The Usage of Extended Non-Linear Filter Kalman in Improvement of Groundwater Simulation Results in Unconfined Aquifer

GMS, is the software that simulate groundwater flow, with using ModFlow based on finite difference method. With comparing the acquired result from GMS and observation data, the existence of error is clear and obvious. One of the methods that decrease simulation error is the usage of filter kalman algorithm. In this study, extended non-linear filter kalman was used in order to improve the simulation results in Birjand unconfined aquifer in southern Khorasan province. The results showed that using this algorithm revealed satisfactory results in prediction of groundwater head. As the mean error (ME), mean absolute error (MAE) and root mean square error (RMSE) are -0.246, 1.125 and 1.341 meter respectively. With influencing extended filter kalman on results these errors decrease to -0.007, 0.015 and 0.019 meter. Also with investigation the results of extended non-linear filter kalman and non-linear filter kalman, it was found extended non-linear filter kalman is better choice for reducing simulation errors.