资源环境科技发展态势分析平台

Many groundwater contaminants react with components of the aquifer matrix, causing a depletion of the aquifer's reactivity with time. We discuss conceptual simplifications of reactive transport that allow the implementation of a decreasing reaction potential in reactive-transport simulations in chemically and hydraulically heterogeneous aquifers without relying on a fully explicit description. We replace spatial coordinates by travel-times and use the concept of relative reactivity, which represents the reaction-partner supply from the matrix relative to a reference. Microorganisms facilitating the reactions are not explicitly modeled. Solute mixing is neglected. Streamlines, obtained by particle tracking, are discretized in travel-time increments with variable content of reaction partners in the matrix. As exemplary reactive system, we consider aerobic respiration and denitrification with simplified reaction equations: Dissolved oxygen undergoes conditional zero-order decay, nitrate follows first-order decay, which is inhibited in the presence of dissolved oxygen. Both reactions deplete the bioavailable organic carbon of the matrix, which in turn determines the relative reactivity. These simplifications reduce the computational effort, facilitating stochastic simulations of reactive transport on the aquifer scale. In a one-dimensional test case with a more detailed description of the reactions, we derive a potential relationship between the bioavailable organic-carbon content and the relative reactivity. In a three-dimensional steady-state test case, we use the simplified model to calculate the decreasing denitrification potential of an artificial aquifer over 200 years in an ensemble of 200 members. We demonstrate that the uncertainty in predicting the nitrate breakthrough in a heterogeneous aquifer decreases with increasing scale of observation.

Plain Language Summary Many aquifers can break down nitrate that was introduced into the groundwater by farmers. Unfortunately, this natural ability is lost over time while the reaction takes place. When modelling this process, one faces the difficulty that many parameters are not or only little known. That's why we have to run many models with different parameters. In existing models, the decreasing ability to break down the nitrate could not be accounted for. In our model, we look at the important factors controlling this process. By that, we can account for the decreasing reaction potential. And we are still able to calculate this so fast that many model runs are possible. We believe that this is an important contribution to the existing models because it helps to make better predictions about nitrate in groundwater and to get a better understanding of what is really happening.