A HYBRID OPTIMIZATION APPROACH TO THE ESTIMATION OF DISTRIBUTED PARAMETERS IN TWO‐DIMENSIONAL CONFINED AQUIFERS¹

Abstract

In using non‐linear optimization techniques for estimation of parameters in a distributed ground water model, the initial values of the parameters and prior information about them play important roles. In this paper, the genetic algorithm (GA) is combined with the truncated‐Newton search technique to estimate groundwater parameters for a confined steady‐state ground water model. Use of prior information about the parameters is shown to be important in estimating correct or near‐correct values of parameters on a regional scale. The amount of prior information needed for an accurate solution is estimated by evaluation of the sensitivity of the performance function to the parameters. For the example presented here, it is experimentally demonstrated that only one piece of prior information of the least sensitive parameter is sufficient to arrive at the global or near‐global optimum solution. For hydraulic head data with measurement errors, the error in the estimation of parameters increases as the standard deviation of the errors increases. Results from our experiments show that, in general, the accuracy of the estimated parameters depends on the level of noise in the hydraulic head data and the initial values used in the truncated‐Newton search technique.

Parameter estimation in distributed hydrological modelling: comparison of global and local optimisation techniques

“…As for the effectiveness of this method, Figure 7 clearly illustrates that the convergence of the local method to suboptimal solutions cannot be avoided (similar results were found by Heidari and Ranjithan (1998)). This confirms the previous results: PEST still has convergence problems for the steady-state model, demonstrated, also in this case, by the very low probability of convergence to the global optimum.…”

“…In this respect, the performance of two different ways of merging the global and local methodologies is analysed and compared. Previous trials to develop hybrid, two-stage optimisation procedures based on merging global and local methods have already been made with RR models (Franchini and Galeati 1997;Kuczera 1997) and with GW models (Heidari and Ranjithan 1998). Apart from these works, according to the knowledge of the authors, no attempts to couple global and local optimisation methodologies have so far been conducted with distributed, integrated models.…”

Inverse modelling using PS-InSAR data for improved land subsidence simulation in Las Vegas Valley, Nevada

“…In order to get reasonable parameter estimates, prior information is commonly used during the regression(Hill and Tiedeman, ). These prior information may be obtained from pumping/recovery test (Heidari and Ranjithan, ). In this study, prior information is used for parameters T and S ke and S kv on the basis of the T and elastic and inelastic storage coefficient distribution map provided by Morgan and Dettinger (), where parameters were calibrated based on groundwater level data, land subsidence data, pumping‐test data and lithologic data interpreted by Plume ().…”

Calibration of subsurface batch and reactive-transport models involving complex biogeochemical processes

“…Notable exceptions have used nonlinear regression [5,24,36], genetic and evolutionary algorithms [12,13,40] or combinations thereof [73]. While application of heuristic algorithms to groundwater flow models is relatively uncommon (exceptions include [44,81,91]) and generally regarded as unnecessary, the numerous interdependent processes involved in SBRT (subsurface batch and/or reactive-transport) modeling suggest that the global-search capabilities of heuristic methods may be advantageous. Therefore, to guide future SBRT calibration efforts, this study presents a systematic evaluation of alternative search algorithms (i.e.…”