Projected climatological changes over the next century are expected to drastically alter the weather patterns of the Northeastern United States. Combined with anthropogenic changes to land use and projected shifts in population distribution and urban development, the future health of aquatic environments within the region remains uncertain.As part of the development of a comprehensive integrated assessment model (IAM) for the Lake Champlain Basin, the behavior of farmer and governmental regulator agents in the Missisquoi Bay Area of Lake Champlain in Vermont was modeled to explore potential changes in human decision-making and land-use under a variety of projected climatological, environmental, and economical scenarios for the region to the year 2040. In particular, we are looking at how farmers within the region may choose to change their land-use practices and adopt or reject agricultural best management practices (BMPs) and how a government regulator may implement taxes on or subsidize farming practices in an attempt to stymie environmental damages to the lake ecosystem.480 farmer agents, corresponding to the agricultural land parcels within the Missisquoi Bay Area, and 1 municipal regulatory agent were included in the model. The behavior of these agents was trained under 14 scenarios,13 projected and 1 baseline/business-as-usual, using deep reinforcement machine learning with double q-learning (DDQN). A comparative analysis of the results of the model under these various scenarios will be discussed, such as those seen in Figure 1, com-paring the mean BMP adoption of farmer agents in the model across 10 runs, trained for both the baseline scenario and a scenario with a 10% increase in extreme weather event intensity.