Statistical analysis of head and transmissivity in natural aquifers: application to structure identification and transport prediction

摘要: Flow and transport in natural aquifers depend on the spatially variable transmissivity T, usually modeled as a space random function. As consequence, all derived quantites, like piezometric head H, are functions, their statistical moments those of T. The present work introduces methodology for identification logtransmissivity statistics that use measured heads. In particular, second order head-logtransmissivity cross-covariance CHY variogram ΓH can be conveniently used parameter inference. is characterized by three parameters: variance σYc integral scale IY correlated residuals, nugget σ2Yn represents variability over small support. Unlike previous studies pursue same objective through first approximation variance, emphasis here highly heterogeneous formations, large values σ2Yc. Under proposed methodology, indentifications achieved fitting experimental variograms with theoretical ones; latter means effective medium validated accurate numerical simulations, based analytic element method. applied to Eagle Valley basin west-central Nevada which numerous measurements available. σ2Yc computed directly, from data, agreed well variograms. A disparity prevails IY. prediction solute behavior aquifer then made inferred data results -xbDagan et al. [4]. Comparison also first-order analysis flow transport.