Hacking the Watershed

The following is a summary of a chapter/article that I am working on for my dissertation. I am hoping that sharing this now will allow me to get some feedback on the general ideas I am working with and the overarching argument. I appreciate any comments.

For the past few years, I have been doing ethnographic research with computer modelers working in the Chesapeake Bay watershed. Primarily, my work has been with modelers who contribute to the Chesapeake Bay Program’s (CBP) effort to reduce nutrient pollution that flows into the Bay. The CBP is a watershed management organization composed of a collaboration between the federal government represented primarily by the EPA, the six states within the watershed (NY, PA, DE, MD, VA, and WV) along with several academic, private, and non-profit partners. CBP modelers along with partners from federal, state, and local agencies as well as from academic institutions in the region have, over the past 30+ years, constructed a large and complex computer model that simulates the watershed and the estuary and provides detailed information about the sources and effects of nutrients on the Bay. The model is sometimes referred to as the Chesapeake Bay Modeling System (CBMS) to indicate the multiple linked models that are part of it. This CBMS is used to help inform decision-making as well as to identify, track, and predict the effects of management practices on the landscape. However, in addition to this, I would like to understand the effect that the process of building and using a computational model like the CBMS has on the institutional relationships that form around environmental management.

I suggest that we can view the CBP modelers as “hackers” as described by Wark and others as those who “produce something new out of the old.” (It is important to note here that the term “hack” typically has a negative connotation in popular culture, indicating the act of breaking into computer systems to steal information. However, within the tech community, this is generally considered a misuse of the term.) In that sense, the modelers hack together the CBMS out of various bits of data and code. However, I argue that, in the process of hacking the model, they also hack the institutional structures that make up the CBP partnership. I have three ethnographic examples that illustrate my point. I won’t go into detail on the examples here. Instead, I provide a brief account of each:

First, the modelers hack the tools and resources of the CBP partnership to build a network of modelers and researchers to help inform and improve the CBMS. They use meetings and workshops to assemble these researchers, to provide information, and modeling methods and code. For example, when trying to resolve issues around the Conowingo Dam, they assembled a workshop to bring together disparate data and theories for modeling dam reservoir infill. This network of modelers and scientists give credibility to the model and the associated management process.

Second, the modelers hack the institutional structures and procedures of the CBP in order to get access to data that wouldn’t be available to them otherwise. For example, in order to get access to USDA farm data that would normally be unavailable to them due to privacy constraints, the modelers forged an agreement that created a data pipeline from the USDA to the USGS to the states and then to the CBP. In the process, the data is transformed into a form that does not breach privacy concerns. In addition to gaining access to data, this process helps to build links between the various institutions involved.

Finally, the modelers hack the different incentive structures of the various institutions involved in the CBP in order to generate data that is needed for the model, but doesn’t yet exist. Regulatory institutions have different incentives (informing management) than research institutions (developing good scientific knowledge). By providing model data, helping to get funding, and collaborating on publications, the modelers are able to encourage people in those institutions to work on research that is needed to improve the model. This not only provides further access to data, but also builds reciprocal relationships between the institutions that help to keep them involved in the process.

Together, these three examples illustrate how modelers at the CBP have become adept at institutional hacking. This process goes far beyond the construction of a complex modeling system. It also shapes the institutional relationships that make up the CBP partnership. Modelers might not actually be “hackers” in the general sense, but understanding how they engage in these forms of “hacking” can help us understand their role in the management process and the role of modeling in management institutions.