Groundwater Level Prediction Based On Rainfall And Groud Water Utilization

  • K. Chinnathambi, Dr. M. Punithavalli

Abstract

Groundwater is a valuable water resource for various human activities. Hence, groundwater should be systematically managed for its sustainable use. The variability in the groundwater level depends on the different factors such as drought, water consumption and desalinating water techniques. The prediction of groundwater is very critical and operational activities undertaken by meteorological services worldwide.This task is more complicated since all decisions are to be taken in the features uncertainty. So, efficient techniques are required for Groundwater Level Predictions (GLP). In this article, regression techniques are used for GLPbased on water usage and rainfall data. From AQUASAT which is a global information system, the information water resources, rainfall and water usage are gathered. The gathered information are processed in various regression techniques such as Non-Linear Support Vector Regression (NL-SVR), Random Forest Regression (RFR) and Improved RFR (RFR) for prediction of groundwater level of upcoming year. The NL-SVR is developed from statistical learning theory and it can be utilized for GLPaccording to the training that utilizes past data. In RFR, a regression tree is constructed where each leaf node is the output of GLP and non-leaf node contains a set of decision rules. IRFR is an improved version of RFR where unbalanced dataset can be handled effectively. Initially in IRFR, important features are chosen based on its importance and those features are used in RFR. After that, F-measure value of each decision is computed as weight that is used to create a prediction model for GLP.  

Published
2020-03-30
How to Cite
K. Chinnathambi, Dr. M. Punithavalli. (2020). Groundwater Level Prediction Based On Rainfall And Groud Water Utilization. International Journal of Advanced Science and Technology, 29(3), 11566 - 11575. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/29827
Section
Articles