Please use this identifier to cite or link to this item:
|Title:||Critical review of applications of artificial neural networks in groundwater hydrology|
|Citation:||Proceedings of the 12th IACMAG: Geomechanics in the Emerging Social & Technological Age, Goa, India, 1-6 October 2008, 2463-2474|
|Abstract:||Artificial Neural Networks (ANNs) are massively parallel distributed processors made up of simple processing units which store knowledge of the phenomenon under consideration without explicit physical consideration, which can be used as and when required. Water resources are important asset of any country and are facing worldwide shortage. In such a situation planned use of groundwater resources can give sustained relief. To avoid over exploitation, an effective groundwater management policy aimed at promotion of efficiency, equity and sustainability is required. Efficient management of ground water resources requires a thorough understanding of the often complex confined, unconfined leaky aquifer systems and its adequate conceptualization for its assessment and predictions. Most of the ground water hydrology problems involve processes which are nonlinear, complex, multivariate with parameters having spatial and temporal variability. These are expressed by complex partial differential equations which are normally solved with approximations. ANNs do not require these governing equations which are constructed on simplifying assumptions. From the review of the literature authors consider that ANNs may provide better alternative to the conventional numerical computational techniques in groundwater hydrology. Main advantages of ANNs can be stated as ease in application, reduced data requirement in some cases lesser computational burden and improved result accuracy than the conventional methods (Jain and Deo, 2004). The technique can be used for function approximation, optimization, system modeling (prediction and forecasting) and pattern recognition (classification problems). This is an attempt to examine applications of ANNs in groundwater hydrology in various problems. Some practical guidelines regarding network design are put forth for new modelers. The technique is critically evaluated by studying its advantages along with demerits for applications in groundwater hydrology with a discussion on its future scope of research.|
|Appears in Collections:||Proceedings papers|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.