Aquifer Water Level Prediction Using Support Vector Machines Method

In this research, a new data-driven model called Support Vector Machine (SVMs) uses the initial water level measurements, production well extractions, and climate conditions to forecast the final water level elevation in multi-time scale (i.e. daily, weekly, bi-weekly, monthly and bi-monthly) at a specific monitoring well. Due to the fact that SVMs approach does not require the explicit characterization of the physical onditions and input parameters, simulation is made based on the easily quantifiable and measurable variables. This study will demonstrate the prediction capability of SVMs compared to that of ANNs in forecasting the aquifer Water Level Elevation (WLE).

Keywords:

Support vector machines , ANNs , Aquifer water level elevation , Climate conditions