An Exploration and Prediction of Rainfall and Groundwater Level for the District of Banaskantha, Gujrat, India
Keywords:
Machine learning, SARIMA, Regression, Groundwater, Rainfall.Abstract
The groundwater level is declining all over the world, especially in India. Some states, such as Rajasthan
and Gujarat, are experiencing very low levels of groundwater. In this study, we explored the rainfall pattern and
groundwater level of the Banaskantha district of Gujarat and predicted a rise in the groundwater level using
SARIMA, multi-variable regression, ridge regression, and KNN regression. The results indicated that the water
table is declining for Banaskantha, but due to an increase in irrigated land and good rainfall in previous years, the
decline in groundwater level is not alarming according to earlier studies. Ridge regression was found to be the
most suitable method, given the constraint of low data availability, especially in this particular study where the
remote area of Banaskantha, Gujarat was considered.



