Modelling tool for predicting and simulating nitrogen concentrations in cold-climate mining ponds

A nitrogen model was developed with the aim to trace nitrogen cycling in a cold climate-mining pond at the Aitik copper mine in northern Sweden. The model contains 10 state variables and 19 nitrogen cycling reactions. The model also includes sediment and physical properties of the pond, such as evaporation, freezing and thawing. The model was written in Mathworks MATLAB and was calibrated and validated using environmental monitoring data for the clarification pond at the Aitik mine. The data used comprised monthly values of nitrogen speciation, phosphorous and water flow. The model accurately predicts ammonium (r2 = 0.84) and nitrate (r2 = 0.82) concentrations in a time series from February 2012-August 2014. The model did not accurately predict nitrate concentrations (r2 = 0.11), presumably due to high oxygen concentration in the pond water that prevented denitrification in the water column. The transport of organic material to the sediment was also limiting denitrification in the sediment. When allowing denitrification in the water column as well as increasing the rate of transport of organic material to the sediment the nitrate prediction capacity increased to a satisfactory level (r2 = 0.54). A sensitivity analysis for the system showed that the most sensitive reactions for the water column were oxic mineralisation as well as the nitrification rate.