I wrapped up the month of August 2026 with a road trip up to Door County, Wis., a picturesque part of the state known for great camping and tourism, strong maritime traditions, and thriving fruit and dairy farms. Like many parts of the United States, dairy farms in Wisconsin are ramping up in size and scale, and some Wisconsin dairy operations have expanded considerably in recent decades.
One such Door County farm, Gilbert Farms, recently received a concentrated animal feeding operation (CAFO) permit to expand to 2,430 animal units. The terms of the permit require more safeguards than similar operations due to the region’s sensitive karst geology, and the permit notably mandates early groundwater monitoring, but public concerns about risks to the county’s groundwater persist. Records already exist in Wisconsin linking agricultural land uses to tainted drinking water—and this region is highly vulnerable given the geology alone (bit.ly/nitrate-report-pdf).
The point of this article isn’t to belabor the shortcomings of managed agriculture or the many other land uses that affect groundwater. Urban areas have long been sources of contaminants such as excess phosphorus, PFAS, heavy metals and more. We now try to reduce rates of surface-water and groundwater pollution through federal and state law, ordinances, green infrastructure and mitigation plans.
Whether the environment is urban or rural, surface water and groundwater interactions are a dominant and increasingly important consideration for long-term water management. As concerns about contamination grow, awareness builds. Developments in remote sensing and other technologies make monitoring and effective oversight possible but proceeding with more caution and a greater appreciation for underlying hydrologic connections also is essential.
Exchange Points and Contaminant Flows
Understanding flow interactions between groundwater and surface water as well as the processes by which pollutants move between the two is important to safeguard drinking water. It’s also important to understand these processes for stream and wetland restoration efforts (given significant baseflows from groundwater) and to better protect groundwater-driven ecosystems.
At the most-basic level, groundwater and surface water interactions are complex and difficult to observe, even when dealing with unmodified landscapes. Groundwater and surface water flow are governed by different dominant processes and are commonly analyzed at different spatial and temporal scales, making their points of exchange difficult to identify and quantify. Surface waters can gain or lose flow from groundwater in multiple ways, depending on location of the water table relative to surface flow (differences in hydraulic head), character of streambed geology, erosion, seepage in the hyporheic zone (the subsurface region beneath and alongside a streambed where surface water and shallow groundwater mix and exchange), heavy rainfall and countless other variables.
Clearly, both surface water and groundwater can transport a slew of diverse contaminants that can be stored in groundwater systems at high concentrations for long periods of time and released much later to surface waters. When contaminants enter subterranean systems, identification of pathways and timing of discharge becomes very difficult.
Monitoring Water Quality
Increasingly, communities and research groups are considering watershed-scale solutions that leverage data from sensors and newer technologies to track changes in a watershed’s surface water and groundwater systems. One of the more ambitious projects I’ve come across is led by researchers at the Stanford Doerr School of Sustainability, which aims within a five-year period to develop a “sensor-informed analytical framework” for the San Joaquin Valley Watershed in California, one of the most productive agricultural regions in the United States.
The project aims to manage surface water and groundwater as a unified system, reducing uncertainty about underlying processes and system interactions through data and integrated analysis. Information pathways are organized through the development of a digital twin and “decision support dashboard” that allows data-based decision-making to guide adaptive management. The project has too many goals to list here, but one is to track changes in water quality at a fine spatiotemporal scale and map where water exchange occurs between surface and groundwater. The degree to which pumping in the watershed affects streamflow also will be quantified.
This project is still in development as multiple data pipelines are worked out, but it’s an important effort to gather facts and use those facts to assess causes of adverse change, evaluate water quality and supply, and modify management practices accordingly.
Groundwater Risks
The Gilbert Farms CAFO permit issued by the Wisconsin DNR requires groundwater quality to be monitored monthly from each of the farm’s three groundwater monitoring wells for the first year of operation, and quarterly samples are required thereafter. Other safeguards—reinforced manure storage and spreading limits, for example—also are in place. But in a setting with shallow soils, highly fractured karst bedrock and a sinkhole already on the property, an extreme rainfall event can test even well-designed safeguards. The engineering question is whether monitoring, storage and land-application controls provide sufficient resilience under those conditions.
I’d argue that as freshwater continues to come under pressure, we strive to adopt a more deliberative and integrated approach that considers what’s at stake with incomplete risk analyses. We increasingly need to ask tough questions, collect data and proactively consider what may be lost. View the water cycle in its entirety—in particular the hidden but critical subsurface connections. 