On identifie simultanément les coefficients d'emmagasinement et de transmissivité hydraulique dans un écoulement souterrain gouvernent par une équation parabolique linéaire. Ces deux paramètres sont supposés être des fonctions constantes par morceaux en espace. Les inconnues du problème sont non seulement les valeurs de ces coefficients mais aussi la géométrie des zones dans lesquelles ces coefficients sont constants. Le problème est formule comme la minimisation d'une fonction de moindres carres calculant la différence entre les mesures et les quantités correspondantes évaluées avec la valeur courante des paramètres. L'objectif principal de ce travail est la construction d'une technique de paramétrisation adaptative guidée par des indicateurs de raffinement. L'utilisation d'indicateurs de raffinement, nous permet de construisons la paramétrisation de façon itérative, on allant d'une paramétrisation à une seule zone à une paramétrisation avec m zones où m est une valeur optimale à identifier. Nous distinguons les cas ou les deux paramètres ont la même paramétrisation et le cas où les deux paramètres ont des paramétrisations différentes. Pour améliorer la résolution du problème inverse d'estimation de paramètres, nous incorporons des estimateurs d'erreurs a posteriori. / We identify simultaneously storage and hydraulic transmissivity coefficients in groundwater flow governed by a linear parabolic equation. Both parameters are assumed to be functions piecewise constant in space. The unknowns are the coefficient values as well as the geometry of the zones where these coefficients are constant. This problem is formulated as minimizing a least-square function calculating the difference between measurements and the corresponding quantities computed with the current parameters values. The main point of this work is to construct an adaptative parameterization technique guided by refinement indicators. Using refinement indicators, we build the parameterization iteratively, going from a one zone parametrization to a parametrization with $m$ zones where $m$ is an optimal value to identify. We distinguish the cases where the two parameters have the same parameterization and different parameterizations.\\ To improve the resolution of the inverse problem, we incorporate a posteriori error estimations.

The overall goal of this dissertation is to determine and develop optimal strategies for inversely calibrating transmissivities (T), elastic and inelastic skeletal storage coefficients (Ske and Skv) of the developed-zone aquifer and conductance (CR) of the basin-fill faults for the entire Las Vegas basin, and to investigate future trends of land subsidence in Las Vegas Valley. This dissertation consists of three separate stand-alone chapters. Chapter 2 presents a discrete adjoint parameter estimation (APE) algorithm for automatically identifying suitable hydraulic parameter zonations from hydraulic head and subsidence measurements. Chapter 3 compares three different inversion strategies to determine the most accurate and computationally efficient method for estimating T and Ske and Skv at the basin scale: the zonation method (ZM), the adaptive multi-scale method and the Differential Evolution Adaptive Metropolis Markov chain Monte Carlo scheme (DREAM MCMC). Chapter 4 outlines a fine-scale numerical model capable of capturing far more hydrologic detail than any previously developed model of Las Vegas Valley The new model is calibrated using high-resolution InSAR data and hydraulic head data from 1912 to 2010. The calibrated model is used to investigate the influence of faults and their potential role on influencing clay thicknesses and land subsidence distributions, and to investigate future trends of land subsidence in Las Vegas Valley. / Ph. D.

From the Proceedings of the 1971 Meetings of the Arizona Section - American Water Resources Assn. and the Hydrology Section - Arizona Academy of Science - April 22-23, 1971, Tempe, Arizona / Much future computer modeling of the responses of groundwater to water development stresses may be poorly done if the errors and limitations of digital models are not fully appreciated by groundwater hydrologists. Two digital models were constructed of the Tucson basin, one with 1,890 nodes of 1/4 square mile area each and one with 509 nodes of 1 square mile each. The starting point for the digital model was the 2-dimensional, linear, parabolic, time-and space-invariant differential equation of incompressible flow through porous media. An explicit finite-difference equivalent was determined, and a set of 1,890 equations were put in implicit form and solved on a computer in less than 20 seconds at a cost of 2.00 dollars. The errors associated with the model are discussed. In deciding what new data collected in the Tucson basin would give the most improvement in the digital model, a statistical decision theory approach was utilized in which expected opportunity loss and expected worth of sample were calculated for 5 variables. The data was computed using about 110 seconds of computer time, costing about 13.00 dollars. This technique has the advantage of including basin dynamics in estimating worth of additional data by means of using the digital model to compute all values of predicted and 'true' water levels included in the loss function.

From the Proceedings of the 1974 Meetings of the Arizona Section - American Water Resources Assn. and the Hydrology Section - Arizona Academy of Science - April 19-20, 1974, Flagstaff, Arizona / The inner basin is a collapse and erosional feature in San Francisco Mountain, an extinct volcano of late Cenozoic age, which lies approximately eight miles north of flagstaff, Arizona. The main aquifer's coefficient of transmissibility is approximately 14,000 gallons per day per foot and the storage coefficient was 0.08. Aquifer boundaries increased rates of drawdown of water levels in the inner basin well field. Inner basin springs which issue from perched reservoirs are not affected by pumpage of inner basin wells. Recharge is greater than the average yield from springs and wells in the basin which has an average of 8,000 acre-feet of water in storage in the principal aquifer. A large amount of water is lost from the inner basin aquifer system via leakage into underlying fractured volcanic rocks. It is believed that a part of this water could be intercepted by pumpage from a well constructed in the interior valley.