Fitness Evaluation While Using Contaminant Transport Models For Environmental Forensic Investigation

Pollution, especially groundwater pollution, is one of the major challenges faced by many countries of the world. Using 'polluter pay' principle in settling pollution issues, though is not a new practice, is gaining wider acceptance these days. Scientific techniques capable of fixing responsibility of pollution and meeting the evidentiary requirements in a court of law are important in making a polluter pay for the damages he has made to the environment. In certain cases of groundwater pollution, using mathematical models is the only option in identifying the pollution source(s) and its characteristics. Use of numerical groundwater contaminant transport models in predicting pollution source characteristics involve using them repetitively along with an optimization algorithm until the concentrations predicted by the model match the observed concentrations within a margin of error. But, this matching procedure becomes complex as the model time and observation time are rarely the same. This paper proposes a method to match the modeled and observed concentrations for designing a fitness function. Further, it also explains how the source characteristics are decoded from the solution given by the optimization technique. MT3DMS was the contaminant transport model and genetic algorithm was the optimization algorithm used in the study.