish water monitoring locations in watersheds that flow through highly populated—and highly polluted—urban areas (where most EJ areas are located). In addition, many of the CSOs monitoring water quality in southwestern Pennsylvania have long been active in addressing the water quality impacts of coal mining, an industry that brought environmental devastation—and rural poverty—to that area decades before shale gas drilling. Thus, when Marcellus Shale development introduced a new threat to water quality, there were already numerous organizations and volunteers who were dedicated to monitoring and protecting their watersheds in the region. This history also appears to have contributed to the mobilization of volunteers in this part of the state, despite the area’s socioeconomic challenges.
What lessons should we take away from this analysis? First, gaps in water quality monitoring are a significant problem. As we state in our paper:
Nearly half of Pennsylvania watersheds are not monitored frequently or continuously by any state, federal, or multistate agency. It may be unreasonable to expect that every watershed should be monitored to the same extent. Some watersheds have more complex problems that require different interventions than others, and some watersheds affect fewer people than others. Nevertheless, the absence of official water quality monitoring in such a large proportion of Pennsylvania’s watersheds is a problem that, in our view, requires public attention. Gaps in monitoring may be obscuring the effects of Marcellus Shale development across the state, especially given signs that regulatory agencies struggle to keep up with the rapid pace of the industry.
This is not just a problem for science – it is also a form of social inequality that leaves some communities ill-equipped to confront watershed problems arising from shale gas development. Regions that lack long-term baseline data are less able to determine how their water quality is changing due to shale gas development. While this negligence may be unintended, these knowledge gaps ultimately can have a disempowering effect for those who contest the actions of big oil and gas companies, or for those who question the findings of regulatory scientists.
Concluding Thoughts from Kinchy
CSOs that mobilize volunteers or monitoring tools to track water quality go a long way toward filling knowledge gaps. However, we have some concerns about relying on citizen scientists when government agencies fall short. We have found that not all communities are equally equipped to obtain knowledge about their watersheds. We suggest that watershed researchers – whether in government agencies, nonprofit organizations, or higher education – should direct their resources and attention not only to EJ areas but also to watersheds where communities struggle to mobilize citizen science efforts.
We also argue that environmental regulators should take advantage of the incredible amount of environmental knowledge that citizen scientists possess, and not just make demands for their data. While CSOs should not be seen as a one-to-one substitute in the work of environmental oversight, these organizations have proven effective in collecting meaningful data in regions impacted by shale gas extraction using relatively inexpensive methods. Volunteers are also the “eyes and ears” on the ground, noticing steam bank erosion, fish kills, methane bubbles, and other unusual watershed events. Evidence of CSO monitoring success tells us that they can play an important role in protecting local watersheds that are not always accessible to regulatory agencies with limited resources.
Appendix: Government Monitoring Programs
|Governmental Water Monitoring Program||Description|
|US Geological Survey (USGS)||Current Conditions and Field Samples: Surface water quality is monitored both sporadically (field samples) and continuously (in-stream devices). Field sample data which did not meet our definition of “frequent” monitoring was excluded.|
|Delaware River Basin Commission (DRBC)||Baseline Monitoring Program for Natural Gas: Surface water quality is monitored both sporadically and continuously. Field sample data which did not meet our definition of “frequent” monitoring was excluded.|
|Susquehanna River Basin Commission (SRBC)||Remote Water Quality Monitoring Network: Surface water quality is monitored continuously. Field samples for particular indicators are also taken at these stations. (SRBC also collects water quality data on a rotating 6-yr. schedule through their Subbasin Survey Program; however, by our definition of “frequent” monitoring, these locations were excluded).|
|Pennsylvania Department of Environmental Protection (PADEP)||Water Quality Network: Surface water quality is monitored on a bi-weekly or monthly basis. Field sample data which did not meet our definition of “frequent” monitoring was excluded.|
|Ohio River Valley Water Sanitation Commission (ORSANCO)||Surface water quality is monitored sporadically.|
For more information, please contact Kirk Jalbert: email@example.com