Decision Tree for Measuring the Interaction of Hyper-Saline Lake and Coastal Aquifer in Lake Urmia

摘要

Lake Urmia is located in the North West of Iran. The hyper saline lake is drying up very fast and more than seventy percent of the water in the lake has vanished in recent years. In this research, the West and South banks of the lake's basin which is known as the West Azerbaijan province of Iran are studied. During the period from March 2001 to August 2011, six pilot stations for ground water near the lake shore were monitored. Correlation, cross-correlation, distribution, and regression analysis were done for lake and pilot stations. Several decision trees were fitted to the model and the most proper one was selected to test the hypothesis. Results show that the North West of the basin is the most interactive part of the ground water and the fitted decision tree model with randomly selected data is performing well.