Estimation of heterogeneous aquifer parameters using centralized and decentralized fusion of hydraulic tomography data

摘要

Characterization of spatial variability of hydraulic properties ofgroundwater systems at high resolution is essential to simulate flow andtransport phenomena. This paper investigates two schemes to invert transienthydraulic head data resulting from multiple pumping tests for the purpose ofestimating the spatial distributions of the hydraulic conductivity, K, andthe specific storage, S_s, of an aquifer. The two methods arecentralized fusion and decentralized fusion. The centralized fusion oftransient data is achieved when data from all pumping tests are processedconcurrently using a central inversion processor, whereas the decentralizedfusion inverts data from each pumping test separately to obtain optimal localestimates of hydraulic parameters, which are consequently fused using thegeneralized Millman formula, an algorithm for merging multiple correlated oruncorrelated local estimates. For both data fusion schemes, the basicinversion processor employed is the ensemble Kalman filter, which is employedto assimilate the temporal moments of impulse response functions obtainedfrom the transient hydraulic head measurements resulting from multiplepumping tests. Assimilating the temporal moments instead of the hydraulichead transient data themselves is shown to provide a significant improvementin computational efficiency. Additionally, different assimilation strategiesto improve the estimation of S_s are investigated. Results showthat estimation of the K and S_s distributions using temporalmoment analysis is fairly good, and the centralized inversion schemeconsistently outperforms the decentralized inversion scheme.