Developing a groundwater flow and transport model with stakeholder feedback from model conception

Abstract:

Like many shallow aquifers in urbanized areas, the shallow dolomite in Chicago’s southwest suburbs is in danger of high chloride concentrations originating from road salt. This aquifer is important for many different water users and also serves as an important ecological wetland habitat where the dolomite is exposed. MODFLOW modelers at the Illinois State Water Survey are working closely with local water users and other stakeholders in creating a solute transport model to investigate historic and future chloride contamination. Stakeholder investment is critical as these model results could inform their decision to find a new water source if chloride concentrations in the aquifer exceed the EPA’s secondary standard of 250 parts per million.

Previous studies in this region have found that after decades of road salt application the overall trend of chloride concentration is that of steady accumulation. To capture this long-term trend, we work with communities to understand the amount of road salt they apply to paved surfaces throughout the winter and convert those measurements to chloride input concentrations for our MODFLOW-MT3DMS model. Additionally, the model features historical salt application and road growth, information that was provided by stakeholders. More recent measurements collected over a shorter time period indicate that concentrations can vary greatly over the span of weeks, even at considerable depths. To address the issue of short-term transience of chloride measurements, communities have been proactive at collecting more frequent measurements, that in turn allows for expanded simulation capability to ascertain whether curious measurements are outliers or part of a greater trend.

In this presentation we will discuss how building this model incrementally with input from communities allows for transparency in the participatory modelling process. Seeing the shallow model in each stage of its development has improved stakeholders understanding of how to evaluate model uncertainties and how the data they share are utilized to answer specific questions pertinent to their community. Likewise, this participatory modelling approach has enhanced the quality of the assumptions underlying model inputs and calibration data for the model output.