Swarm Intelligence Based Inverse Model for Characterization of Groundwater Contaminant Source

Swarm Intelligence (SI) based inverse model is developed for the first time, to recover the release history of a groundwater contaminant plume from the forensic observations. The flow system is assumed to be one-dimensional, with the plume originating from a known single site. The contaminant transport model is solved using explicit finite-difference scheme for the forward solution. The inverse model is developed then by integrating the transport model with particle swarm optimization algorithm (PSO) to reconstruct the pollutant source time history. The inverse model based on standard PSO algorithm is analyzed. Further a global constraint, “smoothness” is imposed to convert the present ill-posed problem into a well-posed problem. To improve the efficiency of the algorithm, a perturbation is introduced to all the particles in PSO and the results were compared with the solver based on standard PSO using synthetic numerical example. The proposed inverse model is tested for the gradual release scenario. The proposed model outperforms its standard counterpart. The present solver is simple and alternative to the existing gradient based approaches.