Author(s): C. Sivapragasam; V. M. Arun; Nitin Muttil
Linked Author(s):
Keywords: Genetic programming; Ordinary kriging; Spatial interpolation
Abstract: Spatial estimation of rainfall has many vital applications in water resources management of a basin. Conventionally, stochastic methods such as kriging are widely used where the performance of the methods crucially depends on how the variogram model is constructed. In the recent past, attempts have been made to replace the traditional variogram models with universal function approximator based models such as Artificial Neural Network (ANN). In this study, a detailed investigation is done to assess the suitability of Genetic Programming (GP) based universal function approximator as a replacement for traditional variogram models and their performance is evaluated in estimating missing rainfall in two stations in Tamarabarani basin in Peninsular India. The study shows that GP based semi variogram seems to be a potential alternative to the conventional models. This model was further utilized to re-design the rain gauge network of the basin.
Year: 2011