Title: Parameter estimation of aquifer transmissivity variogram using WLS, GA and PSO algorithms

Authors: Seiyed Mossa Hosseini

Addresses: Physical Geography Dept., University of Tehran, P.O. Box: 14155-6465, Tehran, Iran

Abstract: This study focuses on the parameters estimation of aquifer transmissivity (T) variogram; sill, nugget effect and range using the weighted least square (WLS), continuous genetic algorithm (CGA) and particle swarm optimisation (PSO) algorithms. Three types of semivariograms including the spherical, Gaussian and exponential have been used to select the best-fitted model for corresponded empirical values. The objective function for this optimisation procedure is minimising the sum of deviation squares between theoretical and the empirical semivariogram values. These models have been applied with the transmissivity data of Qazvin plane aquifer which is located in north of Iran. The kriged values of T over the aquifer indicate that the simulated results based on PSO have the lowest estimation variance and most reliability in modelling of spatial structure of transmissivity compared to other optimisation algorithms (GA and WLS). Results show that if the kriging technique combined with the PSO algorithm, it would promote to a promising tool in estimation of spatial structures of aquifer transmissivity and be effective in groundwater quality and quantity management.

Keywords: parameter estimation; variogram; genetic algorithm; particle swarm optimisation; PSO; transmissivity; Qazvin plane aquifer.

DOI: 10.1504/IJHST.2017.084138

International Journal of Hydrology Science and Technology, 2017 Vol.7 No.2, pp.103 - 123

Received: 11 Mar 2014
Accepted: 15 Aug 2014
Published online: 14 May 2017 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article