At CSULB, researchers trace climate threat rising beneath California

Climate-driven changes to groundwater levels are increasing potential exposure risks across California. This statewide map shows where rising groundwater could mobilize toxic legacy contamination, increasing the risk of vapor intrusion into buildings. Exposure risk is based on groundwater depth, contamination history and proximity to populated areas.

Historically, Hagedorn said, once landowners prove that contamination levels at these sites have abated, they receive closure letters that end regulatory oversight. But those decisions assume conditions underground remain stable.

"What if you have sea-level rise, or a spring tide, or an atmospheric river?" he asked. "Where does it stop?"

Seeing risk statewide

Hagedorn and Custodio combined data pertaining to groundwater depth, known contamination sites and population density throughout California. They then used machine learning and explainable artificial intelligence to show how those factors overlap in a way that makes unhealthy exposure more likely.

"We were able to develop a model that predicts exposure risk," Hagedorn said. "It takes into account recent detections and past detections and weighs them. The result is one consistent database [that gives us] the whole picture."

At the county level, the model reveals concentrated areas of elevated risk tied to shallow groundwater and historic industrial land use.

Perhaps most striking are the heat maps, which translate complex groundwater data into gradients of risk, from cool blues and greens to deep red. The danger is most severe in major coast cities, particularly in the Bay Area, where shallow groundwater, urban development and a history of industry converge. Yet the maps also reveal pockets of vulnerability far inland, where various conditions keep groundwater close to the surface.

The result is a clear picture not only of where contamination is likely to resurface, but who lives above it. Many of the most vulnerable areas overlap with low-income neighborhoods and communities of color, areas that have experienced decades of industrial use and limited cleanup.

The findings will be published this year and presented to state regulators at a meeting of the California State Water Resources Control Board in March.

Turning data into protection

Although Hagedorn has spent years collecting and analyzing groundwater samples, this project did not involve any new field sampling. In fact, the work was done almost exclusively with computers, using geochemical indexing algorithms to identify patterns and determine which combination of factors is at play.

"We're trying to identify the most important predictors for exposure vulnerability," Hagedorn said. "What matters, and what doesn't."

He emphasized that the goal is not to "create hype or hysteria," but to give regulators a clearer way to protect people before exposure occurs - particularly in communities that have long borne the burden of industrial pollution.

Custodio, who is working toward her second bachelor's degree - the first in Environmental Science & Policy - said she appreciates Hagedorn's project for its broader public mission: using science not just to understand risk, but to reduce it before it becomes a crisis.

"Hopefully," she said, "this provides an objective way to identify areas that should be prioritized for cleanup, so that we can allocate resources accordingly."