Application of artificial intelligence for the management of landfill leachate penetration into groundwater, and assessment of its environmental impacts

The current research was an effort to simulate landfill leachate penetration into groundwater using fuzzy logic and neural network modeling approaches. The obtained models were used as efficient tools for predicting leachate penetration and assessment of its environmental impacts. The training procedures were successful for both neural networks and fuzzy models. The train and test models showed over 70 perfect matches between the observed and the simulated values. The coefficient of determination for train model by fuzzy logic was 0.99998, which was even more precise than neural networks. The introduced intelligent models were useful for examining environmental impacts of contaminants because they could simulate the concentration of contaminants with high accuracy. These models could discern the relation between concentration of leachate at a given depth and concentration of leachate in groundwater. The analysis of 14 input variables in the modeling process indicated almost the same results for both modeling approaches. The analysis of contaminants showed that the Molybdenum, Sodium and Chemical Oxygen Demand are the three most important variables in the simulation of leachate penetration into groundwater for the study area. It was observed that heavy metals should be monitored carefully when leachate penetrates into groundwater.