Water-quality improvements in urban watersheds play an important role in the development of sustainable cities because of the benefits to both society and the environment. In the United States, Total Maximum Daily Loads (TMDLs) quantify river pollutant limitations and have been the key regulatory framework to manage these considerations. Accordingly, this paper proposes a model to enhance the capability of decision-makers implementing TMDLs in watersheds. This model is denoted as TMDL-C3, where ``C3" refers to three key challenges of TMDL Implementation: 1) procuring necessary funding for pollution-reduction investments, 2) improving communication and coordination between local agencies, 3) acquiring sufficient data to characterize the watershed and pollutant sources. To address these challenges, we demonstrate how non-cooperative game theory can be used to provide a holistic view of the costs of best management practices (BMP), stakeholder coordination mechanisms, and hydrologic data uncertainty. Our model is a stochastic, mixed complementarity problem, which uses advanced, game-theory techniques from operations research. The merits of the approach are demonstrated in both a stylized watershed as well as for the Anacostia Watershed in Metropolitan Washington, DC. The results indicate that a water-quality, credit-trading market can be an successful mechanism to maximize BMP cost-effectiveness in a heterogeneous watershed with multiple players. However, epistemic uncertainties such as BMP efficiencies and biological endpoints can obscure these benefits if they are overly conservative.
