Organic Pollutant Transport in Groundwater Using Particle Filter
In this study, the Particle Filter approach was introduced for the estimation of BOD concentration and its first-order decay constant. A two-dimensional subsurface transport model with decay factor was developed. Random Gaussian errors were added to the numerical method result to simulate the observation data. Then, the Particle Filter recursive process was applied for the data assimilation. The performance consistency of the Particle Filter approach was evaluated by the root mean squared error (RMSE) for 20 runs. Furthermore, two sensitivity analyses were done to show the effects of observation error and the true value of k on the estimation. The results show that the Particle Filter approach can give accurate estimations of the BOD concentration and the first-order decay constant and that its performance is consistent for 20 runs.
KeywordsRoot Mean Square Error Data Assimilation Particle Filter Observation Error Water Quality Modeling
This work was sponsored by the Department of Energy’s Samuel Massie Chair of Excellence Program under grant number DE-FG01-94EW11425. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of the funding agencies.