Pollution of groundwater by fuel compounds is a serious threat to water quality. Since most groundwater contaminants originate from the soil surface (or close to it), it is crucial to estimate the risk to groundwater quality from surface contaminants. Conventional models for transport of contaminant through the vadose zone are complex and require detailed knowledge on the site physical, chemical and biological properties. This knowledge is often missing or only partially available. In addition,  the conventional models are computationally expensive and require specialized expertise. 

In this project, we are using a machine learning approach to estimate the risk of groundwater pollution from soil contaminants. We use data on soil and groundwater pollution, together with a detailed model for transport of pollutants in the vadose zone to build a machine learning model and to predict the concentration of pollutants in the groundwater.