Forecasting the upwelling phenomenon using an artificial neural network

Abstract

In this paper, we investigate the upwelling phenomenon using data of 97 monitoring stations in Ouargla and El Oued valleys located in the Low Septentrional Sahara south of Algeria. This research paper constitutes a contribution to the morphological, hydrological, hydrogeological study of the water table in order to understand the processes of upwelling groundwater. By using ArcGIS as a mapping tool, we worked on real UTM coordinates in X and Y for real data overlay drawn maps in clear and usable way of this phenomenon. On the other hand, we propose a new method based on neural network to model the level flctuation of the groundwater as well as to predict the evolution of the water table level. The obtained model allows us to warm this harmful phenomenon and plan sustainable solutions to protect the environment. The finding shows that the obtained model provides more significant accuracy rate and it drives more robustness in very challenging situation such as the heterogeneity of the data and sudden climate change comparing to the related research.